Last time we looked at the used electric vehicle market, it was to see what the options are if you’re spending $10,000 or less. Two solid choices emerged quickly: a BMW i3 if you don’t need much range, and a Chevrolet Bolt if you do. Lots of earlier Nissan Leafs made the list, too, but these had limited range and air-cooled batteries to contend with; we also included an assortment of compliance cars and, perhaps for the very brave, a Tesla. But what happens when you grow the budget by 50 percent? What EVs make sense when there’s $15,000 burning a hole in your pocket?
As it turns out, at this price point the planet starts looking a lot more like your own personal bivalve. For starters, the cars that looked good at $10,000 look a lot better in the next bracket up, generally newer model years or with lower mileage than the cheaper alternatives. Which means you can afford the facelifted i3. For model-year 2018 onwards, BMW fitted its electric city car with a larger-capacity battery, which means up to 114 miles (183 km) of range on a full charge, or about 150 miles (241 km) if it’s the one with the two-cylinder range-extender engine. Apple CarPlay and Android Auto might also be built into these i3s, although there are aftermarket solutions now, too.
No aftermarket is required to get CarPlay or Android Auto on any of the Bolts you might buy for under $15,000, which include a mix of pre- and post-facelift (model-year 2022 onwards) cars, although few of the slightly more spacious Bolt EUVs. Like the i3s, expect lower mileage examples, plus all the usual caveats: slow DC charging and seats that can get a bit hard on long drives.
Elgato’s Stream Deck helped popularize using a macro pad among aspiring livestreamers and regular desk jockeys, so it’s only natural for the brand to continue to expand on its lineup. After announcing a suite of new Stream Deck devices last year, Elgato is kicking off 2026 with the Stream Deck + XL, a larger version of the already big Stream Deck XL, and a collection of new audio accessories that should help improve sound mixing in and out of livestreams.
The Stream Deck + XL not only expands on the number of buttons on the regular Stream Deck XL (from 32 customizable buttons to 36 customizable buttons), but also introduces multiple new control options, including a touch strip and six dials. All the controls on the new Stream Deck are customizable, and the presence of the touch strip means you can receive visual feedback about adjustments you’re making while you’re making them. The new macro pad should still work well as a way to quickly end a live stream or queue up a transition, and Elgato says that paired with its new Wave XLR Pro audio interface, the Stream Deck + XL could also become a straightforward tool for sound mixing, too.
The XLR Dock MK. 2 adds an XLR input to the Stream Deck +.
Elgato
The Wave XLR Pro “integrates two fully independent XLR inputs with 80 dB of clean gain each,” Elgato says, along with multiple headphone outputs and “a built-in Maximizer effect” that can maintain consistent loudness. When you’re livestreaming, Elgato’s says the Wave XLR Pro’s “per-channel ducking” can also automatically lower other audio sources when you speak. The XLR Pro is one of several Wave audio products Elgato is announcing alongside the Stream Deck + XL, a lineup which also includes a new condenser microphone and a new way to connect XLR microphones to the company’s smaller Stream Deck. The XLR Dock MK. 2, exclusively compatible with the Stream Deck +, lets you connect professional XLR audio devices directly to your Stream Deck. The Dock “adds 80 dB of clean gain, 48V phantom power, onboard DSP effects, and low-latency VST Insert support,” and can be entirely managed through the Stream Deck’s buttons and dials.
Elgato says both the Stream Deck + XL and the XLR Dock MK. 2 are available to pre-order today for $350 and $130, respectively. The Wave XLR Pro will cost $350 when it launches at some point in Q2 2026.
This article originally appeared on Engadget at https://www.engadget.com/computing/elgatos-stream-deck–xl-adds-a-touch-strip-and-dials-to-an-already-massive-array-of-buttons-195200129.html?src=rss
AMD isn’t the only company launching non-mobile processors at Mobile World Congress 2026. Intel’s got some hardware to show off today, too, in the form of the Xeon 6+ processors that we know better by their codename “Clearwater Forest”. These chips are the successors to Sierra Forest, the all E-core Xeons that Intel launched back in 2024.
This is a free-to-play mobile game set in the universe of The Division, complete with an MMO-style shared open world and both co-op and solo gameplay. It also features an original story that’s set during the early days of the pandemic in NYC, making it a sequel to the first game but a prequel to the second one.
It’s a third-person action RPG with PvP. The loop looks fairly similar to the mainline games. Players should expect plenty of combat and loot to collect, which in turn can be used to upgrade gear.
Pre-registration is already open for the game at the iOS Store and the Google Play Store. Ubisoft has also announced some perks for The Division 2 players. Franchise veterans will receive some loot upon booting up Resurgence, including gear and clothing. Those who dip into The Division 2 after trying the mobile game will get other perks, including Resurgence-themed gear.
Today’s launch date announcement was part of a 10th anniversary showcase for the franchise. Ubisoft also revealed an anniversary-themed season for The Division 2, which includes “global events inspired by the original game” and an in-game hoodie. The company is also making the Warlords of New York expansion free during this period, which has received a new update focused on realistic combat.
