A free version of Gemini Code Assist, Google’s enterprise-focused AI coding tool, is now available globally for solo developers. Google announced today that Gemini Code Assist for individuals is launching in public preview, aiming to make coding assistants “with the latest AI capabilities” more accessible for students, hobbyists, freelancers, and startups.
“Now anyone can more conveniently learn, create code snippets, debug, and modify their existing applications — all without needing to toggle between different windows for help or to copy and paste information from disconnected sources,” said Ryan J. Salva, Google’s senior director of product management. “While other popular free coding assistants have restrictive usage limits, with usually only 2,000 code completions per month, we wanted to offer something more generous.”
That feels particularly targeted at GitHub Copilot, the most direct competitor to Gemini Code Assist, which also provides a free user tier that’s limited to 2,000 code completions and 50 Copilot Chat messages each month. Google is offering up to 180,000 code completions per month by contrast, which it describes as “a ceiling so high that even today’s most dedicated professional developers would be hard-pressed to exceed it.”
Like the enterprise version, Gemini Code Assist for individuals is powered by Google’s Gemini 2.0 artificial intelligence model and can generate entire code blocks, complete code as you write, and provide general coding assistance via a chatbot interface. The free coding tool can be installed in Visual Studio Code, GitHub, and JetBrains developer environments and supports all programming languages in the public domain.
Developers can instruct Gemini Code Assist using natural language, such as asking the coding chatbot to “build me a simple HTML form with fields for name, email, and message, and then add a ‘submit’ button.” It currently supports 38 languages and up to 128,000 chat input tokens in the token context window, which is the amount of text (tokens) that can be processed or “remembered” when generating a response.
The free Individual tier seems pretty expansive, but it doesn’t include all of the advanced business-focused features available in the Standard and Enterprise versions of Gemini Code Assist. If you want productivity metrics, integrations with Google Cloud services like BigQuery, or to customize responses using private code data sources then you’ll need to use Google’s paid tiers.
Microsoft has started testing a free version of Office for Windows that includes ads. Right now, you have to pay for a monthly Microsoft 365 subscription to get access to the full desktop version of Office, but Microsoft has been quietly testing an ad-supported version in certain countries.
Beebom first noticed that the ad-supported version of Office for Windows appeared in India recently, allowing Windows users to access Word, Excel, PowerPoint, and other Office apps without the Microsoft 365 subscription fee.
“Microsoft has been conducting some limited testing. Currently, there are no plans to launch a free, ad-supported version of Microsoft Office desktop apps,” says a Microsoft spokesperson in a statement to PCWorld. While Microsoft claims this is limited testing, the company has specifically engineered its Office apps to now work on Windows with ads, so we may well see this version appear in more markets eventually.
The ad-supported version of Office includes banners that are permanently visible at the side, as well as 15-second video ads that play every few hours, according to Beebom. Microsoft also forces users of this free version of Office to store documents in OneDrive, with support for local file storage disabled.
Microsoft currently only offers free versions of Office on the web, so you have to use a browser to access far more limited versions of Word, Excel, and PowerPoint. This test version of Office for Windows doesn’t include the full features of the apps, either. Word is missing drawing and design tools, line spacing, and more. The free version of Excel doesn’t support add-ins, pivot tables, or macros. PowerPoint is also missing support for dictation, custom slide shows, and other features.
The Logitech Powerplay 2 wireless charging mousepad.
I’ve never reviewed a perfect product, but Logitech’s Powerplay Wireless Charging System comes close. For over three years and counting, I’ve never even had to think about charging my wireless mouse. It’s so dead simple, it feels like magic, and it’s a shame that most people probably can’t afford it at $120.
The good news: Logitech is releasing a new $100 version in March, called the Powerplay 2, and it’s just as easy to set up. Plug in mousepad, snap a magnetic “Charging Coin” into the base of your mouse, then put mouse on mousepad to continuously charge.
The bad news: It’s only $20 cheaper, yet it feels like Logitech made its mousepad more than $20 cheaper to hit that goal.
The mousepad does come with improvements. Logitech boasts it has a 15 percent wider charging area and is thinner at just 3.5mm, and that’s what I see with my review unit. Now, as long as the entirety of my G502 Lightspeed mouse is resting within any corner of the mousepad, the charging indicator lights up, which wasn’t quite true of the original. My calipers do read 3.5mm when I’m measuring the charging base and its thin fabric mousing surface together.
But my calipers also show the charging base is exactly the same 2.7mm thickness as before, and the old mousepad wasn’t all that much thicker: just 4.3mm in total before vs. 3.5mm in total now, a difference I do not feel. And do you leave your mouse all the way at the corners of your mousepad? Again, I’ve spent over three years charging this mouse on the old mousepad without even thinking about it. I never bother to reposition my mouse on my old Powerplay to make sure it’s charging; I just drop it when I’m done using it, and I’ve never once run out of charge.
What do we lose with the Powerplay 2? First, while Logitech has ditched the old micro-USB cable, we’re not getting USB-C. Instead, Logitech’s opted for a fixed cable, so I can’t as easily take the mousepad off my desk on days I need more space there.
