Datasets:

sahithya20

/

tech

like 0

Snapchat now has more than 2 million paid subscribers

Ivan Mehta

Snapchat has on its Snapchat+ program, the company said during its latest earnings report. Snap noted that the paid plan, which costs $3.99 a month, is a part of “diversifying” its revenue pipelines. The social network first launched in June 2022 in countries like the U.S., Canada, the U.K., France, Germany, Australia, New Zealand and the United Arab Emirates. Later, it expanded the subscription service to , Kuwait, Qatar, Oman, Bahrain, Egypt, Israel, Denmark, Norway and the Netherlands. Snap was able to attract . So comparatively it has taken more time for the company to reach 2 million paid users. Snapchat+ offers users features like the ability to pin someone as their No. 1 friend; priority in replies to Snap Star, the company’s program for popular creators; and the ability see “the general direction of travel for where friends have moved recently” if they have turned on location sharing. Snapchat+ In the last quarter, the company introduced new things to the subscription, such as , . According to data from analytics company Sensor Tower, Snap’s subscription was than the legacy Twitter Blue subscription. Up until October’s end, consumers had spent a total of $6.4 million on the Twitter app. In comparison, Snap’s paid plan generated $28 million after its launch, with many folks opting for the $39.99 annual plan. While the Elon Musk-led company has launched that costs $8 per month, there are no estimates of how many people are actually paying for it. Snap registered a revenue of $1.3 billion in Q4 2022, which was marginally higher than $1.298 billion for the same period in 2021. The company noted that its annual revenue jumped 12% from $4.1 billion to $4.6 billion. It also had a 17% bump in daily active users to get to the 375 million mark. However, Snap posted a loss of $288 million in the quarter, having earned a net $23 million in 2021 for the same period. The social media platform said that the company’s revenue could decline by up to 10% year-on-year because of stiff competition from apps like TikTok. It is holding a Snap Investor Day event on February 16 to outline its plan for the future. The company has been pushing actively to achieve revenue growth. Apart from the subscription plan, the social network has also tested . What’s more, a report from Business Insider published Tuesday indicated that Snap could also .

Peacock kills its free tier option for new customers

Aisha Malik

Peacock is no longer offering its free tier to new customers, a spokesperson for NBCUniversal confirmed to TechCrunch on Tuesday. The company says it’s shifting its focus to its Premium offering and that doing so will allow the streaming service to remain competitive in the marketplace. The change was first reported by . The free tier is still available to users who are already on the plan, the spokesperson said in an email. In addition, users who cancel their paid subscriptions will automatically be downgraded to the free tier. Peacock has offered the free tier since its launch in 2020, giving users restricted access to the streaming service’s content catalogue. The tier included a limited amount of content when compared to the paid tiers. New customers will now have to choose between Peacock’s Premium or Premium Plus tiers. The ad-supported Premium tier costs $4.99 per month and includes the full content library, live sports and NBC and Bravo shows after they air on TV. The Premium Plus tier costs $9.99 per month and offers an ad-free experience, offline viewing support and a . Peacock launched around the same time as Disney+, HBO Max and Apple TV+ but was the only one to offer a free tier. Now the company’s focus is shifting. The spokesperson told TechCrunch that Peacock is now focusing on its Premium offering, noting that the paid subscription is more reflective of its brand. The move comes as Peacock recently reported its since its launch. The streaming service added 5 million paying subscribers in its , bringing the total to 20 million, up from the in the previous quarter. The boost in paid subscribers was primarily due to the  , which streamed in Spanish on Peacock Premium and Telemundo.

OpenAI releases tool to detect AI-generated text, including from ChatGPT

Kyle Wiggers

After telegraphing the move in , OpenAI has a tool that attempts to distinguish between human-written and AI-generated text — like the text produced by the company’s own and models. The classifier isn’t particularly accurate — its success rate is around 26%, OpenAI notes — but OpenAI argues that it, when used in tandem with other methods, could be useful in helping prevent AI text generators from being abused. As the fervor around generative AI — particularly text-generating AI — grows, critics have called on the creators of these tools to take steps to mitigate their potentially harmful effects. Some of the U.S.’ largest school districts have ChatGPT on their networks and devices, fearing the impacts on student learning and the accuracy of the content that the tool produces. And sites including from sharing content generated by ChatGPT, saying that the AI makes it too easy for users to flood discussion threads with dubious answers. OpenAI’s classifier — aptly called OpenAI AI Text Classifier — is intriguing architecturally. It, like ChatGPT, is an AI language model trained on many, many examples of publicly available text from the web. But unlike ChatGPT, it’s fine-tuned to predict how likely it is that a piece of text was generated by AI — not just from ChatGPT, but any text-generating AI model. More specifically, OpenAI trained the OpenAI AI Text Classifier on text from 34 text-generating systems from five different organizations, including OpenAI itself. This text was paired with similar (but not exactly similar) human-written text from Wikipedia, websites extracted from links shared on Reddit and a set of “human demonstrations” collected for a previous OpenAI text-generating system. (OpenAI admits in a , however, that it might’ve inadvertently misclassified some AI-written text as human-written “given the proliferation of AI-generated content on the internet.”) The OpenAI Text Classifier won’t work on just any text, importantly. It needs a minimum of 1,000 characters, or about 150 to 250 words. It doesn’t detect plagiarism — an especially unfortunate limitation considering that text-generating AI has been shown to the text on which it was trained. And OpenAI says that it’s more likely to get things wrong on text written by children or in a language other than English, owing to its English-forward dataset. The detector hedges its answer a bit when evaluating whether a given piece of text is AI-generated. Depending on its confidence level, it’ll label text as “very unlikely” AI-generated (less than a 10% chance), “unlikely” AI-generated (between a 10% and 45% chance), “unclear if it is” AI-generated (a 45% to 90% chance), “possibly” AI-generated (a 90% to 98% chance) or “likely” AI-generated (an over 98% chance). Out of curiosity, I fed some text through the classifier to see how it might manage. While it confidently, correctly predicted that several paragraphs from a TechCrunch article about Meta’s and a snippet from an OpenAI support page weren’t AI generated, the classifier had a tougher time with article-length text from ChatGPT, ultimately failing to classify it altogether. It did, however, successfully spot ChatGPT output from a Gizmodo about — what else? — ChatGPT. According to OpenAI, the classifier incorrectly labels human-written text as AI-written 9% of the time. This mistake didn’t occur in my testing, but I chalk that up to the small sample size. OpenAI On a practical level, I found the classifier not particularly useful for evaluating shorter pieces of writing. Indeed, 1,000 characters is a difficult threshold to reach in the realm of messages, for example emails (at least the ones I get on a regular basis). And the limitations give pause — OpenAI emphasizes that the classifier can be evaded by modifying some words or clauses in generated text. That’s not to suggest the classifier is useless — far from it. But it certainly won’t stop committed fraudsters (or students, for that matter) in its current state. The question is, will other tools? Something of a cottage industry has sprung up to meet the demand for AI-generated text detectors. ChatZero, developed by a Princeton University student, uses criteria including “perplexity” (the complexity of text) and “burstiness” (the variations of sentences) to detect whether text might be AI-written. Plagiarism detector is developing its own AI-generated text detector. Beyond those, a Google search yields at least a half-dozen other apps that claim to be able to separate the AI-generated wheat from the human-generated chaff, to torture the metaphor. It’ll likely become a cat-and-mouse game. As text-generating AI improves, so will the detectors — a never-ending back-and-forth similar to that between cybercriminals and security researchers. And as OpenAI writes, while the classifiers might help in certain circumstances, they’ll never be a reliable sole piece of evidence in deciding whether text was AI-generated. That’s all to say that there’s no silver bullet to solve the problems AI-generated text poses. Quite likely, there won’t ever be.

Energy X secures $20M at $120M valuation to slash building sector emissions

Kate Park

  , — Energy X was founded in 2019 by co-CEOs Sean Park and Tom Hong. That’s when the duo pivoted from their first startup — a sustainable architecture crowdfunding platform, called Xquare. “Most building owners are not experts in building management, and they don’t know how to use it because building energy management systems (BEMS) are software-based,” Park said. In addition to the marketplace, “Energy X provides cloud-based BEMS where our AI manages, maintains and optimizes the system at all times without always having to monitor, manage or control directly,” said Park. The new round, led by Shinhan Financial Group, brings the startup’s total funding to approximately $31.5 million. The funds will help Energy X expand its marketplace and energy efficiency tech, grow its team from 86 to 200 employees this year and launch in Japan, Park said. The startup will open an office in Japan in February, Park added.

Egyptian financial services provider MNT-Halan valued at $1B in $400M funding

Tage Kene-Okafor

Egyptian fintech and e-commerce ecosystem has raised up to $400 million in equity and debt financing from local and global investors as it continues to serve underbanked and unbanked customers in the North African country. The round includes $260 million in equity financing and $140 million through two securitized bond issuances secured within the past year, investments that will now see MNT-Halan command a post-money valuation of about $1 billion. A large chunk of the equity, about $200 million, was provided by Abu Dhabi–based Chimera Investments. The investment firm invested that amount in exchange for 20% of the Egyptian digital lender and e-commerce platform, which is also in advanced stages of raising $60 million in additional capital in the coming weeks. Last week, the IFC in the company, but MNT-Halan declined to comment; it’s expected that the remaining financing will come from existing shareholders. In a statement, MNT-Halan says the investments “demonstrate continued confidence in its value proposition, management team, and superior technology.” The company also plans to expand internationally after solid growth in Egypt and progress on the swap agreement between super app Halan and Netherlands-based microlending platform MNT Investments. In 2021, Halan, operating a digital wallet that offered bill payments, e-commerce and ride-hailing as well as micro, nano and consumer loans, entered into a swap agreement with MNT Investments (a microlending platform operating in Egypt with roots dating back to 2010) to provide financing solutions to the underbanked and unbanked. The leveraged buyout deal, which was formed in 2018, saw both companies adopt a new name: MNT-Halan. Headquartered in Egypt, its digital ecosystem connects consumers, merchants and micro-enterprises with business loans, consumer finance, payments, BNPL and e-commerce offerings, all backed by Neuron, its proprietary technology. Last year, MNT-Halan from private equity firms, including Apis Growth Fund II, Development Partners International (DPI) and Lorax Capital Partners, and venture capitalists such as Middle East Venture Partners, Endeavor Catalyst and DisruptTech. At the time, it had served over 4 million and disbursed more than $1.7 billion worth of loans since inception. CEO , who founded the company with , said MNT-Halan continued where it left off and is presently Egypt’s largest lender to the unbanked: Total loans disbursed now exceed $2 billion per the company’s website (MNT-Halan issued loans north of $65 million last month). On average, businesses access $1,000 worth of loans while paying a 25% annual interest on the platform; Nakhla noted the fintech maintains a healthy nonperforming loan ratio without disclosing its figure. The two securitizations, totaling $140 million, that MNT-Halan are behind its impressive lending operations. The fintech’s wholly owned subsidiary, Tasaheel, managed to secure these funds locally via a securitization program with the Commercial International Bank (CIB), Egypt’s largest private sector bank. It can further securitize up to $250 million, the company said. In addition to CIB, participating regional and local financial institutions include Abu Dhabi Commercial Bank, Al Ahli Bank of Kuwait, Al Baraka Bank and National Bank of Egypt. It’s been demonstrated that lending is MNT-Halan’s primary business and main revenue generator; however, what’s interesting about the company is how it has layered a digital ecosystem of products, including e-commerce, FMCG delivery and mobile POS payments that feed its lending operations. To paint a picture: Last June, the five-year-old company , a B2B e-commerce platform that offers FMCG supplies directly to small merchants and retailers with next-day delivery. Nakhla tells me that this acquisition has allowed MNT-Halan to provide loans to these merchants or grocers, who then, in an , act as mobile agents to individual customers who frequent their shops. The company also wants to extend grocery shopping — in addition to other e-commerce stores selling electronics and personal items — to individual customers. “We’re capitalizing on our existing distribution through million-plus customers and adding services within our ecosystem,” said the chief executive. “If you need a loan for your business, we’re going to give you one; you need a loan for consumption, we’re going to give you one; you need to order groceries or buy a mobile phone on our platform, we’ll deliver it to you via our e-commerce stores. Also, we can give them the credit they can use to make all of these purchases within the ecosystem.” MNT-Halan lends to single small business owners or individuals who need lending to manage their businesses. According to the Egyptian startup, its digital ecosystem serves more than 5 million customers in Egypt, of which 3.5 million are financial clients and over 2 million are borrowers. The startup plans to launch a debit card for its customers by the end of March. Nakhla noted that due to the company’s focus on commerce and lending, it’s had to shut down its ride-hailing operations, one of Halan’s core offerings — before the merger — which mostly lagged international mobility outfits like Uber, Careem and inDriver. Meanwhile, MNT-Halan faces competition from , and across its other product offerings in Egypt. “In some sectors, we do have competition. But in the most important sector, we’re the largest, and no one is as advanced in technology or creates a fully-fledged ecosystem for the underbanked. I think this is where we differentiate ourselves from any other player in the market,” said the chief executive when asked about competing players in Egypt, while adding that the company is exploring a couple of mergers and acquisitions to consolidate its position in the country’s fintech and e-commerce space. For MNT-Halan to raise this sum in the current venture capital climate, it had to increase its revenues and open new streams, Nakhla noted in his statement. The fintech claims to have made over $300 million in revenue last year, representing a modest 3.4x multiples on its unicorn valuation which aligns with the present public market calculations as previously reported . On a related note, MNT-Halan is Egypt’s only private billion-dollar company; payments giant Fawry achieved that valuation after going public in 2019 although it’s well off the mark now. “We are thrilled to be part of Egypt’s greatest fintech success story,” said Seif Fikry, CEO of Chimera Abu Dhabi, in a statement. MNT-Halan’s upward trajectory and momentum reflect the management team’s realization of its extraordinary vision to transform a high-touch business by seamlessly infusing an unparalleled proprietary tech platform while increasing product depth for its target customer segment.”

Daily Crunch: Cell network provider Google Fi confirms customer data breach

Christine Hall

As January is coming to a close, the TechCrunch team is firing on all cylinders (do we still say that, in a time of electric cars? What is a better expression these days?), with a wall of amazing content for you to download straight into your brain. We’ve picked the cream of the crop, even as we are further confused as to why there was cream on the crops in the first place. In summary, idiomatic English continues to confound even the biggest language nerds among us. — and There are some people on the internet who don’t want to be found. That seems to be the case for the elusive, mysterious owner of Stripper Web, a 20-year-old forum for exotic dancers and sex workers. With just one week of advance notice, the , erasing the decades-long digital footprint of a community on the margins. ’s feature story tries to get to the bottom of things and is fantastic — give it a read! This January, Germany’s largest vaccine maker, BioNTech, announced that it had agreed to acquire Tunisian-born and London-headquartered AI startup InstaDeep for up to £562 million, including a performance-tied £200 million tranche investment. argues that . Not enough to keep you busy? Well, here’s another handful: / Getty Images SaaS pricing comes in three flavors: the classic sales-led model, free trials that eventually force users to make a decision, or freemium plans that hopefully deliver enough value to keep users coming back. “Given the obvious differences between these models, choosing one should be fairly straightforward,” writes Konstantin Valiotti, product director of growth at PandaDoc. “However, current market conditions do not support having just a single model.” In this TC+ article, he explains how to identify the right time to roll out a freemium plan and, equally importantly, when to. He also includes a tactical framework for developing freemium products that includes use cases for limited and unlimited usage. “Every strategy is unique and depends on the company’s idea of how it wants to proceed,” writes Valiotti. “Therefore, you should consider freemium as an extension of your strategy and see if it is right for you.” Three more from the TC+ team: Finally someone is turning tablets into something you can use other than surfing the internet or watching Netflix. has your look at that turns your tablet or phone into a workstation. Meanwhile, ponders what would happen if China-based . Would it make a difference? And what kind of limitations would it have? Now here’s five more:

Tesla records $204M loss from bitcoin in 2022

Rebecca Bellan

Tesla recorded a $204 million impairment loss in 2022 on its bitcoin holdings, according to regulatory . The loss was offset by $64 million in profits from bitcoin trading, leaving the automaker with a net loss of $140 million. during the first quarter of 2021, stating that it believed in the longevity of the cryptocurrency. At the time, the company said bitcoin was a great place to store cash and still access it immediately, all the while providing a better return on investment than more traditional central banks. In fact, after its initial purchase, Tesla promptly trimmed its position by 10%, making the automaker a quick $101 million. In March 2021, CEO Elon Musk said as payments for Tesla vehicles, causing the price of the crypto to boom. A few weeks later, , expressing concern for the amount of energy needed to mine bitcoin. The price of bitcoin then subsequently sank. Last year, as the value of bitcoin began to plunge, Tesla sold 75% of its bitcoin holdings and used the proceeds to buy traditional currency. The automaker today owns about $184 million worth of bitcoin, and may be holding out for a rebound. The crypto market has experienced a . Today, bitcoin is trading at $23,045.50, which is about 66% below the all-time high of $68,789.63. Despite losing over half its value, bitcoin has remained resilient and may experience a comeback, especially as buyers spooked by the crypto market in general gravitate toward household names and more mature ecosystems, like bitcoin or ether. Tesla disbanded its press department and could not be reached for comment.

