Look Outside is an unexpected cosmic horror masterpiece that shook me to the core

What kind of person would you be in the face of a disaster that’s beyond explanation? Would you help distressed neighbors and open your home even to sketchy-looking strangers in the hope that there’s strength in numbers, or go it alone regardless of how heavily the odds are stacked against you? Would you still brush your teeth every day?

The horror in Francis Coulombe’s RPG Look Outside is all encompassing. There is the cosmic element: something incomprehensible is happening beyond the walls of your apartment building and it’s in your best interest to not even look outside, let alone go there. And as you quickly learn, anyone who has looked or been outside is transformed in unimaginable ways, making for some extreme (and extraordinarily creative) body horror. But in much of the game leading up to its multiple climactic endings, the unease also stems from how it makes you look inside — at the choices you’ve made and the person you’ve become in order to survive.

Despite trying to take an empathetic approach, I still found myself in situations that left my character (and me) wracked with guilt. There is a gnawing sense of doubt that grows over the course of the game, repeatedly making me question whether I’d, say, made a bad call and been too quick to kill that neighbor whose entire head is teeth, or if my hand was really forced into making a very upsetting sacrifice.

After a while, the most unnerving thing is looking in the mirror. And you have to do that a lot in this game, because hygiene affects your stats.

A still from the game Look Outside showing the protagonist, Sam, reflected in a mirror above a sink. He looks haggard and unsettling
Francis Coulombe/Devolver Digital

It all begins with your character, Sam (you can change the name, if you want), waking up after a strange dream with a strong urge to look outside. You are immediately given the choice between satisfying your curiosity and listening to your gut, and you’ll find yourself grappling with that dilemma time and time again.

At this point, you also meet Sybil, the mysterious next-door neighbor who only speaks to you through the wall, with one glaring eyeball peering out of a large crack. Sybil, whom it’s unclear if you should trust, tells you that everything will blow over in 15 days if you just wait it out. You need to scavenge for resources if you’re going to make it that long, though, and once you leave your apartment and get a chance to talk to some other neighbors, you may decide you want to take a more active role in getting to the bottom of the catastrophe.

Some neighbors, particularly a few robed amateur astronomers who appear to be in a cult, seem to know quite a bit about what’s going on, and it’s insinuated that doing tasks for them will help you figure out the how and why of the phenomena around you. Others are more focused on addressing their immediate needs and will try to rope you into their causes: locating missing people, picking up laundry, cleaning the messes left behind by eldritch horrors, etc. There’s a full on war happening somewhere in the building, which you can choose to play a part in if you’re so inclined. Your landlord will unsurprisingly still demand you pay him rent despite the circumstances.

If you choose to play in Normal mode, like I did, Sybil is your only save point, so you’ll have to return home regularly. Easy mode autosaves.

All interactions are turn-based, and as you explore the apartment building, you’ll run into tons of enemies and potential allies — but the line between the two isn’t always easy to distinguish. Sometimes you can only attack or try to escape, which answers the question for you, but other times, you have the opportunity to talk and ask questions. The turn-based nature allows you to take a moment and evaluate each new encounter, but there’s always an air of ambiguity about everyone’s trustworthiness. Even when you’re back at your apartment, where you can shower, rest, do some cooking and crafting, and play video games, people will come along and knock on your door, and you’ll have to make up your mind about whether you should let them in.

The thing is, surviving can be pretty difficult once you really get going if you’re on your own. Enemies will outnumber and overpower you. That’s where it becomes helpful to have a few allies. With as many as three other people in your party, the scale tips heavily in your favor. I took the trusting approach, for better or worse. This resulted in me having a pretty solid group of fighters on my side, but a pair of those same allies kicked me out of my own bedroom and complained about my cooking.

The creature designs come disgustingly, beautifully alive in the pixel art style. Body horror can be really hard to stomach, and something is often lost for me in the process of consuming it when it’s intended to seem realistic — either because I’m hiding behind my hands and only taking tiny peeks through my fingers, or because it ends up achieving the opposite effect and just looks ridiculous without meaning to. But Coulombe’s art equally embraces the horrifying and the absurd, and the effect of that balance is powerful.

Nothing was ever so disturbing that I couldn’t look straight at it, but there were certainly moments that gave me a genuine scare or made my skin crawl. Even the characters that aren’t being transformed, like the protagonist, look a little grotesque, which adds to how unsettling everything feels. But just when the dread would reach a fever pitch, something overtly silly would be thrown in almost as if to splash some cold water on the whole thing and say, yes this is the apocalypse but we’re still human, we still have a sense of humor.

A still from the game look outside showing turn-based combat. A monster with multiple arms pulls its skull-head apart like an accordion with teeth, eyes and flesh stretching between the two halves
Francis Coulombe/Devolver Digital

So much of the joy of playing this game is discovering all the tricks it has up its sleeve, so I won’t get into any descriptions of bosses, puzzles or the building itself, other than to say that the latter has a whole House of Leaves thing going on that is unbelievably frustrating at times, but in a way that only adds to the brilliance of it all. There is no map to guide you, either. The soundtrack, composed by Eric Shumaker, keeps in perfect step with every emotion the environment evokes, and I could probably write an entire separate review about how good it is.

All of this builds up to an absolute cosmic gut-punch of an ending (or endings, there are several) that completely changed the way I felt about the game up until that point. In the end, it becomes something much, much bigger than it once seemed, and the feelings were almost overwhelming. I can’t stop thinking about it.

By now I’ve played Look Outside many hours beyond what a typical run would be, just picking apart every detail and turning over every stone to try and figure out all the secrets, reach all the conclusions. I have died in all sorts of strange ways, and lived to see wildly different fates pan out.

I went into this only with the expectation of cool art and a relatively unique approach to survival horror, and came away shook from what turned out to be one of the best cosmic horror games I’ve played in a while, maybe ever. Look Outside, published by Devolver Digital, is only available on Steam for now, but I sure hope it makes its way to other platforms soon so more people can experience it.