The big one: there’s no more wireless mouse receiver built into the Powerplay 2, a feature I found handy with the original. Now, my mouse requires two full-size USB ports instead of one because I still have to leave the mouse’s dongle plugged into my PC, too. I can’t leave the dongle stowed in the mouse for grab-and-go travel, and I can’t leave it in my laptop and switch between laptop and desktop anymore by turning the mouse on and off and yanking the Powerplay’s plug.
There’s also no more programmable RGB light inside the Logitech G logo. I don’t terribly miss that, since I don’t sync up gamer lights. But the dull black Logitech G feels cheaper; before, the RGB was at least a nice reassuring reminder that my mousepad was properly receiving USB power and ready for action.
And, while I do like the new thinner mousepad that comes with the Powerplay 2, which looks like it might not delaminate from its backing as easily as the original (it’s the one piece of my Powerplay that has deteriorated over the past three years), the original Powerplay came with two mousepads (one hard, one cloth) in the box. Now, you get the one.
(Also, just as FYI, the new Powerplay 2 charging coin doesn’t seem to work with the original pad and vice versa. You can’t mix and match those parts.)
I tried hard to get Logitech to show me more benefits, because the original’s one of my favorite products. Perhaps this one’s so much easier for Logitech to produce that it’ll offer some great discounts, or sell amazing bundles after launch? Or perhaps that 15 percent larger charging area will come in handy for possible additional supercapacitor mice that’d react worse than battery mice if they aren’t getting reliably fed, though the current supercap G309 seems to work fine with the original Powerplay in my early tests.
Logitech wouldn’t comment on future supercap products, wouldn’t hint at sales, and wouldn’t promise its own bundles — though Logitech does “anticipate retailers will be offering bundles shortly after launch,” according to Logitech senior global marketing manager Andrew Siminoff.
The original Powerplay is no longer in stock at major retailers, so I expect it will soon fetch a premium price on eBay. But the Powerplay 2 still seems like a good product that achieves the core goal. Fingers crossed that come Black Friday, we’ll be able to buy a combo pack with it and Logitech’s cheapest compatible mouse — that G309 — for under $100 in total.
The Powerplay 2 should be available on Amazon and Logitech’s site on March 11th.
Disney has had a tumultuous run since Bob Iger’s return as the company’s CEO, which came just two years after he handed the reins over to Bob Chapek. Iger has since worked to undo some of Chapek’s changes as the company contends with a streaming-focused future.
Outside of streaming, Iger is hoping to boost Disney’s slate of films and reinvigorate interest in the Marvel Cinematic Universe. But with Iger’s contract set to expire in 2026, a lot remains unclear about what’s next for the future of Disney. Here are all the major changes Iger has made so far.
When Nvidia originally confirmed that some of its new RTX 50-series graphics cards had a “rare” manufacturing issue that left them missing some promised render units and a slight amount of performance as a result, it only named three affected cards: the RTX 5090, RTX 5090D, and RTX 5070 Ti. But now, Nvidia has confirmed to us that RTX 5080 production was affected by the same issue as well.
“Upon further investigation, we’ve identified that an early production build of GeForce RTX 5080 GPUs were also affected by the same issue. Affected consumers can contact the board manufacturer for a replacement,” Nvidia GeForce global PR director Ben Berraondo tells The Verge.
In response to The Verge’s questions, Berraondo adds that “no other Nvidia GPUs have been affected” — we specifically asked about the upcoming RTX 5070, and he says it’s not affected either. Nor should any cards be affected that were produced more recently: “The production anomaly has been corrected,” he says. In case you’re wondering, he also told us that Nvidia was not aware of these issues before it launched these GPUs.
Here’s the company’s full amended statement:
We have identified a rare issue affecting less than 0.5% (half a percent) of GeForce RTX 5090 / 5090D, RTX 5080, and 5070 Ti GPUs which have one fewer ROP than specified. The average graphical performance impact is 4%, with no impact on AI and Compute workloads. Affected consumers can contact the board manufacturer for a replacement. The production anomaly has been corrected.
While it doesn’t seem like a lot of GPUs were affected, given how few of these GPUs have shipped so far, and Nvidia is also promising replacements, it’s the latest in a line of annoyances with Nvidia’s new cards.
After months of preparation, Epic Games will finally take on Apple in court in a trial that could fundamentally change the makeup of the App Store. The fight dates back to August, when Epic added a direct payment mechanism to its hit battle royale game Fortnite in violation of Apple’s rules. The iPhone maker quickly removed the game from the App Store, and Epic responded shortly after with an antitrust lawsuit aiming to establish the App Store as a monopoly. The case will finally be brought to trial starting May 3rd.
The trial promises to deliver huge revelations about the inner workings of one of the biggest and most influential companies in the world, with testimony from Apple CEO Tim Cook, Craig Federighi, Phil Schiller, Epic CEO Tim Sweeney, and more. We’ve already made some fascinatingdiscoveries from documents published ahead of the trial, and there’s sure to be a lot more news ahead.
The Nintendo-branded Alarmo will yell at you for sleeping in if Mario’s leer of disapproval isn’t enough.
Ongoing scarcity has made it challenging to purchase Nintendoâs adorable Alarmo since its soft launch last year, but it appears those days are behind us. The $99.99 alarm clock is now readily available from third-party retailers in the US, starting with Best Buy. You can also buy it directly from Nintendo without a subscription, with the only caveat being that you must sign in with a Nintendo account.
