Elon Musk posted on X that Tesla will be restarting work on Dojo3, the third generation of its in-house supercomputer project. The Dojo team had been disbanded last year as the company prioritized the AI chips that run on board Tesla vehicles. Musk said the company is returning to the project "now that the AI5 chip design is in good shape."
The purpose of the Dojo project is to process video recordings and other data from Tesla vehicles and use that to train the "neural net" behind the company's Full Self-Driving software. Last year, however, Musk posted on X that "It doesnt make sense for Tesla to divide its resources and scale two quite different AI chip designs. The Tesla AI5, AI6 and subsequent chips will be excellent for inference and at least pretty good for training. All effort is focused on that."
The AI chips Musk is referring to are ones developed for running FSD onboard Tesla vehicles and are not optimized for training. The AI6 chips will be made by Samsung in the company's Texas factory, after it struck a $16 billion agreement with Tesla.
Musk has also claimed a lot of things over the years, and many of those assertions either were misrepresentations or simply didn't pan out. Working against this chip project: Musk said that Dojo3 will be "space-based AI compute," as he and others believe that data centers in orbit are a superior alternative to the land-based behemoths currently being built. The idea is that space provides easier access to the sun's energy, and the cold temperatures there might greatly reduce the power needed, among other benefits. While it's an increasingly popular if entirely speculative idea, experts have their doubts.This article originally appeared on Engadget at https://www.engadget.com/ai/musk-claims-tesla-will-restart-work-on-its-dojo-supercomputer-173127863.html?src=rss
Metas Threads is pulling further ahead of Elon Musks X on mobile, based on recent estimates from analytics firm Similarweb, Forbes reports. In the first stretch of January, Threads averaged roughly 143 million daily active users worldwide on mobile devices, compared with about 126 million for X.
Similarwebs year-over-year snapshot shows Threads growing sharply, up 37.8 percent year-over-year, while Xs daily mobile audience fell 11.9 percent across the same period. The picture is more mixed in the US, where X still holds a narrow edge on mobile. Similarweb data puts X at about 21.2 million daily active US mobile users in early January versus roughly 19.5 million for Threads.
However, Threads US mobile usage has risen substantially faster over the past year, surging almost 42 percent to X's 18 percent. X remains far larger on desktop, where it draws around 150 million daily users or visits worldwide, while Threads web presence sits at just 9 million.
Forbes also reported on Similarweb data for Bluesky, another competing text-based platform started by Twitter founder Jack Dorsey. Dorsey left the board in the summer of 2024, later telling Pirate Wires he believed Bluesky was "literally repeating all the mistakes we made as a company," in reference to Twitter. The social network opened registrations in 2024, and sits now with a daily mobile user base of 3.6 million, which Similarweb says is down 44.4 percent year-over-year.
X has found itself in hot water yet again over xAI's Grok chatbot, which was altering pictures of women on the platform to create lewd images at the request of users without the consent of those pictured. In some cases, the chatbot also altered the images of underage girls. The uproar in response led the company to shut off image generation for nonsubscribers and place firmer guardrails on what types of images can be generated. The delayed action came after weeks of Grok creating tens of thousands of these images, and after the Attorney General of California launched an investigation.This article originally appeared on Engadget at https://www.engadget.com/social-media/threads-has-more-global-daily-users-than-x-on-mobile-for-the-first-time-144936831.html?src=rss
Addressing a disconnect between Gen Z's values and its capabilities, the Levi's Wear Longer Project is a free educational initiative teaching young people how to repair, alter and customize their clothing. Developed in partnership with Discovery Education, the program offers a digital curriculum, in-classroom lessons and community workshops designed to fill what the brand's 2025 research identified as a significant skills deficit: 41% of Gen Z lack basic clothing repair abilities like hemming or patching, compared to less than 25% of older generations who typically learned these skills at home or in school. And 35% of Gen Z say they'd keep garments longer if they knew how to fix them. The program scales across multiple touchpoints, from self-directed online guides to employee-led workshops in Levi's stores, positioning the 150-year-old denim brand as an educator rather than just a retailer. By teaching skills that extend its clothing's lifespan, Levi's is betting that reducing waste and building customer capability can be part of its overall business strategy.TREND BITEThe Wear Longer Project signals how legacy brands can leverage their heritage to address contemporary consumer tensions. Gen Z's simultaneous commitment to sustainability and lack of practical skills creates space for companies to become educators. As millions of wearable garments end up in landfills each year, repair knowledge becomes a competitive differentiator.Levi's isn't simply selling vague promises of durability it's teaching customers to actualize values they already hold but lack the skills to put into practice. For brands grappling with how to authentically engage younger consumers on sustainability, this approach offers a template: identify the gap between aspiration and ability, then fill it with practical tools that reinforce your product's core strengths.