DOJ requests Autopilot, FSD documents from Tesla

Rebecca Bellan

The U.S. Department of Justice has asked Tesla for documents related to its branded Full Self-Driving and Autopilot advanced driver-assistance systems, the automaker disclosed in a securities filing. Tesla said  it “has received requests from the DOJ for documents related to Tesla’s Autopilot and FSD features. . . . To our knowledge no government agency in any ongoing investigation has concluded that any wrongdoing occurred,” Tesla noted in the 10-K filing that was posted Monday. Tesla has been under investigation by the DOJ for at least a year, last fall, citing three people familiar with the matter. It’s unclear if the DOJ’s request for documents is connected to that investigation, which was launched in late 2021 following more than a dozen accidents involving the active use of Tesla’s Autopilot system. Tesla vehicles come standard with a driver-assistance system branded as Autopilot. For an , owners can buy “full self-driving,” or FSD — a feature that CEO Elon Musk has repeatedly promised will one day deliver full autonomous driving capabilities. Neither one of these systems are self-driving. Autopilot and FSD are advanced driver-assistance systems that automate some driving tasks and still require the driver to be ready to take over at any moment. Autopilot keeps the vehicle centered in the lane, can automatically change lanes and maintains the proper distance from other vehicles in traffic. FSD has those features and more, including an active guidance system that navigates a car from a highway on-ramp to off-ramp and can navigate interchanges and make lane changes. Musk’s claims and promises of these systems, as well as the branding, has caught the attention of regulators. The DOJ’s inquiry reflects an uptick in regulator scrutiny of Tesla. The after specific comments and efforts were made to promote the vehicle’s “self-driving” capabilities. The investigation follows a testimony from a Tesla engineer claiming that a and that . Tesla has been investigated and sued by several agencies and individuals for its claims of self-driving. The National Highway Traffic Safety Administration (NHTSA) has opened a number of into Tesla for crashes involving Autopilot; the California Department of Motor Vehicles has accused Tesla of ; and the company for deceitful marketing. All of the attention hasn’t thwarted Musk. During Tesla’s fourth-quarter 2022 earnings call, Musk said “full self-driving is obviously getting better very rapidly.” In the past he has boasted that Tesla was close to “solving” full self-driving.

Spotify’s test of a Friends tab on mobile hints at expanded social ambitions

Sarah Perez

Spotify’s success with , designed for social sharing, may be pushing the company toward building more social experiences directly into its mobile app. The company for many months has been testing different iterations of a “friends activity” tab on its mobile app, and investors have now taken notice. During the streamer’s Q4 call earlier today, the company was asked to clarify some details about its social plans. Though Spotify CEO Daniel Ek declined to comment on the specific feature the investor asked about, he didn’t shoot down the idea of Spotify becoming a more social platform. Instead, he replied that social could become “a meaningful driver of creating an even stickier and more engaging experience” for the company. The exec was answering an investor question about Spotify’s recent tests of a “Friends tab,” which appears in the app’s bottom navigation bar for some subset of Spotify’s users. Though only an experiment at present, the test has gained many positive reviews from members of the test group. On Twitter, for example, Spotify users the company to please to the test group if they didn’t have access to the feature, or when they were removed from the test group and their Friends tab disappeared. Others have been asking Spotify when the feature would roll out more broadly. So far, Spotify has not made any public announcements about its plans to launch a Friends tab on mobile, nor has it responded to testers’ questions. The company, however, had signaled its interest in an expanded set of social features last year when it began This variation offered a dedicated place to view what music friends were streaming on the app as well as what playlists they’d recently updated. Spotify’s , but it has limited user access to that same activity on mobile devices. At the time, Spotify confirmed to TechCrunch it was in the early stages of testing the Community feature and provided no other information about its plans. Those tests have since evolved and the experience has been given its own “Friends” button in a prominent place in the app’s main navigation. According to , the Friends tab includes a “weekly picks” section at the top, in a Story-like format, followed by a feed of friends’ listening activity, much like you’d see in Spotify’s desktop experience. PUMPED about this new friends tab on Spotify Imagine if there was the ability to share songs with friends in-app… — Danny + Desatnik🇨🇦 (@datkeed) This version still may not be the final concept, nor is there any guarantee that the feature will definitely launch. On the investor call, Ek declined to comment on the Friends tab test specifically, saying the company runs a lot of experiments and “what you probably have seen is one of those experiments — and since we’re not committed to rolling that out, I don’t really have much of a sort of comment.” But he didn’t downplay the company’s interest in social overall, suggesting it remains an area of interest. “We’re committed to creating the best audio experience for consumers and creators in the world. And obviously, social could be a meaningful driver of creating an even stickier and more engaging experience,” the CEO clarified. The company, no doubt, has seen the traction its personalized year-end review called Spotify Wrapped brings to its service and wants to know if baking in more social features that are accessible year-round would have a similar impact. As the company noted during earnings, its eighth version of Wrapped broke new records, as 156 million monthly active users engaged with its content — a metric that was up 30% year-over-year. Wrapped also boosted other areas of Spotify’s business, it said. The company saw its highest-grossing merch sales for artists to date during the week Wrapped was live. And it drove a 2.7x increase in visitors to artists’ tour pages, while also growing user engagement across all regions and demographics. While Spotify’s Friends tab isn’t an introspective look at your own listening behavior, it would be another way to engage with friends in a social environment. One earlier version of the Friends tab even showed directly in the app, in addition to tracking new songs friends added to playlists and those that they had “on repeat.” hey pls put me back in the friends tab a/b test, I am begging you 🙏 — Katherine Champagne (@keccers) Spotify needs to build out its own social experiences as the youngest generation of consumers is shifting away from using traditional social networks, where they build out a friend graph, to instead spend more time on entertainment apps, like TikTok — which has proven to have powerful influence over music charts. Although Spotify hasn’t prioritized its Friend Activity features for years, it has maintained a close relationship with Meta for social integrations. Spotify users’ social graphs today continue to rely on Facebook — even though many Gen Z users a Facebook account. The companies also worked together in prior years, including in 2021 when they partnered on a mini-player that  Facebook, however, competed with Spotify on podcasts for a brief period before   to focus instead on its metaverse efforts. Social is an area where Spotify may be hesitant to rock the boat, given Meta’s competitiveness with any company that tries to build its own social graph. But it’s an area that’s ripe for development, as Apple has to build social products around music over the years. Plus, it’s clearly something many users want to see in the app and one that plays into Spotify’s overall goal of offering a personalized experience for music fans. Spotify had no further comments on the Friends tab test.

Meta starts testing ‘members-only worlds’ in Horizon Worlds

Aisha Malik

Meta is starting to test closed spaces called “members-only worlds” in Horizon Worlds, its social VR experience. The company has begun a limited alpha test to give creators the ability to grow and moderate their own communities. Meta has selected a small group of creators to build and obtain feedback about members-only worlds. In a , Meta explained that creators can hand-select members and offer them exclusive experiences. During the alpha test, each members-only world can have up to 150 world members and 25 concurrent visitors at any given time. With members-only worlds, creators can launch a dedicated space to do things like host a book club, gather a gaming group, organize a support group or just hang out with friends and family without having to worry about uninvited guests. Meta “Every community develops its own norms, etiquette, and social rules over time as it fosters a unique culture,” Meta explained in its blog post. “To enable that, we’ll provide the tools that allow the creators of members-only worlds to set the rules for their communities and maintain those rules for their closed spaces. Creators can choose whether or not to share their moderation responsibilities with other trusted group members and decide if they’ll allow members to visit the world without a creator or moderator present.” The idea of members-only worlds in Horizon Worlds is likely a welcome addition for users of the platform. It’s no secret that Horizon Worlds can sometimes create unsafe environments for users. After reports that women were    and  in Horizon Worlds, the company a “Personal Boundary” feature that creates a bubble of space with a radius of two virtual feet around each avatar. The new members-only worlds could be seen as another way for Meta to address these issues. The launch of the new test comes as Meta in Horizon Worlds in September. Personal space gives users a place where they can hang out, play minigames, or invite friends over before heading to an event. Meta said last year that Horizon Worlds will be and mobile in the future. Now the company says the VR experience will be available on these platforms “soon.” By launching Horizon Worlds on more platforms, Meta will make it a lot more accessible, as it’s currently only available on the company’s own Quest VR headsets.

GM invests $650M in lithium mining to lock down EV raw materials

Kirsten Korosec

General Motors said Tuesday it will invest $650 million into Lithium Americas as part of an agreement to develop a mine in Nevada, the latest effort by the automaker to lock down a supply chain of key components needed to produce millions of EVs. The investment in raw materials for batteries is the biggest to date, according to GM. And it’s no wonder. General Motors has a litany of all-electric sedans, SUVs, crossovers and trucks that are coming to market in the next two years, including the , Cadillac Lyriq, Chevrolet Silverado EV, Chevrolet Blazer EV and . GM said in November it expects to generate from sales of its 30 EV models in 2025, with profit margins in the low to mid-single digits. But that can’t happen if supply chain constraints prevent GM from producing the vehicles. The semiconductor chip shortage that kicked off in 2020 is an experience that every automaker is keen to avoid, especially as the industry transitions to EVs. GM and other automakers have made to bring production to the U.S. and ensure enough cells are available. GM has taken it a step further by locking up a supply of lithium, a key component in EV batteries. Lithium Americas estimates the lithium extracted and processed from the project at Thacker Pass mine can support production of up to 1 million EVs per year. Production at Thacker Pass is projected to begin in the second half of 2026. Lithium Americas expects Thacker Pass to create 1,000 jobs in construction and 500 in operations. “GM has secured all the battery material we need to build more than 1 million EVs annually in North America in 2025 and our future production will increasingly draw from domestic resources like the site in Nevada we’re developing with Lithium Americas,” said GM Chair and CEO Mary Barra. “Direct sourcing critical EV raw materials and components from suppliers in North America and free-trade-agreement countries helps make our supply chain more secure, helps us manage cell costs, and creates jobs.” The investment will be split into two tranches. The first will be released to Lithium Americas if certain conditions are met, including a pending ruling in U.S. District Court. The second amount will be made after Lithium Americas separates its U.S. and Argentina businesses, the companies said.

TechCrunch+ roundup: SaaS spending squeeze, tax time tips, freemium frameworks

Walter Thompson

If you thought egg prices were out of control, the SaaS inflation rate is outpacing the rest of the economy. In the U.S., SaaS expenses are growing 3.5x faster than market inflation. In Australia and the U.K., that rises to 5x, according to Eldar Tuvey, founder and CEO of Vertice. “If cutting SaaS costs is a top priority for your business in 2023, ,” he writes. Finding meaningful ways to save can extend your runway and even stave off layoffs: This article contains tactics and strategies for working with vendors to reduce contract length and arrange more favorable terms. “Negotiating each of the contracts that make up your SaaS stack will provide long-term savings by mitigating the effect of rising prices,” says Tuvey. Thanks for reading, Walter Thompson Editorial Manager, TechCrunch+ / Getty Images SaaS pricing comes in three flavors: the classic sales-led model, free trials that eventually force users to make a decision, or freemium plans that hopefully deliver enough value to keep them coming back. “Given the obvious differences between these models, choosing one should be fairly straightforward,” writes Konstantin Valiotti, product director of growth at PandaDoc. “However, current market conditions do not support having just a single model.” In this TC+ article, he explains how to identify the right time to roll out a freemium plan, and equally importantly, when not to. He also includes a tactical framework for developing freemium products that includes use cases for limited and unlimited usage. “Every strategy is unique and depends on the company’s idea of how it wants to proceed,” writes Valiotti. “Therefore, you should consider freemium as an extension of your strategy and see if it is right for you.” / Getty Images For SaaS startups, tax time can create a conundrum. Some states regard software-as-a-service products as, um, services, while others classify them as, er, products. “There’s also the issue of bundling on its own,” according to startup tax accountant Ardy Esmaeili. “SaaS might not be taxed, but it will be when paired with hardware.” To help founders better understand their liability, Esmaeili shares tips on how to identify a company’s physical nexus and lists multiple SaaS categories that states are likely to tax. “Engage an expert as early as you can,” he writes. “Don’t think you won’t have to worry about it yet, because waiting can have big consequences down the line.” Early-stage startups that adopt a product-led growth strategy may not need sales teams to build their customer base. Before his company was acquired by Salesforce, Slack CEO Stewart Butterfield said, “I think we can get away without having a sales team in any kind of traditional way, probably forever.” Today, Slack has an “expansion product team” that integrates product, data science and sales operations. According to Elena Verna, interim head of growth at Amplitude, “you’re starting to have product and sales work very closely together as a unit, like brother and sister.”

Avoiding the pitfalls of OnlyFans with Rosie Nguyen from Fanhouse

Rebecca Szkutak

Welcome back to Found, where we get the stories behind the startups. This week, and catch up with , the co-founder and CCO of , a startup that helps content creators monetize their work. Rosie talked about her initial reluctance to become an entrepreneur but how issues at other content platforms pushed her to launch Fanhouse. She also talked about intentional choices Fanhouse has made to protect its content creator customers and what it was like to raise as a founder without a traditional network. to hear more stories from founders each week. Connect with us:

Thrive Capital believed to be leading new multibillion-dollar investment in Stripe

Mary Ann Azevedo

Founded by Irish brothers John and Patrick Collison (the CEO), Stripe has raised more than $2.2 billion in funding since its 2010 inception from investors such as Allianz (via its Allianz X fund), Axa, Baillie Gifford, Fidelity Management & Research Company, Sequoia Capital, General Catalyst, Base Partners, GV and an investor from the founders’ home country, Ireland’s National Treasury Management Agency (NTMA).

Precision Neuroscience is making brain implants safer, smarter and reversible

Devin Coldewey

Brain researchers have relied on devices called microelectrode arrays for decades, but the technology behind these tools is increasingly outdated. Precision Neuroscience is building a modern alternative that’s not only an order of magnitude better, but far less invasive to put in. With a in the bank, they’re all set to embark on the complex path to market. In order to understand what’s going on in the brain, sometimes an EEG or MRI from outside it just isn’t enough — you need to really get in there. Implanted electrodes have been used for this purpose for a long time, and arrays of them in formation are used to collect information from multiple points within the cortex at the same time. But while an electrode array a couple dozen strong is invaluable in a research setting, it simply isn’t enough for something like a functional brain-computer interface. The information density is too low for the patient to, say, control a prosthetic limb or move a cursor on the screen. And you can’t just add more electrodes: because each one pierces the brain tissue and necessarily causes a small amount of damage, going from an array with 100 to an array with 1,000 will cause 10 times the damage. Precision Neuroscience aims to solve both of these problems with the one major innovation: an ultra-thin electrode array that doesn’t need to pierce the brain at all, yet can collect hundreds of times more data than traditional arrays. It’s the brainchild, so to speak, of Dr. Ben Rapoport, a neurosurgeon by trade who has spent decades working on the idea, and co-founded the company in 2021. (He previously was on the founding team at Neuralink.) “This has been his life’s work,” said Michael Mager, Precision’s CEO. “His view has always been that even for basic functionality you need high electrode density, and the tech has to be deployable in a minimally invasive way, with no damage to the brain. Our hope is to scale to tens of thousands of electrodes — and you can’t just keep penetrating more and more tissue.” The array they’ve developed is called Layer 7, a reference to the fact that the cortex itself has six layers, which the interface sits on top of. A single Layer 7 array is bit bigger than a thumbnail, but it has 1,024 microelectrodes on it, producing a density hundreds of times better than what’s in general use today. And they’re designed to be used in arrays themselves, essentially tiling across a region of the brain. Each array would provide a fast, accurate picture of the activity of the cortical regions it covers. These capabilities and specs are impressive, but it is perhaps even more important that the interface can be implanted without a craniotomy — open brain surgery. Instead, the super-thin film-based Layer 7 can be inserted through a small incision in the skull — still brain surgery, to be sure, but a much less invasive technique that may not even require general anesthesia. It would attach to an exterior control unit, but the dimensions and specs of that device would vary depending on various factors. Two cool Precision Neuroscience employees. You can see the implant on the microscope slide. Precision Neuroscience Avoiding the risk and complications of major surgery is important because the populations who stand the benefit the most from a technology like this are people with existing neural issues. “There are tens of millions of people in the U.S. alone who suffer from stroke, TBI [traumatic brain injury], degenerative diseases… but for those patients there really are no medical solutions we can offer right now beyond physical therapy,” said Mager. “There are two broad use cases,” explained chief product officer Craig Mermel. Stimulation of the brain and a two-way interface is one of them, he said, but still highly experimental. “What we’re doing that has backing from research is more on the ‘record and decode’ side, using it to read info from people with epilepsy or stroke, and translate intent into motor or speech output.” This capability has been studied and successfully demonstrated in other contexts for years, but the holdup is that the implants themselves are “still research grade,” Mermel said. “Nobody has put this into a clinical grade system that patients could potentially benefit from. That this [i.e. Layer 7] doesn’t damage the brain is going to be an incredibly important aspect of our system. Every device will have a lifespan, and you’ll have to replace it; the fact that our interface is reversible and the brain can stay intact reduces the risk to the patient.” By now most readers will be wondering how this compares with Neuralink, the brain-computer interface company funded by Elon Musk. One important difference is that Neuralink’s approach still involves a craniotomy and brain-piercing electrodes — though finer and more sensitive that the ones currently used, and implanted via robot. But Precision Neuroscience considers the company a colleague rather than a competitor. “Honestly, what we say internally is they’re different approaches that will be optimal for different situations,” said Mager. “This is not going to be a winner-take-all market. There’s room for more than one company.” One of the biggest challenges when building a medical device of any kind, to say nothing of a brain implant, is the huge task of proving both the applications and safety before you go to market. And you can’t just build the gadget — it needs to be distributed, supported, documented, etc. “It’s not just the array, but the software — the sophistication of machine learning is a must have to drive really powerful BCI. It’s a full-stack product that requires an interdisciplinary team to develop,” Mager said. “And you have to take it through the FDA regulatory process.” On that side of things the company is taking a two-pronged approach. They are first focusing on short-term and emergency use, such as during a hospitalization — when understanding what’s going on in the brain could be a life-saving technique. They hope to submit their 510K application along these lines to the FDA within the year and be ready to go when the agency gives the green light. Longer term, the plan is proving out the safety of semi-permanent implantation: the kind where someone could use the implant all day every day for a year from home or traveling. That’s a different risk profile and a more stringent approval process. Precision’s Stephanie Rider examines the Layer 7 implant. Precision Neuroscience Such relatively long time horizons are common in medicine but less so in venture-backed startups. Why ask VCs when so many are interested in companies quicker and easier to scale, like software and services? “It was a huge mistake, we should have started a software company. I talk with Craig about this all the time!” joked Mager. “But really, despite the challenges and the time, there’s a group of VC firms that is not insignificant, that is excited in investing in companies looking to have a huge impact in human health and also build a large company — not in two-three years, but 10 years.” Here Mager gave credit to Musk for helping popularize the idea that venture capital can back large, long-term efforts like SpaceX and Tesla, not just fast-scaling software companies that sell in a year or two. A rocket company may not have seemed like a likely endeavor to be backed by venture capital 10 or 15 years ago, he said, but now no one questions it. The same may prove true for neural interfaces — “and we may create some meaningful clinical good in the meantime.” The $41 million B round will enable Precision to continue working toward its FDA clearance and further develop and support the Layer 7 stack, from hardware to training and customer service. The round was led by Forepont Capital Partners. Mubadala Capital, Draper Associates, Alumni Ventures, re.Mind Capital, Steadview Capital, and B Capital Group.

