verge-rss

Image: Netflix

You would think that a sci-fi movie where Jennifer Lopez partners with a smart-ass, sentient mech suit to fight against her evil AI brother would be a little more fun. Alas, Atlas — Netflix’s latest attempt at a hit streaming action movie — takes itself far too seriously. It also fails to really dig into the complexities of the AI debate, despite essentially being a conflict between a friendly AI assistant and machine intent on a doomsday scenario. There are some funny moments, particularly the banter between Lopez and her mechanical companion, but every other part of the movie seems to be fighting against Atlas’ true form. This is a buddy comedy trying too hard to be a serious action flick.
Atlas takes place almost three decades after an uprising that saw an advanced AI bot named Harlan (Simu Liu) help liberate other machines, who then proceeded to bypass their security protocols and start a war with humanity. It’s a setup that echoes plenty of real-world concerns. Except, in this case, the AIs lose, and Harlan heads off-planet to lick his wounds — but not before issuing an ominous threat to the human population. Atlas (Lopez), the daughter of Harlan’s creator who essentially grew up with him as a sibling, spends the ensuing 28 years trying to locate precisely where Harlan went so the threat can be eliminated for good. The movie kicks off when she discovers that location after interrogating the severed head of an AI henchman.

The most important thing you need to know about Atlas is that she has grown to absolutely loathe AI and, by extension, most futuristic tech. She has the same fears many of us do (along with sci-fi characters like Will Smith in I, Robot), which are exacerbated by the fact that the tech around her can be hacked and exploited by Harlan and his associates. At one point, while briefing a group of soldiers, she says, “You can’t trust AI,” while handing out plans printed on paper.
This fear extends particularly to a device called a Neural Link (not to be confused with the Elon Musk-backed Neuralink), which lets a human mind connect directly to an AI companion. It’s a cool idea, but the movie never slows down enough to explore it in depth. Inevitably, Atlas finds herself with no choice but to use a Neural Link to connect to an AI named Smith (Gregory James Cohan) who looks just like Siri and is housed inside of a mech suit ripped right out of Titanfall.

Image: Netflix
This is what Alexa could look like one day.

Contrived as it might be, the relationship between Smith and Atlas is easily the best part of the movie. Atlas is cranky and sarcastic, and because of his adaptive learning capabilities, Smith soon becomes exactly the same. The AI swears and makes jokes, dishing it out to Atlas the same way she does to him. The banter is genuinely funny, to the point that, even though you can see it coming a mile away, their inevitable friendship still feels touching. It’s almost worth watching the whole movie just for its heartwarming finale.
The problem with Atlas isn’t so much that it’s predictable (though that doesn’t help, nor does its painfully generic vision of a sci-fi future). It’s that the movie doesn’t lean into this strength. Outside of Smith and Atlas, everything else about Atlas is self-serious and dull. Harlan is the biggest offender, played with a stilted effect by Liu that makes him more boring than scary. In a future where AI bots can mimic human beings perfectly, it’s confounding that the most advanced machine sounds like an old GPS giving directions. Overall, there’s a lot of wasted potential. In particular, the movie’s premise is a perfect framing for current AI debates — Siri vs. Skynet — but doesn’t take the opportunity to say anything new.
There are already plenty of recent movies that explore a potential AI future with a heavy dose of sincerity, whether it’s The Creator, Dead Reckoning, or even Netflix’s own Jung_E. Atlas adds nothing to that extensive body of work. Even worse, it fails to capitalize on its one defining aspect. The comedic moments are the best part of the movie, and yet they can feel out of place buried under everything else. Atlas was a chance to take an urgent AI conversation and explore it in an approachable Hollywood package. It could’ve been fun and smart — instead, like a lot of AI right now, it’s neither.
Atlas starts streaming on Netflix on May 24th.

Image: Netflix

You would think that a sci-fi movie where Jennifer Lopez partners with a smart-ass, sentient mech suit to fight against her evil AI brother would be a little more fun. Alas, Atlas — Netflix’s latest attempt at a hit streaming action movie — takes itself far too seriously. It also fails to really dig into the complexities of the AI debate, despite essentially being a conflict between a friendly AI assistant and machine intent on a doomsday scenario. There are some funny moments, particularly the banter between Lopez and her mechanical companion, but every other part of the movie seems to be fighting against Atlas’ true form. This is a buddy comedy trying too hard to be a serious action flick.

Atlas takes place almost three decades after an uprising that saw an advanced AI bot named Harlan (Simu Liu) help liberate other machines, who then proceeded to bypass their security protocols and start a war with humanity. It’s a setup that echoes plenty of real-world concerns. Except, in this case, the AIs lose, and Harlan heads off-planet to lick his wounds — but not before issuing an ominous threat to the human population. Atlas (Lopez), the daughter of Harlan’s creator who essentially grew up with him as a sibling, spends the ensuing 28 years trying to locate precisely where Harlan went so the threat can be eliminated for good. The movie kicks off when she discovers that location after interrogating the severed head of an AI henchman.

The most important thing you need to know about Atlas is that she has grown to absolutely loathe AI and, by extension, most futuristic tech. She has the same fears many of us do (along with sci-fi characters like Will Smith in I, Robot), which are exacerbated by the fact that the tech around her can be hacked and exploited by Harlan and his associates. At one point, while briefing a group of soldiers, she says, “You can’t trust AI,” while handing out plans printed on paper.

This fear extends particularly to a device called a Neural Link (not to be confused with the Elon Musk-backed Neuralink), which lets a human mind connect directly to an AI companion. It’s a cool idea, but the movie never slows down enough to explore it in depth. Inevitably, Atlas finds herself with no choice but to use a Neural Link to connect to an AI named Smith (Gregory James Cohan) who looks just like Siri and is housed inside of a mech suit ripped right out of Titanfall.