This article originally appeared on Engadget at https://www.engadget.com/gaming/look-outside-is-an-unexpected-cosmic-horror-masterpiece-that-shook-me-to-the-core-171542211.html?src=rss

Anthropic’s Max Plan offers nearly unlimited Claude usage for $200 per month

Anthropic is joining the ranks of OpenAI in offering a more expensive tier of its flagship chatbot. On Wednesday, the company announced Max Plan. Starting today, you can either pay $100 or $200 per month to use Claude up to 5x or 20x more than you can with Anthropic’s existing Pro plan. The company told Engadget it’s introducing the Max tier in response to the popularity of Claude 3.7 Sonnet. The new hybrid reasoning model, which excels at coding tasks, has been so popular with users, many are asking to use it as much as they want.

With tokens for Claude 3.7 Sonnet costing significantly more than what Google and others charge for access to their lightweight models, Anthropic isn’t quite ready to offer unlimited usage of the new model. However, you’ll find even OpenAI’s $200 per month Pro plan has its own set of restrictions, with monthly limits imposed on “very compute intensive” features like Deep Research.

Part of the reason Anthropic is offering two different price points at the Max tier is so that it can give flexibility to customers as their needs change. Additionally, those same users will gain priority access to new models and capabilities as Anthropic releases them.

This article originally appeared on Engadget at https://www.engadget.com/ai/anthropics-max-plan-offers-nearly-unlimited-claude-usage-for-200-per-month-170032710.html?src=rss

Possessor(s) is a game about hot demons and other things

Heart Machine knows how to make a stylish game, so it’s no surprise that the studio’s latest project, Possessor(s), is sexy as hell. I mean, just look at Rhem, the cosmic demon with a chiseled jawline, aquamarine horns and a fine-cut suit unbuttoned down to his navel, who alternately encourages and negs you throughout the game. Stupid sexy Rhem.

“It’s just fun to have hot characters at the end of the day,” Heart Machine founder Alx Preston said at GDC 2025. Preston and three other Heart Machine developers — producer Myriame Lachapelle, narrative director Tyler Hutchison and writer Laura Michet — played through Possessor(s) while we chatted about the game and the studio.

Hutchison continued the hot-characters conversation by praising Heart Machine concept artist Sophie Medvedeva and adding, “Since we were kind of building out this visual novel, I was trying to leverage the expertise I brought from Dream Daddy to be like, well, let’s make sure these characters are very hot.”

Sometimes, it’s as simple as that.

Possessor(s) is a sidescrolling sci-fi action game set in the ruins of Sanzu, a corporate megalopolis that’s been ripped apart by interdimensional horrors. Sanzu was once a company town controlled by the tech firm Agradyne, but three years ago, a flood of demons escaped the city’s underground laboratories, possessing its citizens and transforming most of them into vicious, mutated creatures. The game spans desolate skyscrapers and abandoned aquariums, and it looks like a daydream of a 1990s anime, splattered in neon and shimmering with a sickly green VHS aura.

It stars Luca, a possessed-but-lucid 16-year-old who’s searching the urban wasteland for any signs of her mother and best friend. The “hot” descriptor in this game is reserved for the demons, but Luca is stylish, with an iridescent coat, slender metal legs and two horns of her own. Luca is bursting with teenage angst and she has an uneasy relationship with Rhem, the demon who’s invaded her consciousness. He’s the only thing keeping her alive, and he needs her in order to finally return home.

“Luca is missing her mom and her best friend, who both got lost when the demons escaped to destroy the city,” Michet said. “So she’s resentful Rhem won’t let her take some time off to find all the people important to her.”

Possessor(s)
Heart Machine

Together, Luca and Rhem explore Sanzu, investigate the true nature of Agradyne and search for Luca’s loved ones. Levels are littered with evil creatures to battle and puzzles to solve, and throughout the game, Luca masters an arsenal of improvised weapons like kitchen knives, a guitar, a hockey stick and a corded computer mouse. Her whip, a power provided by Rhem, allows her to swing across gaps, rapidly reposition and pull in heavy objects and enemies.

Combat in the demo seemed to flow smoothly; it was heavy on visual flourishes, leaping airborne attacks and satisfying slashing moves, with Luca bounding seamlessly across floors, walls and platforms. A few frames of line-drawn placeholder art flashed across the screen at one point, but these were the only indication that Possessor(s) isn’t actually complete yet. It’s due to hit PC and PS5 this year.

“The original premise was something more like Smash Bros., where you can do directional inputs and kind of build out your loadout to feel more like a different character,” Preston said. “Where we’ve ended up is still a bit Smash Bros.-y, in that there’s juggling and bouncing enemies on surfaces and things like that. But we’ve found the spirit of Luca and her character, and what she wants to do.”

Possessor(s)
Heart Machine

Possessor(s) is eye-catching and combat-focused, and it also has a narrative that’s supremely relevant for our times. Boiled down, the premise is that capitalistic megacorporations with an iron grip on society are bad, and young people fighting the system are rad. Interdimensional beings with chiseled pecs and a devastating smoking habit are the hot twist that makes the story uniquely Heart Machine’s.

“I mean, it sucks to live in a hierarchical corporate society where your value is determined by your value to the business,” Michet said. “And Luca will discover this from her teenager perspective.”

The story in Possessor(s) is branching, with multiple paths to follow and a quarantined city of characters to meet. This is a new mechanic for Heart Machine — in fact, the entire game is composed of firsts for the studio, and Possessor(s) is purposefully different from its previous games, Hyper Light Drifter, Solar Ash and Hyper Light Breaker. Hyper Light Drifter‘s moody, 16-bit-inspired isometric action put Heart Machine on the map even before its release in 2016. Solar Ash followed in 2022 and it was a third-person, 3D platformer. Hyper Light Breaker hit the market this January as a third-person, online roguelite with stylized 3D graphics. Meanwhile, Possessor(s) is a hand-drawn, 2.D side-scroller with touches of horror and visual-novel mechanics.

Possessor(s)
Heart Machine

This progression aligns with a mentality of curiosity that Preston outlined to Engadget in 2021. At the time, Heart Machine had fewer than 30 developers and they were still building Solar Ash, unsure of how the switch to 3D would be received. Back then, Preston said the following:

“Hyper Light was a way for me to get started, it was a way for me to be more grounded and put together a crew and understand, can I do this? Can I actually make games? And so having answered that question, then the natural next step for me was something in 3D. Can I put something out there that really opens up the world and makes you feel like you can truly escape into something, a creation that is otherworldly, that you otherwise wouldn’t have the experience of?”