Teal unwraps $8.8M to build out a telehealth platform for women — starting with cervical cancer screening

Natasha Lomas

Female-focused telehealth startup, , is popping up today to announce an $8.8 million seed round with a roster of heavy hitting investors on board — including (Serena Williams’) Serena Ventures, ( Metrodora Ventures, and (Laurene Powell Jobs’) Emerson Collective. The February 2020-founded San Francisco-based startup’s first product will be a service that supports women to collect their own sample for cervical cancer screening in the comfort of their own home. It wants to tackle the problem of women not getting screened — either because the traditional route of going to a doctor’s surgery for a pap smear (using ) is uncomfortable or inconvenient or both. Teal has developed a novel device for women to self collect a sample to mail off for lab analysis. Its websites refers to this device as a “collection wand” — and we gather there’s a sponge involved — but details of what exactly it looks like and how it will function remain under wraps as the startup is still in the process of applying for FDA clearance, per Teal’s CEO, Kara Egan. She also can’t say when exactly they’ll be able to launch a service — as that depends on its application for “de novo” FDA clearance. “We hope to be in the market soon,” is all she’ll say on that. The startup previously raised $1M in pre-seed funding, back in early 2021, which it used to refine the design of the product — working with the IDEO design agency. “What makes our product unique, I would say, is the idea that it’s designed to be very intuitive and increase confidence and the accuracy of the sample,” Egan tells TechCrunch. “And that it gets pap smear — it gets cervical cells.” After a chat with her team members, to confirm what else they can say at this point, she also offers: “This design makes it simple for a woman to collect her own sample quickly and comfortably. The device is inserted similarly to a tampon, the device contains a soft sponge tip which is rotated to collect cells. The whole collection from undressing, reading instructions, collecting, and packaging to send to the lab via the mail should take less than 5 minutes.” Another important detail she can disclose is that Teal’s collection method will allow for samples to be tested for primary HPV and Pap cytology triage — meaning the startup will be able to support follow-on triage of women who do test positive for HPV (aka, the virus that’s linked to cervical cancer). So it can provide a fuller service for cervical cancer screening care. While it’s starting with cervical cancer, the broader mission for Teal is to build out a women’s telehealth platform in the US — which will offer a range of services that traditional healthcare might be happy to hand off to a dedicated female-focused provider. So the core focus for the startup is on developing a fully attentive, female-friendly service wrapper. Egan argues there’s huge potential to create a compelling, modern telehealth service for a population that’s typically been underserved by traditional healthcare. “We know that self collect will increase adherence — without a doubt,” she says. “But there’s an opportunity here to actually create something that women are missing… So much of healthcare is an inconvenience for people — and especially for women who are working and are mothers. They just put themselves last. So we’re kind of like hey — let’s design this and fully cater it to women.” “So many things have been designed by men for women,” she adds. “Women have hated this experience [smear testing] — you run the spectrum of hate it, fear it, literally don’t go because of it and tolerate it, but there’s no one who’s like oh that felt good or that was fine… So we have this opportunity to be like, hey women, for once we’ve built something for you — and also be like let’s design an experience that brings that back into their healthcare, that makes them feel trusted. “Don’t just throw it in a plastic bag and mail it to them. Use this as an opportunity to open the door back up and say it can be better from here on out — and that’s kind of what we think about. This is such an incredible opportunity to do something so important. Truly you can eradicate cervical cancer — with screening coupled with vaccine it doesn’t have to exist. That’s a goal; that really can happen. But — simultaneously — we can actually build something that helps the key decision maker in the household stay healthy and make healthier choices and do it in a modern way.” Egan used to be a VC and she also talks up the sheer commercial opportunity in smartly addressing women’s health. “I spent my time in healthcare investing before too and I’ve never seen an entry point like this. I’ve never seen a situation where it’s all women — 25 to 65 mandated screening — a universally disliked current experience and then an experience that the doctors are also willing to say if you can do it another way for them I’m happy to,” she says. “And it’s something so big and meaningful. It’s cancer — and we can really make a dent in it.” So how did Teal land such a line-up of high profile women investors for its seed round? What tips does she have for other female founders looking for help to get a great idea off the ground? “Having more women out there as decision makers in these funds, obviously, is so helpful,” she notes. “A lot of times women’s health is considered ‘small’ — like, are you kidding?! It’s half the market… So it’s just finding people. And then just like the general advice — it takes a while. Until you find your ones. It does take time even if you have a great [network].” As she chews over the question a little more, Egan lands on another tidbit of fundraising advice for pitching traditional VC that boils down to: Remember you need to pitch a company, even if (maybe if) the cause is great… “Truthfully, the big reason I think I was very successful at it is I was a VC before. So I understand how to create the slides that tell the story from the viewpoint of an investment. And I think that especially, for things like this that can be misconstrued as like a charity effort… So the point is when it comes to investing what I was lucky about is — and why I know this opportunity is do incredible is — I was able to frame it as an investment even though it’s also something so important,” she says. “And I think, sometimes founders, when they’re really working on something so important, it can come off as more of ‘a cause’ than a company.” Commenting on Women keen to be first in line for Teal can join its waitlist to get details

Superstrata e-bike review: Rebel without a cause

Taylor Hatmaker

specimen — there’s no two ways about it. The bike’s concept, borne out in seamless 3D-printed carbon fiber, springs from an equally strange premise. We’ll get into that. Talking to , Superstrata’s bike was crafted not out of a love of cycling, the hunt for a climate-friendly urban transportation solution or any traditional justification for going to all the trouble of making an e-bike. Instead, Vu created the (and its non-electrified counterpart) as a proof-of-concept for . In that light, how the bike wound up makes perfect sense — even if buying one probably doesn’t. “Everyone thought we’re an e-bike company, but we’re not,” Vu told TechCrunch. “We’re an advanced manufacturing company.” We certainly can’t blame e-bike enthusiasts for being confused. Superstrata is deep into what Vu calls “deep tech” — basically manufacturing processes so cutting edge that they haven’t even shown up in consumer products yet. “This isn’t your typical 3D printing system — it’s built for industrial speed and scale,” Vu said. Arevo, Superstrata’s parent company, is exploring aerospace applications in the future and potentially UAVs. Superstrata and Arevo spun up the e-bike as a consumer proof-of-concept to fill the gap while they wade through the regulatory red tape that defines more complex industries. The company’s bike frames are currently printed in Vietnam, though Vu has plans for print farms in the U.S. and Europe to reduce shipping times and generally make its whole carbon fiber printing operation more efficient. TechCrunch So, carbon fiber. Carbon fiber is one frontier of climate technology, promising fuel-saving lightweight materials at the cost of a fairly energy-intensive production process. Superstrata’s silky, unibody bike frames are made out of industrial-grade 3D-printed thermoplastic carbon fiber composite rather than “thermoset” — a more common polymer process. While Vu was eager to dive into the technical advantages and the manufacturing process, at the end of the day what you need to know is: This is an e-bike and its frame is made out of very fancy 3D-printed carbon fiber. Here’s the basic pitch. The Superstrata e-bike is a carbon fiber unibody bike that’s custom printed to meet your preferences, souped up with a 250W pedal-assist motor and shipped in an array of fun prints and colors. The sleek, angular design of the bike’s frame and its conspicuous absence of a seat tube — the part of the frame that would normally flow down from where the saddle sits — is pleasantly futuristic or downright odd, depending on who you ask. While early critics raised alarms about Superstrata’s missing seat tube, which would normally connect the top tube and down tube to form a strong triangle shape to hold the rider’s weight, Vu assured us that the single, seamless piece of carbon fiber is rated up to 750 lb and is more than strong enough to hold a cyclist. Riding the Superstrata, I wasn’t concerned about the strength of the carbon fiber, though the potential for smoothing the ride with adjustments involving a seat tube are a missed opportunity. The frame is strong and the ride is stiff — and that’s just how it is. (The frame’s design is eye-catching even without the missing tube, and does bear some striking similarities to . It’s possible that Superstrata printed those frames but the company isn’t listed anywhere and they do boast a much more traditional geometry.) The missing chunk of frame turns heads, but Superstrata’s other aesthetic choices also set the bike apart visually. The company is all about customization and that applies to the paint too. My review unit was an intense purple, maybe a periwinkle. Honestly, as someone who wears all black most days, the color hurt my soul a bit but my wife found it appealing. Some of the special paint jobs you can order are very cool — including two that look like starry skies and another designed to evoke the aurora borealis — but they’ll run you an extra $1,250. That extra cost would get you most of the way to a more affordable e-bike made by competitors like or . This isn’t a bike for the wallet conscious (most of us). Superstrata really needs to remove these misleading sample images with integrated lights. But: Cool paint. Superstrata That choice and other aspects of the Superstrata feel a bit “It’s one banana, Michael. What could it cost?” I’m not convinced the bike is really designed to be sold to much of anyone and that’s just a weird takeaway to have when reviewing something. Superstrata’s e-bike is obviously meant to present as a luxury product, but the experience of the ride and the deeper design doesn’t exactly give off a luxurious, cohesive vibe. At the end of the day, if you asked, What is special about this bike? the full answer is “the 3D-printed unibody carbon fiber frame.” At this price point — the e-bike starts at $3,500 — Superstrata’s base offering gives you a lot less for your money than what you’d get with a much more fully featured electric bike like the , or even its last-generation models, which float around $2,000. Those tech-forward bikes pack perks like built-in interfaces, companion apps, embedded automatic front and back lights, phone chargers and alarm systems for at or around what you’d pay for a base-price Superstrata bike. A feature that illustrates this well is the fact that Superstrata initially planned to have integrated front and back lights — many of the bikes pictured on its website still misleadingly show this — but that idea was scrapped in the finished version. The lights, the little computer for the electric motor and any bells and whistles are all aftermarket, not integrated into the bike’s design. After speaking to Vu — who was transparent to the point of admitting the whole e-bike idea was just kind of for the hell of it — it didn’t particularly feel like Superstrata was trying to actively mislead people about the headlight situation or anything else. The fact that Superstrata’s website shows a product you literally can’t buy just further demonstrates that it isn’t this company’s . The problem with that is that most people making a purchase this big would feel safer buying from a company that lives and breathes bikes — not carbon fiber. TechCrunch Back to the frame — the meat of the innovation here. The Superstrata bike is custom printed to order and that’s a huge boon for people who are on the extremes of the height spectrum, including adults under 5’2”, who apparently ordered the bike with gusto. For these folks, who are hard-pressed to find properly sized bikes elsewhere, Superstrata’s frame size options are probably genuinely a big deal. If you’ve searched high and low for a bike to fit your unusually small or large stature, Superstrata’s bikes might be a great choice. For anyone who falls in the normal-ish height spectrum, the rest of the customization process is relatively shallow. When relying on a bike for commuting or urban transportation purposes, the real customization options that matter offer utility — things like baskets, front or back racks and tire width. Building the bike, you can choose from an “urban” or “sport” configuration for the bars, yielding a casual upright ride or a more aerodynamic drop bar, road bike style. You can opt for a single speed or multiple gears, an option that most people would probably prefer but one that also adds $500. You can pick whether you’d want tires for the road or for “paths” — maybe gravel and light offroading — but there’s not much detail offered here. Superstrata , but the review unit we had packed Shimano disc brakes, Bafang powering the electric side of the bike and a grab bag of brands for the rest. After that stuff is dialed in, you can input your specific height and measurements for a custom-sized frame. On that count, we couldn’t really make a determination — my review bike was designed for someone a bit taller, though it was still rideable and okay. What else? The experience of riding the bike is fine, but nothing particularly sophisticated. The frame’s design makes for a very stiff ride, so watch out for being jolted by uneven terrain. The battery life is more than adequate for normal needs. Superstrata claims that it will last for 60 miles, but on a higher assist setting, you’re going to get significantly less than that. Still, the battery would serve you for at least 20 miles, even when drawing more power, which is more than fine for most around-the-town needs. It’s also worth mentioning here that the Superstrata e-bike doesn’t have a throttle — basically a button that gives you a burst of speed to power the bike along. A throttle is a really nice way to scoot quickly through dangerous intersections or to get your bike to maintain a higher speed and it’s tough to go back once you’ve used one. The Superstrata e-bike can go up to 20 miles per hour but you’ll have to work for it. On that note, Superstrata’s state-of-the-art frame might be carbon fiber, but the e-bike doesn’t exactly feel like a featherweight. Because the bike is so light in the front and so heavy in the back where the motor lives, it actually feels heavier than it is. It’s also more awkward to carry than a bike that’s uniformly weighted and you’re not going to want to be lugging this thing down more than a few stairs at a time. The Superstrata website claims that the e-bike weighs 24 pounds, but there’s no way this thing weighs that much less than my regular ride, a VanMoof X3 that’s a hefty but evenly weighted 45.8 pounds. Vu noted that the final version’s weight varies, but the e-bike weighs in around 38.5 to 39.6 lb, enough to pretty much obscure the weight savings of the carbon fiber. Ultimately, the Superstrata could be a solid option for someone who falls outside of normal height parameters and desperately wants an e-bike. Superstrata’s bikes might also be a good choice for someone who wants a custom-printed carbon fiber bike frame and is confident about dialing in the rest themselves, though buying the frame alone isn’t an option on the website. In either of these cases, the advanced carbon fiber technology doesn’t come cheap and neither does this bike — especially when compared to competitors building feature-rich, cohesive e-bike experiences. Superstrata should probably focus on its custom unibody carbon fiber frames and let other companies — or people — build the bikes out. And since this whole thing was an experiment anyway, that might very well be what the company plans to do.

Practice your startup pitch on TechCrunch Live with Benchmark and Cambly

Matt Burns

TechCrunch Live is back! The weekly show , and we’re excited to bring back a popular segment. Called Pitch Practice, it should be self-explanatory. Participants have a chance to practice their pitch by presenting to another founder and investor. This show’s guests are fantastic too. You want their feedback on your pitch. Our first guests are  , CEO and co-founder of Cambly, and  , a long-time investor at Benchmark and previously Greylock. They’re the perfect guests to kick off the third season of TCL, and they’re going to give three founders feedback on their pitches. TCL’s mission is still to help founders build better venture-backed businesses. But going into 2023, there’s new urgency behind this mission. TechCrunch Live started in the heady days of 2021, and now in early 2023, the startup world is experiencing new challenges. It’s harder to fundraise, sales cycles are much longer and investors (and their LPs) have different expectations. We’re looking for startup founders who have a well-rehearsed pitch for an early-stage startup. Not selected for this show? No worries; try next week. This segment is a regular feature of TechCrunch Live.