Contrived as it might be, the relationship between Smith and Atlas is easily the best part of the movie. Atlas is cranky and sarcastic, and because of his adaptive learning capabilities, Smith soon becomes exactly the same. The AI swears and makes jokes, dishing it out to Atlas the same way she does to him. The banter is genuinely funny, to the point that, even though you can see it coming a mile away, their inevitable friendship still feels touching. It’s almost worth watching the whole movie just for its heartwarming finale.

The problem with Atlas isn’t so much that it’s predictable (though that doesn’t help, nor does its painfully generic vision of a sci-fi future). It’s that the movie doesn’t lean into this strength. Outside of Smith and Atlas, everything else about Atlas is self-serious and dull. Harlan is the biggest offender, played with a stilted effect by Liu that makes him more boring than scary. In a future where AI bots can mimic human beings perfectly, it’s confounding that the most advanced machine sounds like an old GPS giving directions. Overall, there’s a lot of wasted potential. In particular, the movie’s premise is a perfect framing for current AI debates — Siri vs. Skynet — but doesn’t take the opportunity to say anything new.

There are already plenty of recent movies that explore a potential AI future with a heavy dose of sincerity, whether it’s The Creator, Dead Reckoning, or even Netflix’s own Jung_E. Atlas adds nothing to that extensive body of work. Even worse, it fails to capitalize on its one defining aspect. The comedic moments are the best part of the movie, and yet they can feel out of place buried under everything else. Atlas was a chance to take an urgent AI conversation and explore it in an approachable Hollywood package. It could’ve been fun and smart — instead, like a lot of AI right now, it’s neither.

Atlas starts streaming on Netflix on May 24th.

Read More 

Illustration by Alex Castro / The Verge

Google is now letting users securely share passwords with members of their own family group over the internet, making it easier for everyone in the household to access passwords for shared services like Netflix. The new ability is included in the Google Play services update for May 2024 that’s rolling out now, as reported by Android Authority.
The new password sharing feature just applies to ones that are stored in Google Password Manager, the company’s service that natively stores your passwords and passkeys in Chrome and Android and is linked to your Google account. As of today, the new password sharing feature works on mobile — but apparently not via Chrome on desktop, yet. Once you share a password with one of your family members, a copy of it will be saved into that member’s own Google Password Manager.
You can’t use this feature with people who aren’t in your Google-sanctioned family group (which can only have up to six people). So if you want to share a password with anyone else, you’ll need to use Nearby Share to zap it over in person or use more rudimentary and / or less secure methods.

Illustration by Alex Castro / The Verge

Google is now letting users securely share passwords with members of their own family group over the internet, making it easier for everyone in the household to access passwords for shared services like Netflix. The new ability is included in the Google Play services update for May 2024 that’s rolling out now, as reported by Android Authority.

The new password sharing feature just applies to ones that are stored in Google Password Manager, the company’s service that natively stores your passwords and passkeys in Chrome and Android and is linked to your Google account. As of today, the new password sharing feature works on mobile — but apparently not via Chrome on desktop, yet. Once you share a password with one of your family members, a copy of it will be saved into that member’s own Google Password Manager.

You can’t use this feature with people who aren’t in your Google-sanctioned family group (which can only have up to six people). So if you want to share a password with anyone else, you’ll need to use Nearby Share to zap it over in person or use more rudimentary and / or less secure methods.

Read More 

Anker’s latest wireless chargers work great whether you’re on the move or getting ready for bed. | Image: Anker

I’m in the process of overhauling my entire arsenal of chargers now that the world is finding common ground when it comes to charging tech — whether voluntarily or not. The iPhone 15 finally features a USB-C port, and I anticipate upgrading from my iPhone 12 Pro Max this year to exile Lightning cables from my life for good. Thankfully, though, my last-gen phone is keeping up with the evolving wireless charging landscape just fine thanks to its newly gained support for Qi2.

I’m glad I’ve been forced to wait on adopting the new magnetic charging standard, as that means I can now take advantage of discounts like the one we’re currently seeing on Anker’s Qi2-ready MagGo power banks. The Anker MagGo Power Bank (10K), the 10,000mAh version with a built-in kickstand and an LED display, is available at Amazon right now in select colors starting at just $69.99 ($20 off), which is the first discount we’ve seen since the power bank arrived late last year. It’s also on sale from Anker for the same price through May 27th with offer code WSPEV22ONZXB.

The kickstand model’s 10,000mAh capacity is ideal for bigger devices, as it allows you to squeeze out close to two full charges on an iPhone 15 Pro. The integrated display, meanwhile, can tell you the exact capacity remaining and the total time to charge. You can even use the same USB-C port you use for recharging the power bank to charge various devices at up to 27W speeds, which is useful if you’re pressed for time.
If you can get away with a smaller power bank, you can also save 20 percent on the 6,600mAh model, the Anker MagGo Power Bank (6.6K), which uses the battery bay as a weighty base. The latter is available for $55.99 ($14 off) when you clip the on-page coupon at Amazon, which is a mere $7 more than its all-time low. It’s also on sale at Anker when you apply offer code WSCPV2ZVMS7Z at checkout.

The 6,600mAh alternative doesn’t have a built-in display or reverse USB-C charging, but the stand offers more flexibility since you can adjust the angle between 35 and 65 degrees. Both devices also support passthrough charging, allowing you to charge both the power bank and your phone simultaneously, and the new Standby feature in iOS 17, which can showcase the time, photos, or your calendar when your phone is mounted in landscape orientation.
Since they both use the new Qi2 standard, you can use either to charge any MagSafe-compatible iPhone at speedy 15W speeds. Anker says it can bring an iPhone 15 from empty to 50 percent in just 44 minutes. Without Qi2 or true MagSafe certification, that speed would effectively be halved, as most Qi chargers only offer 7.5W charging speeds (with iPhones).