Nintendo announced its alarm clock in October while we patiently awaited (any) news regarding the Nintendo Switch 2, which is set to launch later this year. Although the Alarmo is certainly not as exciting as a new console, it is unapologetically Nintendo, with a cartoonish look that calls to mind a vintage alarm clock â albeit with a few modern features.
Nintendoâs bright red alarm clock features an illuminated button on top and a rounded face that houses an LCD display. The 2.8-inch panel shows the time/date and will attempt to wake you each morning with scenes and sounds from several iconic franchises. It makes room for visits from beloved characters like Link and Mario, as well as pikmin. There are currently 35 scenes across five franchises available, and Nintendo says yo …
We’re nearing the deadline that Elon Musk imposed for government workers to reply to a mass email about productivity, and the results have been predictably confusing — with even a direct statement from President Donald Trump failing to clear things up.
Government agencies have taken significantly different tacks toward the Musk-promoted email, which he announced to the public midday on February 22nd. Sent by the Office of Personnel Management (OPM), the message demanded all federal employees respond by the end of the 24th with “5 bullets of what you accomplished last week,” and Musk said on X that “failure to respond will be taken as a resignation.” The email reportedly didn’t include this noteworthy detail.
Other agencies have issued more nebulous guidance. In an email obtained by The Verge, Federal Trade Commission Chairman Andrew Ferguson told staff that responses were “voluntary” — but he added that “I enthusiastically responded” to the message and “strongly encourage you to respond as well.”
As of this article’s publication, the White House has done little to clarify the situation. An unnamed administration official said on Monday morning that employees should defer to their agencies’ guidance, reported Politico. An OPM official further told The Washington Post that the office was “unsure what to do with the emails” and had “no plans” to analyze them. Yet more anonymous officials, however, said that workers’ reports would be “fed into an artificial intelligence system to determine whether those jobs are necessary or not,” per NBC News.
Meanwhile, on the same day, Trump said publicly that people who failed to respond would be “sort of semi-fired,” adding that “a lot of people are not answering because they don’t even exist.” He denied that agencies were clashing with Musk by issuing conflicting guidance, saying it was “done in a friendly manner.”
Musk’s email echoed his behavior after taking over Twitter, where he demanded employees do things like print out 50 pages of their recent coding work or write a memo justifying their jobs to receive previously promised company stock. But unlike at Twitter, where he held sole unquestioned control, he’s dealing with formal chains of command and many other stakeholders here.
Still, the whole impossibly tangled situation is conducive to Musk and Trump’s goal of paralyzing the government, letting them instill fear in employees while creating an excuse to fire people as desired. (If they work on nuclear safety or bird flu, maybe they’ll get semi-rehired afterward.)
Like many moves by Musk and his Department of Government Efficiency (DOGE), the latest action ignores existing government structures in a way that may be aimed at avoiding legal or political accountability. The email nonetheless drew an immediate challenge in court. It was included in an amended suit filed by groups including the American Federation of Government Employees, which condemned the email as “thoughtless and bullying … meant to intimidate federal employees and cause mass confusion.”
Ironically for figures who claim to be fighting bureaucratic confusion, Musk and Trump have created one of the most downright kafkaesque scenarios imaginable. We’re now looking at a government order presenting a drastic ultimatum that is never mentioned in the order, in which a response may be either mandatory or forbidden, and failing to respond may or may not get you simultaneously fired and not fired. Also, you may not actually exist.
NORTH DAKOTA, UNITED STATES – 2017/02/22: Defiant Dakota Access Pipeline water protectors faced-off with various law enforcement agencies on the day the camp was slated to be raided. | Photo: Getty Images
A pivotal trial over the embattled Dakota Access Pipeline opens today that could have grave consequences for protests in the US and the future of the environmental group Greenpeace.
Members of the Standing Rock Sioux and more than 500 other tribes protested the development of the pipeline alongside demonstrators who joined from across the US nearly a decade ago. Legal battles are still in motion, even after oil started flowing through the pipeline that runs from North Dakota to Illinois in 2017.
The company that operates Dakota Access, Energy Transfers, is suing Greenpeace for $300 million in a lawsuit that goes on trial this week. Energy Transfers claims that Greenpeace supported protestersâ âunlawful acts of trespassâ and property destruction to stop construction. It also alleges that the organization spread false information about the company and concerns about the pipelineâs impact on the environment and cultural sites to the public and to banks financing the project.
âThis directly impacts everybody, not just Standing Rock, not just Greenpeace.â
Paying that amount in damages would be equivalent to about 10 times Greenpeace USAâs annual budget, according to …
SpaceX believes it has determined why Starship exploded during its seventh test flight. | Image: SpaceX
SpaceX believes it knows what caused the explosion during the seventh test flight of its Starship spacecraft and Super Heavy booster on January 16th, 2025. Fires in the aft section of Starship, located between the bottom of its liquid oxygen tank and rear heat shield, caused “all but one of Starship’s engines to execute controlled shut down sequences” leading to a loss of communication and eventually the spacecraft’s safety system triggering its own destruction.