Privacy assistant Jumbo tears down its paywall

Romain Dillet

, an app that lets you control your privacy on the web, is hitting the reset button — sort of. While the company is still focused on privacy and security, users can now download and use all features for free as the premium subscription is gone. In addition to this pricing update, Jumbo’s newest version now includes free identity theft insurance for people based in the U.S. “Something we didn’t anticipate and that we’re tying to fix today is that a paid product creates an important barrier to entry,” Jumbo founder and CEO Pierre Valade told me. Valade previously founded , a popular calendar app that was acquired by Microsoft. Jumbo’s flagship feature is a dashboard that lets you control your privacy settings across various online services. You can use the app to connect to your Facebook, LinkedIn or Instagram account and adjust privacy settings, such as the visibility of posts you’re tagged in. For social networks in particular, Jumbo can delete and archive old posts. For instance, you can use the app to delete tweets that are older than a certain threshold. Jumbo saves everything on your phone in a local storage area called the Vault. And Jumbo stands out from other privacy assistants as it doesn’t rely on APIs to control your online accounts. Instead, Jumbo works more or less like a web browser in the background. Everything happens on your device (which is great for privacy) and the startup isn’t limited to what’s possible with official APIs. In many ways, Jumbo exposes privacy settings to people who don’t know that these settings exist and don’t understand what they’re supposed to do. Sure, anyone can already check and delete Google Maps activities, past web searches and YouTube keywords. But Jumbo centralizes all these settings in a simple consumer app. It regularly scans your account activity and even tells you that you should probably enable two-factor authentication to improve your account’s security. At first, Jumbo that consumer subscription was the only business model that could turn Jumbo into a sustainable business without any compromise. It’s easy to understand how companies make money when there’s a paid subscription. “We reached 25,000 paid subscribers and we realized that it was quite small for a consumer subscription business,” Valade said. “And there was another issue — churn.” Every year, around 40% of Jumbo users would keep their subscription active. If you’ve run a subscription business, you know that this isn’t a bad churn rate. But it means that the company had to spend money on marketing and paid installs to compensate this churn rate. Even with paid subscriptions, Jumbo wasn’t turning a profit. Hence, today’s pivot. Going forward, Jumbo will be a free consumer app with a business offering coming later this year. This hasn’t been an easy decision. Jumbo is a smaller company today, with around 25 employees — the marketing team has been laid off. Valade told me that at some point he thought about calling it a day and selling the company to the highest bidder. “We realized that we were limiting our growth rate because we offer a paid product. The best business model on the internet is B2B software as a service,” Valade said. The startup raised some fresh funding in a $17 million round led by Index Ventures with some existing investors and several angel investors also participating. The company has reached a post-money valuation of $77 million. Jumbo Tearing down the paywall is one thing, but how will people hear about Jumbo now? Pierre Valade is adding identity theft insurance to the app, and turning this insurance product into a social feature. When you download the app, Jumbo now offers identity theft insurance for free in the U.S. with up to $25,000 in coverage through . Identity theft is a big issue in the U.S. with malicious people opening unauthorized credit cards under someone else’s name and other wrongdoings that can cost you a lot of money. Currently, identity theft insurance products are mostly paid products. For example, Norton offers for $125 per year while plans start at $15 per month. Jumbo’s new insurance product pairs well with the rest of the app as Jumbo alerts you in case of a new data breach that may contain some personal information (using data). Usually, when you’re aware of a new data breach, you don’t really know what you’re supposed to do. Every time you invite a friend or a family member to Jumbo, it increases your insurance coverage by $25,000 with a hard cap at $1,000,000. Essentially, it encourages Jumbo users to invite other people to the app. As for the business offering? “Today, we don’t really know the feature set [for businesses]. Right now, I’m really focused on demonstrating that the free product can truly grow significantly faster than before,” Valade told me. Some companies could pay for Jumbo to encourage their employees to set up two-factor authentication on their personal accounts, for instance. I asked about other business-oriented security startups like . “We sell simplicity,” he told me. “Phishing training is not really consumer friendly.” Let’s see how this B2B pivot will play out, as Jumbo is still very much a consumer app for now. But it seems like Valade feels better about his startup’s positioning now. “We were frustrated with the fact that we had to oversell the product to consumers,” he said. “You end up scaring people in order to convince them, and we were becoming uncomfortable with that.”

How to cut your SaaS spending by 30% in 2023

Eldar Tuvey

of affairs paints a concerning picture for SaaS buyers. Our data shows that the SaaS inflation rate is around than the general market inflation rate. Specifically, SaaS expenditure in the U.K. and Australia is currently growing at a rate five times higher than market inflation — and in the U.S., a substantial 3.5 times more. These would be worrying figures in any economy, but for CFOs attempting to drive growth during an economic downturn, soaring software costs should be ringing alarm bells. One of the reasons that SaaS vendors are able to increase their prices year after year is that so many obscure their pricing information. As a result, buyers lack the insight to negotiate best-in-class deals on their software contracts. Without a frame of reference on what other companies pay, many are quoted higher rates for enterprise-level software subscriptions. So much so that our data indicates that as many as 90% of companies are overpaying for their SaaS products by 20%-30%. Vendor pricing is rarely set in stone, and buyers have more purchasing power than they realize when they approach negotiations and are equipped with leverage. When your company reaches out to sales teams to inquire about SaaS solutions, there is often room to negotiate the contract. Finance and procurement teams need to be aware of the negotiation tactics that can be used to attain SaaS products at a better price and avoid the common pitfalls associated with software negotiation. If you’re looking to cut your software costs in 2023, these are the strategies that are recommended for securing SaaS contracts with long-term value. The first step in gaining negotiation leverage is to allow yourself time. Ensure that you approach a vendor to renew or take on a brand new contract far before you plan to onboard with the new product. Before anything else, you’ll need to understand your needs and comprehensively research which tools are best equipped to fulfill them. To help you on your way, put together a short list of tools and consider: This way, you can enter negotiations with the knowledge and time you need to secure the best deal. If you fail to take this time, your supplier could realize that you need a quick turnaround and respond with inflexible terms. Though there is no one-size-fits-all solution, we typically advise that the average company begins the procurement process six to eight months before contract renewal.

When to build a freemium plan and how to get it right

Konstantin Valiotti

users go through when using your product directly affects how the product is evaluated and how it is perceived. The business model you choose strongly influences the entire funnel, enabling or preventing you from engaging certain types of users. This article will examine whether the freemium model is right for your business and how it compares to two other models: free trial and sales driven. Conventional wisdom says that you should choose one model and stick with it. The following diagram shows a choice between the three typical models: Freemium, free trial and sales-led. Konstantin Valiotti Freemium models allow you to target a wide audience and retain them in your product even if users do not have a budget for the solution. The core idea is to create value for the end user and turn the resulting credibility into revenue by selling them something more advanced. The free-trial model creates a time limit and positions the product in a way that shows off its full value, but requires a decision at the end of the trial period. Users can no longer extract value from the product on a particular day and will have to judge whether the perceived value outweighs the purchase cost. The sales-led model routes all new inquiries to a sales representative. The first impression of the product and the anchors related to pricing are set in a conversation. The sales-led model introduces friction at the top of the funnel by prompting users to have a conversation, but it aims to significantly simplify the path to conversion whereas a full self-service experience would have too much friction. Given the obvious differences between these models, choosing one should be fairly straightforward. For example, you should choose the sales-oriented model if you work in a market with big deal sizes and complex products. However, current market conditions do not support having just a single model. Asana offers a free plan, a trial of the paid plan and a “Contact Sales” motion. At HubSpot, you can use the free products or talk to sales about advanced product packages. Calendly has a free plan, a trial, and a “Contact Sales” motion. The list goes on. Introducing freemium is an investment that you should evaluate in the same way as other projects. It requires (sometimes significant) engineering investment, marketing support and organizational changes in sales, finance and operations. There are three good reasons to introduce a freemium model: When you have a free trial, your newly acquired audience will have to decide if they want to pay right after the trial expires. There will be an audience that loves your product and has the budget to pay, and they will do so unless a competing offer is more attractive to them. However, there will always be many users who are either skeptical about the product itself or do not have the budget to pay. When faced with the dreaded paywall, they will go through all reasons to postpone the decision. The longer they hesitate, the less likely they will buy the solution.

What do recent changes to state taxes mean for US SaaS startups?

Ardy Esmaeili

a majority of businesses plan to fully adopt software as a service (SaaS) by 2025, and if the past is any indicator, that means state legislatures are working hard to capture revenue from this new sales stream. As with many U.S. laws and regulations, tax laws regarding SaaS and continue to evolve. Currently, some states consider SaaS to be software while others categorize it as a service. In addition, some states tax all services regardless of type, and more than 20 have a way to target SaaS. At least four states (New York, Pennsylvania, Texas and Washington) are aggressively pursuing SaaS. There’s also the issue of bundling — on its own, SaaS might not be taxed, but it will be when paired with hardware. In the early days of a startup, there’s a tendency to think that the only tax worry would be an audit in the future, the likelihood of which is low. However, tax issues become a problem when you’re fundraising or facing due diligence for mergers and acquisitions. The party conducting due diligence will be focused on sales and use tax, as any liability could transfer to the buyer. We saw this with a new client recently — they hadn’t performed a risk assessment and the buyer identified almost $1 million dollars in tax liability. This reduced the purchase price significantly. Startups think they’ll have lots of time to get to this point, but they actually need to focus on it right away. Any negligence, if identified, could exclude a company from any statute of limitations. While no business is exempt from taxes, it’s critical for startups to understand when they’re liable for tax, and if offering a SaaS solution, how each set of local laws applies. To identify which states you’ll owe sales taxes to, first establish your nexus by determining your physical or economical presence. You can determine your physical nexus by examining which states you have employees, office, property or agents in. Are you “maintaining, occupying or using permanently or temporarily, directly or indirectly, an office, place of distribution, sales or sample room or place, warehouse, server, storage place or other place of business?” Or is there an “employee, representative, agent or salesperson working in the state under the authority of the company on a temporary or permanent basis?” An economic nexus is established for sellers “not having physical presence in the state.” In this case, the state will collect sales tax from customers and remit if the seller meets a set level of sales or number of transactions in that state. With broad definitions like these, it’s easy to see how complex taxes can become.

Peacock tops 20M subscribers in Q4 as losses widen

Lauren Forristal

Comcast-owned streaming service Peacock had its best quarterly result since its 2020 , adding five million paying subscribers in its to bring the total to 20 million, up from the in the previous quarter. In Q1 2021, Peacock had 9 million paid users. Peacock gets much of its success from its sports programming. The boost in paid subscribers was primarily due to the , which streamed in Spanish on Peacock Premium and Telemundo. The streamer also now has exclusive next-day rights to NBC and Bravo shows. “Looking ahead and based on our experience to date, we expect our subscriber cadence will follow our content launches, which will fall more heavily in the second half of ’23, and we continue to see positive trends in engagement churn and ARPU,” Comcast CEO Brian Roberts said during today’s earnings call. In 2023, Peacock Premium subscribers will get to watch the French Open tournament. The company is also in talks with partners to make available on the platform next year. Last month, NBC Universal to become the airline’s official streaming partner, giving customers access to Peacock shows and movies. Although Peacock nearly tripled in revenue to $2.1 billion, its loss widened again compared with the previous year. The company noted an adjusted EBITDA loss of $978 million, compared with a loss of $559 million in 2021. Comcast also reported $541 million in severance costs, including $182 million related to NBCUniversal. And while Peacock had its most impressive quarter to date, it’s still lacking when compared to its streaming competitors, like with , and Disney+ with . Paramount+ grew to in its third quarter. During the earnings call, the company reassured investors that its streaming strategy is just fine the way it is. “We like what we’re doing. We had a phenomenal year getting to 20 million paid subs from less than a year ago, and we see this coming year as the peak year,” said CFO Mike Cavanagh. “We made a decision to invest in Peacock. It’s very clear that we picked the right business model at this point given where we are,” added Jeff Shell, CEO of NBCU. “We’ve been clear from the start that we’re going to see a return on that investment.”

Where should sales sit in product-led companies?

Anna Heim

Product-led sales is a model in which the product, not traditional marketing, helps companies understand who might be their next big customer. Think of a freemium dev tools company, for instance: Instead of tracking which CTO downloaded their latest white paper, they look for organizations that already have dozens of employees engaging with their product on a daily basis.

Elon Musk is being investigated by the SEC for Tesla self-driving claims, report says

Darrell Etherington

Tesla CEO Elon Musk is facing scrutiny by the U.S. Securities and Exchange Commission (SEC) regarding his specific comments and efforts to promote the automaker’s claims regarding its “self-driving” capabilities, . The SEC investigation into Musk is part of its overall efforts to determine whether Tesla has run afoul of its rules in promoting its FSD and Autopilot offering. The SEC doesn’t typically comment on any ongoing investigations prior to formally filing suit, and has not commented on this case in particular. But recent revelations may explain why Musk is in their crosshairs when it comes to Tesla “self-driving” technology: Last week, testimony given by a senior engineer on the Tesla team working on its Autopilot software revealed that a video the company released in . Reporting by . Of course, the SEC’s domain isn’t safety claims, but it does take issue with public companies or company executive officers making forward-looking claims that are false or misleading. That’s apparently what they’re concerned about here — Musk has often suggested FSD would attain essentially driver-free navigation capabilities in timelines that have not ended up proving accurate. Based on what the SEC determines following its investigation, we could see lawsuits or other consequences for Musk, including limitations on his future activity as an officer of a public company if they choose to pursue enforcement of any violations they find.

Tesla engineer testifies that 2016 video promoting self-driving was faked

Rebecca Bellan

Tesla faked a 2016 video promoting its self-driving technology, according to testimony by a senior engineer reviewed by . The video, which shows a Tesla Model X driving on urban, suburban and highway streets; stopping itself at a red light; and accelerating at a green light is still on and carries the tagline: “The person in the driver’s seat is only there for legal reasons. He is not doing anything. The car is driving itself.” CEO Elon Musk used the video as evidence that Tesla “drives itself” by relying on its many built-in sensors and self-driving software. Yet according to Ashok Elluswamy, director of Autopilot software at Tesla, the video was staged using 3D mapping on a predetermined route, a feature that is not available to consumers. In his July deposition, which was taken as evidence in involving former Apple engineer Walter Huang, Elluswamy said Musk wanted the Autopilot team to record “a demonstration of the system’s capabilities.” Elluswamy’s statement also confirms and provides more details on what anonymous former employees in 2021. While there appeared to be no legal ramifications for Tesla following the NYT’s investigation, an on-the-record testimony from a current employee could cause trouble for the automaker, which is already beleaguered by and surrounding its Autopilot and Full Self-Driving (FSD) systems. (To be clear, neither system is actually self-driving. They are advanced driving-assistance systems that automate certain driving tasks, but as Tesla has made clear on its website, drivers should stay alert and keep their hands on the steering wheel when the systems are engaged.) When electric truck maker and eventually admitted to faking a video of its fuel cell–powered Nikola One semitruck prototype — Nikola had actually placed the truck on a small hill, allowing gravity, not the motor, to do its thing — state and federal investigations were launched into both Nikola and its chairman and founder, Trevor Milton. Milton was found guilty on charges of . Tesla’s fake video was created “The intent of the video was not to accurately portray what was available for customers in 2016. It was to portray what was possible to build into the system,” Elluswamy said, according to a transcript of his testimony seen by Reuters. Musk promoted the video at the time, tweeting Tesla vehicles require “no human input at all” to drive through urban streets to highways and eventually to find a parking spot. Neither Musk nor Tesla, which has disbanded its press office, responded in time to TechCrunch’s request for comment. The revelation comes at a time when Tesla is facing for multiple fatal crashes involving its Autopilot system, as well as a criminal for claims Tesla made about Autopilot. Just this week, a Tesla that had FSD engaged into a BC Ferries ramp in Canada, totaling the vehicle. Regarding the 2018 crash that killed Huang, the National Transportation Safety Board concluded in 2020 Tesla’s “ineffective monitoring of driver engagement” had contributed to the crash, which the board said was likely caused by Huang’s distraction and the limitations of the system. While Tesla does tell its drivers to pay attention to the road, there are ways drivers can fool the system to make the car believe they were paying attention, said Elluswamy. Many drivers even go so far as to buy Tesla counterweights on , which can be placed on steering wheels to mimic the weight of human hands that are otherwise engaged while the car is in movement. Even amid regulator scrutiny and reports of crashes, to customers across North America.

T-Mobile says hacker accessed personal data of 37 million customers

Lorenzo Franceschi-Bicchierai

In a financial filing on Thursday, T-Mobile revealed that a hacker accessed a trove of personal data belonging to 37 million customers. The telecom giant said that the “bad actor” started stealing the data, which includes “name, billing address, email, phone number, date of birth, T-Mobile account number and information such as the number of lines on the account and plan features,” since November 25. In , T-Mobile said it detected the breach more than a month later, on January 5, and that within a day it had fixed the problem that the hacker was exploiting. The hackers, according to T-Mobile, didn’t breach any company system but rather abused an application programming interface, or API. “Our investigation is still ongoing, but the malicious activity appears to be fully contained at this time, and there is currently no evidence that the bad actor was able to breach or compromise our systems or our network,” the company wrote. This is the eighth time T-Mobile was hacked since 2018. The most recent incident was in 2022, when a group of hackers known as Lapsus$ , which gave them the chance to carry out so-called SIM swaps, a type of hack where hackers take over a victim’s phone number and then try to leverage that to reset and access the target’s sensitive accounts such as email or cryptocurrency wallets. T-Mobile has 110 million U.S. customers. A spokesperson for T-Mobile did not respond to a request for comment.

Musk oversaw misleading 2016 video saying Tesla drove itself

Rebecca Bellan

Musk published a on Tesla’s website on October 19, the day before the went up, that said all Tesla cars from that day forward would ship with the hardware necessary for full self-driving capability. His emails to staff that month discussed the importance of a demonstration drive to promote the system. Musk’s direct involvement in the video, and subsequent promotion of Tesla vehicles’ abilities to drive themselves, comes at a time when the executive’s reputation and trustworthiness are increasingly at stake. In addition to his Twitter distractions, Musk also promised during Tesla’s Q3 investor call that the automaker would have an “epic end of year,” yet Tesla ended up . On top of that, the . Musk is also poised to take the stand this week in a class-action lawsuit from shareholders who say his claiming that funding had been secured to take the company public — it wasn’t — caused them to lose potentially billions of dollars. The jury will determine whether Musk knowingly, and thus fraudulently, claimed secured funding when it hadn’t been. On October 11, Musk sent an email under the subject line “The Absolute Priority” letting the Autopilot team know that he had canceled his plans for the upcoming weekend to work on the video. “Just want to be absolutely clear that everyone’s top priority is achieving an amazing Autopilot demo drive,” Musk said in the email, according to Bloomberg. “Since this is a demo, it is fine to hardcode some of it, since we will backfill with production code later in an OTA update,” he wrote, referring to the use of a 3D digital map that the Model X used to follow a pre-determined route. “I will be telling the world that this is what the car *will* be able to do, not that it can do this upon receipt,” he continued. Despite that promise, the internal emails show that Musk himself asked the Autopilot team to open the video with the words: “The person in the driver’s seat is only there for legal reasons. He is not doing anything. The car is driving itself.” Then when Musk promoted the video , he wrote: “Tesla drives itself (no human input at all) thru urban streets to highway to streets, then finds a parking spot.” We won’t be tech snobs and say that getting a vehicle to drive semi-autonomously, even if it is a pre-determined route, wasn’t an impressive feat for an automaker in 2016. But it’s the principle of the thing, of knowing that it wasn’t actually driving itself yet saying that it was. Tesla, some argue, should have disclosed that so as to not mislead customers into thinking its tech was further along than it was. “Tesla also maybe could have mentioned that in the filming of the video, the Model X drove itself into a fence,” according to Ashok Elluswamy, director of Autopilot software at Tesla who testified details about the video. Elluswamy’s deposition was taken as evidence in involving former Apple engineer Walter Huang. The lawsuit alleges that errors by Autopilot, and Huang’s misplaced trust in the capabilities of the system, caused the crash. State and federal agencies and customers have also called out Tesla for falsely promoting the capabilities its driver assistance systems, Autopilot and Full Self-Driving (which is not actually fully self-driving), even though Tesla does advise its drivers to stay alert and focused while the systems are engaged. Last July, the of falsely advertising its systems, something a handful of Tesla customers also alleged in a against the company. Additionally, the National Highway Traffic Safety Administration is actively investigating two crashes related to Autopilot. Tesla is also potentially facing a from the Department of Justice over its self-driving claims. Tesla has defended itself, saying that its “failure to realize a long-term, aspirational goal is not fraud,” according to a November motion to dismiss the complaint from customers suing for deceptive marketing. In a conversation last month, Tesla said its leg up over other automakers as it aims to solve full self-driving is that the car is “upgradeable to autonomy,” something that “no other car company can do.”