If you weren't able to shut down your Windows 11 device recently, Microsoft has rolled out an emergency fix addressing a couple of critical bugs that popped up with its latest January 2026 Windows security update. The latest "out-of-band" update repairs an issue for some Windows 11 devices that would only restart when users tried to shut down or hibernate. The same update restores the ability for Windows 10 and Windows 11 users to log into their devices via remote connection apps.
Microsoft said the inability to shut down or hibernate affected Windows 11 devices using Secure Launch, a security feature that protects a computer from firmware-level attacks during startup. As for the remote connection issue, Microsoft explained in its Known issues page that credential prompt failures were responsible when users tried to log in remotely to affected Windows 10 and 11 devices.
According to WindowsLatest, some lingering issues with the January 2026 Windows security update are still affecting users, like seeing blank screens or Outlook Classic crashing. Back in October, Microsoft had to issue another emergency fix for Windows 11 related to the Windows Recovery Environment. For those still hesitant to upgrade to Windows 11, Microsoft is allowing you to squeeze some more life out of Windows 10 by enrolling in Extended Security Updates.This article originally appeared on Engadget at https://www.engadget.com/computing/microsoft-issues-emergency-fix-afer-a-security-update-left-some-windows-11-devices-unable-to-shut-down-192216734.html?src=rss
Washington state residents may soon be forced to produce IDs before getting onto websites with pornographic content. Within the state's House of Representatives, Rep. Mari Leavitt introduced House Bill 2112, which is informally known as the Keep Our Children Safe Act. Similar to the initiatives seen in other states, the bill proposes to restrict access to "online sexual material harmful" to anyone under 18.
In practical terms, those living in Washington state could see websites asking for digital identification or demanding the user go through an age verification system that requests a government-issued ID. If a website that has more than one-third of its content being "sexual material harmful to minors" is found not following these rules, the state's attorney general can pursue steep civil penalties.
If those restrictions sound familiar, it's because many other states have also passed similar constraints. Washington state's proposed bill is very similar to Texas' age verification law that went into effect in September 2023 and was recently upheld by the US Supreme Court. Like the Texas law, several groups expressed disapproval of the bill during the public hearing at the House committee level. As reported by The Seattle Times, groups including the ACLU, Lavender Rights Project and the Northwest Progressive Institute warned of privacy risks related to potential data breaches and the loose definition of "sexual material harmful to minors" in the bill's language.This article originally appeared on Engadget at https://www.engadget.com/cybersecurity/washington-is-the-latest-state-pursuing-an-age-verification-law-for-porn-sites-174423529.html?src=rss
Six years after the announcement of plans to build Atari Hotels in eight cities across the US, including Las Vegas, only one now seems to be moving forward, in Phoenix, Arizona. The Las Vegas deal ultimately "didn't come to fruition," spokesperson Sara Collins told Las Vegas Sun this week, and Atari Hotels is putting its focus into the Phoenix site "for the time being."Phoenix was always meant to be the first site, followed by other hotels in Austin, Chicago, Denver, Las Vegas, San Francisco, San Jose and Seattle. But Las Vegas is now apparently off the table, and there haven't been any signs of life around the other planned locations. The FAQ on the Atari Hotels website notes, "Additional sites, including Denver, are being explored under separate development and licensing agreements." The Atari Hotel project was announced in 2020 just before the onset of the COVID-19 pandemic and consequently experienced development delays. Construction on the Phoenix hotel, which was supposed to break ground in 2020, is expected to begin late this year, with its opening now planned for 2028. But maybe don't hold your breath. According to a December press release, the company is still trying to raise $35 million to $40 million to fund the "playable destination" for gamers in Phoenix.This article originally appeared on Engadget at https://www.engadget.com/gaming/the-plan-for-a-gaming-themed-atari-hotel-in-las-vegas-has-reportedly-been-scrapped-214212269.html?src=rss
Amazon is reportedly adding Teresa Palmer (The Fall Guy, Warm Bodies, Hacksaw Ridge) to its pantheon of Norse gods for its God of War TV show adaptation. As first reported by Deadline, Palmer will play Sif, Thor's wife and eventual leader of the Aesir, in the live-action adaptation. It may not carry as much weight as the casting of the video game's protagonist that was revealed earlier this week to be Ryan Hurst, but it could hint at the direction the TV show may take.
View this post on Instagram
While Sif plays a minor role in the God of War Ragnarok game, the early casting confirmation could indicate that the showrunner, Ronald D. Moore, may give the character a more involved role. In God of War Ragnarok, Sif is known as Odin's diplomat before the events of Ragnarok, where she becomes the new leader of the Aesir, one of two tribes of Norse gods. Notably, Amazon's adaptation is still missing the casting confirmations for Atreus, Thor, Odin and many other Norse gods seen in the video game. Even so, the God of War TV show has already secured at least two seasons.