The Division 1 and The Division 2 are both on sale right now, with massive discounts up to 90 percent on PC and 85 percent on PlayStation and Xbox. Today’s update, however, didn’t give us any new information regarding the recently-announced Survivors DLC for The Division 2. We don’t know much about this, except that it’s being described as “an updated take on the survival extraction experience.”
This article originally appeared on Engadget at https://www.engadget.com/gaming/the-long-delayed-the-division-resurgence-hits-ios-and-android-on-march-31-193905948.html?src=rss
The National Consumers League has released its annual fraud report detailing the many scams that most consumers have fallen for. As you’d expect, AI is playing a large role in how the threat landscape has evolved this past year. While the kinds of scams fraudsters are using haven’t changed much, the effectiveness of certain types of scams
I tend to go out of my way to avoid using Meta AI, but today, I gave it a fair shake. That’s because, as reported by Bloomberg, Meta’s AI service is now testing a shopping assistant. The idea is to compete with similar services from AI platforms like ChatGPT or Gemini, where users tap into the power of generative AI to search the web for product recommendations. That’s all fine and well in theory, but in execution, Meta AI’s shopping assistant is a bit of a disaster—even if it is only in testing.
Meta AI’s shopping suggestions are useless
When I tried accessing Meta AI’s web app logged out, I didn’t see the shopping assistant. But once I signed into my Meta Account, a number of additional options appeared—including “Shopping.” When you click the option, a new “Shopping research” option appears in the model selector, alongside “Thinking” and “Fast.” While you can type your own prompt into the search bar, a number of suggestions appear below, too. At this time, I see things like “Help me find a signature scent,” “Show me mid-century modern living room furniture,” or “Find a one-of-a-kind gift for my friend.” Finally, AI is here to help you spend money when you can’t think of anything to buy.
I was pretty intrigued by that last prompt. Meta AI doesn’t know my friend, and yet, it’s going to find me a “one-of-a-kind” gift they’ll love? I had to see what the bot would come up with. After Meta AI thought for a while (11.6 seconds, or so it tells me), the bot indeed confirmed my suspicions, asking for more info about my friend. It needed details like hobbies and interests, budget, and special memories I could incorporate into the gift. You would think that some gift ideas of my own would pop up by just thinking through these questions myself, but I needed the AI’s help, so I offered the following made-up answers to test it out: “Their hobbies and interests are music, biking, and going to the movies. My budget is $100. We went on vacation to Hawaii and we missed our flight home.”
The AI thought for 45.8 seconds, before returning with its “one-of-a-kind” gift ideas. Its favorite was something that combined my “friend’s” interest in music and our fake vacation to Hawaii: a ukulele. One-of-a-kind, indeed. Meta AI reasoned that, “a Ukulele Starter Kit is the perfect way to bring the island vibes home. The Kala Learn to Play Ukulele Concert Starter Kit is right at your budget and comes with everything they’d need to start, including lessons and a tuner. It’s a great hobby for a music lover and a direct connection to Hawaii.” Other unique gifts included three different types of portable speakers. Thanks.
“For the movie buff,” Meta AI suggested a scratch-off poster of 100 movies everyone should see. Or, as a “subtle nod to our trip,” a movie poster of film shot in Hawaii. This, I’ll admit, is a bit more unique, even though I’d never actually buy that. After scrolling past some bike accessories, though, I really hit gold:
“A Custom Hawaiian Shirt would be hilarious,” said the generative AI chatbot. Credit: Lifehacker
Meta AI isn’t much more helpful if you search for products yourself
After failing to find a perfect gift for my fake friend, I gave a custom prompt a try next: “I need a new couch for my living room.” This was a bit more helpful. The bot returned a list of five different couches, each with a description of the design and brand, in addition to a carousel of 12 couches of varying prices. To narrow it down, the bot asked me some questions about my home and interests, like how big my living room was and what style I was looking for. I sent the bot the following: “My living room is 200 sq. feet. I want a modern leather couch under $2000.” This returned what seemed to be a relevant list of couches, each within budget and style. I also learned about some new brands I didn’t know before the search, like Article and Poly & Bark.
Finally, I asked Meta AI for help finding a new MacBook. I said I needed something fast, but under $1,200. It thought, then brought back three decent options: a 13-inch M4 MacBook Air for $999, a 15-inch M4 Air for $1,199, or an M3 MacBook Air with extra storage for $1,030. There’s nothing wrong with those suggestions, but there were two issues I found with the result. First, the link for that third MacBook Air recommendation didn’t actually go to a listing, but the homepage for Abed Tahan, a store based in Lebanon. A quick search of the site returned results for the M3 MacBook Air in question, but it was more expensive than Meta AI said it was, and the store doesn’t ship to the U.S. Second, I asked Meta AI whether these results were the latest newest models, and it confirmed they were—despite the fact that Apple had announced new M5 MacBooks this morning. If the bot was working on a limited knowledge base, that’d be one thing, but it performs web searches with each query. It should be able to find this information.