The goal of the seventh test flight was to test several upgrades SpaceX made to its heavy-lift rocket system. After a successful launch and completing a full duration burn, the Super Heavy booster separated from Starship and triggered a boostback burn designed to return it to the launch site. Following a landing burn, the Super Heavy booster was successfully caught mid-air by the launch tower at Starbase for the second time.
The mission didn’t go quite as well for Starship. Approximately two minutes after the spacecraft ignited its second stage Raptor engines following separation, a flash was observed in the aft section of Starship known as the attic, followed by sensors detecting a pressure rise from a leak. Two minutes after that, a second flash was observed, followed by sustained fires in the attic section that eventually “caused all but one of Starship’s engines to execute controlled shut down sequences and ultimately led to a loss of communication with the ship.”
Post-flight analysis indicated that Starship’s Autonomous Flight Safety System triggered a self-destruct approximately three minutes after the ground crew lost contact with the spacecraft. According to SpaceX, the most probable cause of the incident was vibrations that were much stronger during the flight than had been experienced during testing. That resulted in increased stress on the propulsion system’s hardware and, eventually, a propellant leak that “exceeded the venting capability of the ship’s attic area and resulted in sustained fires.”
The explosion created falling debris that looked more like a meteor shower over the islands of Turks and Caicos to several tourists who shared videos of the aftermath on social media. Although SpaceX says all the “debris came down within the pre-planned Debris Response Area,” the Federal Aviation Administration briefly slowed and diverted several flights in the area on January 16th as a result of the incident.
As part of the investigation into the explosion involving SpaceX, the FAA, NASA, the National Transportation Safety Board, and the U.S. Space Force, the company conducted a 60-second static test fire with the Starship that will be used on an upcoming eighth flight. Following the results of that test, SpaceX has made hardware changes to fuel feedlines as well as adjustments to propellant temperatures and operating thrust targets.
The company has also added additional vents and a “new purge system utilizing gaseous nitrogen” to the attic section of Starship designed to make that area more robust to propellant leakage.
SpaceX currently plans to launch an eighth test flight of Starship on February 28th, 2025, but is still “working with the FAA to either close the mishap investigation or receive a flight safety determination, along with working on a license authorization to enable its next flight of Starship.”
An image of the call completed by AT&T and AST SpaceMobile.
Verizon and AT&T have each announced milestones in making cellphone-to-satellite video calls in partnership with satellite company AST SpaceMobile.
Verizon has completed its first cellphone-to-satellite video call, while AT&T has completed its first using satellites that will be used as part of a commercial network. There’s lining up competition to T-Mobile’s arrangement with SpaceX and Starlink on satellite-to-cell service which launched a public beta for messaging via satellite earlier this month. AT&T and Verizon have said that T-Mobile and SpaceX’s offerings would harm their networks.
Verizon pulled off “a live video call between two mobile devices with one connected via satellite and the other connected via Verizon’s terrestrial network connection,” according to a company press release. In AT&T’s case, “AT&T and AST SpaceMobile have successfully completed another video call by satellite to an everyday smartphone over AT&T spectrum,” per AT&T’s press release.
Both phone companies relied on AST’s constellation of five BlueBird satellites that were launched last September for the tests. AT&T’s initial video call test happened in June 2023.
Amazon is holding a press event this week, where we expect it to finally launch its “new” Alexa. This could be the beginning of a major shift in how we use generative AI in our homes, or it could be a big disappointment.
The latter seems likely, based on the delays and persistent rumors that the voice assistant is struggling with its revamp. But I’m hoping we’ll at least end up somewhere in the middle — with a smarter, more useful Alexa, if not the “superhuman assistant” Amazon has promised.
The event, scheduled for 10AM on Wednesday, February 26th, in New York City, is being hosted by Amazon’s new devices and service chief Panos Panay, which is a strong hint there’ll be new hardware. The flagship fourth-gen Amazon Echo speaker is way past due for an upgrade, and with smart glasses being so hot right now, I could see Alexa getting cozier on our faces.
Here’s a look at what we expect from the event, what not to expect, and what we hope is coming. Remember to tune in to The Verge’s live blog on Wednesday.
A “new” Alexa
I expect Amazon to announce the long-awaited arrival of its new Alexa, which has been rearchitected and infused with generative AI — courtesy of its LLMs, including Titan and reportedly some of Anthropic’s Claude.
First announced in the fall of 2023, the revamp of Amazon’s once trailblazing AI voice assistant has been a long time coming. While it’s fine to set a timer or play music and turn your lights off, the current Alexa has struggled to find purpose in a world dominated by ChatGPT, Gemini, and their ilk.
The new “Remarkable Alexa,” as it’s reportedly called, should understand natural speech, interpret context, respond to multiple requests in a single command, and take action on your behalf with either deeper API integrations and / or genuine agentic abilities.
All of this means that we should be able to talk to Alexa without using clunky nomenclature and get more useful responses (assuming Amazon has managed to squash Alexa’s reported need to show off).
For smart home control, we should be able to say a command like, “Alexa, turn out the lights, lock the back door, and turn the thermostat to Sleep — oh, and play sleep sounds in the bedroom,” and the Assistant will do it all.