Google Fi says hackers accessed customers’ information

Carly Page

Google’s cell network provider has confirmed a data breach, likely related to the recent at T-Mobile, which allowed hackers to steal millions of customers’ information. In an email sent to customers on Monday, obtained by TechCrunch, Google said that the primary network provider for Google Fi recently informed the company that there had been suspicious activity relating to a third-party support system containing a “limited amount” of Google Fi customer data. The timing of the notice — and the fact that Google Fi uses a combination of T-Mobile and U.S. Cellular for network connectivity — suggests the breach is linked to . This breach, disclosed on January 19, allowed intruders access to a trove of personal data belonging to 37 million customers, including billing addresses, dates of birth and T-Mobile account details. The incident marked the eighth time T-Mobile has been hacked since 2018. In the case of Google Fi’s breach, Google says the hackers accessed limited customer information, including phone numbers, account status, SIM card serial numbers and information related to details about customers’ mobile service plans, such as whether they have selected unlimited SMS or international roaming. Google said that the hackers did not take customers’ personal information or payment card data, passwords, PINs or the contents of text messages or calls. While some emails told customers that there is “no action required,” at least one Google Fi customer claimed in that their disclosure said that their phone number had been briefly hijacked, known as . Google reportedly told the customer that the intruders had transferred their number for close to two hours, during which they “could have involved the use of your phone number to send and receive phone calls and text messages.” This technique is used by hackers to gain access to a victim’s other online accounts that are protected by the same, albeit hijacked phone number. TechCrunch asked Google whether it could confirm that the incident was linked to the recent T-Mobile breach but has yet to receive a response. It’s not immediately clear how many Google Fi subscribers have been affected by the breach. Google hasn’t made public how many cell subscribers it has in total. In its email to customers, the company said it is working with the as-yet-unnamed network provider to “identify and implement measures to secure the data on that third-party system and notify everyone potentially impacted.” It added that there was no access to Google’s systems or any systems overseen by Google.

Lost your crypto amid Chapter 11 bankruptcy filings? You’re probably not getting it back

Jacquelyn Melinek

access to your money when the crypto firm holding your assets filed for bankruptcy, then you’re probably out of luck getting it back. As Chapter 11 bankruptcy proceedings move forward for several big-name crypto companies, those who lost funds are surely hoping to get all — or at least some — of their money back. Lawyers and experts shared their thoughts with TechCrunch on what these cases could mean for creditors and what may happen to those who saw their money disappear overnight. Earlier this month, , a subsidiary of the crypto conglomerate Digital Currency Group ( ), filed for Chapter 11 bankruptcy. Genesis is the latest crypto-focused entity to join the Chapter 11 bankruptcy club alongside , , , and — all of which filed mid- to late 2022. For the most recent Chapter 11 filers, Genesis to its top 50 unsecured creditors, while FTX owes its top 50 unsecured creditors . The bankruptcy filings have redacted the majority — if not all — of the identifying information for the parties involved. One of FTX’s biggest unsecured creditors is owed more than $226 million, and the company could have , according to earlier bankruptcy filings. So it’s safe to say that a lot of people are heavily invested in the outcome of these bankruptcy cases, as their funds, ranging from small amounts to millions of dollars, are involved. But it’s not certain if they’ll ever see the deposited funds again. What will happen to creditors “really depends on the mix of assets and liabilities of the company as well as the prospects of the same company exiting bankruptcy,” Jason Allegrante, chief legal and compliance officer at Fireblocks, said to TechCrunch. “If the business is otherwise healthy but has experienced a liquidity shock, for example, there is still a chance that the business can recover and generate revenue,” meaning creditors may be reunited with some of their funds. Secured creditors will have priority “if and when assets are distributed,” Joel Telpner, chief legal officer at Input Output Global and special counsel at Sullivan & Worcester, said to TechCrunch. “All other creditors stand in line after the secured creditors are first paid. If it’s a company with shareholders, then if there’s anything left, it’ll go to shareholders.”

Apple HomePod (2023) review

Brian Heater

dark night of the . A few years back, the — and, frankly, why not? We’ve smartened our phones and watches. Why shouldn’t our homes be the next step? For decades, many looked longingly at home automation. Smart blinds that opened with your alarm clock might as well have been magic — or, at the very least, the domain of those with money to burn. Suddenly, however, a new generation of the technology arrived with the promise of a (relatively) affordable entry point into home control. Smart lights, smart smoke detectors, smart locks, smart doorbells — and at the center, the smart speaker, the consummate hub and portal into this brave new world. But the best laid plans are still beholden to market forces (among others). The looming recession of the last 12 or so months hasn’t been especially kind to the smart speaker. As is now well documented, a disproportionate number of , as it stared a $10 billion annual loss in the face. We don’t have exact numbers for Google’s Nest division, though the company appeared to freeze hiring ahead of its own massive culling. None of that factors in all of the quick reversals (Microsoft Cortana) and aborted starts (Samsung Bixby) from a few years back. The cadence of hardware releases generally appears to have slowed across big corporations, owing to economic headwinds and supply chain shortages. It’s all a very strange way of setting the scene for a rare Apple reversal. Just shy of two years after , the original premium smart speaker has arisen from the ashes. But why? And why now? The official line from Apple seems to be, simply, because the people asked for it. As far as that explanation goes, I will say this: The most recent Apple iteration is a company that is, in fact, more responsive to users’ needs and desires. We’ve seen the clearest manifestation of that in the , reversing previous foot dragging by improving some features (keyboard, webcam), dropping unwanted ones (Touch Bar) and bringing back old favorites (MagSafe, SD card slot). Even so, this move is odd on the face of it. In , the company told the press that it was discontinuing the original HomePod to “[focus] our efforts on HomePod Mini.” It seemed Apple had lost faith in the notion of a premium smart speaker. And it wasn’t alone. The has been MIA for six years, and while the Echo Studio got a software upgrade last year, the device certainly hasn’t been a focus for Amazon. It seemed that, in a world of Nest Minis and Echo Dots, Apple was putting its eggs in the $99 Mini basket. Brian Heater If the as the company is calling it, had arrived in its current form after its predecessor, I doubt anyone would have blinked an eye. Nothing about the product’s presentation screams “radical departure.” In fact, the speaker looks remarkably like the model that debuted in 2018. As it is, you might need to consult a spec sheet to pick out the key differences between the two Apple devices, which are separated by roughly half a decade. The 2023 Apple HomePod is 0.2 inches shorter than its predecessor, with a height of 6.6 inches (the radius is the same, at 5.6 inches). It’s also lighter, dropping from 5.5 pounds to 5.16. That’s not an insignificant difference, and certainly if we were talking laptops, it would be a big deal. Although the HomePod is designed for one to pick it up and move it around the house, my sneaking suspicion is a great majority of users will clean off a spot on a table top or shelf and rarely — if ever — move it. After all, there’s no battery power on board. If there were, it would be a distinctly different product. As it stands, it needs a constant power source. In fact, . The 2018 model had what one might deem a technically detachable cable, but it fell firmly into the standard corporate-speak that recommended users visit a licensed Apple certified repair specialist. That also currently appears to be the standard line with regard to repairability, meaning that, in spite of the recent availability of user repair kits for the iPhone and MacBook, I wouldn’t attempt to crack it open at home if you’re worried about potentially voiding the warranty. This time, mercifully, the cord is designed to be detached by the user. In fact, it ships unconnected, in its own little box. Once connected, you can pull it off with one, firm yank. The cables maintain the same woven design you’ll find across most of the latest Apple products. Unlike the HomePod Mini, however, it ends in a standard wall plug (three-pronged here in the U.S.), rather than USB-C. My biggest issue here from a cable perspective is that it’s too short. It’s five feet long — a full foot shorter than the Mini cable. I set up two HomePods as a stereo pair, currently flanking my desktop display. The near one (right) reached just fine. The other required an extension cord. A minor complaint, but an easy potential fix for future products. The good news here is that, while Apple won’t be selling a longer version of the cable, it’s an industry-standard plug, so you’ll be able to buy a longer one to swap in. I suspect that, before long, third parties will be offering their own matching woven fabric versions that come in six-foot lengths. Brian Heater There are no other ports on the product. I’ve had a Google Home Max sitting on my desk since…well, since . It’s a fine workhorse of a smart speaker, with an auxiliary in port. It’s great for hard wiring to a PC or, in my case, a turntable. It’s not necessary, exactly, but it’s a great option to have when you want it. Beyond these small things, you’d be hard pressed to tell the old and new HomePods apart by looks alone — the white one, at least. Apple swapped the first gen’s grey color for something it calls “midnight.” In this instance, at least, the company can be forgiven for what isn’t an especially clarifying name. It’s probably too wordy to call it “mostly black, but blueish in certain lighting.” It’s a nice color, though of the two, I’m still leaning toward the white. I might have gone with a more straightforward black or dark gray, but, to my eyes at least, it’s mostly black most of the time. The best thing about the new color, however, is that it’s made from entirely recycled fabric. There isn’t the same variety you’ll find with the Mini. Nor is there anything that pops quite like that line’s orange and yellow options. That’s long been Apple’s M.O. — “fun” vibrant colors for the cheaper models and “serious” muted options for the high-end models. The latter makes sense here, as I’m sure most people are looking for something a bit more staid that fits in with its surroundings in the home. Though, again, more options are rarely a bad thing. I’ve always liked the HomePod’s design. It’s the best-looking smart speaker. As such, it’s not especially surprising Apple didn’t fix what wasn’t broken, as tempting as it might have been to truly reinvent the thing ground up after five years. The mesh fabric remains, a crisscross diamond shape also found on the Mini that the company says is optimized for sound. Certainly, it’s distinct and, one might even suggest, iconic. It’s rigid but does have a slight give when a little pressure is applied. The white color is prone to getting dirty. After sitting on my desk for a few days, the bottom has already become slightly discolored from dust. Predictably, Apple . The speaker’s cylindrical shape slowly tapers on the top and bottom. Below is a silicone pad that creates a bit of friction, while still allowing for the system to be nudged into position (a change from the 2018 model’s convex foot). There’s an internal suspension system connected to the woofer that allows the speaker to output loud, deep bass sounds without intentionally moving the rest of the speaker in the process. Up top is the familiar touch display — the most striking bit of the HomePod design language. Apple The illuminated touch surface up top has been expanded by 6x. That means the shiny touch display and its underlying lighting system effectively reach edge to edge. The glow for playback and the cloud of colors for Siri have always been a clever touch that’s nice to look at. Apple clearly wants Siri to serve as the primary interface here. It’s a smart speaker at the center of a smart home, so all of that logically tracks. There is limited touch control as well. A single tap will play or pause the music and a pair of fixed buttons turn the volume up and down. There’s little doubt in my mind that, when Apple began discussing bringing HomePod back, they explored a lot of ideas, including fundamental changes to the system. I suspect during those conversations, the notion of a more full-fledged touchscreen was raised a few times, at least. After all, in spite of their stubborn refusal to bring one to the Mac, Apple (and its suppliers) has gotten quite good at . Again, this is another one of those decisions that would have fundamentally altered the product. Suddenly your smart speaker is a smart screen — albeit one that is still a lot more speaker than screen. I still really like the Nest Hub. It’s a great little piece of hardware, whose usage has evolved for me over the years. It makes sense as a control panel for smart home functionality and a visualizer for alarm features and voice functionality like weather and sports scores. These days, I probably use it most as a companion piece for Spotify. It’s a small thing, but I really like having the album art and song information on a screen that also lets me skip and repeat tracks. It’s easy to see how similar functionality would be nice on a HomePod, though its design choices (a 360 speaker that covers most of its surface area, but for a small display on top) severely limit such things. I certainly won’t go so far as to see that Apple has created some platonic ideal speaker design, but I wholly understand why the team tasked with bringing HomePod back found it difficult — and in many respects unnecessary — to reinvent the wheel. From a pure marketing strategy standpoint, however, killing the HomePod, only to bring it back in a nearly identical form factor nearly three years later, is baffling. But we’re not here to dwell on marketing decisions. Brian Heater One nice thing about fast-forwarding three years is the HomePod has picked up some tricks from its fellow Apple products along the way. Those start with setup. The processor has been updated from the A8 (which debuted on the iPhone 6 in 2014) to the S7 (which powers the Apple Watch Series 7). Mind, Apple has since , but that chip has the same processing power as its predecessor, with the primary distinction being the accelerometer and gyroscope that enable crash detection — something that isn’t especially relevant to a smart speaker. One unfortunate product of the S7’s inclusion is that the HomePod has Wi-Fi 4 on board. The latest hardware products support Wi-Fi 6e. The iPhone 14 supports Wi-Fi 6. Wi-Fi 7 is set to debut in 2024. WiFi 4 debuted the year Barack Obama was elected to his first term. The 2018 HomePod 1 supported Wi-Fi 5, which debuted four years prior. If you’ve got an iOS, pairing is simple. In fact, it’s a lot like pairing an Apple Watch. Hold the device close to your new speaker, wait for confirmation and then hold the phone up so it captures the light pattern on top of the device inside a circle on your display. Pick up a second HomePod and the process of making a stereo will bring the entire setup process to around two minutes or so. Note, however, that you need an iPhone for setup, even if you expect to rely entirely on AirPlay. That’s a bit of a disappointment, but not entirely unexpected, given Apple’s general approach to hardware interoperability. Apple A note on pairing: HomePods can only be paired with their same model. It seems perfectly reasonable that you can’t make a stereo pair between, say a HomePod and HomePod mini. Listen, I get it. The hardware, tuning and size of the two devices are so dramatically different, a balanced sound is a near impossibility. The inability to pair HomePod 1 and HomePod 2, on the other hand, remains something of a mystery as I write this (more later), but Apple’s statement on the matter boils down to the idea of “giving a user the best possible experience.” Here we are, once again, at the tricky crossroads of “best experience” and user control. I would argue that the inability to pair the two product is — in itself — a less than best experience. Apple ran out of HomePod 1 stock. That means that if you have a single first-gen device and want a stereo pair, you’ll have to buy another first-gen, definitely not through Apple and likely refurbished. It’s not the end of the world, by any means, but it does put the customer in something of an awkward position. One nice feature borrowed from the Mini is handoff. This utilizes the S7 chip and ultra-wide band (UWB). You’ll need to hop into settings to enable the feature, but it’s worth it. Start a song with Apple Music on your iPhone, hold it near the HomePod and it will start playing there, accompanied by a satisfying haptic fist bump. Move the phone near the speaker again and you can transfer it back. I really like this feature. It’s a good example of how nicely if you make your own devices, software and chips. It’s also surprisingly receptive. In fact, I found myself having to disable it while the HomePods are on my desk, otherwise it will accidentally trigger when I’m using the iPhone two feet from the speakers. Apple I’m a longtime Spotify subscriber who moved over to Apple Music for the purposes of this review (shout out to ). You can use it through your Airplay-enabled device (also a good way of using a stereo pair as an entertainment system without Apple TV), but the experience just isn’t as good as it is with Apple’s software (though the company says the sound quality should be unaffected). Apple builds ecosystems. It’s worked this way for decades, and these days, any hardware maker with the necessary resources has more or less followed suit. As someone who writes about tech for a living, I think it’s important not to rely on a single company — though I understand why people do. It’s often a smoother experience. The biggest thing I miss about Spotify thus far (aside from being far more used to the UI) is the way it syncs song playback in real time across devices. It’s a great feature and losing it is a bit annoying, as I’m not great at keeping track of what was playing where. Among other things, it’s great to be able to view what’s now playing on your computer and phone screen. There were also instances where I would wait too long after playing a song on my computer, tap the top of the HomePod to pick up where I left off and have Siri instead choose something from a playlist of songs it thinks I’d like. Apple Point in favor of Apple Music after a few days back on the platform: Lossless audio, spatial audio (still a bit on the fence about its usefulness here — more below) and a few of the folks who jumped ship from the competition as fallout from the whole Joe Rogan/vaccine misinformation situation — in particular. Also, Apple Music pays artists a lot more (relatively speaking) per stream than Spotify (though still not enough, in my humble opinion). While the exterior of the product remains largely unchanged, the internals have received a significant overhaul, beyond the inclusion of a new chip. I asked Apple for a rough percentage of how much the insides have changed, but didn’t get a straight figure, beyond that the changes were “significant.” The woofer and tweeter arrays have both been updated. Interestingly the tweeter figures have dropped from seven to five. It’s worth remembering that music has always been a major driver for Apple. It was something Steve Jobs was famously passionate about, and it’s something that has remained an essential part of the company’s DNA. The original HomePod was, arguably, the first smart speaker to prioritize the speaker over the smart. That’s likely partially due to the fact that Siri wasn’t built with the smart home first, like Google Assistant or Amazon Alexa. Even more pertinent, however, is that Apple is all about a premium experience, and that continues to set HomePod apart from most. Give me a sec, I’m gonna quote Apple directly here: HomePod delivers incredible audio quality, with rich, deep bass and stunning high frequencies. A custom-engineered high-excursion woofer, powerful motor that drives the diaphragm a remarkable 20mm, built-in bass-EQ mic, and beamforming array of five tweeters around the base all work together to achieve a powerful acoustic experience. The S7 chip is combined with software and system-sensing technology to offer even more advanced computational audio that maximizes the full potential of its acoustic system for a groundbreaking listening experience. “Groundbreaking” is, perhaps, a bit hyperbolic. But yeah, in spite of its relatively compact size, the HomePod 2 sounds great. It’s big, bold and dynamic. It fills a room easily and without distortion. It gets loud — like real loud. Like, I didn’t push it to the limit because I live in a New York City apartment surrounded by neighbors who (thankfully) don’t test speakers for a living. Really, in my current living situation, two HomePods is probably overkill, but I can’t argue with the value add of a stereo pair. If you made me choose a single word to describe the sound here, I would probably go with “full.” The highs are high, the lows are low, the mids are…you get the picture. This is, of course, best experienced with songs available in high bit-rates or a lossless format. This, of course, is one of the places where Apple Music shines relative to Spotify. I’ve long believed that years of listening to compressed music has made it difficult for the average listener to pick out the difference between, say, lossless and a high-bit rate. But paired with the new HomePod, digital streaming starts to sing. The separation is great, especially with a stereo pair, which lends a sense of space to the music. Drums, in particular, sound clean and clear, though the low end can overwhelm, and, if you’re like me and placed your speakers on the desktop in front of you, you can really feel elements like the kick drum. That’s sometimes nice and sometimes too much. Apple’s position favors its own in-house tuning. It’s a delicate balance, but I tend to favor more control and would love the ability to fiddle around with EQ sliders in the home app. For an out of the box experience, however, it’s hard (if not impossible) to beat HomePod. You’re up and running within seconds, and the sound is rich and full. I’ve picked up details I’ve missed in familiar songs and, listening to things like podcasts, even noted some things (like mic movement), which probably weren’t intended to be heard. The company notes that while the focus of the press material has (understandably) primarily been about music, that it’s also tuned things for the human voice. In fact, an upgrade that recently dropped via will improve the quality of voice on the system. The feature has also rolled out to the first gen HomePod, which bodes well for Apple’s continued support for the product five years after its initial release. Other features will be arriving down the road via new software, including the ability to hear and alert you when a smoke or fire alarm goes off (unfortunately, unlike , broken glass detection will not be available at launch), which is coming in the spring, and “remastering” of ambient sounds like waves and other white noise. The latter will also be available for home routines, so you can have it automatically trigger before bed. Apple I’m still not wholly sold on Spatial Audio being much more than a fun gimmick. I understand how it will eventually be important for augmented reality (that , right?), but I’ve yet to be sold on its broader import. Until now, I’ve largely thought of it as being tied to headphone head tracking, providing the user with a sense of fixed orientation for different sound elements. With the HomePod, Apple extends the concept to apply to fixed speakers. It effectively builds on stereo sound, offering more precise location and a feeling of space. The number of albums that have been mastered — or remastered — for the feature is lacking at the moment, but Apple Music curates them all into one handy spot. Contemporary pop is well represented — unsurprising, as that seems to be an ongoing focus for the company. There are some older albums in there, too. I’ve been listening to Lil Wayne, Michael Jackson and McCoy Tyner. I’d say overall, it sounds like a slightly elevated stereo separation. “Room sensing technology” is a concept that’s probably familiar if you’ve set up any of the higher-end smart speakers, but here’s a refresher, again, from Apple: With room sensing technology, HomePod recognizes sound reflections from nearby surfaces to determine if it is against a wall or freestanding, and then adapts sound in real time. Precise directional control of its beamforming array of five tweeters separates and beams direct and ambient audio, immersing listeners in crystal-clear vocals and rich instrumentation. The system has a built-in accelerometer that detects when it’s being moved. The next time you play music, it will effectively use on-board mics to detect whether that sound is being reflected back and then extrapolate its position relative to a wall. This uses the four-mic far-field array designed for trigger ‘hey Siri’ (which it does quite well, even when the music is bumping). From there, it will adjust the tuning accordingly, determining whether to incorporate those reflections into the sound. The 360-degree design means that you can also place it free-standing in the middle of a room and get good sound on all sides. Brian Heater , the HomePod 2 $50 cheaper than the first generation’s starting price (though it, too, had dropped to $299 before it went gently into that good night). That’s certainly not cheap. At $600, a stereo pair is very much an investment and one that makes the most sense if your situation ticks off the following boxes: Remember that spiel about the long, dark night a few thousand words back? I firmly stand by it. It’s a weird moment for the smart home and a strange time to be launching (or relaunching) a smart speaker. On the other hand, maybe it’s a great time. After all, neither Google nor Amazon seem to have a clear strategy at the moment — unfortunately, given that late-last-year’s arrival of Matter is set to bust things wide open, particularly when coupled with Thread (which, for the record, was available on HomePod 1). I’m not going to dive too deep into Matter. The long and short of it is it’s a protocol jointly developed by some of the industry’s biggest names, including Apple, Amazon, Google and Samsung. It effectively serves as a universal standard, meaning that companies that make smart home accessories don’t have to choose between HomeKit and the rest. You can develop for everything, everywhere, all at once. I did , centered around an interview with Jon Harros, the CSA’s (Matter’s governing organization) director of Certification and Testing Programs. He told me, “The IoT started reaching a point where it became obvious to have that reality of the billions of sensors and connected devices that we all know is possible. They all have a major slice of the pie. They’re all doing very well, but the size of the pie could grow orders of magnitude. You’re now not talking about shipping millions of products, you’re talking about shipping billions.” It’s a nice bit of harmony from vicious competitors, and it’s a boon to the end user. Many existing accessories are backward compatible and will be receiving over the air updates in the coming year or so (waiting with bated breath here for my favorite smart home accessory, the Nest Protect smoke detector). Meanwhile, plenty of new products that support the standard are already shipping. Home, the same app that lets you control HomePod, is also the central control center for connected devices. To add one, open the app, scan the product’s QR code and wait for the system to connect the two. Once everything is loaded in Home, you can start building scenes and routines, tied to things like your morning alarm or coming home from work. The HomePod also now features built-in temperature and humidity sensors (another trick borrowed from the Mini), which can be used to, say, trigger your smart air conditioner if it gets too hot or cold (I also like being able to ask Siri how warm it is in my apartment versus the outside). Apple Strategically speaking, this feels like the perfect time for the HomePod to strike back. Amazon and Google’s defenses are down, and focusing on Matter support specifically and the smart home broadly should help Apple gain some traction. Apple is always focused on that “just works” experience, and the new standards bring the connected home a heck of a lot closer to that goal. The long-rumored HomeOS continues to be long rumored, even as the company continues to drop hints as to its existence. If/when it does finally materialize, it’s a pretty safe bet the HomePod experience will be foundational. Ditching the original HomePod to the focus on the Mini was — and remains — a weird move. It’s nice to see the big speaker back and better than before — if not radically changed in the way it might have been had the company continued to issue regular updates. A permanent price drop to $299 is certainly nice, though the system still feels a bit aspirational compared to the competition — and doubly so for a stereo pair. I have quibbles with the HomePod 2: Apple Music has its shortcomings, Wi-Fi 4 is ancient and the lack of backward stereo compatibility with the first-gen sticks out like a sore thumb. But the HomePod 2 works well, looks nice and sounds great. It’s going to fit perfectly for a select cross-section of consumers. It continues to not be for everyone, and there’s a part of me that strongly suspects that wouldn’t have it any other way.