While there are some apparent plusses to Meta AI’s shopping assistant, like its ability to find furniture that fits a particular room size and style, I feel pretty confident that I won’t be using it. The fact that some products links don’t actually work, and that it can’t reliably give you the most up-to-date products on the market, defeats the entire purpose of a shopping tool. I’ll be sticking to my usual shopping research: a traditional search engine combined with real user experiences.
Google unveiled its new features coming both to the Pixel line of devices and to the broader Android ecosystem this spring. Two of the more exciting additions for all Android devices are coming through the Find Hub. This can now be used to share real-time locations within Google Messages conversations or to send tracking information to airlines in instances of lost luggage.
With both of those features, users can turn off their sharing any time. Specific to the luggage issue, Google has partnered with 10 airlines to integrate Find Hub locations as part of their baggage recovery process. Using the luggage recovery does require the missing item to have a Find Hub-compatible tag or network accessory.
Another aesthetic update coming to Android is the ability to create a personalized Calling Card. You can choose a photo, font and color for your information that will be displayed to friends within Phone by Google. The short-form video trend is also coming to Google Play. The app storefront will now have a feed of Google Play shorts. It’s being positioned as a discovery tool for seeing different programs in action.
Google (modified)
The March Pixel Drop will bring improvements to the Circle to Search feature on smartphones. Going forward, it should be easier to use Circle to Search either to virtually try items on yourself or to shop for fashion from videos or social media posts. Magic Cue, an AI feature that debuted on the Pixel 10, can now also deliver restaurant recommendations. Pixel smartphones will also get some new personalized looks. Users can create AI-generated custom icons on the home screen with six different styles.
On the smartwatch side, Pixel devices will have new alerts if your phone is lost or left behind, as well as access to Express Pay. Pixel watches will also be getting some new safety alerts for earthquakes, and Satellite SOS is rolling out to more users across Europe, Canada, Hawaii and Alaska.
This article originally appeared on Engadget at https://www.engadget.com/mobile/smartphones/google-introduces-find-hub-updates-and-more-ai-tools-in-the-march-pixel-drop-190000773.html?src=rss
Accenture is acquiring Downdetector parent company Ookla from Ziff Davis in a $1.2 billion deal to bolster its network analytics and visibility tools for telecoms, hyperscalers, and enterprises. “The deal, which will transfer all of Ziff Davis’s Connectivity division to Accenture, includes Ookla’s Speedtest, Ekahau, and RootMetrics,” notes The Register reports: “Modern networks have evolved from simple infrastructure into business-critical platforms,” said Accenture CEO Julie Sweet in a canned statement. “Without the ability to measure performance, organizations cannot optimize experience, revenue, or security.” Ookla is meant to let them do just that.
Data captured at the network and device layer are used to enhance fraud prevention in banking, smart homes monitoring, and traffic optimization in retail, Accenture said. Ookla’s platform, which lets user’s test their own connectivity speed, captures more than 1,000 attributes per test, and provides the foundation for those analytics, Accenture said.
The ECRO World Tour is back for 2026 with a 77-race season kicking off Saturday, March 7 with Chasing Bianche. If you missed the original announcement last March, ECRO is a comprehensive racing platform built on top of Zwift that brings the structure, strategy, and integrity of professional cycling to virtual racing.
If you’ve followed the journey from the original Chasing Yellow in 2022 through Chasing Tour 2023 and Chasing Yellow 2024, you know the trajectory. Each year brought more structure, more competition, and more riders. But what ECRO has built over the past year goes well beyond a bigger race calendar. The entire platform has been rebuilt from the ground up, and the list of new features heading into the 2026 season is significant!
Let’s take a look at what’s new…
The Pro Sports Experience
What makes ECRO unique starts with making virtual cycling feel like a real sport with real stakes.
Team Divisions and Competition
ECRO’s team system goes beyond simple clubs. Teams can consist of up to 12 riders across the five categories, and the top 6 finishers per category contribute to team standings. This creates genuine strategic decisions about roster construction: do you stack one category or spread talent across the board?
The platform tracks team standings throughout the season, with points accumulating across all 77 races on the calendar, organized into four seasonal phases separated by transfer windows.
The Transfer Market
This is where ECRO starts to feel less like a race series and more like a sports league. The platform features a live transfer market where team managers can scout available riders, review their race history and power profiles, and extend contract offers.
Each rider has a virtual market value based on their performance and category. When teams sign riders, contract terms include a prize split percentage that determines how virtual earnings are divided between rider and team. Competitive riders command higher percentages, while developing riders might accept lower splits for the exposure and team support.
Free agents are visible on the market with detailed scouting information. When a rider signs, gets released, or sees a market value change, it shows up in a real-time activity feed that the whole community can follow.