Amazon’s new Alexa will be tuned in to your smart home and its capabilities
Amazon’s new Alexa should also be tuned in to your smart home and its capabilities. At CES this year, I spoke with companies working on integrations with the new Alexa, using the new developer tools Dynamic Controller and Action Controller that Amazon announced in 2023. Then Amazon said it was working with GE Cync, Philips Hue, GE Appliances, iRobot, and Roborock on features that would allow the Assistant to better understand what you want devices to do. For example, say “Alexa, the floor is dirty,” and it will send out your robot vac.
GE Cync’s Carmen Pastore confirmed to The Verge that the smart lighting company is working on integrating what he called “Amazon Alexa Reflex” to simplify lighting scene control with natural language voice commands.
This is where Alexa can differentiate itself. The voice assistant could bring value if it can fuse its current capabilities with generative AI-powered improvements. However, reports suggest this has been a challenge, with the new Alexa prone to hallucinating or refusing to turn on lights. It’s also an area competitors Apple and Google, who are tackling the same challenge with Siri and Google Assistant, are reportedly struggling with.
A new flagship Echo speaker and better access to Alexa on the go
Thanks to the billions of cheap Alexa-enabled devices in people’s homes, Amazon has a head start in the smart home. However, the flagship Echo fourth-gen smart speaker is now four years old. The company has said its new Alexa will run on all current hardware, but I expect the improved voice assistant will come with a shiny new home — especially considering Panay, the new devices and services chief, has impressive hardware chops, courtesy of his time at Microsoft building the Surface line.
The stage is set for an Echo speaker with a new design, improved processing power, local control, and more sensors
When it announced the new Alexa in 2023, Amazon launched the Echo Hub smart home controller and the third-gen Echo Show 8 smart home display. While we might see an Echo Hub 2.0, the stage is set for a new Echo speaker with a new design, improved processing power, local control, and more smart home sensors.
While the home is Alexa’s comfort zone, Amazon could continue pushing us to use Alexa on the go with new Echo Frame smart glasses and a third generation of the flagship Echo Buds, making the smarter Alexa accessible wherever we are.
New Fire TV features
At the 2023 event, Amazon showed off several new AI features on its Fire TV line, including an improved Alexa search and generative AI screensavers, along with a new soundbar and souped-up Fire TV sticks. While new capabilities are likely, it’s also possible we’ll see new, more powerful Fire TV hardware, perhaps with Thread and Matter functionality built in, to help power deeper integration between Fire TV, Alexa, and the smart home. We’ll have to wait and see.
No firehose of crazy gadgets
We probably won’t see a slew of new devices. There wasn’t a traditional fall hardware event in 2024. Instead, Amazon has announced a steady flow of new products over the past few months, including new Kindles and two new Echo Shows. And earlier this month, Ring announced its first 2K-capable security camera, and Eero expanded its line of Wi-Fi 7 routers. All of this points to this event being just about Alexa and ways to communicate with the AI.
Alexa for a price
Amazon has said publicly that it’s considering charging for the new Alexa, with reports suggesting a price between $5 and $10 a month. Some have said it will be free for a limited time.
Reuters reported that Amazon could generate $600 million annually if just 10 percent of its users paid $5 per month for the service. Considering that Amazon reportedly lost over $25 billion on its Alexa division, this would be a much-needed boost for the product.
Making Alexa really useful is Amazon’s biggest hurdle
A subscription-based Alexa would be a first for the company, but it’s fairly common with AI services. In the smart home, we’ve already seen features like AI-powered video search from Ring and Google Home behind paywalls.
But will you pay for a better Alexa? If it can deliver on its promises and more — maybe. The new Alexa needs to create enough value for users. One area it can do this is by solving specific problems. For example, I tested the Skylight Calendar, whose AI assistant could manage my household’s calendar for me. It costs $80 a year but is genuinely useful.
Making Alexa really useful is Amazon’s biggest hurdle. It doesn’t have the personal context that competing assistants like Siri and Google Assistant have by being embedded in your phone. If Amazon can find a way to connect to that personal data, combined with the context it has about your home, it could get there. It’s a big if, and Amazon has a huge trust and privacy mountain to climb to get there.
Annapurna Interactive is hosting its first showcase of the year, and it’s an important one. Last year, Annapurna Interactive’s entire staff resigned en masse after negotiations to spin off the respected indie game publisher into its own independent entity broke down. Since then, those employees have formed their own company and even have a new crop of games to manage since it took over publishing duties for Take-Two’s former indie label Private Division. But where does that leave Annapurna?
After the resignations, the company issued a statement assuring players that its games would still be published as normal. With this showcase (which you can watch in full here), we will get more details on what to expect from Annapurna in the first half of 2025.
To a T
To a T is the latest release from Katamari Damacy creator Keita Takahashi, and damn, if it ain’t adorable like most of his games. The game tells the story of Teen, a young person with a… we’ll say “unique” posture, and it’s up to you to help them navigate the challenges that come with being born stuck in a permanent “T” pose. For the showcase, To a T got a special music video a la “It’s Bugsnax!” featuring the vocal stylings of Steven Universe creator Rebecca Sugar. To a T launches on May 28th on Xbox, PS5, and PC, and a demo is available now on Steam.
Morsels
Morsels is a creature collector roguelite with a creepy / cute grossout vibe. Use your powers of transformation to turn into the right monster for the job and battle against evil cat creatures. (Finally, some cozy catgame counterprogramming!) No release date on Morsels yet, but expect it to launch sometime in 2025 on PC and Nintendo Switch.