Since 2021, Heart Machine has endured layoffs and launched a Patreon, and today it has 58 people on staff. Preston remains at the studio’s core, still testing new concepts.

At GDC 2025, he said, “We’ve done overhead stuff with Drifter back in the day, and then fully 3D was a big jump. I wanted to do something sidescrolling, because there’s a lot of different, unique challenges there that are fun to tackle. That perspective for how we do art, and mixing our 3D stuff with a 2D plane, there’s a lot of cool opportunities there. Something like Inside was a good point of inspiration, seeing what a smaller team can do with atmospheric chops. And we’re pretty good on those fronts.”

Possessor(s) is due to hit PC and PS5 in 2025, published by Devolver Digital.

This article originally appeared on Engadget at https://www.engadget.com/gaming/possessors-is-a-game-about-hot-demons-and-other-things-170016014.html?src=rss

The Last of Us will return for a third season

Fret not, Fireflies, The Last of Us will continue on HBO beyond its second season. Just before the hit adaptation of Naughty Dog’s games returns to the network and Max for its next batch of episodes, HBO has officially greenlit a third season of the show.

It would have been a major shock if TLOU wasn’t going to return beyond this latest seven-episode stint. Season two doesn’t cover all of the shocking events of The Last of Us Part 2. The show’s creators and showrunners, Neil Druckmann of Naughty Dog and Craig Mazin, have long been adamant that it would take at least two seasons to adapt that overly long misery simulator (still a very good game, though). Plus, the first season was a huge success, as tens of millions of people watched Joel and Ellie fend off infected and other enemies as they trekked across the remnants of America.

“We approached season two with the goal of creating something we could be proud of,” Mazin said in a statement. “The end results have exceeded even our most ambitious goals, thanks to our continued collaboration with HBO and the impeccable work of our unparalleled cast and crew. We look forward to continuing the story of The Last of Us with season three!”

HBO hasn’t given a timeline for when we can expect the third season, but there should hopefully be a smaller gap between seasons this time around. Season two was delayed by the Hollywood writers’ and actors’ strikes in 2023. Plus, it’s hard to imagine that viewers will be willing to wait so long to find out what happens after [redacted].

Speaking of which, season two of TLOU premieres on HBO and Max this Sunday, April 13. Engadget’s Nathan Ingraham reviewed the season, and found it did an admirable job of adapting (at least some of) the non-linear source material while resolving certain issues from the first season.

This article originally appeared on Engadget at https://www.engadget.com/entertainment/tv-movies/the-last-of-us-will-return-for-a-third-season-163801196.html?src=rss

All the Companies That Have Paused US Sales Due to Tariffs (So Far)

“Tariff” might be the defining word of 2025. The Trump administration’s crusade to tariff trading partners around the world has upended markets and shaken trust in the United States. The situation isn’t resolving anytime soon, either: Following the U.S.’s decision to impose a 104% tariff increase on Chinese goods, China reciprocated by raising its tariffs on American goods from 34% to to 84%.

In short? It’s a mess, and no one can predict where we go from here. Markets, of course, loathe uncertainty, so when the global economy gets this shaky—especially when that instability is the direct result of one administration’s actions—companies act to protect their interests.

If you, like me, are an American, it’s our country that’s throwing these tariffs at nations around the world. As such, it shouldn’t come as a huge surprise to see companies pause sales and business in the U.S. for the time being. While there aren’t a ton of organizations running with this playbook, there are some notable names that have done so already.

Nintendo

Last week was like Christmas Day for us Nintendo fans. The company spent an hour walking gamers through the new Switch 2, announcing new games, upgraded play styles, and, for the first time, in-game chat (including video calls).

Switch 2 preorders were scheduled to start April 9, and seeing as the console has been hyped for years, preordering the console would be your best chance of picking one up.

That’s no longer the case. Following the tariffs, Nintendo announced Switch 2 preorders were delayed in the U.S., though the console’s official June 5 release date remained unchanged. It’s not clear if or when Nintendo plans to make preorders available to Americans, or whether the company will hike the price of the Switch 2, but seeing as the console alone starts at $450, any increases will make this an expensive handheld.

Framework

Framework makes laptops you can easily mod and upgrade yourself—the antithesis of most tech companies these days. Unfortunately, U.S. customers are going to have a hard time getting these machines in the near future.

Following Nintendo’s lead, Framework announced on Tuesday a pause on sales for specific laptops. The company says that, following the April 5 tariffs, they will not sell some of the base-model Framework Laptop 13 machines in the U.S., including the Ultra 5 125H and Ryzen 5 7640U models.

Then on Wednesday, the company announced a pause on U.S. preorders for the new Laptop 12. Framework is also holding back on announcing a price for this machine, as it will likely cost more than originally intended.

Razer

Razer, a major laptop manufacturer, has taken things even further: The company has essentially stopped all laptop sales to the U.S., and has paused sales on some of its accessories as well. If you browse the various laptops for sale on the company’s site, you’ll only find “Notify Me” buttons, rather than any option to purchase.

You can still purchase Razer laptops from other vendors, but I imagine those machines could sell out quickly given the company’s sales pause, or even jump in price.

Jaguar Land Rover Automotive

If you’re in the market for a luxury vehicle, you might want to hold off on buying a Jag or a Land Rover. Jaguar Land Rover Automotive is pausing shipments of vehicles to the U.S., following Trump’s 25% import tax on vehicles entering the country.

Basic Fun

You might not know the name Basic Fun, but you probably know its products. The company makes toys like Care Bears, Lincoln Logs, K’nex, and Tonka Trucks, but has paused shipments of its products that are made in China. The company’s chief executive, Jay Foreman, told the New York Post that he “cannot risk putting any product on the water that might incur [high] tariffs as implemented and threatened by the president.”

Swiss watches

The Post also reports that Swiss watchmakers, such as Audemars Piguet, Breitling, and Rolex, are pausing shipments as well. For context, Trump imposed a 31% tariff on Switzerland.

Small businesses

The big-name businesses like Nintendo will make headlines, and will be easy to track when they pause sales or preorders to the U.S. But what’s more difficult to identify are the number of small business pausing or stopping their sales due to these tariffs.