And covering it all is Nigel Cadence, ECRO’s AI-generated transfer market correspondent. Every day, Nigel delivers a sportscaster-style audio recap of the latest signings, releases, and market moves, complete with a British accent and the energy of Sky Sports News on deadline day. Think of it as your daily transfer briefing: who signed where, what they’re worth, and what it means for the teams involved. The audio clips are posted to Discord and available on the platform’s transfer market page. Listen to the Tuesday, March 3 update:
ECRO$ and Team Economics
Teams manage virtual wallets funded by their share of rider earnings. The first 12 contract signings are free, but after that, teams need sufficient ECRO$ to cover new signings. When riders leave, teams recoup their market value. It’s a closed economic system that rewards smart roster management and long-term thinking.
Team Tactics
New for the 2026 season, team managers can publish race tactics for their riders before events. Managers can set team-wide strategies as well as individual instructions for specific riders, adding another layer of coordination to race day.
Making Racing Fair
Fair racing has always been the elephant in the virtual cycling room. Without physical proximity, how do you know someone’s power numbers are legitimate? ECRO has attacked this problem from multiple angles, building what may be the most comprehensive verification system in e-cycling.
Equipment Registration and Verification
Every ECRO rider must now register their equipment on the platform, including the make, model, and serial number of their smart trainer or power meter. The process works like two-factor authentication: riders photograph their equipment with a unique verification code displayed on screen, and the platform’s AI (powered by Google Gemini) cross-references the serial number in the image against what the rider submitted. If both match with high confidence, the equipment is auto-approved. If something doesn’t add up, it goes to manual review.
Serial numbers are unique to each rider. If someone tries to register a serial number that’s already tied to another account, the system flags it. It’s a simple concept borrowed from real-world equipment tracking, but it’s the first time anything like this has been implemented at scale in virtual cycling.
Dual Recording Validation
For top-category riders, ECRO requires dual recording, a practice pulled straight from UCI regulations. Riders record the same effort on two independent devices (typically a smart trainer and a separate power meter), and the platform compares the data. Discrepancies between the two recordings raise flags for further review.
The Performance Analysis Engine
This is where things get technical. ECRO has built a multi-detector integrity analysis system that runs automatically on every race result. The engine currently includes six detection modules:
Sticky Watts Detection identifies flat-top power patterns, where a rider’s power output remains suspiciously constant for extended periods. This follows the methodology established by the Fair E-Racing Association (FERA) and is one of the strongest indicators of manipulated data.
Power Anomaly Detection performs statistical analysis on power output, flagging impossibly sudden jumps (like zero to 400 watts instantly) or unnaturally stable power output that doesn’t match normal human physiology.
Heart Rate Correlation checks whether a rider’s heart rate patterns align with their power output. High-power efforts with flat heart rates, or missing heart rate data at critical moments, both raise flags.
Cadence Correlation looks for similar mismatches between pedaling cadence and power, including the suspicious scenario of producing significant power with zero cadence.
Microbursting Detection identifies burst-and-coast pedaling patterns that can exploit Zwift’s physics engine, where riders alternate between extreme sprints and coasting in rapid cycles.
Physiological Limits (ZADA) compares power output against known human physiological limits across different durations, flagging results that exceed what’s physically possible.
Each detector assigns flags at three severity levels (info, warning, or critical), and results flagged as suspicious are held from standings until reviewed. Riders can also flag results through a community-driven VAR (Verification Adjusted Results) system, similar to video review in professional sports.
Season Phase Locking
To address sandbagging (riders intentionally underperforming to race in a lower category), ECRO locks riders into their category after their first race in each season phase. Categories are based on ZwiftRacing.app‘s vELO rating system, with five tiers ranging from E (entry-level) through A (elite). Once you race in a phase, your category is locked until the next transfer window opens.
Platform and Tools
The Organizer Platform
One of ECRO’s less visible but most significant developments is its multi-tenant organizer platform. Race organizers beyond ECRO can use the platform to run their own competitions, inheriting ECRO’s categorization engine, results processing, and integrity detection without building any of it themselves. Organizers manage their own branding, calendars, scoring rules, and standings, while riders maintain a single account that works across all competitions.
Race Intelligence
Before each event, riders can access AI-powered race intelligence that includes predicted finishes based on vELO ratings, course suitability scores (analyzing whether you’re a better fit for a flat sprint stage or a mountain climb), and head-to-head records against other signed-up riders. The platform classifies courses as sprint, climbing, rolling, or mixed, then matches them against individual power profiles to identify who has the advantage.
Route Database
ECRO has catalogued 305+ Zwift routes with detailed visual profiles. Each race on the calendar includes elevation data, distance, and course characteristics, so riders know exactly what they’re getting into before they sign up.
Mobile App
The full ECRO platform is available as a mobile app, giving riders access to their dashboard, race calendar, results, transfer market, and equipment verification from their phone. Team managers can scout riders, send contract offers, and publish race tactics on the go.
Discord Integration
The ECRO Discord bot connects directly to the platform, offering slash commands for checking profiles, viewing results, pulling standings, and finding upcoming races. A built-in support system lets riders create tickets directly from Discord, with an AI-powered agent that can answer common questions and escalate to human admins when needed.