NetApp, a specialist in cloud data management, says it will lay off 8%, or around 960 people, citing economic climate

Ingrid Lunden

, one of the big players in cloud data management, today that it would lay off 8% of its staff, citing “macroeconomic challenges and the reduced spending environment” in the current market. The company is estimated to employ about 12,000 people globally, so this will work out to around 960 people impacted. NetApp said that it would begin the process this quarter and is expected to take a charge of $85 million to $95 million related to that. We have contacted the company to ask which product lines or types of jobs might be impacted. It looks like the cuts will be in multiple geographies, including EMEA and Asia-Pacific. San Jose-based NetApp is listed on Nasdaq and has a market cap of just over $14 billion. Like a lot of tech stocks, NetApp has seen its shareprice rollercoaster over the last year and overall drop in value over that time. “Companies are facing an increasingly challenging macroeconomic environment, which is driving more conservatism in IT spending. We are not immune to these challenges,” CEO George Kurian wrote in a memo to employees today. “Against this backdrop, we must be agile, deliver on our near-term commitments, while positioning ourselves for long-term success. This means sharpening our strategy to focus on the areas of our business best positioned for growth, adapting our cost structure to reflect focus and market conditions and raising the bar on our performance. Having successfully navigated similar challenges together with you before, I am confident that sharp focus on our strategy and strong execution will enable us to capture the opportunity ahead.” Indeed, NetApp is not a stranger to layoffs. In 2016, also under Kurian as CEO, the company laid off first 12%, and then a further 6%, of employees within months of each other. This time around, it is part of a bigger wave of reductions across all of tech, covering not just a number of major enterprise vendors, but those in consumer, too. Other recent layoffs have been announced at (12,000), (18,000), (500), SAP (3,000), IBM (3,900) and more. Counting this recent round at NetApp, there have been more than 76,000 layoffs in the technology sector this year alone, according to . That’s a massive and very disconcerting rate. All of 2022 had 159,684, but we’re only two months into the year. NetApp beat on both earnings and revenues, and set guidance of net revenues in the range of $1.525 billion to $1.675 billion and non-GAPP earnings of $1.25 – $1.35. Let’s see how the company performs in Q3 when it reports on February 22.

Stripper Web, a 20-year-old forum for sex workers, is shutting down. No one knows why.

Amanda Silberling

people on the internet who don’t want to be found. That seems to be the case for the elusive, mysterious owner of Stripperweb, a 20-year-old forum for exotic dancers and sex workers. With just one week of advance notice, the forum’s unknown owner announced that the website will shut down on February 1, erasing the decades-long digital footprint of a community on the margins. “This place was a digital dressing room where I could listen in on thousands of interesting, profane, disgusting, inspirational, challenging and painfully true experiences,” said a forum moderator, who uses the screen name Optimist, in an email to TechCrunch. “I didn’t feel alone and isolated anymore. This was a digital home that accepts us all, no matter [how] broke, busted, or disgusted.” The news of Stripperweb’s closure broke last week with a simple notice posted to the top of the website, which is decorated in innumerable different shades of pink. “For over 20 years, Stripperweb has been one of the best resources for exotic dancers and webcam models on the internet. We’ve made the difficult decision to close Stripperweb effective February 1st. We urge you all to check out AmberCutie’s Forum as a possible new home after we close. Thank you to everyone who has made this such an amazing community.” Longtime members and moderators acted quickly, driven by a slew of unanswered questions. Who posted the message? Why now? And why choose a seemingly random forum as a refuge? Taja Ethereal, a Stripperweb moderator who uses the screen name PhatGirlDynomite!!!, took action immediately and offered to buy the site outright. “The first offer was $12,000… I’m losing my head right now. He wouldn’t even come to the table for $12,000, to be paid out in 24 hours, in cash,” she said. “Even if this person is wealthy, why would you leave money on the table unless the site has already been purchased?” Taja Ethereal said she knows that the owner saw her offer — but she still doesn’t know who the owner is, or if the site was already sold. None of the moderators know, even those like her and Optimist who have been on the site for decades, putting in years of labor to educate newer dancers about staying safe and getting paid. Time is ticking. Some sex workers have swiftly taught themselves the Python programming language, writing scripts to scrape decades of data from Stripperweb for posterity. Others are diving deep into the earliest history of the forums, searching for clues as to who might hold the key to the most comprehensive, collaborative archive of what life is like for sex workers in the twenty-first century. A Stripperweb member named cutiecam, who has been on the forum for more than seven years, explained it succinctly: “This is a literal digital museum.” Another user, neverendingkneebruises, wrote on the forum, “I think almost every single dancer [came] across this website when they were starting, no joke… This site helped me get into the industry with standards and boundaries!” Yet even in the face of five-figure cash offers and intense sleuthing, the owner of Stripperweb and their motivations remain a mystery. stripperweb is shutting down next week?! The depth of knowledge & advice on this forum is incalculable! An immense communal & cultural loss for dancers, esp new girls looking for a solid resource to start. — marla cruz (@prolepeach) For those in the know, the cultural value of Stripperweb’s poppy, retro pages has been obvious since its infancy. Now a criminology professor at Kwantlen Polytechnic University in British Columbia, Dr. Lisa Monchalin wrote a master’s thesis about Stripperweb in 2006, analyzing conversations from the forums to investigate the motivations of professional strippers. While sex work is often assumed to be a financial last resort, the conversations on Stripperweb proved otherwise, according to Monchalin’s study. Stripperweb is an appealing source for academic research, as it provides such an unfiltered, honest look into a historically stigmatized profession. For the dancers who found each other on the forum, it’s also one of the only places they can talk openly about the challenges, victories and nuances of their daily lives. Despite the forum’s name, Stripperweb evolved to include a spectrum of different types of sex work, including camming, acting in porn clips and operating phone sex hotlines. “Losing Stripperweb is losing our communal memory,” Marla Cruz told TechCrunch. Cruz is a dancer who turned to the forums for guidance when she entered the industry in 2017. “You could mention Stripperweb in any dressing room, and dancers would chime in about their experiences on the forum, or what they learned from the forum. It was that ubiquitous, that everybody kind of knew what it was, but for some of them, it was like, ‘Stripperweb is the Holy Grail of strippers’ knowledge. It’s my Bible.’” The early lore of Stripperweb’s founding has already been lost to the detritus of the internet. But in that 17-year-old master’s thesis, Monachlin inadvertently preserved a now-forgotten relic of the early forum, which explains how and why the site was founded. The story of Stripperweb begins in 2001. Pryce, a college student who programmed websites as a side gig, heard from his dancer friends that they couldn’t find a good, central community online. “One day […] Pryce mentioned creating a better site – one that was positive, full of quality advice, and one that gave exotic dancers new tools to help them do their job better and make more money,” reads the origin story reprinted in Monchalin’s thesis. Sure enough, a user who goes simply by “Pryce” served as the admin of the forums for many years. A WHOIS domain lookup reveals that the website is owned by someone who uses the email , but that address — as well as every other Stripperweb email address — is no longer active. In a post commemorating the 10-year anniversary of Stripperweb in 2012, Pryce echoes the founding story captured in Monchalin’s thesis. “StripperWeb first opened to the public on January 4th, 2002. Things have certainly changed since then,” Pryce wrote. “I was just a college kid trying to figure out what the Internet could be used for. At the same time, a group of exotic dancers were in need of a virtual hang out to call home. After 6 months and a LOT of late nights, we had a whopping 74 members! That was incredible in those days. Fast forward 10 years, thanks to the growth of the Internet, social media, and all of you posting – we now get 30-50 new members DAILY.” Pryce wrote that a man named Wayne, “an experienced consultant with access to quality resources,” would lead a team of developers and administrators who would take over his role. Members also recall an admin named Bob managing the forums after Pryce’s departure. Pryce, Wayne and Bob did not respond to requests for comment, sent via email and Stripperweb’s direct messaging system. As the forum approaches its final days, one section of that 10-year-old post sticks out. Pryce offers thanks to someone he calls “The Other Owner,” in quotes. “Thank you so much for your support over the past 6 years. You offered to carry most of our financial burden at a critical time,” Pryce wrote. “Since then, no matter what, you have stayed true to your words and continued to help significantly. I respect the loyalty you have shown. Additionally, at times when the site needed extra funds to complete hardware upgrades or software updates, again you contributed. I thank you on behalf of the community.” That 10-year commemorative post was Pryce’s last on the forum; his profile shows that he has not logged in since 2013. Taja Ethereal said she believes that this “other owner” is still in control of the site. Optimist remembers this time on Stripperweb as “a long, strange tale.” These days, the forum is overwhelmingly dominated by conversations among actual sex workers. But according to Optimist, the forum used to have a robust, bright blue counterpart to the pastel pink pages where so many dancers have found refuge. “The site used to be twice the size it is now because there was a blue board to match the pink board,” she told TechCrunch. “The idea was that the customers had their own area, and we had our area.” Some men on “the blue side” simply exchanged tales of their experiences in clubs; others tried to arrange meetups with the dancers, who then asked Pryce to close off certain areas of the forum for privacy reasons. But Optimist says he refused. In some threads from the blue board — which still exist, mostly inactive, in the deep recesses of Stripperweb — men discuss whether or not it’s okay to track down girls from the forums at their clubs. Others debate whether getting a dancer’s number means she’s romantically interested, or if she just wants to text you to come in on a slow night. On other forums, the men compare notes about their “free” sex lives. Taja Ethereal still refers to male customers who intrude on Stripperweb conversations as “blues.” Optimist added, “[Pryce] claimed it was a stripper support site that he and his stripper girlfriend created for the community. But we had no control over it or [means of] protecting ourselves beyond using avatars and fake names.” The “pinks” and “blues” as seen on Stripperweb, documented on January 30, 2023. Stripperweb A lot changed over the years on Stripperweb. When the forum first opened in 2002, Internet Explorer held 95% of the market share for web browsers, Mark Zuckerberg was an unknown high schooler, YouTube didn’t exist and the cell phone was the Nokia 6610, which could only hold 75 text messages. The sex industry itself has changed a lot too. “There used to be a lot of money in stripping,” Optimist said. “Average girls were walking into good regional clubs and walking out with at least $1,000 per shift on average. So we had women putting their ‘vanilla’ careers aside to make two to four thousand dollars a week.” When the 2008 recession hit, the industry was sent into a tailspin. Though Cruz was not in the industry at the time, she learned about the economic impact on strip clubs from old Stripperweb forums, as well as elder dancers from work. “The site has been running since the early aughts, and what’s interesting is that if you actually look through the years, you can kind of see the ebbs and flows of engagement,” Cruz said. “You see these two really big breaks in 2008 for the recession, and then 2020 under COVID.” In the midst of the 2008 downturn, Page Six wrote of a “lap deficit” and that “Wall Street’s financial crisis has trickled down to Manhattan’s mammary meccas.” In other words, fewer big spenders could pay for lap dances. “I worked at a club where there were a handful of 40- to 50-year-old dancers who had been dancing for 20 years,” Cruz said. When COVID hit in 2020, the older dancers reflected on their past experiences with a sudden economic shift. “COVID is like this generation of strippers’ 2008. It was a big bomb that went off and decimated everything, and all the dancers were kind of like, left in this wake to fend for themselves.” As strip clubs closed down for the pandemic, some dancers took their talents online, working as livesteaming camgirls or as creators on adult clip sites. Naturally, the pandemic generated a boom in online sex work. The platform OnlyFans became a household name, bringing in . The company has only . Cruz theorized that Stripperweb’s closure could be linked to the changing state of the industry, but there are a number of possible explanations, including the increasingly suffocating legislative changes to online sex work. In 2018, former President Donald Trump signed the Fight Online Sex Trafficking Act ( ) into law after it received bipartisan support in the Senate and the House of Representatives. The legislation overrides Section 230, the statute that renders social platforms immune to liability for what users post — under FOSTA, social platforms can be indicted for enabling illegal sex work. Laws like FOSTA are positioned as ways to curb sex trafficking, but in practice, the policy has been shown to have . In 2020, Senator Elizabeth Warren (D-MA) proposed a study on the secondary effects of SESTA/FOSTA on sex workers. “Sex workers have reported a reduced ability to screen potential clients for safety, and negotiate for boundaries such as condom use, resulting in reports of physical and sexual violence,” the bill . “Many sex workers have turned to street-based work, which has historically involved higher rates of violence than other forms of transactional sex.” To comply with the law, many credit card processors have cracked down on adult payments, even for legal online sex work. “Liability is not just about whether or not something’s legal or illegal,” Cruz told TechCrunch. “If you’re a legal sex worker — if you do OnlyFans, or if you’re a stripper — and you have an account with Bank of America, for example, Bank of America can decide to shut down your account. You’re at high risk because you’re associated with adult services.” Cruz thinks that whoever owns Stripperweb might have chosen to shut it down due to potential liability concerns under FOSTA, but the timing seems strange, since the law has been in effect for over four years. “It could be something to do with FOSTA, like the hammer’s coming down,” said Taja Ethereal. “Every time you turn around, something’s becoming more difficult to do, like people losing their payment processors. We tried to be legit with adult payment processors — they are shrinking, shrinking, shrinking.” She added that one company she works with recently got banned from using Wells Fargo. “It seems like the noose is getting tighter and tighter.” From advice for dealing with antagonistic customers, to guidance on how to navigate changing legislation, Stripperweb has continually served as a resource for workers in the adult industry. While some community leaders try to hunt down the owner of the forum to negotiate, others have resigned to the possibility of losing the forums forever. In the final days before the February 1 deadline, some denizens of Stripperweb have worked tirelessly to preserve its 20 years of forums as an everlasting archive, writing their own programs and leveraging tools from the Internet Archive to salvage as much of their history as possible. Taja Ethereal set up two computers to run for 12 hours at a time, scraping as much data from Stripperweb as possible. She’s not the only one. “A lot of people are archiving it, and then I have other people who I have hooked up with from Twitter. They’re sending these .zip files,” she told TechCrunch. “So basically, it looks like I’m gonna end up with two hard drives filled with files.” Within days, much of the website was to the Internet Archive’s Wayback Machine, though it’s not a complete copy of the site. Once all of the data is retrieved, another Herculean task awaits: organizing 20 years of unwieldy, retro internet forums. Many of these conversations are merely historical relics — one thread about stripper safety from the early 2000s recommends that dancers invest in a cell phone — yet some advice from the forums is timeless. “The most chilling thing about the site being destroyed instead of sold to a loving supportive group of industry women, is that for some newbie cammers and dancers, that site is the most support they’ve ever had,” Optimist said. “They literally come on to chat like we’re sisters or like elder family members.” When the site shuts down, there won’t be a central place for forum members to go. The site’s vague notice about the shutdown offers a forum operated by a sex worker named AmberCutie as an alternative, but Stripperweb users seem unenthused, because her forum focuses specifically on camming, rather than other trades like dancing. It’s not clear why the site now points people toward that particular forum, but some users say it reflects how out of touch the owner is with what the community needs. AmberCutie did not respond to a request for comment. Taja Ethereal wonders if the repository of advice on Stripperweb is exactly what led to its downfall. Dancers share tips on which clubs to avoid, what they should expect to be paid and how they know if they’re getting ripped off. She describes the website like an unofficial union. Last year, dancers at Star Garden in North Hollywood before receiving approval from the National Labor Relations Board to conduct an election to join the Actors’ Equity union. If they win their vote, they would become the first unionized strip club since 2013, when the Lusty Lady club in San Francisco closed. Taja Ethereal says she became suspicious when she and other moderators wanted to upgrade sections of Stripperweb where dancers compare notes about which clubs are safe to work at, but their requests were denied or ignored. “Why wouldn’t you want a resource so people can find out what’s a good club, what’s a bad club?” she said. “Other people have said that maybe that’s because strip club owners don’t want us talking about their clubs.” As the community prepares for Stripperweb to close, the members of the forum are looking for a place to go — Taja Ethereal started a website called , almost an homage to the forum she’s spent more than a decade moderating. As the mysterious, unknown owner lets the forum disappear, its members lament the loss of an accessible place for new sex workers to learn how to stay safe and take care of themselves. “Right now, there are women living in precarious situations who need help getting hired, getting their hustle down before they end up homeless. Girls fly into a city on a ‘Strip Trip’ and need to know what to do,” Optimist said. “They’ll look up on Wednesday and 20 years of aunties who left their best advice will be gone.”