In general, these tariffs are going to hit the smaller players in our economy hardest. These companies don’t have the capital to either pay the higher fees on goods they’ve purchased from other countries, or to stop sales while they figure out their next moves. They’ll have to raise prices to stay afloat, but then contend with consumers who may be hesitant to shop, since, well, everything is getting more expensive.

Many of these companies have placed orders before the tariffs went into effect, but by the time the goods arrived in the U.S., multiple rounds of tariffs ballooned the fees. Sarah Wells, founder of Sarah Wells Bags, owed an additional $15,000 just to get her goods from customs. On the other hand, Emily Ley, owner of Simplified, has sued the Trump administration over these tariffs. Ley says the tariffs will cost her an extra $630,000 in fees this year.

This is an ongoing situation. I will continue to add companies to this list as they announce pauses in sales and business.

World Record Eggshell Carvings

The Guinness World Record for largest display of carved eggs was recently awarded to Chinese artist Ma Chao for the 1,040 intricately carved eggshells he’s produced. The eggs feature characters from the 1986 television series Journey To The West (adapted from Wu Cheng’en’s 16th-century novel of the same name), and Chao carved one egg a day for roughly the last four years. I wonder how many breaks he had during that time. Both eggs and mental. Personally I had at least a couple dozen of each in that time.

You Can Turn Off Apple’s New AI Features in Mail for Mac

With macOS Sequoia 15.4, Apple fully rolled out all of its new Apple Intelligence features to the Mail app. As such, when you open Mail after updating to the latest OS, the app will look quite different. You’ll see new “Mail Categories,” a Priority Messages box, and email summaries, the latter two of which are powered by AI. It’s not something many asked for, and you might be longing for the Mail app you’re used to. But there’s good news here: If you don’t like these features, you can easily disable or customize them to suit your needs.

How to disable Mail Categories on Mac

Mail Categories on Mac.

Credit: Khamosh Pathak

Apple first added Mail Categories to the iPhone back in iOS 18.2. After a few rounds of updates, it’s now front and center on the Mac, too. This feature automatically sorts all your emails into four neat boxes: Primary, Transactions, Updates, and Promotions.

Apple’s positioning for this feature is a bit weird. While Apple is using on-device processing to sort out these emails, it’s not actually an Apple Intelligence feature. So, turning off Apple Intelligence on the Mac won’t instantly get rid of this feature, either.

The idea is that Apple does all the heavy lifting to figure out which emails are most important and useful, and presents them to you in the Primary box. It mostly works, but it isn’t always accurate.

Disable Mail Categories on Mac.

Credit: Khamosh Pathak

If you want to see all your emails at any given time, swipe right on the trackpad or Magic Mouse while hovering on the Categories tab to switch to the All Mail tab. This tab presents your messages in chronological order, as you’re used to.

But that might not be enough. If you want to disable the Categories view altogether, click the three dotted Menu button at the top of the Mail list, then uncheck Show Mail Categories. After that, the Mail list will be back to how it used to be.

How to recategorize email

In case you do like this feature and want to keep it around, here’s a tip for when things go a bit wrong. In case the Mail app incorrectly categorizes an email, you can take a bit of manual control and point the app in the right direction.

Categorize Sender in Mail.

Credit: Khamosh Pathak

Right-click on an email in the Mail list, then choose Categorize Sender. Here, choose the category where you want the email to end up in.

How to disable Priority Messages

When you receive a time-sensitive email, or something that Apple deems significant (like an email from your company, or your family member) it will show up in a new Priority box on top of the Mail list.

Disable Priority Mail box.

Credit: Khamosh Pathak

If you don’t like this feature, there’s a quick way to disable it as well. Click the three-dotted Menu button on top of the Mail window, then uncheck Show Priority Messages.

How to disable AI summary previews

AI Summary in Mail app.

Credit: Khamosh Pathak

Apple now provides a short summary for each email instead of showing you the first couple of lines. This too, is hit or a miss. Just like Notifications Summaries on iPhone, these Mail summaries can sometimes go horribly wrong, as the app changes the meaning of the message completely.

Disable AI Summary in Mail for Mac.

Credit: Khamosh Pathak

If you don’t like this feature, here’s what to do: Go to Mail > Settings (Command + Comma) from the menu bar, and switch over to the Viewing tab. Here, uncheck Summarize Message Previews to get rid of Mail summaries.

How to disable all Mail AI features

Lastly, there’s the nuclear option. To turn off all AI features in Mail, you can simply disable Apple Intelligence on your Mac. In one fell swoop, you can get rid of the Priority mail box, AI summaries, and the Summary button in emails, as well as hidden features like Apple Intelligence Writing Tools and Smart Replies options.

Disable Apple Intelligence on Mac.

Credit: Khamosh Pathak

To do this, open System Settings, go to Apple Intelligence & Siri and disable the Apple Intelligence toggle. From the popup, click Turn Off Apple Intelligence to confirm.

Of course, when you do this, you’ll also miss out on the new Siri design, Writing Tools, Gemmoji, Image Playground and more. But if you don’t care for these AI features anyway, better to just keep them off completely.

My Favorite Amazon Deal of the Day: The Bose SoundLink Flex

We may earn a commission from links on this page. Deal pricing and availability subject to change after time of publication.

Bose makes speakers that have been dominating their space since their release—and when it comes to portable mid-size rugged speakers, the Bose New Soundlink Flex has been one of the best since its release in 2021, and is still very competitive three years later. Right now, you can get it for $119 (originally $149), the lowest price it has been according to price-tracking tools.

The Bose New SoundLink Flex is the middle sibling between the oldest (but smallest) Bose SoundLink Micro, and the youngest (but biggest) Bose SoundLink Max. The Flex also strikes a good middle ground between them as far as portability and sound, making it the best portable speaker for most people.

The Flex has an IP67 rating, which means you get complete water and dust-proof protection. The speaker even floats in the water, so you can bring it with you in the pool. While it doesn’t have full EQ customization in its companion app, the sound it offers is clear with rich bass depth and clear highs out of the box, according to PCMag’s review. It also has a speakerphone, so you can take calls with it and even speak to it, thanks to its microphone. Although it’s older, it comes with a USB-C charger, so it will charge quickly. At 3.6 by 7.9 by 2.1 inches (HWD) and 1.3-pound, it is light and portable. The Bluetooth 4.2 supports SBC codec, but not AAC or AptX, so it is more suitable for Android users than for Apple users, but either works.

