Posts with «author_name|will shanklin» label

TikTok’s AI efforts reportedly exploit loopholes to use premium Nvidia chips

The US has banned companies like Nvidia from selling their most advanced AI chips to China since 2022. But if loopholes exist, profit-hungry corporations will find and exploit them. The Information published a bombshell report on Thursday detailing how Oracle allows TikTok owner ByteDance to rent Nvidia’s most advanced chips to train AI models on US soil.

ByteDance, which many US lawmakers believe has direct ties to the Chinese government, is reportedly renting US-based servers containing Nvidia’s coveted H100 chips from US cloud computing company Oracle to train AI models. The practice, which runs against the spirit of the US government’s chip regulations, is technically allowed because Oracle is merely renting out the chips on American soil, not selling them to companies in China.

The US government has cracked down on exporting the chips to China as an extension of the tensions between the two nations. The Biden Administration fears the nation could use advanced AI for military or surveillance purposes or to gain an economic upper hand. The US government passed bipartisan legislation in April that will force ByteDance to either sell its US operations or face a ban. But ByteDance still has until early next year to close a deal, and it’s suing the US government, which could delay enforcement.

Although ByteDance is training its models in the US, “it could be difficult to stop them from sending the models they produced back to their headquarters in China,” according to US-based cloud providers and a former Nvidia employee who spoke to The Information. Quite the loophole, indeed.

ByteDance’s Project Texas initiative, which the company claims siloes off TikTok’s US operations from its Chinese leadership to allay US fears, is at the heart of the arrangement. However, former ByteDance employees have described Project Texas as “largely cosmetic,” as they claim the company’s US wing regularly works closely with its Beijing-based leadership.

ByteDance isn’t the only Chinese company looking to game the rules. The Information says Alibaba and Tencent are discussing similar arrangements to gain access to the sought-after chips. Those deals could be harder to squash because they have their own US-based data centers and wouldn’t have to rent servers from American companies.

US cloud computing company Oracle reportedly enables ByteDance’s training of AI models in the US.
Oracle

Not every company has been as willing as Oracle to skirt the law’s intent. “Two small American cloud providers” reportedly turned down offers to rent servers with Nvidia’s H100 chips to ByteDance and China Telecom because “they seemed to go against the spirit of U.S. chip restrictions.” However, Oracle, cofounded by American businessman Larry Ellison and run by current CEO Safra Catz, apparently found the opportunity for profit through technically legal workarounds too tempting to pass up.

The US Commerce Department, the bureau that could close the loophole, may already be aware of the practices. Earlier this year, the department proposed a rule that would require US cloud providers to verify foreign customers’ identities and notify the US if any of them were training AI models that “could be used in malicious cyber-enabled activity.” However, the Commerce Department recently said most cloud providers disapproved of the proposal, claiming “the burden of additional requirements might outweigh the intended benefit.” In the meantime, the proposed rule, which could theoretically plug the loophole, remains in limbo.

But even if the US manages to shut down that exploit, The Information says it wouldn’t cover Chinese cloud providers like Tencent and Alibaba from buying Nvidia’s chips and using them to train AI models in their own US-based data centers. The Commerce Department will have its hands full figuring this one out as business and defense interests wrestle for control.

This article originally appeared on Engadget at https://www.engadget.com/tiktoks-ai-efforts-reportedly-exploit-loopholes-to-use-premium-nvidia-chips-173432988.html?src=rss

YouTube is putting new age restrictions on gun videos

YouTube quietly snuck in a policy change that will age-restrict some gun-related videos and bar others altogether. Content featuring homemade and automatic firearms will be banned for viewers under 18, while tutorials for removing safety devices will be prohibited regardless of age. The new policy will take effect on June 18.

“Starting June 18, 2024, certain content showing how to remove safety devices will be prohibited,” a disclaimer on YouTube’s firearms policy page now reads. “Content showing the use of homemade firearms, automatic firearms, and certain firearm accessories will be age restricted.”

In a statement to Engadget, YouTube spokesperson Javier Hernandez wrote, “These updates to our firearms policy are part of our continued efforts to maintain policies that reflect the current state of content on YouTube. For example, 3D printing has become more readily available in recent years so we’re expanding our restrictions on content involving homemade firearms. We regularly review our guidelines and consult with outside experts to make sure we are drawing the line at the right place.”

YouTube added that the prohibitions will apply to the real use of firearms and won’t pertain to video games, film clips or other artistic content. The platform may also make exceptions for content that’s in the public interest, like military or police footage, news or warzone videos.