InstaDeep’s acquisition is a classic case of an African startup gone global 

Tage Kene-Okafor

largest vaccine maker BioNTech announced that it had Tunisian-born and London-headquartered AI startup InstaDeep for up to £562 million, including a performance-tied £200 million tranche investment. InstaDeep’s deal — subject to regulatory approval and expected to close in the first half of this year — is quite intriguing, for a few reasons. First, when completed (at $682 million, adjusted in U.S. dollar terms), it’ll become the largest acquisition deal involving an African or Africa-focused startup, besting prices bargained for Sendwave, DPO Group and . Second, unlike the other high-profile acquisitions, InstaDeep isn’t a fintech. And third, although early believers who witnessed InstaDeep’s growth from a local firm to a global startup knew it had enough exit options, they didn’t think the acquisition would happen this fast, said , senior partner at AfricInvest, one of InstaDeep’s earlier investors, on a call with TechCrunch. In 2019, InstaDeep raised an $8.5 million Series A at a $30 million valuation, according to sources familiar with the round, which AfricInvest led with participation from New York–based Endeavor Catalyst and a broad range of business angels in the global AI industry. The investment was AfricInvest’s first involvement in an AI startup, a decision based on InstaDeep’s founders selling a global vision to the Pan-African private equity firm. “InstaDeep happened to be quite different from other companies in our pipeline as they were actually into deep tech versus applying technology to a certain sector, where basically, you become an operator in that sector. They were developing specific technology that could impact many sectors,” noted Jilani on InstaDeep’s pioneering tech. “And it was also interesting, especially in Africa, where such companies are quite rare. And so when we had discussions with Karim over his vision and strategy, we quickly realized that InstaDeep could transform from an African leader in AI to a global player.” InstaDeep utilizes advanced machine learning techniques, including deep reinforcement learning in applications within an enterprise environment that cuts across various industries such as biotech, transportation, electronics manufacturing and logistics. Ultimately, this helps companies optimize the decision-making process and improve efficiency. and founded the startup in Tunis in 2014 with “two laptops, $2,000, and a lot of enthusiasm,” CEO Beguir . The bootstrapped company — which didn’t receive outside capital until 2018 — depended on original , which led to the startup being discovered by specialized clients who later became partners and investors, such as DeepMind, Google and its future acquirer BioNTech. As InstaDeep’s clientele grew globally, so did its team. The company has 240 staff across Tunis, London, Lagos, Dubai, Berlin, Cape Town, Paris, Boston and San Francisco. Also, InstaDeep’s ambition to become a global company made it move its headquarters from Tunis to London, which some , thus neglecting its African roots. “InstaDeep is a global company, but in terms of origins and like the company’s early days, there’s no doubt that we’re African,” Beguir told me on the call. “One of the reasons we founded InstaDeep was to show that there was real potential and opportunity for AI in Africa. So we want people to see us as a deep tech African startup gone global, which sends a powerful message of hope for the space.” If anything, InstaDeep has proven that an African company with African talent can successfully serve clients globally while building a talent bridge corresponding to that growth. On the other side of the table are somewhat naive views that argue InstaDeep’s “Africanity.” Tunisia, due to its inhibiting government policies, is an — excluding InstaDeep, Tunisian startups raised $17 million last year, according to a report by VC firm Partech. As such, most startups have had to domicile abroad to access funding. Also, InstaDeep’s influence in building AI talent on the continent isn’t discussed enough. Last year, the upstart played a notable role in helping to organize and nurture Africa’s AI ecosystem via and , hackathons and events with thousands of AI talents and 400 researchers in attendance. Most importantly, an African startup serving clients outside the continent doesn’t make it less African; in fact, founders should be encouraged to build software and AI businesses that present better exit opportunities than e-commerce, logistics and payments, sectors that international companies only consider when expanding into a new region. Tech Safari The ripple effect of InstaDeep building global-first is that it has put the Tunisian tech ecosystem and, more broadly, the AI industry in Africa under the radar with the news of its acquisition. Yet, it’s too early to assume that because of that, it’ll suddenly open the sluice of venture capital in Tunisian tech or Africa’s AI market, which currently lags several industries as hotbeds of investments on the continent. There is potential, though, particularly with the applications of the technology in various sectors such as agriculture and manufacturing; startups like South Africa’s and have raised significant funding for this — however, patience will be required before any breathtaking activity occurs. To my question on whether InstaDeep is an outlier, Begiur expressed optimism that more success stories from Africa’s deep tech and AI community would be told sooner rather than later, especially as the venture capital market has turned red-hot for AI-based innovation. When this happens, the CEO says he hopes that founders and investors reinvest back into the space, something InstaDeep and AfricInvest intend to act on moving forward. “I believe that AI is a huge opportunity for Africa and I’ve been vocal about it. We often see AI as a technology and a competition between developed countries. In reality, AI is essential for Africa’s success in the 21st century, and the reason is that it is the transformational technology of our time; I think you’ll see so many examples these days from GPT and beyond of its disruptive potential,” Beguir, who is half-Tunisian and half-French continued. “But importantly, the barrier to entry to AI is much lower than, let’s say, technologies of the past that were classically associated with legacy companies and strong superpowers. As such, it is a great opportunity for the continent.” Last January, InstaDeep , over 12x what it raised in its previous priced round. Such was the proactive interest of new investors, including Alpha Intelligence Capital, CDIB, Google and BioNTech, its new owner with whom it started working with in 2019 and launched  the following year to deploy the latest advances in AI and ML to develop novel medicines for a range of cancers and infectious diseases. Following the investment, InstaDeep was looking to make some acquisitions to ramp up its data collection capabilities to complement its AI systems before BioNTech swooped in with the acquisition offer, virtually leaving most of the growth financing untouched. “That was crazy. Frankly, we [InstaDeep and early investors like AfricInvest] did not expect that to happen,” expressed Ben Jilani, whose firm may be sitting on a conservative 10x+ exit multiple based on independent calculations. InstaDeep exited at a higher valuation than what it commanded for its Series B, according to Beguir. According to a statement on the acquisition, BioNTech and InstaDeep have already developed multiple end-to-end AI-based applications trained on public and proprietary datasets across various scientific domains. These include projects to enhance neoantigen selection, ribological sequence optimization for BioNTech’s platforms, and the development of an Early Warning System to detect and monitor high-risk SARS-CoV-2 variants based on their ability to escape immune defenses announced last . “With BioNTech, we have developed a partnership over the years and completed many successful projects together. We see great opportunities to build the next generation of immunotherapies and become the leader in biopharma and AI. I believe this is an exciting time, and we will have more to share in coming months,” Beguir said about the acquisition without divulging new information while adding that InstaDeep will use its Series B funding and exit money to scale its teams and capabilities across Africa and globally. “It’s a continuation of what we’ve done in many ways,” he added.

Entocycle grabs $5 million for its insect breeding technology

Romain Dillet

Even if insects don’t sound appealing to you, black soldier flies could play an essential role in the food chain in the coming years. In particular, these flies’ larvae can become an important source of proteins for livestock and fish. That’s why is raising another $5 million in a Series A funding round led by , a European climate-focused VC firm. and are participating in the round as well. is also investing in Entocycle. This new French VC firm is partnering with current and former athletes to invest in tech companies. In addition to contributing money, those sports professionals help startups with team-building advice and mentorship. In that case, Antoine Dupont, Nikola Karabatic, James Haskell and Antoine Brizard are investing in Entocycle. While Entocycle has been , the team of 21 people will now try and commercialize its products. In particular, with this new injection of cash, Entocycle plans to iterate on its flagship product — the Entocycle Neo. It’s a hardware module that can be used in insect farms to monitor and collect data on the health and productivity of black soldier flies. The Entocycle Neo uses optical sensors combined with a software solution that analyzes images and accurately measures production. By automating these processes, Entocycle hopes it can increase productivity in insect farms. Using the company’s modules should lead to higher-feed conversion rates and lower mortality. Similarly, Entocycle has developed a fly cage with built-in climate control. The idea is that Entocycle can help companies in the food industries get started with black soldier fly larvae so that they can secure their supply of proteins. And this is key to understanding the appeal behind Entocycle. Switching to insect-based proteins could drive down soybean production and imports, as well as deforestation — indirectly. Larvae are a low-carbon alternative as they can be produced anywhere. Insect farms could also integrate into the waste management cycle as black solider flies gobble down food waste. Entocycle

Fairphone nabs $53M in growth capital for ‘sustainable’ consumer electronics

Natasha Lomas

Dutch social enterprise , which makes modular and — the claim is — more sustainable and ethical consumer electronics, has nabbed a chunk of funding to continue scaling a circular-economy-aligned smartphone business. The €49 million (~$53 million) “growth capital” investment — from an international consortium of impact investors, led by new shareholders Invest-NL, the ABN AMRO Sustainable Impact Fund and existing investor Quadia via its Regenero Impact Fund — is more than the startup has raised since being founded, back in 2010. (Fairphone had previously raised $40.7 million, per — spread across nine funding rounds, drawing on a range of sources from crowdfunding to VC and debt.) Other existing Fairphone investors DALHAP, DOEN Participaties and PDENH also participated in the new raise — while a number of other investors, including PYMWYMIC and over 1,000 crowdfunders, exited at this point. While Fairphone said it’s using some of the new funding to settle some existing debt. But the headline claim for the investment is growth. The European startup is well positioned to capitalize on opportunities flowing from the bloc’s push for a green transition by encouraging a shift to circular business models, with a top-line goal for the European Union to be carbon neutral by 2050. EU regulators are also for mobile devices and other consumer elections — a future requirement which Fariphone is already fulfilling, thanks to its modular, repairable-by-design devices, putting it well ahead of the (waste-generating) industry curve. Fairphone said the new funding will be used to strengthen its brand positioning — and create further awareness around fairness and sustainability in the electronics industry. Additionally, it said it wants to accelerate integration of fair and recycled materials into its full product portfolio — saying for example that it will be extending its mining “value chain” programs in Africa & South America, and fair wage programs in Asia. It will also be funnelling funds into product development and improved customer service — including to keep pushing the envelop on . Fairphone recently launched its — which contain recycled plastic, Fairtrade gold in the supply chain and an extended battery life vs rival products. We’ve asked Fairphone for its latest sales metrics and will update this report with any response. Fairphone’s CFO, Noud Tillemans, told us the company sold around 120,000 devices last year — up from around 88,000 in 2021 and 23,000 in 2018. “This is more equity than we raised ever before in total,” he confirmed. “The €49 million is pure equity. Some of it will be used to settle existing loans. I can only share the majority is used for growth. Some initial shareholders, investing 5-10 years ago, were keen to exit.” “We are excited to support Fairphone’s growth ambitions, as a truly circular lighthouse case within the electronics industry,” added Elisabeth Storm de Grave, Principal at Fairphone’s new investor, Invest-NL, in another supporting statement. “With its unparalleled approach to creating ethical products with both people and planet in mind, Fairphone sets new standards for the entire industry. Together, we are disrupting a short-term way of thinking that the world can no longer afford, creating a sustainable and fair future for all stakeholders’” While Hanna Zwietering, of the ABN AMRO Sustainable Impact Fund, pointed to what she couched as a “growing trend towards conscious consumer behavior” — lauding Fairphone as “a frontrunner in the sustainable electronics industry” she said has “proven it can develop high-quality modular and fair smartphones in most competitive markets”.