The 2026 Season
The ECRO World Tour 2026 features a 77-race calendar organized into four phases, with transfer windows between each phase where categories unlock and teams can restructure their rosters. The season opens March 7 with Chasing Bianche and includes single-day races, multi-stage tours, and seasonal series that span the full year.
Category thresholds have been updated for 2026: A (2000+ vELO), B (1700-1999), C (1400-1699), D (1100-1399), and E (0-1099). Each race offers five time slots to accommodate riders across different time zones.
Registration and Pricing
Rider License: $12.99 for the 2026 season
Team License: $19.99 for the 2026 season
Anyone can participate in ECRO events on Zwift, but only licensed riders have their results counted toward official standings. Licenses must be registered before event participation for results to count.
Apple unveiled a new MacBook Air today, and apart from the new M5 chip, things don’t look remarkably different. Sure, it’s getting a mild refresh, but maybe not in the way most people would want. Namely, it’s more expensive — a $100 price bump across all models. In exchange, the MacBook Air M5 does get faster performance and double the storage.
If you placed the MacBook Air M4 and the M5 in front of me, I wouldn’t be able to tell which was which. The M5 offers an overall CPU/GPU performance boost, along with some extra storage and double the file transfer speeds. Still, I didn’t expect a radical change, but I wish we got more for that extra $100.
It’s not worth jumping to the MacBook Air M5 if you’ve got the M4 unit already, but if you’re working with an older model, then it may be worth the switch, especially if you’re a professional. Like the rest of Apple’s new suite of products, pre-orders start at 9:15AM ET on March 4. I’d wait until full reviews are published before committing to spending money, though. Still, if all you care about are the specs and what they mean, we’ve got a pretty good idea of all that.
MacBook Air M5 vs. MacBook Air M4: Performance and battery life
Naturally, the biggest difference between the MacBook Air M5 and the MacBook Air M4 is their chipsets. While the price did go up, we also got double the storage in the MacBook Air M5, jumping from 256GB to 512GB at the base configuration for both sizes. Upgraded configurations start at 1TB.
According to Apple, the MacBook Air M5’s unified memory is 28 percent faster than the M4’s, and the AI performance is 4x faster. Casual users likely won’t feel the impact of the overall performance gains. (You could potentially get more use out of it compared with an M4 chip.) But Apple claims that folks who are rendering 3D in Blender will see a 50 percent increase in speed. That’s pretty significant if you’re a professional looking for a relatively affordable premium laptop with some oomph.
The MacBook Air got double the storage, but what you might overlook is that it also got double the speed. In theory, the read and write performance should be much faster. In real-world use, you won’t have to wait as long when you’re copying files, importing photos or videos and doing AI-enabled tasks. It could also potentially decrease boot times, meaning your MacBook might be faster at waking up after shutting down, but if so, the improvement may be negligible.
The MacBook Air M5 is also equipped with Apple’s new N1 chip. This enables the latest standards in connectivity, like Wi-Fi 7 and Bluetooth 6. Unless you’re a power user, it’s unlikely you’ll be using this chip to its full capacity, which can take up 46 gigabits per second of internet speed. If you’re paying your ISP for those rates, you can probably afford a MacBook Pro, which might better suit your needs (you speed demon).
The battery life of the MacBook Air M5 remains the same as its predecessor. In my experience reviewing laptops, Apple has been killing it with its battery life scores (with the M4 lasting over 18 hours on Engadget’s battery test), so I wouldn’t be too shocked if its claims are accurate. Those are: 18 hours of video streaming and 15 hours of wireless web surfing. I’ve also tested the MacBook Air M4, and that latter time is on point.
MacBook Air M5 vs. MacBook Air M4: Design, display, audio
Similar to the iPad Air M4 announcement this week, the MacBook Air M5’s design, display and audio remain unchanged despite the overall price increase. Apart from being frustrated by the higher cost, I was satisfied that everything that’s here is already pretty solid.
There’s a decent array of colors, with the MacBook Air available in either blue, silver, beige and black. Although, I will always advocate for more, and brighter, colors especially since these hues are so tame.
The MacBook Air M5 continues to live up to its name with a thin and light design. The 13-inch model comes in at 11.97 x 8.46 x 0.44 inches and 2.7 pounds, while the 15-inch stacks up to 13.4 x 9.35 x 0.45 inches and 3.3 pounds. Unfortunately, since there’s no change in design, we’re stuck with only two Thunderbolt 4 ports and a headphone jack.
I’d like to complain that the MacBook Air still doesn’t have an OLED panel, but I can’t, since even the Pro MacBooks still have yet to feature one. (There are laptops, like the ZenBook 14, that cost around the same as the MacBook Air and manage to sport an OLED display, so it’s not unreasonable.)
Despite that, the MacBook Air’s displays are plenty vibrant. Both models come with a Liquid Retina LED panel at 224 ppi, and emit up to 500 nits of brightness. It makes working outdoors possible, but since it’s a glossy panel, I’d still recommend staying inside (or finding shade). The MacBook Air 13 has a 13.6-inch, 2,560 x 1,664 panel, while the 15 is equipped with a 15.3-inch, 2,880 x 1,864 screen. I’ve seen both the 13-inch and 15-inch versions of the MacBook Air M4 in person and they’re pleasantly bright.
