ArduinoMania

AI normally needs to be trained on existing material to detect objects, but Meta has a way for the technology to spot items without help. The social media giant has published a "Segment Anything" AI model that can detect objects in pictures and videos even if they weren't part of the training set. You can select items by clicking them or using free-form text prompts. As Reutersexplains, you can type the word "cat" and watch the AI highlight all the felines in a given photo.

The model can also work in tandem with other models. It can help reconstruct an object in 3D using a single image, or draw from views from a mixed reality headset. Effectively, Segment Anything can limit the need for additional AI training.

Both the AI model and a dataset will be downloadable with a non-commercial license. That is, creators can't use it for products. This is primarily for research and expanding access to the technology. Right now, Meta uses somewhat similar tech to moderate banned content, recommend posts and tag photos.

The developers acknowledge that the existing model is flawed. It might miss finer details, and isn't as accurate at detecting the boundaries as some models. And while Segment Anything can handle prompts in real-time, it bogs down when demanding image processing is involved. Some more specialized AI tools are likely to outperform this model in their respective fields, Meta says.

You aren't about to see this AI in robots or other devices where fast, accurate object detection is (usually) vital. However, models like this may still help in situations where it's impractical to rely exclusively on training data. A social network could use the tech to keep up with a rapidly growing volume of content. If nothing else, this shows that Meta wants to generalize computer vision.

Meta is no stranger to sharing AI breakthroughs, such a translator for unwritten languages. With that said, there's pressure on the company to show that it's as much of a powerhouse in the category as tech heavyweights like Google and Microsoft. It's already planning generative AI "personas" for its social apps, and inventions like Segment Anything show that it has a few advantages of its own.

If you’ve been waiting for Apple’s 2022 iPad to go on sale, you can get one now for $50 off its usual price. Amazon currently offers the tablet starting at $399, its lowest price yet.

The $50 discount applies to all colors, storage and networking variants of the 10.9-inch tablet. Color options include silver, pink, blue and yellow; you can order it in 64GB and 256GB tiers. In addition to the deal on the base model, you’ll pay $549 for either the WiFi / 256GB variant or the WiFi + cellular / 64GB option. The 256GB / WiFi + cellular model costs $699.

The 10th-generation iPad, launched in October, has an updated design reminiscent of the more expensive iPad Air and iPad Pro — losing the Home button and shifting Touch ID to the edge-facing power / sleep button. Its 10.9-inch display is also significantly larger than its predecessor, and it includes a faster A14 Bionic chip that’s plenty powerful for most iPad-related tasks. Additionally, Apple repositioned its front-facing camera to be centered in landscape mode, making your gaze appear more natural on video calls. It also supports the first-generation Apple Pencil (including its awkward charging setup) for jotting notes or sketching.

In Engadget’s review, the iPad’s battery life exceeded Apple’s 10-hour estimate, lasting 11 hours and 45 minutes while playing a movie from the iTunes Store. In addition, although it doesn’t support the Magic Keyboard that works with Apple’s premium tablets, you can add the Magic Keyboard Folio accessory for $229, a $20 discount. Despite a few tradeoffs, the iPad’s sale price addresses many of our reservations about it, which centered around its overall value.

Follow @EngadgetDeals on Twitter and subscribe to the Engadget Deals newsletter for the latest tech deals and buying advice.

Utah recently passed two laws that would drastically change how teens in the state are able to use social media. The new laws will impose strict rules for how companies handle teenagers’ accounts, including provisions requiring parental consent, and mandates for in-app parental controls and curfew features.

But among the most controversial aspects of the law is age verification. It requires companies like Snap, Meta and TikTok to confirm the ages of their youngest users in order to enforce the other age-based restrictions. Under the rules, which are set to take effect next March, large platforms will no longer be able to simply allow teens to enter their own birthday at sign-up. Instead, they would need to go through some other process, like providing a copy of an I.D, before they could access their accounts.

While Utah is the first state to enact such a law, it’s unlikely to be the last. Arkansas, Ohio, Connecticut and Minnesota are all considering social media laws with either explicit age verification requirements or other age-based restrictions. At the federal level, Senator Josh Hawley has proposed a bill that would prohibit teenagers under 16 from using social media entirely, and require social media companies to independently verify the ages of their users. Even the US Surgeon General has suggested that 13 may be “too young” for teens to use social media.

The proposed laws are part of broader reckoning around how social media is impacting its youngest users. For years, lawmakers, armed with teenage finstas and incriminating research, have made youth safety a central part of their effort to regulate Big Tech. Along the way, they’ve also proposed laws that would rein in algorithms, make it more difficult to post and limit apps’ more “addictive” features.