The SoundLink Flex is perfect for those who want to take their speaker with them anywhere, thanks to its small size and light weight. And for the price, it offers some of the best audio available.

Instagram is testing Reels that are locked behind a secret code

Instagram is testing a feature that locks Reels behind a secret code and a provided hint, according to a report by TechCrunch. This looks like a novel way for creators to interact with fans, as they can share exclusive reels with people who are more likely to know the answer to these clues.

As an example, a creator could lock a Reel with a hint like “my favorite food” and rabid fans would, of course, know the answer. The Meta-owned social network has been testing the feature on its Design account. It shared a locked Reel that prompts viewers to enter a secret code based on a hint that read “1st # in the caption.”

A design post on Insta.
Meta / Screenshot from Engadget

The first hashtag mentioned in the caption was “Threads.” Upon entering the correct code, viewers access a Reel that announces that the Design account is launching a profile on Threads. 

This could be used by brands to advertise new products or by creators trying to amp up engagement. Personally, I could see it as a place to hide in-jokes with friends, but maybe that’s just me.

Engadget has reached out to Instagram for more information regarding this feature. We’ll update this post if we find out anything regarding availability. This could be just a test, with no actual plans for a general rollout.

This is also quite similar to another Instagram tool called Reveal. This feature lets creators post a hidden Story for followers to uncover by sending a DM.

This article originally appeared on Engadget at https://www.engadget.com/social-media/instagram-is-testing-reels-that-are-locked-behind-a-secret-code-162142807.html?src=rss

Sony adds three more speakers to its ULT lineup, bass boost button included

Last year, Sony rebranded its portable speaker lineup under the ULT umbrella with new models classified as either Field or Tower. The former are the more “traditional” Bluetooth options while the latter are the party box, karaoke machine sort of devices. Just like 2024, the company has three new entries in the ULT lineup this time around with the ULT Field 3, ULT Field 5 and ULT Tower 9. Those first two fill the mid-range gap between the current ULT Field 7 and ULT Field 1 while the last item is a smaller alternative to the ULT Tower 10.

The ULT Field 3 will replace the SRS-XE300. As the name suggests, Sony’s ULT button is among the updates, a control that adds instant access to ULT1 mode for bass boost. The company also expanded the customizable EQ and added its Sound Field Optimization. Inside, a single woofer and a tweeter are angled slightly in the center while two passive radiators are positioned at the ends of the speaker. The ULT Field 3 works in both vertical and horizontal orientations where the XE300 was only vertical.

There’s a new Party Connect button to easily sync with other Sony speakers and you can charge your devices via the USB-C port. The company kept the 24-hour battery life the same as the XE300 but improved the quick-charge feature to give you two hours of use in 10 mins. The ULT Field 3 remains IP67 waterproof, dust proof and shock proof like its predecessor, however this new speaker is also rust proof. A removable shoulder strap is among the additions too, but there’s one major item Sony retained. The ULT Field 3 is the same price at the XE300 at $200. 

Sony's ULT Field 3 is a mid-sized Bluetooth speaker with bass boost and party mode buttons.
ULT Field 3
Sony

The ULT Field 5 isn’t quite at large as the ULT Field 7, but it’s still hefty. This speaker will replace the SRS-XG300 in Sony’s arsenal, clocking in at $330 — $20 less than that predecessor. Here you’ll get a single woofer and two tweeters angled in the middle of the unit with passive radiators at either end. You’ll also be able to use a second ULT mode, ULT2, that “emphasizes powerful sound” in addition to the ULT1 bass boost. A 10-band EQ and Sound Field Optimization help maintain sound quality and that new Party Connect button is on this larger speaker too. Like the ULT Field 3, you can use the ULT Field 5 in either vertical or horizontal positions, and similarly there’s a removable shoulder strap. 

Battery life remains up to 25 hours on the ULT Field 5, just like the XG300. Once again, Sony improved quick charging to give you 30 more minutes of use in a 10-minute session. This speaker has lights at the ends though, which can impact longevity. Sony says you’ll reach that 25-hour mark using ULT modes with the lights off, but using both will drop that figure to 10 hours. There are three different light modes though, so there are options available. On the back of the ULT Field 5, there’s a panel with aux input, USB-C charging in/out, a battery care button and a switch for the lights. That’s protected by a sealed flap, which helps the speaker achieve its IP67 rating. 

That brings us to the ULT Tower 9. The party box joins the ULT Tower 10 on Sony’s current roster, replacing the SRS-XV900. Expanded EQ, Sound Field Optimization and two ULT modes are among the additions here, as are a redesigned handle and wheels for better portability. Sony improved the lighting for 360-degree projection as well as making the top control panel splash proof. Besides the Party Connect button from the new ULT Field speakers, the ULT Tower 9 has a dedicated option for a stereo pair. The top panel also houses controls for lighting, volume, power and two inputs for microphones (one of which can be used for a guitar). 

Sony's ULT Tower 9 is a party box speaker with improved lighting and better portability.
ULT Tower 9
Sony

Like the XV900, you can use Sony’s TV Sound Booster feature to enhance living room audio on the ULT Tower 9. The speaker has that ULT button, which gives you two modes for either bass boost or “powerful sound.” The driver arrangement inside includes four tweeters, two of which are pointing out the back of the speaker, along with two mid-range units and a single woofer. The wireless version of the ULT Tower 9 is the same price as its predecessor at $900, but Sony is also offering a wired option that must remain plugged into an outlet for $750. If you opt for wireless, battery life is the same on that Tower 9 as the ULT Field 5: up to 25 hours with only ULT mode and up to 10 hours with both ULT mode and lighting active. 

Lastly, Sony revealed the ULT Mic, which is a $150 set of two wireless microphones and a transmitter that are designed to be used with the company’s party speakers. This new accessory pack works with both the newer ULT Tower speakers in addition to the ULT Field 7. You can also use it with any older Sony speakers that support a microphone jack and receiver (like the XV900, XV800, XV500, XP700 and XP500). The mics are rechargeable with up to 20 hours of battery life and they feature a shock proof design. What’s more, the company equipped these devices with noise reduction for hands, breath and wind, plus there’s a Duet Assist tool to compensate for uneven vocal volumes. 