The speakers on both the MacBook Air M4 models produced loud and clear sound, albeit with middling bass. The MacBook Air M5 features the same speaker system, with the 13-inch supporting four speakers and the 15-inch carrying six, so it stands to reason the new laptops will deliver similar experiences on sound.
macOS Tahoe and Apple Intelligence
No changes to macOS Tahoe here. Everything you get with the MacBook Air M4 you’ll get on the M5. That includes the controversial Liquid Glass design and Apple Intelligence features like Live Translation in Messages. Speaking of Apple Intelligence, like I said earlier, you will likely benefit from faster processing thanks to the M5 chip. Apple claims AI tasks will be 4x faster, so even those who casually use AI may notice the difference.
If you’d like to compare for yourself exactly what changes the MacBook Air M5 brings over its predecessor, we’ve compiled this table to make it easier on your eyes.
MacBook Air M5 vs. MacBook Air M4: Specs at a glance
Spec
MacBook Air M5
MacBook Air M4
Price
$1,099 (13-inch), $1,299 (15-inch)
$999 (13-inch), $1,199 (15-inch)
Processor
M5
M4
Display
13.6-inch: Liquid Retina, LED, 2,560 x 1,664, 224 ppi
15.3-inch: Liquid Retina, LED, 2,880 x 1,864, 224 ppi
13.6-inch: Liquid Retina, LED, 2,560 x 1,664, 224 ppi
15.3-inch: Liquid Retina, LED, 2,880 x 1,864, 224 ppi
This article originally appeared on Engadget at https://www.engadget.com/computing/laptops/macbook-air-m5-vs-macbook-air-m4-whats-changed-beyond-the-apple-silicon-185256565.html?src=rss
As part of today’s MacBook Pro update, Apple has also unveiled the M5 Pro and M5 Max, the newest members of the M5 chip family.
Normally, the Pro and Max chips take the same basic building blocks from the basic chip and just scale them up—more CPU cores, more GPU cores, and more memory bandwidth. But the M5 chips are a surprisingly large departure from past generations, both in terms of the CPU architectures they use and in how they’re packaged together.
We won’t know the impact these changes have had on performance until we have hardware in hand to test, but here are all the technical details we’ve been able to glean about the new updates and how the M5 chip family stacks up against the past few generations of Apple Silicon chips.
X will suspend creators from its revenue sharing program if they post AI-generated videos depicting armed conflicts without disclosing they were made with AI. Head of product Nikita Bier announced the policy change on March 3, saying first-time violators will be cut off for 90 days and repeat offenders would be permanently removed from the program.
Today we are revising our Creator Revenue Sharing policies to maintain authenticity of content on Timeline and prevent manipulation of the program.
During times of war, it is critical that people have access to authentic information on the ground. With today’s AI technologies,…
The policy is notably narrow, applying only to creators enrolled in the platform’s revenue sharing program and only to AI-generated videos of armed conflicts, not AI content in general or non-monetized accounts. Violations will be flagged through Community Notes, X’s crowd-sourced fact-checking system, or by detecting metadata from generative AI tools. Bier framed the change as necessary “during times of war,” though the current conflict unfolding between the United States, Israel and Iran has not been formally, or at least not legally, declared a war. Of course, the US has not formally declared war since 1942.
The quality of AI video generation has progressed at a rapid pace, and generated content has become almost indistinguishable from real footage for most viewers. X already watermarks images and videos generated by its Grok chatbot but has not previously required users to disclose AI-generated content. The platform is separately testing a broader AI labeling toggle that would let users mark any post as containing synthetic content, as first reported by Social Media Today, though X has not shared a timeline for that feature.
This article originally appeared on Engadget at https://www.engadget.com/social-media/x-to-require-ai-labels-on-armed-conflict-videos-from-paid-creators-citing-times-of-war-183631400.html?src=rss
We may earn a commission from links on this page. Deal pricing and availability subject to change after time of publication.
The 32-inch ASUS ROG Strix OLED XG32UCWG gaming monitor has earned stellar review for its excellent visuals, refresh rate flexibility, and other impressive gaming-focused features. If you’ve been looking to upgrade your setup, right now is a good time: It’s $200 off, reaching its lowest price ever—$799 (originally $999).
This 32-inch monitor has a 4K UHD WOLED panel that delivers high contrast and deep blacks, making games, movies, and TV shows look more vivid. Compared to a traditional OLED panel, WOLED (White OLED) uses a white light source with filters to provide deeper blacks, faster response times, and more uniform images across a range of lighting conditions. The TrueBlack Glossy coating is better at reducing glare than a matte screen while delivering a clearer picture (though some reviews claim it’s still not bright enough to mitigate intense glare if bright light hits the screen directly).