Anker’s latest wireless chargers work great whether you’re on the move or getting ready for bed. | Image: Anker

I’m in the process of overhauling my entire arsenal of chargers now that the world is finding common ground when it comes to charging tech — whether voluntarily or not. The iPhone 15 finally features a USB-C port, and I anticipate upgrading from my iPhone 12 Pro Max this year to exile Lightning cables from my life for good. Thankfully, though, my last-gen phone is keeping up with the evolving wireless charging landscape just fine thanks to its newly gained support for Qi2.

I’m glad I’ve been forced to wait on adopting the new magnetic charging standard, as that means I can now take advantage of discounts like the one we’re currently seeing on Anker’s Qi2-ready MagGo power banks. The Anker MagGo Power Bank (10K), the 10,000mAh version with a built-in kickstand and an LED display, is available at Amazon right now in select colors starting at just $69.99 ($20 off), which is the first discount we’ve seen since the power bank arrived late last year. It’s also on sale from Anker for the same price through May 27th with offer code WSPEV22ONZXB.

The kickstand model’s 10,000mAh capacity is ideal for bigger devices, as it allows you to squeeze out close to two full charges on an iPhone 15 Pro. The integrated display, meanwhile, can tell you the exact capacity remaining and the total time to charge. You can even use the same USB-C port you use for recharging the power bank to charge various devices at up to 27W speeds, which is useful if you’re pressed for time.

If you can get away with a smaller power bank, you can also save 20 percent on the 6,600mAh model, the Anker MagGo Power Bank (6.6K), which uses the battery bay as a weighty base. The latter is available for $55.99 ($14 off) when you clip the on-page coupon at Amazon, which is a mere $7 more than its all-time low. It’s also on sale at Anker when you apply offer code WSCPV2ZVMS7Z at checkout.

The 6,600mAh alternative doesn’t have a built-in display or reverse USB-C charging, but the stand offers more flexibility since you can adjust the angle between 35 and 65 degrees. Both devices also support passthrough charging, allowing you to charge both the power bank and your phone simultaneously, and the new Standby feature in iOS 17, which can showcase the time, photos, or your calendar when your phone is mounted in landscape orientation.

Since they both use the new Qi2 standard, you can use either to charge any MagSafe-compatible iPhone at speedy 15W speeds. Anker says it can bring an iPhone 15 from empty to 50 percent in just 44 minutes. Without Qi2 or true MagSafe certification, that speed would effectively be halved, as most Qi chargers only offer 7.5W charging speeds (with iPhones).

Read More 

Illustration: Hugo Herrera / The Verge

The IRS Free File program will be around for another five years. The agency has announced that it’s extending the program through October 2029, allowing people to file their taxes with one of the IRS’s partners for free.
The Free File program got more attention recently after the IRS changed its rules in 2020 to prevent tax preppers from hiding their free services. Both TurboTax and H&R Block left the Free File program after that, and TurboTax owner Intuit paid out $141 million in restitution because it had “cheated millions of low-income Americans out of free tax filing services,” according to New York Attorney General Letitia James.

IRS Free File has been around for more than 20 years now. For 2023, taxpayers with an adjusted gross income of $79,000 or less could get access to online guided tax software through Free File, while those with over $79,000 can use the IRS’s Free File Fillable Forms.
This year, the IRS is partnering with 1040Now, Drake (1040.com), ezTaxReturn.com, FileYourTaxes.com, On-Line Taxes, TaxAct, TaxHawk (FreeTaxUSA), and TaxSlayer to offer free preparation.
The IRS says it received around 2.9 million tax returns through Free File in 2024, as opposed to the 2.7 million it received at the same time last year. Alongside Free File, the IRS started testing its Direct File service this tax season. The program lets you file your taxes for free directly through the IRS, but it’s only available in 13 states for now.

Illustration: Hugo Herrera / The Verge

The IRS Free File program will be around for another five years. The agency has announced that it’s extending the program through October 2029, allowing people to file their taxes with one of the IRS’s partners for free.

The Free File program got more attention recently after the IRS changed its rules in 2020 to prevent tax preppers from hiding their free services. Both TurboTax and H&R Block left the Free File program after that, and TurboTax owner Intuit paid out $141 million in restitution because it had “cheated millions of low-income Americans out of free tax filing services,” according to New York Attorney General Letitia James.

IRS Free File has been around for more than 20 years now. For 2023, taxpayers with an adjusted gross income of $79,000 or less could get access to online guided tax software through Free File, while those with over $79,000 can use the IRS’s Free File Fillable Forms.

This year, the IRS is partnering with 1040Now, Drake (1040.com), ezTaxReturn.com, FileYourTaxes.com, On-Line Taxes, TaxAct, TaxHawk (FreeTaxUSA), and TaxSlayer to offer free preparation.

The IRS says it received around 2.9 million tax returns through Free File in 2024, as opposed to the 2.7 million it received at the same time last year. Alongside Free File, the IRS started testing its Direct File service this tax season. The program lets you file your taxes for free directly through the IRS, but it’s only available in 13 states for now.