Wheel World
There’s something about the kinds of presentations that give a glimpse into the people and lives that make these games that really endears me to both the studio and the game. I’m not a bike person, but after seeing the behind-the-scenes presentation for Wheel World, I want to be. Wheel World is an open-world magical cycling adventure where you can ride, race, customize your bike, and ultimately save the world. It launches summer 2025, but you can check out the demo now during Steam’s Next Fest.
Faraway
According to its one-man development team, Faraway can be played with just one hand. In it, you play as a shooting star destined to flicker out but that keeps itself shining brightly by traveling the stars and making constellations. The game consists of a “story” mode that procedurally generates in response to players’ actions and skill levels. There’s also a puzzle mode designed to challenge players in their journey across the stars. Try out Faraway during Steam’s Next Fest or wait for the game’s release on PC later this year.
Geometric Interactive
Geometric, developer of the phenomenal puzzle platformer Cocoon, is back with some lo-fi vibes to carry orbs to. The studio is working with iam8bit to release a collector’s edition of the game as well as Cocoon’s soundtrack on vinyl. The studio also teased that it’s working on a completely new game but, of course, didn’t share any details yet.
Lushfoil Photography Sim
As the name suggests, Lushfoil is all about taking pictures using true-to-life photography techniques. Explore different locations around the world, including the torii gates of Kyoto, Japan, the lush forests of Italy, and even the mountainous Annapurna region in Nepal. Each location is rendered in stunning detail and features photography objectives and collectible cameras that’ll encourage players to strive for the perfect shot. Lushfoil has a demo available on Steam and is slated to launch on April 15th, 2025, on Xbox, PlayStation, and PC.
Wanderstop
Wanderstop is a cozy adventure sim all about tea and navigating difficult life changes, but mostly about tea. Collect ingredients for tea, brew it in an elaborate contraption that looks more like a mad scientist chemistry kit, and thwart the designs of an apparently nefarious bird. Wanderstop is the first game from Ivy Road and is set to launch on Steam and PS5 on March 11th.
Skin Deep
Skin Deephas one hell of a pitch. According to the developers at Blendo Games, the game is essentially “Die Hard with more comedy.” You play as Nina Pasadena, a legendary assassin turned “insurance commando” tasked with saving a starship from a space pirate invasion using whatever tools and weapons she can scrounge along the way. Based on the trailer, Skin Deep feels like a cross between Hitman and Skibidi Toilet, which, honestly, sounds like a neat vibe. You can test that vibe for yourself in the demo or wait for the game’s launch on April 30th.
Sayonara Wild Hearts
Sayonara Wild Hearts was one of the breakout hits of 2019, and it’s back with next-gen updates and a brand-new game mode exclusively for PS5. Remix Arcade allows players to chase high scores in infinitely replayable, progressively faster, and more difficult sound bites with no loading. Sayonara Wild Hearts on PS5 is out now and available as a free upgrade if you have the PS4 version.
Google has announced new searchable transcripts for videos you store in Google Drive. The feature is rolling out to all Google Workspace users, starting today and finishing by March 26th.
The transcripts look a lot like those you’d see on YouTube. If one is available, it’ll appear in a sidebar next to the video as blocks of timestamped text, with each block highlighted as it’s spoken in the video you’re watching. The search bar for the transcript sits at the top of this sidebar, and clicking on specific text blocks will take you to that moment in the video. You can see the transcripts by clicking the settings icon at the bottom of the video and then choosing “Transcript.”
The transcript option is only available if the video already has captions, as indicated by the CC button that shows up in the controls along the bottom of the Google Drive video player. If a video doesn’t have captions, you can create them by right-clicking a video file in Drive, choosing “Manage caption tracks,” and then “generate automatic captions.”
Anthropic is releasing Claude 3.7 Sonnet, its first “hybrid reasoning model” that can solve more complex problems and outperforms previous models in areas like math and coding.
In addition to a new model, Anthropic is also releasing a “limited research preview” of its “agentic” coding tool called Claude Code. While Anthropic already powers AI coding tools like Cursor, it’s pitching Claude Code as “an active collaborator that can search and read code, edit files, write and run tests, commit and push code to GitHub, and use command line tools.”
Claude 3.7 Sonnet is available starting Monday in the Claude app and for developers through Anthropic’s API, Amazon Bedrock, and Google Cloud’s Vertix AI. The model costs the same to run as its predecessor, 3.5 Sonnet, at $3 per million input tokens and $15 per million output tokens.
While OpenAI and others offer separate so-called reasoning models, Anthropic product research lead Dianne Penn tells The Verge that the company wanted to simplify the experience of using a model. “We fundamentally believe that reasoning is a feature of the AI rather than a completely separate thing,” she says, noting that Claude shouldn’t take long to answer the question “What time is it?” versus responding to a more complex prompt like, “plan a two-week trip to Italy while considering the weather in late March.”