As NYC public schools block ChatGPT, OpenAI says it’s working on ‘mitigations’ to help spot ChatGPT-generated text

Kyle Wiggers

New York City public schools have restricted access to , the AI system that can generate text on a range of subjects and in various styles, on school networks and devices. As this morning and confirmed to TechCrunch by a New York City Department of Education spokesperson, the restriction was implemented due to concerns about “[the] negative impacts on student learning” and “the safety and accuracy” of the content that ChatGPT produces. “While the tool may be able to provide quick and easy answers to questions, it does not build critical-thinking and problem-solving skills, which are essential for academic and lifelong success,” the spokesperson told TechCrunch via email, adding that the restricted access came in response to requests from schools. It’s not a ban per se. The New York City public school system is using the same filter for  that it uses to block other apps and websites — for example, YouTube and Facebook — on school property. Individual schools can request to have ChatGPT unblocked, and the spokesperson said that the New York City Department of Education would “welcome” the opportunity to have a conversation with OpenAI, the startup behind ChatGPT, about how the tool could be adapted for education. When reached for comment, an OpenAI spokesperson said the company is developing “mitigations” to help anyone spot text generated by ChatGPT. That’s significant. While TechCrunch reported recently that OpenAI was with a watermarking technique for AI-generated text, it’s the first time OpenAI has confirmed that it’s working on tools specifically for identifying text that came from ChatGPT. “We made ChatGPT available as a research preview to learn from real-world use, which we believe is a critical part of developing and deploying capable, safe AI systems. We are constantly incorporating feedback and lessons learned,” the OpenAI spokesperson said. “We’ve always called for transparency around the use of AI-generated text. Our policies require that users be upfront with their audience when using our API and creative tools. . . We look forward to working with educators on useful solutions, and other ways to help teachers and students benefit from AI.” has an aptitude for answering questions on topics ranging from poetry to coding, but one of its biggest flaws is its ability to sometimes give that sound convincing but aren’t factually true. That led  from sharing content generated by the AI, saying that ChatGPT made it too easy for users to flood the platform with dubious answers. More recently, the International Conference on Machine Learning, one of the world’s largest AI and machine learning conferences, a prohibition on papers that include text generated by ChatGPT and other similar AI systems for fear of “unanticipated consequences.” In education, the debate has revolved largely around the cheating potential. Perform a Google search for “ChatGPT to write school papers,” and you’ll find plenty of examples of educators, journalists and students testing the waters by wielding ChatGPT to complete homework assignments and standardized essay tests. Wall Street Journal columnist Joanna Stern ChatGPT to write a passing AP English essay, while Forbes staffer Emma Whitford it to finish two college essays in 20 minutes. to The Guardian, Arizona State University professor Dan Gillmor recalled how he gave ChatGPT one of the assignments he typically gives his students and found that the AI’s essay would’ve earned “a good grade.” Plagiarism is another concern. Like other text-generating AI systems, ChatGPT — which is trained on public data, usually collected without consent — can sometimes regurgitate this information verbatim without citing any sources. That includes factual inaccuracies, as alluded to earlier, as well as biased — including racist and sexist — perspectives. OpenAI continues to introduce filters and techniques to prevent problematic text generations, but new workarounds pop up every day. Despite those limitations and issues, some educators see pedagogical potential in ChatGPT and other forms of generative AI technologies. In a recent piece for Stanford’s Graduate School of Education website, Victor Lee, associate professor of education at Stanford, noted that ChatGPT may help students “think in ways they currently do not” — for example, by helping them discover and clarify their ideas. Teachers may benefit from ChatGPT as well, he posits, by generating many examples for students of a narrative where the basic content remains the same but the style, syntax or grammar differ. “ChatGPT may [allow] students to read, reflect and revise many times without the anguish or frustration that such processes often invoke, [while] teachers can use the tool as a way of generating many examples and nonexamples of a form or genre,” Lee said in a . “Obviously, teachers are less delighted about the computer doing a lot of legwork for students. And students still need to learn to write. But in what way, and what kinds of writing? A . . . side effect of this new medicine is that it requires all of us to ask those questions and probably make some substantive changes to the overarching goals and methods of our instruction.” In any case, the New York City public schools policy, which appears to be the first of its kind in the country, will surely force the conversation at school districts elsewhere. As use of the tech grows — ChatGPT had over a million users as of December — and have begun piloting tools to detect the use of AI-generated text in student submissions. Some educators might choose to embrace them, while others, like Lee, instead encourage the use of ChatGPT as an assistive writing tool.

Why Africa had no unicorns last year despite record fundraising haul

Tage Kene-Okafor

scene was met with fanfare in 2021: Venture capital investments in the region totaled between $4 billion and $5 billion and produced five unicorns. In my piece this progress, I predicted there would be more unicorns in 2022. Those predictions proved to be way off the mark by year’s end. Data from market insights trackers and reveal that funding raised by African startups exceeded $5 billion (including undisclosed deals) in 2022 — a slight percentage increase from the figures reported in 2021 despite a global pullback in VC funding. And yet, no unicorns popped up throughout the year, compared to five in 2021. That fact may appear insignificant because, at the end of the day, private valuations don’t pass an actual test till startups go public. However, producing no unicorns despite raising more venture capital suggests it’s perhaps too early to assume African markets are mature enough to consistently pop out private billion-dollar companies like their Global South counterparts: India, Southeast Asia and Latin America. That said, 2022 was peculiar. The global economic downturn and venture capital crunch ensured that every region produced fewer billion-dollar companies than the previous year. Globally, 216 unicorns were minted in 2022, per , compared to 541 in the previous year. In , 22 companies became unicorns last year, compared to 46 in 2021. While in Latin America got their horns in 2021, that figure last year. Unlike Africa, these regions raised way less venture capital in 2022 than in 2021, so it makes sense that their unicorn numbers dropped. For example, in India, the number of unicorns dropped by more than half as . Latin America and Southeast Asia also compared to 2021, though the drop in unicorns indicates more damage. So what happened in Africa in 2022 that made it so … weird?

TBD Health is rolling out at-home sexual healthcare to all US states

Haje Jan Kamps

Reproductive rights is a hot-button topic for much of the political spectrum in the U.S. Against that backdrop. is not-so-quietly making patient-focused and inclusive sexual healthcare available and accessible to everyone in the U.S., as it raises a $4.4 million round to scale up further. “At a time when women’s rights are being challenged by our government, it was critical for us to build a sex-positive healthcare company that makes it easy for people to prioritize their health,” said Daphne Chen, co-founder and co-CEO of TBD Health in an interview with TechCrunch. “As someone who has experienced firsthand the judgment that doctors and clinicians pass when it comes to sexual health, it’s our mission to create a safe space for people to seek inclusive, trauma-aware care, as well as treatment and resources on their own terms, regardless of where they live.” The company is extending operations in the context of a huge amount of change. At one end of the scale, people are more likely to have non-traditional relationships where multiple sexual partners are a thing, and sexual safety and testing remains important. At the other extreme, a huge swathe of the U.S. has woefully lacking sex ed and access to sexual health clinics. The CDC reports that 20% of people in the U.S. have an STI, and that young people are not as good at preventing them as perhaps they ought to be. TBD Health is pushing a new line of services it refers to as digital sexual health. “We have a couple of services that are available today. One of which is our at-home STI testing, which we’re really excited about because it’s a protocol that allows you to manage your sexual health from the comfort and privacy of home,” says Chen. “What that looks like is we’re able to send you a kit that’s customized to your specific needs and your risk profile to your home, you’re able to self-collect your samples (urine, blood as well as any swab samples that might be needed), which then gets shipped to our partner lab and you get your results in just a few days. All results are reviewed by our clinical team and a customized care plan is created just for you. And we’re also really excited about the fact that your relationship with TBD doesn’t end after your test result comes back. We are creating an ongoing relationship, and we’re doing a lot of follow-up. We really believe that as your sex life changes your sexual health care needs do too. And so we want to stay on top of that and make sure that we’re able to serve you no matter where you are in your sexual health journey.” We spoke with the company , when it was just starting to gather speed with its at-home services, in addition to its in-person clinic in Las Vegas. Today, the company told TechCrunch it has raised a $4.4 million round of investment from Tusk Venture Partners, with participation from Springdale Ventures, Human Ventures, Expansion VC, Starbloom Capital, Hyphen Capital and The Community Fund. “More and more we’re seeing local sexual health clinics nationwide shuttering due to lack of funding and resources. This leaves an immense strain on those who urgently need care,” said Stephanie Estey, co-founder and co-CEO of TBD Health. “We seek to be a solution for those living in sexual health care deserts and beyond by offering a more accessible and approachable way to taking care of your sexual health.” Of course all eyes are on Roe v. Wade’s impact on delivery and access to sexual health services. “Since Roe v. Wade was overturned last year, it’s critical people must have accessible options to reproductive and sexual healthcare,” said Bradley Tusk, managing partner at Tusk Venture Partners in a statement provided by TBD. “We are proud to back TBD Health as they establish new standards for care through its STD testing kits, availability of emergency contraceptives with next-day shipping nationwide, and an in-person care hub in Las Vegas.” The company has invested a lot into ensuring its services are trauma-informed and radically inclusive, which means that its clinicians are able to meet complex sexual health situations with understanding and kindness. “The true differentiator of TBD is the care that comes along with the experience. All of our clinicians are trained in sex positivity, they have sensitivity training, and they are trauma-aware. And so with TBD, you are not just getting a test result. You’re getting a partner. So what does that actually look like?” Estey asks rhetorically. “That looks like counseling. That looks like making sure you have the next steps. That looks like making sure that you have resources to think about an STI result both for yourself as well as your partners. We really walk you through the entire process. That is dramatically lacking in healthcare today, which feels very transactional. TBD is all about putting the human and compassionate approach back into the experience and making sure that people feel empowered to take ownership of their sexual healthcare.” You can .

The Team Slide is the most important slide in a startup pitch deck

Haje Jan Kamps

graduates and becomes a “real” company, with hundreds of employees and the first inklings of large amounts of revenue, you could argue that a Team Slide is less important. Yes, the top leadership team still has to be good, but if the company is growing rapidly, getting new customers and delivering a good product, it’s kind of a self-fulfilling prophecy: If the team is able to grow the company, at least they aren’t completely helpless. But in the early stages of a company, things are a little different, and that’s where a good story around a company’s team is crucial. And yet, I am surprised how often startup teams are wasting the opportunity of telling the story about their team. Spoiler alert: For early-stage companies, investors are looking for exactly three things: Five Flute was confident enough to put its Team Slide as its second slide. And it’s easy to see why: This is what an extraordinary team looks like. . The ugly truth is that if you have a good idea, chances are that there are 10 other teams working on the same idea. It’s unlikely that you are special enough to have a 100% unique idea. It’s unlikely that you are the first founders to pitch a particular idea to a group of investors. So the question morphs into this: Why should the investors back you and your team? That is the question a Team Slide has to answer. In this post, I’ll show you how.

Plugable’s new dock turns your tablet or phone into a workstation

Haje Jan Kamps

Laptops and tablets are getting better and more powerful by the day but our fingers aren’t getting any more precise, and for extended writing and work, it’s still more helpful to have a keyboard and a larger screen at hand. These types of products , but what’s new about ‘s offering is the sleek portability and build quality that makes your phone and tablet so portable in the first place. The USB-C docking station is called UDS-7IN1, and enables the users of supported devices to turn their phones and tablets into workstations. The device adds an HDMI port to connect a monitor, alongside a couple of USB 3.0 (5Gbps) ports, SD and microSD ports, an audio input/output jack, and 100W USB-C pass-through charging to keep everything powered and on charge. The stand is built of aluminum, and folds down to a pretty compact little package to take with you on your travels. Supported devices are most modern USB-C tablets, and select phones, and the device starting today.

Would Baidu’s answer to ChatGPT make a difference?

Rita Liao

Baidu, China’s top search engine provider and robotaxi developer, is apparently working on its own counterpart to ChatGPT. The news, first reported by and , sent Baidu’s stock price rising on Monday to reach its highest point since September. A spokesperson for Baidu declined to comment on the reports. But it wouldn’t be surprising that Baidu, which bills itself as the pioneer in China’s artificial intelligence field, is stepping up to build the Chinese equivalent of today’s most powerful chatbot. The question is how big a difference the tool can make, and where its limitations lie. A driving force shaping China’s tech development over the past few years is the rise of digital sovereignty, which refers to a country’s ability to control its own “ ” and can include autonomy in critical software and hardware in the AI supply chain. Episodes of U.S. export bans on China have pushed Beijing to further call for tech independence in areas ranging from semiconductors to basic research on AI. As OpenAI’s ChatGPT shows the potential to disrupt sectors from education and news to the service industry, China’s tech leaders and policymakers are likely pondering how AI can also be used to drive productivity at home. China naturally wants its homegrown ChatGPTs, not just to secure control over how data flows through such tools but also to create AI products that better understand local culture and politics. That’s where things get interesting. Like all other channels of information in China, the Baidu chatbot will no doubt be subject to local regulations and censorship rules. As we earlier, the firm’s text-to-image application, ERNIE-VilG, already rejects politically sensitive prompts. But conversational AI handles much more complex inquiries than image generators — how will Baidu walk the line between censorship confinement and leaving enough freedom and creativity to its bot? Also important to machine learning performance is the undergirding algorithms. According to The Wall Street Journal, Baidu adapted a “core breakthrough” that Google developed in 2017 and open-sourced, an algorithm that has also powered ChatGPT. Most likely, though, there are other key pieces of proprietary algorithms that Baidu has acquired or developed to form the backbone of its chatbot. Hardware plays another important role in training large-scale neural networks. U.S. chip sanctions against China are posing a threat to China’s AI industry as companies that power supercomputers and large data centers. Baidu, however, believes the chip ban has a , as we reported. In the near term, the company “already stocked enough [chips] in hand.” Lastly, the success of Baidu’s ChatGPT alternative depends in part on continuous data training through user feedback, such as giving a thumb-up or -down to the machine’s responses. In other words, the more people using it, the better the AI assistant understands how to respond to humans. Ella Zhang, founder and CEO of text-to-image startup (who previously was working on ), reckons that Chinese-language chatbots “might not see the same strong demand yet as English ones because China still enjoys relatively cheap labor.” So instead of subscribing to expensive AI software and finetuning it to carry out customer service tasks, a Chinese company might simply hire a team of human staff for affordability and convenience. Things might change in a few years, though, as China gradually loses its labor advantage in a .

Instagram’s co-founders introduce a new social app…for news reading

Sarah Perez

Can lightning strike twice? That’s apparently the question being raised today with the public introduction of the next social app built by Instagram’s co-founders, Kevin Systrom and Mike Krieger. The duo have launched a new venture to explore social apps, according to a report published in , which includes the debut product , a personalized news reader. The app itself is not yet publicly available but offers a waitlist where interested users can sign up. As described, it sounds like a modern-day twist on Google Reader, a long-ago RSS newsreader app that . Except in this case, Artifact is described as a newsreader that uses machine learning to personalize the experience for the end user, while also adding social elements that allow users to discuss articles they come across with friends. (To be fair, Google Reader had a similar feature, but the app itself had to be programmed by the user who would add RSS feeds directly.) Artifact will first present a curated selection of news stories, The Verge’s article notes, but these will become more attuned to the user’s interests over time. Some of the articles will come from big-name publishers, like The New York Times, while others may be from smaller sites. Other key features will include comment controls, separate feeds for articles posted by people you follow alongside their commentary and a direct message inbox for discussing posts more privately. The concept seems as if it has some overlap with one of Twitter’s bigger use cases around discussing news. It also arrives at a time when Twitter users are considering new options after the app’s acquisition by Elon Musk, who has made changes to the app’s roadmap and policies, some longtime users in the process. But as described, Artifact doesn’t sound completely original — not only does it seem like a modern twist on a Google Reader-type experience, it would go up against various other news reading apps, both new and older, which include personalization elements, like , SmartNews and Newsbreak. It also sounds similar to and its newer which offers a combination of news reading, curated recommendations and comments. Even , launching a way for its readers and writers to chat in-app. Overseas, the model has seen success with ByteDance’s Toutiao, but a U.S. version . And of course, the new app would compete in many ways with the social giant Meta, too, which Instgram’s co-founders left back in 2018. Facebook and to a lesser extent Instagram and WhatsApp, today serve as portals where billions interact and engage with news and information, amid their updates from friends, family, groups and businesses they follow. That means no matter how polished or differentiated Artifact may become, it could still face a host of competition in the market, where consumers also already have built-in news apps available with Apple News and Google News. According to The Verge’s report, the duo believes the recent leaps made in machine-learning technology could help give Artifact an edge, however, similar to how algorithmic recommendations have played a role in elevating TikTok to become a dominant app. But while TikTok’s personalized For You feed is arguably addictive, the video app’s growth was seeded by record-breaking marketing spend on its user acquisition efforts — even reaching $1 billion per year in 2018, . A startup, even from remarkable founders, may not have the same amount of fuel to throw on the fire. And news reading in and of itself seems to be a bit of a passé market to chase in an era when younger Gen Z users are often now turning to entertainment apps like TikTok to stay informed on news and world events, too. It’s wading into a polarized news ecosystem, too, with the founders promising to make the  “subjective” and “hard” calls over the content on its network. That said, it’s difficult to count out the success of those who built Instagram, which, was one of the largest social tech acquisitions of its time and has shaped the way the world engages with social media, for better or for worse. As an early-stage product, Artifact is still being developed and is not yet monetized, but a revenue share with publishers was mentioned as a possible option. ( ?) The app’s individual success may or may not ultimately matter, though, given the founders intend on testing other new social products through their new venture, it seems. Currently, Artifact’s website is taking sign-ups from those who have U.S. (+1) phone numbers.