The change comes a year after the Tech Transparency Project (TTP), a nonprofit watchdog group, called out YouTube for recommending gun-related content to several “child” accounts the organization set up to see how easily the platform’s algorithms nudged underage users towards gun videos. The researchers set up four accounts, two posing as nine-year-old boys and another pair pretending to be 14-year-old boys. The accounts watched playlists of videos about video game franchises like Halo, Grand Theft Auto, Lego Star Wars and Roblox, and the team monitored the accounts to see what recommendations popped up.

Lo and behold, YouTube allegedly recommended content about weapons and shootings. “These videos included scenes depicting school shootings and other mass shooting events; graphic demonstrations of how much damage guns can inflict on a human body; and how-to guides for converting a handgun to a fully automatic weapon,” TTP wrote at the time.

Other recommended videos featured a young girl firing a gun and tutorials for converting handguns into fully automatic weapons. Some of the content was monetized with ads.

Manhattan District Attorney Alvin Bragg, recently famous for convicting a 77-year-old who had trouble staying awake in court, chimed in soon after TTP’s report last year. The DA asked to meet with YouTube CEO Neal Mohan to discuss why the platform allowed video tutorials for “ghost guns,” firearms assembled using 3D-printed parts or kit components.

The Tech Transparency Project applauded the policy change but warned the real test would be in how stringently YouTube enforces it. “YouTube’s policy changes to age-restrict gun content are a step in the right direction, given that firearms are the number one cause of death for children and teens in America, but it’s not clear why it took the company so long to address the issue,” TTP Director Katie Paul wrote in a press release. “As always with YouTube, the real proof of change is whether the company enforces the policies it has on the books. Until YouTube takes real action to prevent videos about guns and gun violence from reaching minors, its policies remain empty words.”

Update, June 6, 2023, 3:19 PM ET: This story has been updated to add a statement and additional info from YouTube.

This article originally appeared on Engadget at https://www.engadget.com/youtube-is-putting-new-age-restrictions-on-gun-videos-190942704.html?src=rss

Twitch is raising US subscription prices for the first time

Twitch is joining Spotify, Max, Peacock, Crunchyroll, EA and other content services in everyone’s favorite corporate trend of raising subscription prices (almost as fun as the parallel trend of Big Tech layoffs). The Amazon-owned company said on Tuesday that Twitch Tier 1 subscriptions in the US will increase from $4.99 to $5.99 on July 11. This is the first time the monthly cost has gone up for American subscribers.

“As part of our efforts to help creators build and grow their communities worldwide, the following countries received subscription price adjustments as a part of Local Subscription Pricing,” the company wrote in a support article.

In a separate X reply, the company clarified that streamers will still earn the same 50 to 70 percent through Twitch’s revenue-sharing program, so they will earn more per subscription (likely the rationalization for the questionable “It’s for the creators!” framing). However, streamers’ earning extra revenue depends on Twitch’s subscriber numbers staying the same or increasing. An unpopular price hike could lead to a loss of paying subscribers if enough people shirk the increase.

Twitch had warned this day would come. When the company raised subscription prices in Canada, Australia, Turkey and the UK in February, Chief Monetization Officer Mike Minton added that a US subscription increase would “probably” arrive sometime this year. And here we are.

The company has had a rough 2024, and we aren’t even at the halfway point. Twitch laid off a reported 500 employees in January to “cut costs” and “build a more sustainable business” as CEO Dan Clancy admitted the company wasn’t profitable. For good measure, it cut how much creators earn from Prime subscriptions. Then, late last month, it removed every member of its Safety Advisory Council, replacing them with “Twitch Ambassadors,” which sounds an awful lot like community volunteers.

This article originally appeared on Engadget at https://www.engadget.com/twitch-is-raising-us-subscription-prices-for-the-first-time-193204538.html?src=rss

How to watch Summer Games Done Quick 2024

Summer Games Done Quick (SGDQ) is back for more charity speed runs of classic and contemporary games. The 2024 edition of the marathon will stream from Minneapolis from June 30 to July 6 as turbo-gamers take on Alan Wake II, Halo 3 and the Super Mario RPG remake (among many others). The event’s first speedrunning dog will even make his triumphant return, taking on an SNES baseball game from 1994.

The festivities start with a pre-show event at 12:30 PM ET on June 30, followed by a run of the 1997 platform game Yoshi’s Story at 1 PM. The weeklong event wraps on July 6 with a quick play-through of the new Super Mario RPG remake for Switch at 11:05 PM, followed by a special finale.

The rest of the week is filled with old-school classics, recent AAA titles, oddball novelty games and excruciatingly difficult custom mods. The latter can be especially fun to watch as gamers execute well-timed jumps that would take most of us hours of practice to perform once, only they do a series of them on the first try on cue in front of a live audience.