Penn says that Claude 3.7 Sonnet performs noticeably better on “agentic coding,” finance, and legal tasks. While Claude still lacks real-time web search like other models, version 3.7’s knowledge cut-off date of October 2024 is more up to date. Anthropic is also allowing developers to help steer how the model “thinks” via its scratchpad and even dictate exactly how long it takes to respond. “Sometimes the developer just needs to say it shouldn’t take more than 200 milliseconds to answer this question,” says Anthropic’s VP of product, Michael Gerstenhaber. “And that’s a product decision.”
Inside Anthropic, employees have used the new model to build front-end website designs, interactive games, and even spend up to 45 minutes on coding work by “building test sets and editing test cases back and forth iteratively,” according to Penn.
She says that the company also tests its models on their ability to advance through an old-school Pokémon video game by mapping the model’s API to a controller scheme. Claude 3.5 Sonnet couldn’t get out of Pallet Town at the beginning of the game while version 3.7 was able to defeat multiple gym leaders.
As Elon Musk showed with Grok-3 last week, the AI model race is moving incredibly fast. For now, Anthropic appears to be in the lead again thanks to Claude 3.7 Sonnet’s performance gains. Its release also suggests that, rather than offer standalone reasoning models, the industry is moving toward a future where one model can do everything.
Along with a new Andor trailer showcasing some of the action Cassian (Diego Luna) is going to get into in the show’s second and final season, Disney announced today that the upcoming episodes (there are 12 total) will be released as four weekly chapters. In a statement about the new season, creator and executive producer Tony Gilroy teased that that, because it follows Cassian’s adventures during the four years leading up to Rogue One, the show is going to feel even bigger this time around.
“One of the great thrills of making Andor is the scale of the story and the number of characters we’re able to meet — ordinary people, Imperial overlords, passionate revolutionaries,” Gilroy said. “They are real people making epic decisions, all of them staring down questions with terrifying consequences. Cassian’s journey is the soul and spine of our story, but it’s the choir that makes the show.”
The trailer definitely makes Andor’s second season seem like its stakes are going to be raised to meet its scale as Cassian marches towards a fate we’ve already seen play out. The weekly episode was part of what made the first season work so well, but these new batch drops might be a refreshing change of pace when Andor season 2 premieres on April 22nd.
The Ember Mug 2 is costly for a drink receptacle, but it may be the last one you buy.
Warm beverages are a winter staple, particularly when the flu is making the rounds. Keeping your favorite drinks warm is a battle if you’re not one to chug; however, the Verge-favoriteEmber Mug 2 lets you sip and savor your beverage of choice at the ideal temperature all day long. And right now, Amazon and Best Buy are selling the 14-ounce smart mug in select colors starting at an all-time low of $89.99 ($60 off).
The Ember Mug 2 hides heaters in a stainless steel body to keep your drinks uniformly warm at any temperature you want — so long as that temperature is between 125 degrees to 145 degrees Fahrenheit. It can last up to 80 minutes per charge, but it also runs indefinitely if you set it on the included charging coaster, which has a configurable LED to indicate its charging status. You can also use Ember’s app on iOS or Android to set the temperature and create presets for your favorite drinks.
Washing is a cinch, too, thanks to its IPX7 rating, which means you can dunk it in up to a meter of soapy water, with a scratch-resistant coating that should keep it looking neat. It’s not dishwasher safe, however.
Monday’s deals continued
If you’re running out of storage space on your Xbox Series X / S, the 2TB WD_Black C50 is on sale for $199.98 (about $260 off) at Amazon, B&H Photo, and SanDisk’s online store, which is about $10 more than its best price to date. If you’re working with a smaller budget, you can also pick up the 1TB version for $109.99 ($48 off) from Amazon, B&H Photo, and SanDisk, which is also $10 more than its record low. Unless you’re an exceptionally heavy gamer, either should be enough to harbor your daily rotation of Series X / S games. Western Digital’s officially licensed expansion cards are one of two that are fast enough to play those games directly from your console’s storage, so it’s worth the investment if you’re tired of constantly redownloading or moving games around.
Prime members can get the Shargeek 170 Power Bank for $99 ($69 off) at Amazon and directly from Sharge, which is an all-time low. The 24,000mAh charger is doubly unique in appearance, with a prism-shaped exterior and a translucent window that exposes some of its electronics. That window also offers a clear view of the dot-matrix LED display that communicates charging status and capacity. It has two 140W USB-C Power Delivery 3.1 ports and a single USB-A port. All three are exposed, yet the charger bears an IP66 rating for protection against dust and water, making it an ideal option for those who frequent the great outdoors.
Now through February 28th, Woot is offering steep discounts on a host of popular games and quality accessories, including a number of first- and third-party Nintendo Switch titles. You can get the excellent Super Mario RPG for an all-time low of $29.99 (50 percent off), for example, along with Donkey Kong Country Returns HDfor $44.99 ($15 off). 8BitDo’s Ultimate 3-mode Controller — which works wired on Xbox and wirelessly on Windows and mobile — is also down to $39.99 ($30 off) right now. It’s the newest version with Hall effect sticks and triggers, too, and ships with a charging dock.
Balatro, the hit roguelike with poker imagery and poker hands but no actual gambling, no longer has a PEGI 18 rating in Europe following an appeal, PEGI announced today.
LocalThunk, the creator of Balatro, has publicly shared his frustrations the rating — the game was originally given a PEGI 3 rating, but it was bumped to PEGI 18 shortly after its launch in February 2024 because of the game “contains prominent gambling imagery and material that instructs about gambling,” as explained by publisher Playstack.