Read More 

Illustration by Cath Virginia / The Verge | Photos by Getty Images

Imagine this: you’ve carved out an evening to unwind and decide to make a homemade pizza. You assemble your pie, throw it in the oven, and are excited to start eating. But once you get ready to take a bite of your oily creation, you run into a problem — the cheese falls right off. Frustrated, you turn to Google for a solution.
“Add some glue,” Google answers. “Mix about 1/8 cup of Elmer’s glue in with the sauce. Non-toxic glue will work.”
So, yeah, don’t do that. As of writing this, though, that’s what Google’s new AI Overviews feature will tell you to do. The feature, while not triggered for every query, scans the web and drums up an AI-generated response. The answer received for the pizza glue query appears to be based on a comment from a user named “fucksmith” in a more than decade-old Reddit thread, and they’re clearly joking.
This is just one of many mistakes cropping up in the new feature that Google rolled out broadly this month. It also claims that former US President James Madison graduated from the University of Wisconsin not once but 21 times, that a dog has played in the NBA, NFL, and NHL, and that snakes are mammals.

Look, Google didn’t promise this would be perfect, and it even slaps a “Generative AI is experimental” label at the bottom of the AI answers. But it’s clear these tools aren’t ready to accurately provide information at scale.
Take Google I/O’s big launch of this feature, for instance. The demo was highly controlled, and yet, it delivered a questionable answer about how to fix a jammed film camera. (It suggested they “open the back door and gently remove the film”; don’t do that unless you want to ruin your photos!)
It’s not just Google; companies like OpenAI, Meta, and Perplexity have all grappled with AI hallucinations and mistakes. However, Google is the first to deploy this technology on such a large scale, and the examples of flubs just keep rolling in.
Companies developing artificial intelligence are often quick to avoid taking accountability for their systems with an approach much like a parent with an unruly child — boys will be boys! These companies claim that they can’t predict what this AI will spit out, so really, it’s out of their control.
But for users, that’s a problem. Last year, Google said that AI was the future of search. What’s the point, though, if the search seems dumber than before?

AI optimists argue that we should embrace the hype because of the rapid progress made so far, trusting that it will continue to improve. I really do believe that this technology will continue to get better, but focusing on an idealized future where these technologies are flawless ignores the significant issues they currently face — and allows companies to continue delivering subpar products.
For now, our search experiences are marred by decade-old Reddit posts in the pursuit of incorporating AI into everything. Many idealists believe we are on the brink of something great and that these issues are simply the growing pains of a nascent technology. I sure hope they’re right. But one thing is certain: we’ll likely witness someone putting glue on their pizza soon, because that’s the nature of the internet.

Illustration by Cath Virginia / The Verge | Photos by Getty Images

Imagine this: you’ve carved out an evening to unwind and decide to make a homemade pizza. You assemble your pie, throw it in the oven, and are excited to start eating. But once you get ready to take a bite of your oily creation, you run into a problem — the cheese falls right off. Frustrated, you turn to Google for a solution.

“Add some glue,” Google answers. “Mix about 1/8 cup of Elmer’s glue in with the sauce. Non-toxic glue will work.”

So, yeah, don’t do that. As of writing this, though, that’s what Google’s new AI Overviews feature will tell you to do. The feature, while not triggered for every query, scans the web and drums up an AI-generated response. The answer received for the pizza glue query appears to be based on a comment from a user named “fucksmith” in a more than decade-old Reddit thread, and they’re clearly joking.

This is just one of many mistakes cropping up in the new feature that Google rolled out broadly this month. It also claims that former US President James Madison graduated from the University of Wisconsin not once but 21 times, that a dog has played in the NBA, NFL, and NHL, and that snakes are mammals.

Look, Google didn’t promise this would be perfect, and it even slaps a “Generative AI is experimental” label at the bottom of the AI answers. But it’s clear these tools aren’t ready to accurately provide information at scale.

Take Google I/O’s big launch of this feature, for instance. The demo was highly controlled, and yet, it delivered a questionable answer about how to fix a jammed film camera. (It suggested they “open the back door and gently remove the film”; don’t do that unless you want to ruin your photos!)

It’s not just Google; companies like OpenAI, Meta, and Perplexity have all grappled with AI hallucinations and mistakes. However, Google is the first to deploy this technology on such a large scale, and the examples of flubs just keep rolling in.

Companies developing artificial intelligence are often quick to avoid taking accountability for their systems with an approach much like a parent with an unruly child — boys will be boys! These companies claim that they can’t predict what this AI will spit out, so really, it’s out of their control.

But for users, that’s a problem. Last year, Google said that AI was the future of search. What’s the point, though, if the search seems dumber than before?

AI optimists argue that we should embrace the hype because of the rapid progress made so far, trusting that it will continue to improve. I really do believe that this technology will continue to get better, but focusing on an idealized future where these technologies are flawless ignores the significant issues they currently face — and allows companies to continue delivering subpar products.

For now, our search experiences are marred by decade-old Reddit posts in the pursuit of incorporating AI into everything. Many idealists believe we are on the brink of something great and that these issues are simply the growing pains of a nascent technology. I sure hope they’re right. But one thing is certain: we’ll likely witness someone putting glue on their pizza soon, because that’s the nature of the internet.

Read More 

VideoCardz obtained this first — very modified —image of the Ally X. | Image: VideoCardz

I exclusively revealed what the Asus ROG Ally X handheld gaming PC will be all about and even touched an early prototype, but Asus wouldn’t talk specs ahead of its June 2nd release. But now, VideoCardz has obtained the key missing details — battery, weight, ports, thickness, storage, and RAM.