This article originally appeared on Engadget at https://www.engadget.com/entertainment/tv-movies/amazons-live-action-god-of-war-adaptation-adds-teresa-palmer-201604602.html?src=rss
In another attempt to reduce our attention spans to mush, TikTok has released the PineDrama app, which offers serialized drama series that are roughly a minute per episode. As first spotted by Business Insider, the app is designed exactly like TikTok, but instead of trendy dance videos, you can scroll through and watch "micro dramas."
For those new to the category, micro dramas are bite-sized TV shows shot in vertical video and available in minute-long episodes. Don't expect any nominations for Best Original Screenplay with series like The Officer Fell For Me or Married to my past life's nemesis, since they typically offer soap opera vibes with cliffhangers that keep users scrolling to the next episode. The app is designed to keep people on it with a Discover tab, a place to save favorites and the ability to react in real time alongside other viewers.
Right now, the micro dramas on PineDrama are all free to watch and don't have any ads. It's unclear if TikTok will introduce any costs or ads to the app, since other micro drama options like DramaBox or ReelShort have a paid structure. Late last year, TikTok also introduced a way to watch micro dramas within its own app, with a section called Minis. It's not the first time we're seeing shorter TV show formats, since Quibi made waves with a format of episodes that were less than 10 minutes long. However, maybe even 10 minutes was too long since the startup eventually called it quits after eight months.This article originally appeared on Engadget at https://www.engadget.com/entertainment/tv-movies/tiktoks-latest-spinoff-app-feels-a-lot-like-quibi-but-with-shorter-and-cornier-content-185702010.html?src=rss
We now have some idea of what's at stake in the longstanding feud between Elon Musk and OpenAI. As first reported by Bloomberg, the latest filing, as part of a lawsuit that accuses the AI giant of abandoning its non-profit status, claims that Musk is owed anywhere between $79 billion and $134 billion in damages from the "wrongful gains" of OpenAI and Microsoft.
Musk claimed in the filing that he's entitled to a portion of OpenAI's recent valuation at $500 billion, after contributing $38 million in "seed funding" during the AI company's startup years. Along with providing "roughly 60 percent of the nonprofit's seed funding," Musk offered recruiting of key employees, introductions with business contacts and startup advice, according to the filing. The monetary estimate comes from C. Paul Wazzan, a financial economist who's serving as Musk's expert in the case. According to Wazzan's calculations, OpenAI earned between $65.5 billion and $109.43 billion in wrongful gains, while Microsoft saw between $13.3 billion and $25.06 billion.
The lawsuit between Elon Musk and OpenAI dates back to March 2024, when the xAI CEO first filed a legal action claiming that OpenAI violated its non-profit status. Musk later added Microsoft as another defendant and even tried to get an injunction when OpenAI announced efforts to reorganize its corporate structure. Besides this suit, Musk has named OpenAI in another legal battle, accusing the company, along with Apple, of monopolistic practices that prevent xAI from getting a fair shot in the App Store.This article originally appeared on Engadget at https://www.engadget.com/big-tech/elon-musk-is-looking-for-a-134-billion-payout-from-openai-and-microsoft-171824945.html?src=rss
California Attorney General Rob Bonta has sent a cease and desist letter to xAI, days after his office launched an official investigation into the company over reports that Grok was generating nonconsensual If youll recall, xAI and Grok have been under fire for taking images of real individuals and putting them in revealing clothing like bikinis upon random users requests. Bontas office demands that xAI immediately cease and desist from creating digitized sexually explicit material when the depicted individual didnt consent to it or if the individual is a minor. It also demanded that xAI stop facilitating or aiding and abetting the creation or publication of digitized sexually explicit material of nonconsenting individuals and persons under 18 years of age. X changed its policies after the issue broke out and prevented the Grok account from being able to edit images of real people into revealing clothing. xAI also moved Groks image-generating features behind a paywall and geoblocked paying users ability to edit images of real people into bikinis, but only in regions where its illegal.In his announcement, Bonta said xAI developed a spicy mode for Grok to generate explicit content and used it as a marketing point. The California AG also said that Grok-generated sexual images are being used to harass both public figures and ordinary users. Most alarmingly, news reports have described the use of Grok to alter images of children to depict them in minimal clothing and sexual situations, Bontas announcement reads. The actions above violate California law, including California Civil Code section 1708.86, California Penal Code sections 311 et seq. and 647(j)(4), and California Business & Professions Code section 17200, it said. The states Department of Justice now expects to hear from xAI on the steps its taking to address these issues within the next five days. xAI has been producing deepfake nonconsensual intimate images used to harass women & girls across the internet.This is unacceptable. Were demanding immediate answers on their plan to stop the creation & spread of this content.Well use all tools at our disposal to keep pic.twitter.com/2elGTV0wRq Rob Bonta (@AGRobBonta) January 16, 2026
This article originally appeared on Engadget at https://www.engadget.com/ai/california-ag-sends-cease-and-desist-to-xai-over-groks-explicit-deepfakes-140000574.html?src=rss