The ULT Field 3, ULT Field 5, ULT Tower 9 and ULT Mic are all available now.

This article originally appeared on Engadget at https://www.engadget.com/audio/speakers/sony-adds-three-more-speakers-to-its-ult-lineup-bass-boost-button-included-160008038.html?src=rss

Fruit flies can be made to act like miniature robots

Even the tiniest of living things are capable of some amazing forms of locomotion, and some come with highly sophisticated sensor suites and manage to source their energy from the environment. Attempts to approach this sort of flexibility with robotics have taken two forms. One involves making tiny robots modeled on animal behavior. The other involves converting a living creature into a robot. So far, either approach has involved giving up a lot. You’re either only implementing a few of life’s features in the robot or shutting off most of life’s features when taking over an insect.

But a team of researchers at Harvard has recognized that there are some behaviors that are so instinctual that it’s possible to induce animals to act as if they were robotic. Or mostly robotic, at least—the fruit flies the researchers used would occasionally go their own way, despite strong inducements to stay with the program.

Smell the light

The first bit of behavior involved Drosophila‘s response to moving visual stimuli. If placed in an area where the fly would see a visual pattern that rotates from left to right, the fly will turn to the right in an attempt to keep the pattern stable. This allowed a projector system to “steer” the flies as they walked across an enclosure (despite their names, fruit flies tend to spend a lot of their time walking). By rotating the pattern back and forth, the researchers could steer the flies between two locations in the enclosure with about 94 percent accuracy.

Read full article

Comments

Google unveils Ironwood, its most powerful AI processor yet

Google has unveiled a new AI processor, the seventh generation of its custom TPU architecture. The chip, known as Ironwood, was reportedly designed for the emerging needs of Google’s most powerful Gemini models, like simulated reasoning, which Google prefers to call “thinking.” The company claims this chip represents a major shift that will unlock more powerful agentic AI capabilities. Google calls this the “age of inference.”

Whenever Google talks about the capabilities of a new Gemini version, it notes that the model’s capabilities are tied not only to the code but to Google’s infrastructure. Its custom AI hardware is a key element of accelerating inference and expanding context windows. With Ironwood, Google says it has its most scalable and powerful TPU yet, which will allow AI to act on behalf of a user to proactively gather data and generate outputs. This is what Google means when it talks about agentic AI.

Ironwood delivers higher throughput compared to previous Google Tensor Processing Units (TPUs), and Google really plans to pack these chips in. Ironwood is designed to operate in clusters of up to 9,216 liquid-cooled chips, which will communicate directly with each other through a newly enhanced Inter-Chip Interconnect (ICI).

Read full article

Comments

Road deaths fell below 40,000 in 2024, the lowest since 2019

A rare spot of good news today: For the second year in a row, US roads got a little safer. The National Highway Traffic Safety Administration published its early estimate of road deaths in 2024; 39,345 people lost their lives, which is a 3.8 percent decrease from the 40,901 deaths that occurred on US roads in 2023.

The problem started with the pandemic; although road traffic dried up, the death rate leapt by 20 percent.

There’s no single cause, and studies have identified multiple contributing factors: empty roads designed to practically encourage speeding, little to no enforcement of traffic laws by the police, a general sense of fatalism in the face of public health restrictions that few Americans had ever contemplated in recent times, and car companies making big trucks and SUVs with high hoods, which are much more deadly to pedestrians and other vulnerable road users in a crash.

Read full article

Comments

Scientists Recreate Brain Circuit in Lab For First Time

Scientists have recreated in a laboratory the sensory pathway that transmits feelings of pain to the human brain, in a breakthrough that could lead to better treatments. Financial Times: A team at Stanford University in California is the first to combine different neurons grown from human stem cells into a functioning brain circuit in a lab dish. Their experiments, published in Nature on Wednesday, illustrate scientists’ rapid progress in replicating living tissues and organs through synthetic biology.

When the Stanford scientists exposed the brain circuit they had created to sensory stimulants, they observed waves of electrical activity travelling along it. The molecule that makes chilli peppers hot, capsaicin, immediately induced a strong response.

[…] The synthetic brain circuits could be used to screen for better-targeted therapies for pain that tone down excessive waves of neurotransmission, without affecting the brain’s reward circuitry as opioids do, project leader [Sergiu] Pasca said. The assembloids themselves cannot be said to “feel pain,” he emphasised: “They transmit nervous signals that are processed by a second pathway going deeper into the brain and giving us the aversive, emotional component of pain.”


Read more of this story at Slashdot.

Mushrooms Paint A Picture Using Their Electrical Activity

Previously seen letting mushrooms play a keyboard, Youtuber BionicandtheWires is back with another video, this time allowing a colony of oyster mushrooms to paint a picture. The mushrooms’ natural fluctuations in electrical activity is transformed into brush movement via attached solenoids. That painting, wow — way better than anything my nephew has ever done. Maybe next time he should employ a mushroom robot if he wants to take home anything but a purple ribbon from the middle school art show.

Xbox’s big summer showcase is set for June 8 at 1PM ET

There’s a big Xbox showcase event scheduled June 8 at 1PM ET. This is happening the same weekend as Summer Game Fest, which begins on June 6. Xbox has typically held these kinds of live presentations alongside Summer Game Fest ever since E3 was sent to a farm upstate to live with other discontinued industry conferences.

We don’t know what will be shown at the annual Xbox Games Showcase. The company says it will bring us a “look at upcoming titles from across our first-party studios, in addition to incredible new titles from our third-party partners.” It’s a digital-only event, so there will be no crowd to hoot and holler at reveals.

As a guess, I’d expect some new info on that new Fable entry, which was recently delayed until 2026. We could also get new trailers for the Perfect Dark reboot and Ninja Gaiden 4. Other possibilities include Gears of War: E-Day and Hideo Kojima’s OD. The sky truly is the limit. Xbox owns a lot of studios and, of course, has numerous partnerships with third-party devs.