A few runs worth eyeing include a late-night Alan Wake 2 speed-through (July 2 at 12:49 AM ET), The Legend of Zelda: A Link to the Past (July 4 at 10:35 AM), Halo 3 (July 4 at 7 PM), Metal Gear Solid 4: Guns of the Patriots (July 6 at 1:21 AM) and Super Mario World (July 5 at 5:50 PM). If the original SNES Mario adventure isn’t your thing, MrMightyMouse will perform a run of the game’s ROM mod Grand Poo World 3 on July 5 at 11:26 PM ET.

Games Done Quick / Wes "Fish" Chan

Peanut Butter the Shiba Inu, forever etched in the GDQ record books with his run of Gyromite at Awesome Games Done Quick 2024, will return at this summer’s event. He’ll help his human companion, JSR_, play the SNES title Ken Griffey, Jr. Presents Major League Baseball on July 4 at 8:35 PM. Who needs fireworks when you can watch the grand spectacle of a gamer dog playing a sports title from the 16-bit era for ham and cheese?

The event will raise money for Doctors Without Borders. The nonprofit offers medical and humanitarian care to people in over 72 countries affected by crises like war, disease, natural disasters, and inadequate healthcare. Last year’s SGDQ brought in over $2.2 million. The GDQ series has raised over $45 million for various charities. 

You can check out the week’s schedule here and stream SGDQ 2024 on the Games Done Quick Twitch channel, which is also embedded below. The fun kicks off on June 30 at 12:30 PM ET.

This article originally appeared on Engadget at https://www.engadget.com/how-to-watch-summer-games-done-quick-2024-172007208.html?src=rss

The 2024 Moto Edge combines a 6.6-inch OLED screen with mid-range pricing

Motorola unveiled the 2024 Moto Edge on Tuesday. The $550 mid-range phone is slightly more expensive than the Pixel 8a but has different strengths and weaknesses. It will be available starting on June 20.

The 2024 Motorola Edge combines the mid-range Snapdragon 7s Gen 2 processor with a 6.6-inch pOLED display (much larger than the Pixel 8a’s 6.1-inch display) at 2400 x 1080 resolution (402ppi). It supports a 144Hz variable refresh rate and a 360Hz touch rate that’s only available when using the handset in Game Mode.

The 2024 Moto Edge has a big 5,000mAh battery and fast 68-watt wired “TurboPower” charging. It also has 256GB of built-in storage and 8GB of RAM.

Motorola

Its main camera is a 50MP shooter using Sony’s LY7-700C sensor. The phone’s backside includes a 13MP ultra-wide lens, while a 32MP sensor sits on the front.

As companies tend to do in 2024, Motorola pointed out the handset’s AI capabilities, in this case when using Google Photos (Magic Editor, Magic Eraser, Photo Unblur and Google Auto Enhance). Of course, that pales in comparison to the Pixel 8a’s full suite of Tensor G3-powered AI tools, also including Audio Magic Eraser, Best Take and unlimited uses of the Magic Editor.

The phone is 7.99 mm (0.31 inch) thick and weighs 174g. Its back is made of blue synthetic leather, which should provide a reasonably sophisticated look and feel without killing any cows. It has an IP68 water and dust resistance rating, meaning it can withstand up to half an hour of submersion in 1.5 meters of water.

Beginning on June 20, the 2024 Moto Edge will be available unlocked for $550 on Motorola’s website, Amazon and Best Buy. The company says “subsequent availability” will roll out later at T-Mobile, Metro by T-Mobile, Spectrum, Consumer Cellular, Straight Talk, Total By Verizon and Visible.

This article originally appeared on Engadget at https://www.engadget.com/the-2024-moto-edge-combines-a-66-inch-oled-screen-with-mid-range-pricing-130055585.html?src=rss

Samsung sues Oura to block Oura from suing Samsung over the Galaxy Ring

Samsung has filed a lawsuit against Oura to try to head off intellectual property disputes as the Galaxy Ring launch draws near. The suit notes that Oura has used its patent portfolio to sue smaller wearable tech competitors and has hinted it may do the same against the much larger Samsung. Welcome to the weird modern world of mega-corporations suing startups to prevent them from filing suits of their own.

“Oura’s actions and public statements demonstrate that Oura will continue asserting patent infringement against other entrants into the U.S. smart ring market, including Samsung,” the lawsuit, first reported on by The Verge, reads. “Oura’s immediate response to the announcement of the Galaxy Ring was to point to the purported strength of its intellectual property portfolio.”