While we already knew the Ally X was shooting for double the battery life by including a larger pack, VideoCardz has leaked marketing materials that confirm it’s literally doubling the capacity to 80 watt-hours, up from the original 40Wh pack. If true, that would be the largest battery we’ve seen in a handheld, surpassing the boutique Ayaneo Kun’s 75 watt-hour pack.
And yet, the Ally X only weighs an additional 70g (2.5oz) and is just 5mm thicker, a bit thinner than my hands-on estimate. It’s 36.9mm (1.45 inches) thick in total, versus the 32mm (1.27 inches) of the original.
That’s partially due to a thinner fan design: 23 percent smaller, with 50 percent thinner fins, according to the leak, yet with 10 percent increased airflow.
Aside from battery, the most welcome spec might be the addition of a second USB-C port with USB4 speeds. VideoCardz says it replaces the proprietary eGPU port that Asus included previously but doesn’t say if we can charge from both the top and bottom now. (I would expect so since Asus taped up both the top and bottom of the engineering prototype I touched to keep me from sussing ports out.)
As we’ve reported, the Ally X shouldn’t have much increased performance over the original, with the same Ryzen Z1 Extreme chip and seven-inch 120Hz VRR screen, but VideoCardz also corroborates the rumor that it’ll come with 24GB of faster LPDDR5 memory, giving it an additional 8GB of overhead to share with the GPU that could possibly lead to a slight improvement in games.
Last but not least, it looks like I was right about a new D-pad — though the leak doesn’t seem to go into the rest of the Ally X’s importantly revised ergonomics. We don’t have any corroboration on price, which another leak suggested would be $799.
Asus didn’t immediately respond to a request to confirm the specs.

VideoCardz obtained this first — very modified —image of the Ally X. | Image: VideoCardz

I exclusively revealed what the Asus ROG Ally X handheld gaming PC will be all about and even touched an early prototype, but Asus wouldn’t talk specs ahead of its June 2nd release. But now, VideoCardz has obtained the key missing details — battery, weight, ports, thickness, storage, and RAM.

While we already knew the Ally X was shooting for double the battery life by including a larger pack, VideoCardz has leaked marketing materials that confirm it’s literally doubling the capacity to 80 watt-hours, up from the original 40Wh pack. If true, that would be the largest battery we’ve seen in a handheld, surpassing the boutique Ayaneo Kun’s 75 watt-hour pack.

And yet, the Ally X only weighs an additional 70g (2.5oz) and is just 5mm thicker, a bit thinner than my hands-on estimate. It’s 36.9mm (1.45 inches) thick in total, versus the 32mm (1.27 inches) of the original.

That’s partially due to a thinner fan design: 23 percent smaller, with 50 percent thinner fins, according to the leak, yet with 10 percent increased airflow.

Aside from battery, the most welcome spec might be the addition of a second USB-C port with USB4 speeds. VideoCardz says it replaces the proprietary eGPU port that Asus included previously but doesn’t say if we can charge from both the top and bottom now. (I would expect so since Asus taped up both the top and bottom of the engineering prototype I touched to keep me from sussing ports out.)

As we’ve reported, the Ally X shouldn’t have much increased performance over the original, with the same Ryzen Z1 Extreme chip and seven-inch 120Hz VRR screen, but VideoCardz also corroborates the rumor that it’ll come with 24GB of faster LPDDR5 memory, giving it an additional 8GB of overhead to share with the GPU that could possibly lead to a slight improvement in games.

Last but not least, it looks like I was right about a new D-pad — though the leak doesn’t seem to go into the rest of the Ally X’s importantly revised ergonomics. We don’t have any corroboration on price, which another leak suggested would be $799.

Asus didn’t immediately respond to a request to confirm the specs.

Read More 

Kia’s Soul-ful EV designs continue. | Image: Kia

Kia introduced the finalized version of its EV3 small electric SUV today, and with it comes a big new feature: AI.
The vehicle was first introduced as a futuristic concept vehicle last year, and now it will launch in South Korea this July, followed by Europe in the second half of the year. Kia representative Karla Gonzalez tells The Verge that EV3 is also confirmed for the US — but there’s no timing announced.
The EV3 will be the first production EV to get Kia’s voice-activated AI assistant that first popped up in the Kia K4. TechCrunch reports the assistant “is built off ChatGPT” but is heavily modified to let users plan trips, find entertainment, and more, according to Kia’s head of customer experience design, Pablo Martinez Masip. The feature will launch first in Korea and later in Europe.

Image: Kia
The EV3 looks a lot like the EV9 inside.

The EV3 has an estimated 600km of driving range (about 373 miles) based on the European WLTP system — but the US EPA estimate is usually lower in comparison. That’s based on the EV3 long-range model that has an 81.4kWh battery, while a “Standard” model is also planned that has a 58.3kWh battery.
Kia will also make a sportier-looking “GT-line” model. The EV3 can accelerate from zero to 100km in 7.5 seconds, and its powertrain is based on Hyundai’s E-GMP platform.
As far as other tech inside the car, it comes with a similar screen setup found in other Kia EVs: two 12.3-inch screens for the instrument cluster and infotainment; a five-inch panel for HVAC; and a 12-inch heads-up display. There are also premium streaming apps built in courtesy of LG’s webOS-based auto platform. You can also download themes for the dashboard, including NBA-inspired ones.

On the outside, the design of the EV3 is heavily inspired by the large, boxy, and Soul-ful Kia EV9 and is designed to feel spacious for five passengers. Kia made it so you can fit personal items like backpacks on the ground and includes a sliding table that can accommodate a laptop. Since the EV3 has no gas engine up front, you’ll get a 25-liter (1,526 cubic inches) frunk, which is pretty small but consistent with the company’s other models.

Kia’s Soul-ful EV designs continue. | Image: Kia

Kia introduced the finalized version of its EV3 small electric SUV today, and with it comes a big new feature: AI.