A native 4K, 165 Hz refresh rate allows for high-detail gaming and multimedia use, while its Frame Rate Boost mode offers versatility for playing different kinds of games. For example, FHD at 330 Hz also makes it a solid choice for competitive shooting games and high-FPS play. The monitor has an overall 0.03 ms response time, which is best-in-class, and minimizes motion blur. Other notable features include NVIDIA G-SYNC and AMD FreeSync support, and reduced burn-in risk due to OLED Care Pro and the Neo Proximity Sensor.
If you’re looking for a gaming monitor with dual-mode features, nearly instantaneous response time across a range of refresh rates, low input lag, and inky blacks, the 32-inch ASUS ROG Strix OLED XG32UCWG monitor for gamers is a great choice . While it doesn’t offer 240 Hz+ at 4K like some elite models, it delivers a lot for this $800 price point.
Sure, Blue Prince has the most name recognition among the indie games that had a surprise debut on Nintendo Switch 2 today. However, I implore you not to overlook two other wonderful indies that hit Nintendo’s consoles in the wake of this morning’s Indie World showcase.
Minishoot’ Adventures — from SoulGame Studio and co-publisher IndieArk — is a game I absolutely loved when it first emerged on Steam in 2024. It’s an RPG that adopts the format of top-down Legend of Zelda games. At the same time, it’s a twin-stick bullet-hell shooter. The two-person team at SoulGame blended those genres to great effect.
Here, you play as a ship. Exploration is key, as you’ll need to find upgrades for your ship in order to access new areas. SoulGame injected a ton of personality into Minishoot’ Adventures as well. Errant apostrophe in the name aside, it’s a real gem.
I’m perplexed as to why Nintendo only showcased Öoo in the Japanese version of the Indie World stream. It’s utterly brilliant and by far the best new game I played in 2025. It’s so exquisitely designed by Nama Takahashi (ElecHead) that a 20-second trailer is all that’s needed to show how this puzzle platformer works.
You play as a caterpillar who has a bomb — and later a second one — that can be used to blow stuff up. The bombs are also used to hold down buttons and to propel the caterpillar upwards or sideways.
There are no instructions here. Takahashi (who made the game with the help of Tiny Cactus Studio and Tsuyomi) helps you figure out what to do solely through intelligent level design. The puzzles aren’t overly difficult, but they got my brain matter working just enough that I smiled whenever I found a solution. The visuals, music and sound design are supremely charming too.
This is a truly wonderful game. It doesn’t outstay its welcome either, as it can be completed within a couple of hours.
Öoo is available now on Switch for $11. It’s coming to PS5 on March 11 and Xbox Series X/S later this month.
This article originally appeared on Engadget at https://www.engadget.com/gaming/nintendo/two-fantastic-indie-games-not-named-blue-prince-also-just-arrived-on-switch-182648197.html?src=rss
Approved proposal reverses earlier stance, even as survey highlights bigger frustrationsThe Go team has approved generic methods, reversing a longstanding position in the language’s FAQ. The proposal, from Go co-designer Robert Griesemer, now moves to implementation.…
Microsoft partnered up with ASUS to release the excellent ROG Xbox Ally X last year, to do battle with the SteamOS powered Steam Deck. Because it has jumped into the handheld gaming PC battle a bit late it has been working to continually to play catch up by adding new features, the latest of which is Highlight Reels, now available to Xbox
Apparently, AI data centers are capable of sucking less (power, that is). A recent UK trial demonstrated that they can adjust their energy demands dynamically without disrupting critical workloads. This contrasts with data centers’ current approach of always-on power draw, which can strain grids and drive up prices for everyone.
Over five days in December 2025, more than 200 simulated “grid events” tested a London data center’s ability to adjust its energy use on the fly. The trial used software from Emerald AI, which was involved in the study. Other partners included NVIDIA, National Grid, Nebius and the nonprofit Electric Power Research Institute.
In each simulated grid event, the data center successfully adjusted its energy use to the requested level. It reduced power draw by up to 40 percent, while critical workloads continued to run as normal throughout the trial.
The data center successfully reacted to spikes in demand during soccer match halftimes. In one case, it reduced its power draw by 10 percent for up to 10 hours. It also managed to cut its demand quickly: One event saw the data center reduce its load by 30 percent in only 30 seconds.
The study will serve as a blueprint for a 100MW “power-flexible AI factory” that NVIDIA plans to operate in Virginia. “This trial proves that NVIDIA-powered infrastructure can act as a grid-aware asset, modulating demand in real-time to support stability,” Josh Paker, NVIDIA’s sustainability lead, wrote in a statement. “By making AI workloads responsive, we accelerate deployment while reducing the need for costly grid upgrades.”
The organizations involved in the study say they’ll share their data with the AI industry, regulators and policymakers to try to influence their approach. Fortunately, we don’t need to hope that data center operators’ altruism (ha) will lead to their cooperation. Agreeing to curb usage during peak demand could be good for their balance sheets and lead to faster approvals for new data center grid connections. “We would love to get to a point where we can get customers on the network in two years, and this is part of that,” Steve Smith, president of National Grid Partners, toldBloomberg.