But the latest crop of laws have instead zeroed in on parental consent and age-based restrictions, rather than addressing structural issues like data privacy. Irene Ly, policy counsel for Common Sense Media, a nonprofit that advocates for child safety online, says the shift is happening in part because lawmakers have been unable to pass comprehensive privacy bills.

“Privacy legislation seems to have a lot more sticking points,” she tells Engadget. “It's hard to find a compromise on all the facets of regulating tech.” But lawmakers have been able to find more broad support — at least at the state level — for age-based restrictions and parental consent requirements, particularly in states that have passed other laws emphasizing “giving rights to the parents.”

But experts warn that focusing on age-based restrictions won’t address the core safety issues lawmakers say they want to solve. And age verification measures, like those in Utah, pose a significant threat to the privacy of all social media users, not just teens.

Privacy advocates, like the Electronic Frontier Foundation (EFF), say that there’s no way to enforce age verification requirements without requiring that all users submit to the checks. “It's not just the privacy of young people that's at stake, it's everyone,” Jason Kelley, associate director of digital strategy for the EFF, tells Engadget, noting that a previous attempt to require age verification was struck down by the Supreme Court more than a decade ago. ”Confirming that everyone is the age they say they are is not possible without confirming every single person's age.”

For example, Utah’s law states “the social media company shall deny access to the account” for any “Utah account holder fails to meet the verification requirements.” That means even adult social media users could face being locked out of their accounts if they fail to provide a copy of an ID or submit to another kind of age check.

Figuring out how to apply these laws only in specific states would also be problematic, according to industry groups. “Although the proposed legislation purports to apply only to Utah residents, platforms cannot know which users are Utah residents without first verifying their identity,” Ari Cohn, free speech counsel for TechFreedom, a think tank that’s received funding from Meta and Google, said in a statement. “This legislation would be a nationwide mandate that Utah is not permitted to impose.”

Even figuring out how to verify users’ ages could prove tricky. Many minors don’t have a driver’s license or government-issued ID. Instagram has tested an AI face-scanning tool that claims to be able to accurately estimate users’ ages based on their facial features (experts have raised doubts about the accuracy and ethical implications of using these tools at scale). But Kelley, of the EFF, says that any form of age verification exposes users to additional data privacy risks.

“It's so easy to find examples of these companies taking advantage of data that they explained was going to be collected for one purpose and using it for another,” Kelley says. For example, it wasn't that long ago that Meta and Twitter both admitted to using phone numbers originally collected for two-factor authentication for targeted advertising. Kelley says there could be an even greater risk of something similar happening with any age verification system. “We have no way of knowing whether they're doing that with identity verification information. Whether that's a selfie, a shared driver's license, a call through an API to a credit company — we just don't know.”

Common Sense Media has raised similar concerns. Ly says that laws addressing more fundamental aspects of social media platforms would be more effective than attempting to lock out teens of a certain age.

“If you can implement some key changes to these companies, like limiting how much data they're collecting and what they're using it for, and then making changes to how their platform is designed, that will create a healthier experience,” she said. “It wouldn't necessitate prohibiting teens from being on the platform altogether.”

Sony’s massively successfulGod of War Ragnarök just got a pretty significant upgrade with a New Game+ that is much more than a simple rehash with tougher enemies. This NG+ is packed with features to squeeze more life out of last year’s PS4 and PS5 title. For the uninitiated, a New Game+ is a new adventure that unlocks when you beat the game. In this case, all of your equipment, weapons and skills carry over from your previous save. 

The changes start with several new armor sets that bring unique gameplay mechanics to the table. The Zeus armor increases your magic and spell damage but places you at greater risk for incoming damage. The Spartan Armor, as the name suggests, eliminates all perks and stat boosts, increasing the challenge and leaving no room for error.

Both the Armor of the Black Bear and Ares Armor take the opposite approach, giving players unique perks, like increased defense and an increase in rare item drops. There are also plenty of new customization options for pre-existing armor types if you want to change up your look.

The level cap is increased for both enemies and the player, which is fairly common in an NG+. Any level nine equipment, including weapons and shields, can convert into “Plus” versions with more progression options. Converting equipment also gives you special coins that are used to purchase new Enchantments.

Speaking of new Enchantments, the Engravings and Badge powers let you apply perks and stats typically reserved to a specific armor set to an amulet, so you get all of the benefits without having to actually wear the armor. The system is designed for experimentation to find the best combo for your unique playstyle. Certain enemies like Berserker Souls will also now drop new stat boost enchantments. Finally, there are Burdens, enchantments that apply a negative perk to amp up the challenge.