The vehicle was first introduced as a futuristic concept vehicle last year, and now it will launch in South Korea this July, followed by Europe in the second half of the year. Kia representative Karla Gonzalez tells The Verge that EV3 is also confirmed for the US — but there’s no timing announced.

The EV3 will be the first production EV to get Kia’s voice-activated AI assistant that first popped up in the Kia K4. TechCrunch reports the assistant “is built off ChatGPT” but is heavily modified to let users plan trips, find entertainment, and more, according to Kia’s head of customer experience design, Pablo Martinez Masip. The feature will launch first in Korea and later in Europe.

Image: Kia
The EV3 looks a lot like the EV9 inside.

The EV3 has an estimated 600km of driving range (about 373 miles) based on the European WLTP system — but the US EPA estimate is usually lower in comparison. That’s based on the EV3 long-range model that has an 81.4kWh battery, while a “Standard” model is also planned that has a 58.3kWh battery.

Kia will also make a sportier-looking “GT-line” model. The EV3 can accelerate from zero to 100km in 7.5 seconds, and its powertrain is based on Hyundai’s E-GMP platform.

As far as other tech inside the car, it comes with a similar screen setup found in other Kia EVs: two 12.3-inch screens for the instrument cluster and infotainment; a five-inch panel for HVAC; and a 12-inch heads-up display. There are also premium streaming apps built in courtesy of LG’s webOS-based auto platform. You can also download themes for the dashboard, including NBA-inspired ones.

On the outside, the design of the EV3 is heavily inspired by the large, boxy, and Soul-ful Kia EV9 and is designed to feel spacious for five passengers. Kia made it so you can fit personal items like backpacks on the ground and includes a sliding table that can accommodate a laptop. Since the EV3 has no gas engine up front, you’ll get a 25-liter (1,526 cubic inches) frunk, which is pretty small but consistent with the company’s other models.

Read More 

Image: Samsung

iFixit just broke up with Samsung over its shaky commitment to right to repair, but things may be worse than we thought. Samsung is apparently forcing independent repair shops to give up your personally identifying information — and report you for using aftermarket parts.

404 Media has now obtained a leaked copy of a contract between Samsung and an independent service provider, which states that for every repair, an independent repair shop must send Samsung your name, address, telephone number, your phone’s unique serial number and IMEI, your “customer complaint,” and all the details of your repair.
What’s more, it tells repair shops they must strip any aftermarket parts out of your phone — even though it’s likely perfectly legal for you to use third-party parts, and companies have gotten in trouble for suggesting that they would void your warranty.

Image: 404 Media

404 Media says it’s “verified the authenticity of this contract,” and I’ve independently seen a copy myself.
Samsung did not respond to 404 Media’s request for comment; I’ve been in touch with Samsung for the past two hours, and it has not yet confirmed or denied the report.
It’s not just independent repair shops, by the way — Samsung might be adding you to a database every time you buy an official replacement part. iFixit CEO Kyle Wiens confirms his company shared some customer information with Samsung, too. If you bought a genuine Samsung part, iFixit made you agree to let Samsung have your email and a list of every genuine part you’ve purchased:

Image: iFixit
“I understand that iFixit will provide Samsung with my email and Samsung genuine parts purchase history to provide integrated customer care support.”

“We do not require this information for any other partnerships, and do not share customer information with any other OEM,” says Wiens.
Wiens says he can’t verify whether Samsung’s independent service provider contract is the current agreement that Samsung makes independent repair shops sign, as Samsung has never provided him with a copy, and the version I’ve seen has the year redacted. (A source tells me it’s from 2023.) It’s possible that Samsung has already replaced its language — or will replace it before July 1st, when California’s and Minnesota’s right-to-repair laws go into effect.

That’s because Minnesota’s law, in particular, probably wouldn’t let Samsung enforce such a contract — not only does it require manufacturers to provide parts under “fair and reasonable terms” but it also specifically bans “that a part be registered, paired with, or approved by the original equipment manufacturer.”
(It also bans “imposing a substantial obligation to use or restrict the use of the part” and “imposing any additional cost or burden that is not reasonably necessary.”)
Right-to-repair laws differ by state, though: California’s law, for example, is not so clear-cut about what “fair and reasonable” means. And not every repair shop knows its rights or is eager to fight Samsung, which controls 25 percent of the US smartphone market and is far and away the top Android provider in the United States.
Nathan Proctor, senior campaign director for right to repair at the US Public Interest Research Group, says he doubts that independent repair shops are actually stripping aftermarket parts out of people’s phones, no matter what Samsung asks. That’s a lot of work.
But he does believe that shops are giving up customer information — because Samsung will generally know when its genuine parts are being bought and installed and thus be able to track which repair shops aren’t complying.
It’s not clear (and 404 Media does not say) whether any indie shops make customers clearly acknowledge that Samsung will get their personal information.

Image: Samsung

iFixit just broke up with Samsung over its shaky commitment to right to repair, but things may be worse than we thought. Samsung is apparently forcing independent repair shops to give up your personally identifying information — and report you for using aftermarket parts.

404 Media has now obtained a leaked copy of a contract between Samsung and an independent service provider, which states that for every repair, an independent repair shop must send Samsung your name, address, telephone number, your phone’s unique serial number and IMEI, your “customer complaint,” and all the details of your repair.

What’s more, it tells repair shops they must strip any aftermarket parts out of your phone — even though it’s likely perfectly legal for you to use third-party parts, and companies have gotten in trouble for suggesting that they would void your warranty.

Image: 404 Media

404 Media says it’s “verified the authenticity of this contract,” and I’ve independently seen a copy myself.

Samsung did not respond to 404 Media’s request for comment; I’ve been in touch with Samsung for the past two hours, and it has not yet confirmed or denied the report.