One game that won’t be at the official presser is The Outer Worlds 2. That’s because it’s getting its very own livestream that follows the showcase. This game was originally teased all the way back in 2021, so it’s high time for a release date. Xbox promises a look inside developer Obsidian and trailers that reveal “new gameplay, details and developer insights.” Interestingly, The Outer Worlds 2 is no longer an Xbox exclusive, as it’s also coming to PS5.

This article originally appeared on Engadget at https://www.engadget.com/gaming/xbox/xboxs-big-summer-showcase-is-set-for-june-8-at-1pm-et-152001789.html?src=rss

Claude isn’t a great Pokémon player, and that’s okay

If Claude Plays Pokémon is supposed to offer a glimpse of AI’s future, it’s not a very convincing showcase. For the past month and counting, Twitch has watched Anthropic’s chatbot struggle to play Pokémon Red. Across multiple runs, Claude has failed to beat the nearly 30 year old game. And yet for David Hershey, the project’s lead developer, the showcase has been a success.

“I wanted some place where I could understand how Claude handles situations where it needs to work over a very long period of time,” Hershey explains to me over a video call. As part of his day job at Anthropic, Hershey works on the go-to-market team where he helps the company’s clients create their own agents (more on those in a moment). He first began working on Claude Plays Pokémon as a side project around the time Anthropic released 3.5 Sonnet last June.

As you can probably guess from the name, the project was partly inspired by Twitch Plays Pokémon, which debuted in 2014 and saw 1.16 million participate in a crowdsourced attempt to beat Pokémon Red using only the inputs viewers typed into the stream’s chatbox. Hershey wasn’t the first Anthropic employee to try to mold Claude into a Pokémon League Champion, but the project took on a life of its own right around the time he got involved.

In the early days of the project, it was a big deal when Claude managed to leave Red’s home and find Professor Oak. “I spent some ungodly number of hours tinkering to get it to make that kind of progress,” Hershey tells me. He would update his co-workers on Claude’s progress in an internal Slack channel. At that point, most of the company wasn’t paying attention, and it wasn’t something Anthropic planned to share with the world.

However, Hershey has made it a habit to revisit the project with each new major model release from Anthropic, starting with the upgraded version of Claude 3.5 Sonnet last fall and again more recently with 3.7 Sonnet. “It’s the way I go to see ‘What is this new model?’ ‘How does it work?’ ‘What can I learn about it?'” Hershey explains. And with Claude 3.7 Sonnet, the version of Claude playing the game right now, it was the first time “you could squint and see signs of life.”

Inside Anthropic the hope was that Claude would become better at trying different strategies and adjusting its approach when things didn’t go according to plan. With Pokémon Red, the company saw Claude do those things in real-time. “[Claude 3.7 Sonnet] spends less time stuck on assumptions,” says Hershey. “You’ll still see it make a guess and then spend some number of hours believing that’s true and making dumb decisions in the meanwhile, but previous models would kind of go on doing that forever.”

A chart showing the progress in playing Pokemon Red.
Antrhopic

And you can, quite literally, see Claude develop and run with those assumptions. Each ploddingly slow move in the game is preceded by a paragraph of text output from the AI — “I’ve encountered a wild ZUBAT while trying to navigate to (24,24). As per my strategy, I should flee from this battle to conserve resources” — followed by one single button press. Then it reassess the game state and does that all over again.

If you’ve been watching Claude fumble through Pokémon Red as a fan of the game, a model that spends “less time stuck on assumptions” appears minor, especially when the chatbot will frequently get stuck in areas like Viridian Forest, sometimes for days, due to the maze-like level design. Nonetheless, it is a milestone for the type of AI system that Claude 3.7 represents.

Like a lot of recent frontier AI systems, Claude 3.7 Sonnet is a reasoning model, meaning it’s designed to tackle problems by breaking them down into smaller pieces. “A lot of our customers care about how effective Claude is an agent,” explains Hershey. For the uninitiated, agents or agentic AIs are systems that are designed to plan and carry out complicated tasks without human supervision. Right now, most people think of AI as a blank chat box waiting to answer a question, but chatbots are only the consumer face of the industry; agentic systems represent an incremental but important step towards the promise of artificial general intelligence.

From that perspective, there are a couple of things that make Claude Plays Pokémon interesting. First, there’s the surprising fact Hershey delegated a lot of the programming that made the project possible to Anthropic’s coding agent including an overlay that allows Claude to make sense of Pokémon Red’s game world.

Second, and more importantly, Claude was not pretrained to play Pokémon Red. The chatbot knows some fundamentals about the game, such as the name of each gym leader and the order the player must beat them in, but it doesn’t have hundreds of years worth of game knowledge like some specialized AI systems. “You can throw a model at a game with no preparation, no guidance and it can learn everything itself,” he says. “I aim to be as close to that side as possible.”

Hershey had to give Claude some help. I already mentioned the overlay that allows it to interpret Pokémon Red’s interface. Pixel art is something all AI systems struggle with, and 3.7 Sonnet is no expectation. As humans, our imagination does a great job of filling in the details suggested by just a few pixels. What’s more, Claude doesn’t “see” the way we do.

If you watch it closely, you’ll notice each time it moves the player character, it will make a few inputs before reevaluating its position. Between those frames, Claude doesn’t have any sensory input. It can’t see Red walking, nor does it “hear” when its inputs cause him to crash into a tree or some other obstacle. Claude’s “poor vision” is one of the primary reasons it struggles with the game; in fact, Hershey had to give the chatbot a way to read the game’s memory so it was less likely to get confused if it misinterpreted the screen.

If the goal of the project was for Claude to beat Pokémon Red, that would have been easy. Hershey could have programmed a route through the game for the chatbot to follow, but at that point all he would have been testing is how well Claude follows a rigid set of instructions. “Claude is pretty good at that,” Hershey says. “I knew that. We all knew that.”

Instead, in leaving Claude to its own devices, the new model has shown it’s better at planning, coming up with new strategies and ultimately trying something different when its assumptions prove to be wrong. One of the more novel solutions Claude developed during its third run through the game was to intentionally cause all of its Pokémon to faint so that it could escape from Mt. Moon.

Still, Claude could be a lot better at both short- and long-term planning. In the same example I just mentioned, Claude deleted all of its notes on Mt. Moon after respawning at a nearby Pokémon Center, incorrectly believing it had successfully navigated the cave. One of its more promising runs ended after Claude failed to recognize it needed to talk to Bill to progress the game. It got stuck in an endless loop of bad decision making.