Endgame bosses have been redesigned, the sparring arena has been expanded with new enemies and there’s even a new black-and-white mode for those looking for a more cinematic experience. The free NG+-focused update is available now, making this a good time to return to ancient Scandinavia and cause some brutal mayhem.

A boatload of HBO series and other Warner Bros. Discovery (WBD) shows and movies are now available to watch for free on Roku. The Roku Channel now includes 14 ad-supported channels that are streaming more than 2,000 hours of WBD content, including shows like Westworld, The Nevers and Raised by Wolves, all of which vanished from HBO Max a few months ago.

In January, WBD struck deals with Roku and Tubi for free, ad-supported streaming (aka FAST) channels. Three of the cable-style channels debuted on Tubi in early February, but it took a little longer for them all to land on Roku.

The channels are each focused on different areas. There's one for fan-favorite shows like the canceled Westworld and Nikita (WB TV Watchlist), another centered around docuseries such as How It’s Made and How the Earth Works (WB TV How To), a channel for classic movies (WB TV At The Movies) and even one for baking competition series (WB TV Sweet Escapes). Licensing out content for streaming on FAST channels will give WBD another revenue stream as it tries to improve its bottom line — WBD posted a net loss of $2.1 billion for the last three months of 2022.

The channels are arriving on Roku just before WBD folds Discovery+ content into HBO Max as part of a combined streaming service sometime this spring. Discovery+ will remain as a standalone streaming service in the US. It's believed that the souped-up HBO Max won't immediately become more expensive. WBD raised the price of the ad-free HBO Max plan for the first time in January.

The company has yet to reveal the name of the new streaming service, though reports indicate that it will simply be called "Max" (shudder). We won't have to wait much longer to find out, though. WBD plans to spill the beans at an event on April 12th.

Stellantis says the upcoming Ram 1500 REV will offer up to a 500-mile range, longer than any electric pickup available. The automaker officially announced the electric truck at the New York International Auto Show today after teasing an ultra-sleek concept in January (which the version you can buy looks nothing like) and a production reveal in February.

The standard version of the Ram 1500 EV will include a 168 kWh battery that lasts 350 miles; you’ll need to opt for the premium 229 kWh battery option to reach the 500-mile estimate. By comparison, the Ford F-150 Lightning maxes out at a 320-mile range, while the Chevy Silverado EV can last an estimated 400 miles. Meanwhile, Tesla claims its long-delayed Cybertruck will match the new Ram with a 500-mile range.

The Ram 1500 EV also supports quick charging. Stellantis says it can add up to 110 miles of range with only 10 minutes of charging.

As far as performance, the truck can go from 0 to 60 mph in as little as 4.4 seconds and has 654 horsepower and 630 lb-ft of torque. Its other specs include a towing capacity of up to 14,000 lbs and a maximum payload of up to 2,700 lbs. In addition, it has a frunk with 15 cubic feet of storage and bidirectional charging (including powering another vehicle, charging a home during power outages and sending power back to the grid). Finally, it can handle up to two ft of water fording.

Stellantis hasn’t yet announced pricing, and deliveries aren't scheduled to start until the fourth quarter of 2024. After opening (and quickly closing) reservations in January, you can reserve one again with a (refundable) $100 down payment.

Online alcohol recovery startups Monument and Tempest got caught sharing confidential user data with advertisers without their consent, as originally reported by TechCrunch. Everything came to light after an internal review revealed a data breach impacting 100,000 users, forcing the companies to issue a formal disclosure to the user base. The violations started in 2017 and were ongoing until last month's review.

Monument and Tempest started as two entirely different platforms, but the former acquired the latter several months back. Parent company Monument confirmed not only the data breach but that the companies shared private information with advertisers via a notification filed with California’s attorney general. Data shared with advertisers, without user consent, includes patient names, dates of birth, email addresses, postal addresses, phone numbers, insurance information and more.

Unfortunately, that is just the beginning. In a cruel insult to those seeking recovery, the companies also shared data related to appointment information, assessment information and survey responses, which includes alcohol consumption data. Monument continues to tout its commitment to privacy on its website, saying that survey responses are “protected" despite the recent disclosure

The companies blame third-party tracking systems for the issue, stating that they have removed the offending tracking codes from their websites. The companies do not admit to sharing this information on purpose to increase profits, indicating that the tracking pixels provided by third parties did the deed all on their own.

Though this is an especially egregious example, it is important to remember that most companies have a less-than-pristine record regarding data privacy, even in the case of medical records. There is a near-endless list of similar violations, like the time a mental health startup shared patient information without consent and when Meta was caught with its own hand in the digital cookie jar. Be careful out there folks.