It’s not just independent repair shops, by the way — Samsung might be adding you to a database every time you buy an official replacement part. iFixit CEO Kyle Wiens confirms his company shared some customer information with Samsung, too. If you bought a genuine Samsung part, iFixit made you agree to let Samsung have your email and a list of every genuine part you’ve purchased:

Image: iFixit
“I understand that iFixit will provide Samsung with my email and Samsung genuine parts purchase history to provide integrated customer care support.”

“We do not require this information for any other partnerships, and do not share customer information with any other OEM,” says Wiens.

Wiens says he can’t verify whether Samsung’s independent service provider contract is the current agreement that Samsung makes independent repair shops sign, as Samsung has never provided him with a copy, and the version I’ve seen has the year redacted. (A source tells me it’s from 2023.) It’s possible that Samsung has already replaced its language — or will replace it before July 1st, when California’s and Minnesota’s right-to-repair laws go into effect.

That’s because Minnesota’s law, in particular, probably wouldn’t let Samsung enforce such a contract — not only does it require manufacturers to provide parts under “fair and reasonable terms” but it also specifically bans “that a part be registered, paired with, or approved by the original equipment manufacturer.”

(It also bans “imposing a substantial obligation to use or restrict the use of the part” and “imposing any additional cost or burden that is not reasonably necessary.”)

Right-to-repair laws differ by state, though: California’s law, for example, is not so clear-cut about what “fair and reasonable” means. And not every repair shop knows its rights or is eager to fight Samsung, which controls 25 percent of the US smartphone market and is far and away the top Android provider in the United States.

Nathan Proctor, senior campaign director for right to repair at the US Public Interest Research Group, says he doubts that independent repair shops are actually stripping aftermarket parts out of people’s phones, no matter what Samsung asks. That’s a lot of work.

But he does believe that shops are giving up customer information — because Samsung will generally know when its genuine parts are being bought and installed and thus be able to track which repair shops aren’t complying.

It’s not clear (and 404 Media does not say) whether any indie shops make customers clearly acknowledge that Samsung will get their personal information.

Read More 

Image: Hugo Herrera / The Verge

Buy now, pay later services — or BNPL for short — are now beholden to some of the same rules as credit card providers. On Thursday, the Consumer Financial Protection Bureau (CFPB) issued an interpretive rule that classifies BNPL services as credit card providers and requires them to investigate disputed purchases, among other things.
BNPL services let customers purchase a product and then pay for it in a series of interest-free installments. Under the new guidance, BNPL services like Klarna, Afterpay, and Affirm must provide refunds for returned products or canceled services and send periodic billing statements to customers. The CFPB’s decision comes after launching an inquiry into BNPL services, which found that BNPL “is often used as a close substitute for conventional credit cards.”
Some BNPL services say they’re already meeting the criteria laid out by the CFPB. In a post on its blog, Klarna said the company works to protect customers by covering refunds, investigating customer disputes, and providing purchase information. However, the company still takes issue with the CFPB’s classification of BNPL as credit cards.

“The CFPB’s announcement is a significant step forward in regulating BNPL, which Klarna has actively called for over many years,” Klarna spokesperson John Craske tells The Verge. “But it is baffling that the CFPB has overlooked the fundamental differences between interest-free BNPL and credit cards, whose whole business model is based on trapping customers into a cycle of paying sky-high interest rates month after month.”
Meanwhile, Affirm CEO Max Levchin said in a thread on X that the company is “pleased that the Bureau is promoting consistent industry standards (many of which already reflect how Affirm operates).”
BNPL services have risen in popularity as credit card interest rates continue to skyrocket. (That popularity has even led Apple to create a BNPL service of its own.) However, there are reports suggesting that BNPL could lead some people to overspend and rack up debt. As pointed out by this report from CNBC, it’s hard to tell just how many people are in the red because BNPL services don’t need to report loans to credit reporting agencies.

Image: Hugo Herrera / The Verge

Buy now, pay later services — or BNPL for short — are now beholden to some of the same rules as credit card providers. On Thursday, the Consumer Financial Protection Bureau (CFPB) issued an interpretive rule that classifies BNPL services as credit card providers and requires them to investigate disputed purchases, among other things.

BNPL services let customers purchase a product and then pay for it in a series of interest-free installments. Under the new guidance, BNPL services like Klarna, Afterpay, and Affirm must provide refunds for returned products or canceled services and send periodic billing statements to customers. The CFPB’s decision comes after launching an inquiry into BNPL services, which found that BNPL “is often used as a close substitute for conventional credit cards.”

Some BNPL services say they’re already meeting the criteria laid out by the CFPB. In a post on its blog, Klarna said the company works to protect customers by covering refunds, investigating customer disputes, and providing purchase information. However, the company still takes issue with the CFPB’s classification of BNPL as credit cards.

“The CFPB’s announcement is a significant step forward in regulating BNPL, which Klarna has actively called for over many years,” Klarna spokesperson John Craske tells The Verge. “But it is baffling that the CFPB has overlooked the fundamental differences between interest-free BNPL and credit cards, whose whole business model is based on trapping customers into a cycle of paying sky-high interest rates month after month.”

Meanwhile, Affirm CEO Max Levchin said in a thread on X that the company is “pleased that the Bureau is promoting consistent industry standards (many of which already reflect how Affirm operates).”

BNPL services have risen in popularity as credit card interest rates continue to skyrocket. (That popularity has even led Apple to create a BNPL service of its own.) However, there are reports suggesting that BNPL could lead some people to overspend and rack up debt. As pointed out by this report from CNBC, it’s hard to tell just how many people are in the red because BNPL services don’t need to report loans to credit reporting agencies.