This article originally appeared on Engadget at https://www.engadget.com/ai/ai-data-centers-could-reduce-power-draw-on-demand-study-says-180628982.html?src=rss
An anonymous reader quotes a report from the BBC: India’s Supreme Court has threatened legal consequences after a judge was found to have adjudicated on a property dispute using fake judgements generated by artificial intelligence. The top court, which was responding to an appeal by the defendants, will now examine the ruling given by the lower court in the southern state of Andhra Pradesh. The Supreme Court called the case a matter of “institutional concern” and said fake AI-generated judgements had “a direct bearing on integrity of adjudicatory process.”
[…] Coming down sternly against the fake judgements, the top court last Friday stayed the lower court’s order on the property dispute. It said the use of AI while making judgements was not simply “an error in decision making” but an act of “misconduct.” “This case assumes considerable institutional concern, not because of the decision that was taken on the merits of the case, but about the process of adjudication and determination,” the top court said. The court said it would examine the case in more detail and issued notices to the country’s Attorney and Solicitor General, as well as the Bar Council of India.
Anthropic is taking steps to make it easier to switch to Claude. While the AI chatbot app is popular among developers and vibecoders, it’s also been infamous for keeping its more advanced features behind a paywall (and for rate-limiting free users). But now, Claude is finally catching up to ChatGPT and bringing its memory feature to free users.
The move follows Claude overtaking ChatGPT in the App Store to become the #1 most downloaded free app in the U.S., so the timing makes sense. As for what could have caused the sudden interest in the app, OpenAI recently announced that it will be working with the U.S. Department of Defense (unofficially titled the Department of War), a day after Anthropic CEO Dario Amodei expressed concern about unrestricted AI use by governments.
Alongside the new free memory feature, Claude is also introducing a free import tool to help you bring your AI context along with you when moving from other chatbots. Technically, it’s more of a guided prompt to feed into other bots, but the idea is that it can help new users avoid blank-slate syndrome. Within 24 hours of using the tool, Claude will theoretically know all the personal details you’ve previously shared with the chatbots you’re importing from, including special instructions, your career, and your ongoing projects, making it easier to converse with Claude.
How to enable memory in Claude for free
While Claude’s memory feature is available for free for all users, it’s not enabled by default. Let’s fix that. Open the Claude website or the app, click the Profile icon, and go to Settings. Here, in the Capabilities section, you’ll see a new Memory section up top. Enable the Generate memory from chat history feature. Now, Claude will automatically start remembering key details about your life as you share them. According to Anthropic, Claude “will automatically summarize your conversations and create a synthesis of key insights across your chat history (not including chats in projects). This synthesis is updated every 24 hours and provides context for every new standalone conversation.”
Say you’re a dentist and you ask Claude for dental implants research; it will know that you’re learning about implants the next time you ask a related question.
Credit: Khamosh Pathak
Of course, this does mean that Claude will start remembering your personal data, too, or at least your personal context. Claude does offer a couple of ways to get around this. If you try to disable memory (from the same menu where you enabled it), you’ll see two other options. For a less severe workaround, you can use the Pause memory option to stop the chatbot from creating new memories while keeping its current memories intact. Or, you can choose the Reset memory option to permanently delete all memories, including project-specific memories. That way, you can manually dump what Claude knows about you every once in a while.
Credit: Khamosh Pathak
While you’re in the Memory settings, you can also use the new Import feature. Click the Start Import button to bring up the new menu. Up top, you’ll see a prompt that you’ll have to copy. After that, paste it into ChatGPT or Gemini to snag your memories from these bots. You’ll get your results in a Markdown file. Back in Claude, paste the Markdown file into the textbox below the prompt you copied and click Add to memory. Claude will synthesize it, and it will add its data to its memory file.
Audible just launched a new budget-friendly Standard plan that costs $9 each month. Despite the lower price, this plan still includes a monthly audiobook credit. These credits allow users to download pretty much any audiobook from the library, including most brand-new releases. This was a perk previously locked to the Premium Plus plan, which is $15 per month.
Subscribers also get access to a curated library of content sourced from the Audible Originals library and Wondery+. All of this will be available without ads.
There are some caveats, especially when compared to the pricier Premium Plus subscription. That plan lets subscribers keep downloaded audiobooks forever, but the same isn’t true of the Standard plan. Audiobooks will float into the ether once a subscription has been cancelled. This is only a big deal, however, for those who like to re-listen to their favorites.
Audible
The Premium Plus plan provides some other perks not available to Standard users. Subscribers get regular access to sales and discounts, and there’s a library of audiobooks and podcasts to peruse. This is sort of like Kindle Unlimited, but for audiobooks. It’s not a comprehensive catalog, but it works in a pinch.
Amazon says the new plan is “projected to bring millions of new consumers into the Audible experience over the next year.” It’s available right now in the US, the UK, Canada, Germany, France and Australia. The company says it’s currently testing the tier in additional marketplaces throughout the world.
This article originally appeared on Engadget at https://www.engadget.com/entertainment/streaming/audibles-new-plan-is-9-a-month-and-still-includes-an-audiobook-credit-174641878.html?src=rss