Google wants to make it as easy to scrub an app account as it is to create one. The company has announced that Android apps on the Play Store will soon have to let you delete an account and its data both inside the app and on the web. Developers will also have to wipe data for an account when users ask to delete the account entirely.

The move is meant to "better educate" users on the control they have over their data, and to foster trust in both apps and the Play Store at large. It also provides more flexibility. You can delete certain data (such as your uploaded content) without having to completely erase your account, Google says. The web requirement also ensures that you won't have to reinstall an app just to purge your info.

The policy is taking effect in stages. Creators have until December 7th to answer questions about data deletion in their app's safety form. Store listings will start showing the changes in early 2024. Developers can file for an extension until May 31st of next year.

The changes come several months after Apple instituted a similar rule for App Store software. In both cases, the companies are concerned about privacy violations and the ensuing fallout — they don't want users to fall prey to data breaches because they couldn't easily delete accounts or sensitive info when they stop using an app.

This also follows growing efforts by regulators to demand more control over services. The Federal Trade Commission (FTC) recently proposed rule changes requiring easy ways to cancel subscriptions and memberships. While the FTC is focused more on unwanted charges than privacy, the message to app makers is clear: provide more control of accounts or face repercussions.

Twitter has added a label to the main account of NPR to designate the public broadcaster as "US state-affiliated media." Until now, such labels have typically been reserved for state-run organizations such as RT and Sputnik in Russia and China's Xinhua News Agency. The labels appear on every tweet from accounts they're applied to.

“We were disturbed to see last night that Twitter has labeled NPR as ‘state-affiliated media,’ a description that, per Twitter’s own guidelines, does not apply to NPR," the broadcaster's CEO John Lansing said in a statement. "NPR and our member stations are supported by millions of listeners who depend on us for the independent, fact-based journalism we provide. NPR stands for freedom of speech and holding the powerful accountable. It is unacceptable for Twitter to label us this way. A vigorous, vibrant free press is essential to the health of our democracy.”

As Mediaite points out, Twitter has edited its guidelines regarding the "state-affiliated media" label since applying it to NPR's account. "State-financed media organizations with editorial independence, like the BBC in the UK or NPR in the US for example, are not defined as state-affiliated media," the guidelines page read on Tuesday, according to a Wayback Machine snapshot. By Wednesday morning, Twitter had removed the text "or NPR in the US." Twitter no longer has a communications department that can be reached for comment. 

Elon Musk, who took over Twitter in October, noted the change to NPR's account. In response to a user who gave him acknowledgment for the move, Musk tweeted a portion of the state-affiliated media policy that reads "State-affiliated media is defined as outlets where the state exercises control over editorial content through financial resources, direct or indirect political pressures, and/or control over production and distribution." Musk wrote that the definition "seems accurate" with regards to NPR.

On Tuesday, Musk responded to a tweet that criticized NPR over a report suggesting that “European right-wing politicians [are] lobbing a conspiracy theory that elites want people to eat bugs.” He replied with an exclamation point. The user who posted the thread went on to claim that "NPR is worse than the propaganda of Maoist schoolchildren during the cultural revolution.”

NPR, which is an independent non-profit, says that on average, less than one percent of its annual operating budget comes from government grants. Over the last five fiscal years, around 70 percent of its revenues have come from corporate sponsors and core and programming fees paid by member organizations. Meanwhile, Voice of America, a broadcaster that is owned by the US government, does not have a state-affiliated media label on its Twitter account.

The perpetrators of the ransomware attack against Oakland have leaked more of the data from the hack. The city has confirmed that Play, the hacker group claiming responsibility for the ransomware, has shared a second batch of info on the dark web. While officials aren't sharing more details, The Oaklandsidesources say the dump was 600GB and included confidential Oakland Police Department files (including disciplinary records), council members' communications and city staff's medical records.

The first release from early March was a comparatively modest 10GB, but included city employee rosters and police records. The incident was serious enough that the Oakland police union is now demanding $25,000 per officer for the damage done through the leak. The union also wants to pressure the city into tightening its security.

The February attack prompted the city to declare a state of emergency and take its network offline in a bid to limit the damage. That left many non-emergency services unavailable, including the systems to issue licenses and permits. Oakland closed some buildings and warned of delayed responses to non-urgent complaints. Access to some systems came back in late February.

Play group has been linked to numerous attacks, including Rackspace and the Belgian city of Antwerp. The gang first emerged in June of last year, when BleepingComputer forum users reported attacks attributed to the outfit.

The city still hasn't outlined Play's demands. However, the second wave of leaks suggests Oakland isn't bowing to pressure to pay the ransom. That's not a surprising decision. While the exposed data increases the risk of fraud, the city risks encouraging more ransomware attacks if it gives in.