Read More 

Image: ESA

Stunning new images from the Euclid space telescope show a dazzling array of starry sights, released to celebrate the first science published from the telescope’s early release observations.
The mission from the European Space Agency, launched last year, aims to elucidate the mysteries of dark matter and dark energy — forces that comprise most of the known universe but that are largely unknown since they are so difficult to study. Scientists know that dark matter must exist because they can observe its gravitational effects, for example, but they can’t observe it directly because it doesn’t interact with light.
The solution is an extremely sensitive instrument like Euclid, which can observe distant galaxies to see dark matter’s effects on them. “We have to accurately measure the shapes of over 1.5 billion distant galaxies,” explained Valeria Pettorino and René Laureijs, Euclid project scientists at the European Space Agency. But these galaxies can be faint and tiny, and when viewed from Earth, they are blurred by our planet’s atmosphere. As Euclid sits above the atmosphere in space, it can detect and measure these galaxies with less interference.
The result is images that are four times sharper than those taken from telescopes on the ground, showing a high level of detail over a wide area of the sky. Using data from just 24 hours of observations, they show objects including galaxy clusters, star-forming regions, and a large nearby spiral galaxy called NGC 6744.

As well as the careful measurements of galaxies for dark matter research, the telescope will also provide data on other astronomical objects. “We will be able to also search for objects that were very difficult to detect before Euclid such as free floating planets, ultra-cold stars, brown dwarfs, galaxies with very low surface brightness or very high redshift quasar populations,” Pettorino and Laureijs said. “Euclid may also reveal new objects which have never been observed before.”
Stellar nurseries like Messier 78 will be searched for objects like free-floating or “rogue” planets, which don’t orbit a star but, rather, wander alone through the cosmos. There are thought to be trillions of these rogue planets in our galaxy, but they are hard to study since they often don’t have nearby objects that they interact with. With Euclid, rogue planets down to four times the size of Jupiter can be observed — that’s fairly large compared to planets in our Solar System but tiny in terms of how small the objects are compared to stars.
And at the other end of the mass scale, Euclid will also observe entire groups of galaxies, like the cluster Abell 2390, which contains more than 50,000 galaxies. Such clusters are so massive that they bend space-time, creating an effect called gravitational lensing, in which light from distant galaxies is warped and shows up as a distorted shape like a disk or an arc. This allows researchers to study the amount of dark matter present in the cluster since the greater the amount of dark matter, the greater the mass and therefore the lensing effect.

The first images from Euclid have exceeded expectations, Pettorino and Laureijs said, with results from just a day’s worth of observations already showing millions of different objects. Over the next six years, Euclid will peel back even more of the universe as it surveys a third of the entire sky.
“This is the start of an extraordinary journey towards a fundamental understanding of the Universe we live in,” Pettorino and Laureijs said.

Image: ESA

Stunning new images from the Euclid space telescope show a dazzling array of starry sights, released to celebrate the first science published from the telescope’s early release observations.

The mission from the European Space Agency, launched last year, aims to elucidate the mysteries of dark matter and dark energy — forces that comprise most of the known universe but that are largely unknown since they are so difficult to study. Scientists know that dark matter must exist because they can observe its gravitational effects, for example, but they can’t observe it directly because it doesn’t interact with light.

The solution is an extremely sensitive instrument like Euclid, which can observe distant galaxies to see dark matter’s effects on them. “We have to accurately measure the shapes of over 1.5 billion distant galaxies,” explained Valeria Pettorino and René Laureijs, Euclid project scientists at the European Space Agency. But these galaxies can be faint and tiny, and when viewed from Earth, they are blurred by our planet’s atmosphere. As Euclid sits above the atmosphere in space, it can detect and measure these galaxies with less interference.

The result is images that are four times sharper than those taken from telescopes on the ground, showing a high level of detail over a wide area of the sky. Using data from just 24 hours of observations, they show objects including galaxy clusters, star-forming regions, and a large nearby spiral galaxy called NGC 6744.

As well as the careful measurements of galaxies for dark matter research, the telescope will also provide data on other astronomical objects. “We will be able to also search for objects that were very difficult to detect before Euclid such as free floating planets, ultra-cold stars, brown dwarfs, galaxies with very low surface brightness or very high redshift quasar populations,” Pettorino and Laureijs said. “Euclid may also reveal new objects which have never been observed before.”

Stellar nurseries like Messier 78 will be searched for objects like free-floating or “rogue” planets, which don’t orbit a star but, rather, wander alone through the cosmos. There are thought to be trillions of these rogue planets in our galaxy, but they are hard to study since they often don’t have nearby objects that they interact with. With Euclid, rogue planets down to four times the size of Jupiter can be observed — that’s fairly large compared to planets in our Solar System but tiny in terms of how small the objects are compared to stars.

And at the other end of the mass scale, Euclid will also observe entire groups of galaxies, like the cluster Abell 2390, which contains more than 50,000 galaxies. Such clusters are so massive that they bend space-time, creating an effect called gravitational lensing, in which light from distant galaxies is warped and shows up as a distorted shape like a disk or an arc. This allows researchers to study the amount of dark matter present in the cluster since the greater the amount of dark matter, the greater the mass and therefore the lensing effect.

The first images from Euclid have exceeded expectations, Pettorino and Laureijs said, with results from just a day’s worth of observations already showing millions of different objects. Over the next six years, Euclid will peel back even more of the universe as it surveys a third of the entire sky.

“This is the start of an extraordinary journey towards a fundamental understanding of the Universe we live in,” Pettorino and Laureijs said.

Read More