The lawsuit claims the Galaxy Ring doesn’t infringe on Oura’s patents. However, in justifying its suit, it lays out a pattern of what it frames as aggressive IP protection by the Finnish startup. It lists cases where Oura sued smaller competitors like Ultrahuman, Circular and RingConn “as soon as, or even before, they entered the U.S. market.”

The document also cites Oura embarking on a media tour immediately following the Galaxy Ring announcement, touting the company’s “over 150 patents.” It specifically calls out patent-related quotes published by TechCrunch and a CNBC interview where Oura CEO Tom Hale hinted the company may use its IP portfolio against Samsung.

Daniel Cooper for Engadget

Samsung’s legal filing essentially tries to paint Oura as a patent troll, claiming many of the Finnish company’s patent disputes have been for features common to the entire category of smart rings, like electronics, sensors, a battery and scores that weigh health metrics. That approach conjures memories of Samsung’s old patent disputes with Apple. A common theme in those decade-old courtroom battles was Samsung accusing the iPhone maker of holding bogus patents that should never have been granted because they used obvious technologies or methods shared by the entire industry. (It worked with mixed results in those cases.)

Samsung filed its new lawsuit against Oura in the Northern District of California, San Francisco Division. Oura is based in Finland but has a US wing of its operations based in Delaware, including offices in San Francisco with more than 50 employees.

The lawsuit reveals extra detail about Samsung’s Galaxy Ring, which the company first showed off in a render in January before revealing physical models at the Mobile World Congress in February. The document says Samsung only finalized the Galaxy Ring’s design in “mid-May 2024” and plans to enter mass production in mid-June.

It adds that the Galaxy Ring will arrive in the US “in or around August of this year,” which aligns with expectations that the company will launch it at a summer Unpacked event.

This article originally appeared on Engadget at https://www.engadget.com/samsung-sues-oura-to-block-oura-from-suing-samsung-over-the-galaxy-ring-203353759.html?src=rss

Leaked Google database reveals its secret privacy and security failures

A collection of leaked internal Google privacy cases provides a rare glimpse into the company’s volume and handling of breaches, accidents and other incidents. 404 Media obtained and pored through the database, which covers thousands of internally flagged privacy and security issues from 2013 to 2018. 

Google verified the trove’s authenticity with Engadget but claimed some of the reports were related to third-party services or didn’t end up being cause for concern. “At Google employees can quickly flag potential product issues for review by the relevant teams,” a company spokesperson wrote to Engadget. “When an employee submits the flag they suggest the priority level to the reviewer. The reports obtained by 404 are from over six years ago and are examples of these flags — every one was reviewed and resolved at that time. In some cases, these employee flags turned out not to be issues at all or were issues that employees found in third party services.”

404 Media writes that, when taken on an individual level, many cases only impacted a few people or were fixed quickly. “Taken as a whole, though, the internal database shows how one of the most powerful and important companies in the world manages, and often mismanages, a staggering amount of personal, sensitive data on people’s lives,” 404 Media’s Joseph Cox wrote.

Examples include a potential security issue where a government client of a Google cloud service had its sensitive data accidentally transitioned to a consumer-level product. Google’s internal report added that, as a consequence, a US-based location for the data was “no longer guaranteed for this customer,” according to the report.

Google

In 2016, another case flagged a glitch in Google Street View, where a filter in the service’s transcription software designed to omit captured license plate numbers failed to do its job. “As a result, our database of objects detected from Street View now inadvertently contains a database of geolocated license plate numbers and license plate number fragments,” the report acquired by 404 Media details. (Oops!) That report said the data was purged.

Another incident highlighted a case where a bug in a Google speech service accidentally captured and logged an estimated 1,000 hours of children’s speech data for about an hour. That case report claimed the team deleted all of the data.

Other cases in the database range from “a person” modifying customer accounts on Google’s ad platform to manipulate affiliate tracking codes to YouTube recommending videos based on users’ deleted watch histories. One report even highlights how a Google employee (unintentionally, according to the report) accessed Nintendo’s private YouTube videos and leaked info ahead of the video game company’s announcements.

The full report from 404 Media, which details more of the internal reports, is worth reading for anyone curious about the types of privacy and security incidents a company of Google’s magnitude faces — or causes itself — and how it addresses them.

This article originally appeared on Engadget at https://www.engadget.com/leaked-google-database-reveals-its-secret-privacy-and-security-failures-183232983.html?src=rss

Sony's Astro Bot is getting the Mario-like adventure it deserves

Sony’s lovable robo-mascot is getting a new game. During its PlayStation State of Play on Thursday, Sony dropped a trailer for the upcoming PS5 title Astro Bot. The 3D Mario-esque title, a sequel to the beloved pack-in, Astro’s Playroom, arrives on September 6.