“Moving forward, I don’t know how useful this will be internally as a benchmark. It’s possible that with a small, tiny set of skills, Claude gets a little bit better and beats the game, and then the benchmark is not that interesting,” Hershey admits. “It could also be the case that there are things I don’t quite understand yet about what’s going to make our next model good, and then we’ll still be learning a lot more incremental things along the way.”

As for what happens next, Hershey says he doesn’t have a long-term strategy for Claude Plays Pokémon. “I’ve just spent so much time — my wife would say too much time — staring at this thing,” he says, laughing. I also get the sense Hershey’s not quite ready to close the book on the project. “I would imagine whenever a new model comes out, I’ll be playing Pokémon with it, and I will probably show the world that too.”

Until then, Anthropic, following a recent reset, continues to stream Claude Plays Pokémon on Twitch. The project has been successful enough to inspire an independent developer to program a Gemini Plays Pokémon stream, and if I had to guess, we’ll see more imitators before long.

This article originally appeared on Engadget at https://www.engadget.com/ai/claude-isnt-a-great-pokemon-player-and-thats-okay-151522448.html?src=rss

The AI Therapist Can See You Now

New research suggests that given the right kind of training, AI bots can deliver mental health therapy with as much efficacy as — or more than — human clinicians. From a report: The recent study, published in the New England Journal of Medicine, shows results from the first randomized clinical trial for AI therapy. Researchers from Dartmouth College built the bot as a way of taking a new approach to a longstanding problem: The U.S. continues to grapple with an acute shortage of mental health providers. “I think one of the things that doesn’t scale well is humans,” says Nick Jacobson, a clinical psychologist who was part of this research team. For every 340 people in the U.S., there is just one mental health clinician, according to some estimates.

While many AI bots already on the market claim to offer mental health care, some have dubious results or have even led people to self-harm. More than five years ago, Jacobson and his colleagues began training their AI bot in clinical best practices. The project, says Jacobson, involved much trial and error before it led to quality outcomes. “The effects that we see strongly mirror what you would see in the best evidence-based trials of psychotherapy,” says Jacobson. He says these results were comparable to “studies with folks given a gold standard dose of the best treatment we have available.”


Read more of this story at Slashdot.

The Murderbot TV show trailer is here

We finally have the trailer for Murderbot, the Apple TV+ show based on Martha Wells’ seven book series, The Murderbot Diaries. It features Alexander Skarsgård as a security unit that is less interested in protecting humans and more excited about catching up on its favorite soap operas.  

As Sarsgaard says in the trailer, “I was built to protect and obey humans. And humans are idiots. But now that I’ve hacked my programming, I can do whatever I want. As long as they don’t find out.” But, things take a turn when it gets chosen for a dangerous assignment. The security unit privately refers to itself as Murderbot, a fact that becomes more than awkward once the humans learn about the name.

Apple first announced the sci-fi adaptation, from About a Boy creators Chris and Paul Weitz, in late 2023. The first two of ten Murderbot episodes will come to Apple TV+ on May 16. The next eight episodes will debut every Friday. 

This article originally appeared on Engadget at https://www.engadget.com/entertainment/tv-movies/the-murderbot-tv-show-trailer-is-here-150017588.html?src=rss

Surprise, Google Workspace is adding more AI tools to Docs, Sheets, Chat and other apps

In an incredibly shocking turn of events, more generative AI features are coming to the Google Workspace suite of productivity apps by way of Gemini. Docs, Sheets, Chat, Meet and newcomer Vids are among the tools that are getting new AI capabilities that perhaps some folks will find useful.

Google is taking a cue from audio overviews in NotebookLM to bring new Gemini-powered audio features to Docs. You’ll soon have the option of listening to full audio versions of your documents or just the highlights in what Google calls a podcast-style overview. Those enrolled in the Workspace alpha testing program can start trying out these audio features in the coming weeks.

By the end of June, alpha testers will have access to an AI writing coach of sorts. The “help me refine” feature is said to offer suggestions on bolstering an argument, ensuring consistent formatting, enhancing the structure of the text and clarifying key points.

A similar feature is coming to Sheets later this year. The “help me analyze” tool will seemingly point out trends you might have overlooked, offer pointers on where to dig deeper for more insights and whip up charts to visualize the data.

In Meet, you’ll soon be able to ask Gemini to recap parts of a meeting you missed because you were running late, clarify certain decisions or topics, generate a recap in the format of your choice and help refine a point before you raise a virtual arm to join the conversation. This so-called “personal in-meeting advisor” will be generally available in Meet before the end of June.

As for Chat, Google will soon offer the option to mention @gemini in a conversation to receive a summary that includes open questions, important decisions that were made and lays out next steps. It could be a useful way to distill everything that happened in a bustling group chat. This capability will be available through the Labs early-access program in the coming weeks.

In Vids, you’ll soon be able to generate what Google says are “high-quality, original” clips using the Veo 2 model (perhaps ignoring the fact that generative video AI models mash together footage that they were trained on and arguably don’t create anything truly original at all). Workspace alpha can get their hands on this in the coming weeks.

Elsewhere, there’s another new feature called Google Workspace Flows. Google bills this as “a brand-new way to automate and orchestrate work across your apps, powered by AI.” It employs Gems, customizable AI agents that can be created with Gemini, to take care of specialized tasks, such as researching, analyzing and generating content.

The company says that Workspace Flows can, for instance, address a customer support ticket by analyzing the issue, researching a solution and drafting a reply that the support team can review before sending. Gems can mine your Google Drive files for context. Google’s also working with third-party partners to integrate Workspace Flows into their apps and tools. Google is now rolling out Workspace Flows in the alpha program.

Lastly, as of today, Google will allow Workspace users to choose where Gemini processes their data. So if you’re in the European Union and don’t want your data to travel to the US (or vice versa), that’s now an option. Google makes it clear that this is to help comply with regulations such as the EU’s GDPR.

This article originally appeared on Engadget at https://www.engadget.com/ai/surprise-google-workspace-is-adding-more-ai-tools-to-docs-sheets-chat-and-other-apps-143644125.html?src=rss