Now, Balatro will have a PEGI 12 rating following an appeal submitted by Sold Out Sales & Marketing, which rebranded to Fireshine Games in 2022 and publishes Balatro’s physical edition. “The Complaints Board concluded that, although the game explains the various hands of poker, the roguelike deck-building game contained mitigating fantastical elements that warranted a PEGI 12 rating,” PEGI says.
PEGI is also changing the PEGI 18 rating for Luck Be A Landlord, a roguelike with slot machine mechanics but similarly has no real-world gambling, to PEGI 12 an appeal submitted by Fangamer. “The Complaints Board reasoned that the same arguments as above applied: although the game features a slot machine mechanic, there were no specific transferable gambling skills and the game can therefore be rated PEGI 12.”
PEGI is going to make more classification criteria to address games with gambling imagery but that may not need the PEGI 18 rating. “At this moment, any teaching or glamorisation of simulated gambling automatically leads to a PEGI 18 rating,” PEGI says. “On the basis of these appeals, the PEGI Experts Group will develop a more granular set of classification criteria to handle gambling themes and the simulation, teaching and glamorisation of gambling in different age categories, which will now include 12 but also keep 18 as an age category for games that simulate gambling typically played in casinos and betting halls.
“This is a good step from PEGI”
“After an appeal from my publisher, Balatro has been reclassified by PEGI from 18 plus to 12 plus,” LocalThunk says in a post on Bluesky. “This is a good step from PEGI — bringing nuance to their ratings criteria that used to be 18 plus or nothing. I hope this change will allow developers to create without being unfairly punished.”
“Playstack is pleased to confirm the PEGI Complaints Board’s decision to reclassify Balatro to a PEGI 12 rating, following appeals from Playstack and our physical distributor, Fireshine Games,” Playstack communications director Wout van Halderen says in a statement to The Verge. “We applaud the PEGI Experts Group on their commitment to develop a more granular set of classification criteria as indie games continue to stretch into new ideas and concepts. We believe it is a sign of a healthy classification board that it evolves in line with cultural expectations and continue to ensure that audiences understand the type of content present in games.”
“I’m happy that PEGI finally made the right call here and lowered Luck be a Landlord’s age rating,” says Dan DiIorio, the solo game developer at TrampolineTales, in a post on Bluesky. “I just hope that this sets a standard and other games are treated the same way Balatro and Luck be a Landlord were. Now give games with loot boxes (actual gambling) higher age ratings!”
Update, February 24th: Added statements from LocalThunk, TrampolineTales’ Dan DiIorio, and a revised statement from Playstack.
Google is planning to move away from sending six-digit authentication codes through SMS messages as a two-factor authentication tool for Gmail, Forbes reports. Instead, over the next few months, QR codes will be rolled out as a replacement for SMS codes in an attempt to “reduce the impact of rampant, global SMS abuse,” Gmail spokesperson Ross Richendrfer told Forbes.
Google uses SMS codes to both verify that it’s dealing with the same person who created or owns a Gmail account, and as a deterrent to prevent criminals from creating “thousands of Gmail accounts in order to distribute spam and malware,” Richendrfer says.
But while SMS codes are better than using no two-factor authentication at all, the approach comes with its own security risks. Criminals can trick or force users into sharing a code that’s been sent over SMS messaging, and users may not have immediate access to the device receiving the codes. It’s also dependent on each carrier’s own security practices, and how diligent their support teams are at preventing fraudsters from getting access to someone else’s mobile number.
The move will also help Google avoid a more recent scam called traffic pumping or toll fraud. “It’s where fraudsters try to get online service providers to originate large numbers of SMS messages to numbers they control, thereby getting paid every time one of these messages is delivered,” Richendrfer and Google’s Kimberly Samra explained to Forbes.
Once the change has been implemented, instead of verifying phone numbers by sending a six-digit code, a QR code will be presented that users can scan using the camera app on their smartphones. It will eliminate the risk of users being tricked into sharing codes since they no longer exist, and it takes security risks introduced by phone carriers, including unwanted SIM swapping, out of the equation entirely.
Dan Bongino at the 2022 Fox Nation Patriot Awards. | Photo by Jason Koerner/Getty Images
President Donald Trump announced Sunday night he has named conservative podcaster Dan Bongino as the FBI’s new deputy director. Bongino has worked as a New York police officer and US Secret Service agent, but has no FBI experience.
Bongino is well-known in conservative media, thanks to credits like having hosted a Fox News show and contributed to conspiracy website InfoWars. (Another one-time Fox News presenter, Pete Hegseth, serves as the US Secretary of Defense.)He was also among the right-wing commentators whose popularity sparked a fight over Facebook’s CrowdTangle analytics tool, ultimately resulting in its shutdown. His outsized onlineinfluence saw him named one of the top election conspiracy “misinformation superspreaders” in research by global human rights group Avaaz, The New York Times reported in 2020.
The new acting director currently hosts Rumble podcast The Dan Bongino Show. Trump said in his Truth Social announcement that podcasting is “something he is willing and prepared to give up in order to serve.” On his podcast today, Bongino said he’s leaving the show, calling it “a lot to walk away from.”