The trailer sees our pal Astro wading through puddles full of pink flamingos, grappling across platforms like Bionic Commando and summoning a PS5 console to help fight a giant green alien in a UFO (as one tends to do). In a nod to old-school PlayStation, the trailer even includes a cameo from PaRappa the Rapper.

Sony describes Astro Bot as a “supersized adventure” that spans over six galaxies and 80 levels as you search for your lost crew. You’ll have over 15 new abilities, and, like in its predecessor, it will serve as a showcase for the various sensors and haptics in the DualSense controller.

Astro Bot will be available for pre-order on June 7 ahead of its September 6 release date.

This article originally appeared on Engadget at https://www.engadget.com/sonys-astro-bot-is-getting-the-mario-like-adventure-it-deserves-225716801.html?src=rss

Microsoft and CWA forge labor neutrality agreement covering all ZeniMax workers

Microsoft and the Communications Workers of America (CWA) have forged a labor neutrality agreement for all ZeniMax workers. The CWA says the deal is similar to the one from 2022 covering Activision Blizzard’s workforce, which went into effect in April of this year.

“Thousands of our ZeniMax co-workers now have a free and fair path to organize together for better working conditions,” Page Branson, a Senior Quality Assurance (QA) Tester, wrote in a CWA press release.

In 2022, the Xbox maker signed a formal agreement with CWA to respect Activision Blizzard employees’ right to organize. Set to take effect 60 days after Microsoft’s acquisition closed, it didn’t become official until earlier this year. It also extends Microsoft’s early 2023 recognition of ZeniMax QA workers’ right to unionize, and it now includes every eligible employee at ZeniMax.

The CWA says the new agreement means Microsoft will take a neutral approach when employees express interest in joining the union. The workers can communicate with their peers and union reps about membership “in a way that encourages information sharing and avoids business disruptions.”

According to the CWA, workers will have access to a decision-making process that respects their privacy if they wish. In addition, it includes a dispute resolution and arbitration process in the case of any disagreements between the union and Microsoft.

“When we organized our union under a similar legal agreement the process was clear and management did not try to influence anyone’s decision and the company did not try to interfere with the voting process,” Branson wrote. “There is strength in numbers, and as our numbers grow at ZeniMax, at Microsoft, and in the video game industry, we will gain the respect we deserve and raise the standards of working conditions for everyone across the video gaming industry. When we benefit, the consumer and the company will ultimately benefit with us and help keep this industry stable for current and future workers.”

This article originally appeared on Engadget at https://www.engadget.com/microsoft-and-cwa-forge-labor-neutrality-agreement-covering-all-zenimax-workers-212643324.html?src=rss

Spotify now says it will refund Car Thing purchases

Spotify says it has quietly begun offering refunds for its soon-to-be-bricked Car Thing. The company told Engadget on Thursday that, as of last Friday, customers with proof of purchase (like an emailed invoice) can contact customer service and get their money back for the vehicle streaming device.

Spotify has taken some heat for its announcement last week that it will brick every Car Thing device on December 9, 2024. The company described its decision as “part of our ongoing efforts to streamline our product offerings” (read: cut costs) and that it lets Spotify “focus on developing new features and enhancements that will ultimately provide a better experience to all Spotify users.”

TechCrunch reports that Gen Z users on TikTok have expressed their frustration in videos, while others have complained directed toward Spotify in DMs on X (Twitter) and directly through customer support. Some users claimed Spotify’s customer service agents only offered several months of free Premium access, while others were told nobody was receiving refunds. It isn’t clear if any of them contacted them after last Friday when it shifted gears on refunds.

Others went much further. Billboard first reported on a class-action lawsuit filed in the US District Court for the Southern District of New York on May 28. The suit accuses Spotify of misleading Car Thing customers by selling a $90 product that would soon be obsolete without offering refunds, which sounds like a fair enough point.

It’s worth noting that, according to Spotify, it began offering the refunds last week, while the lawsuit was only filed on Tuesday. If the company’s statement about refunds starting on May 24 is accurate, the refunds aren’t a direct response to the legal action. (Although it’s possible the company began offering them in anticipation of lawsuits.)

Regardless, anyone reaching out to customer service with a valid receipt should be offered a refund. If you have tried since then and have been denied, we recommend you contact them again.

This article originally appeared on Engadget at https://www.engadget.com/spotify-now-says-it-will-refund-car-thing-purchases-193001487.html?src=rss