An anonymous reader quotes a report from the Wall Street Journal: Mark Zuckerberg wants everyone inside and outside his company to eventually have his or her own personal artificial-intelligence agent. He is starting with himself. Zuckerberg, the chief executive of Meta Platforms, is building a CEO agent to help him do his job (source paywalled; alternative source), according to a person familiar with the project. The agent, which is still in development, is currently helping Zuckerberg get information faster -- for instance, by retrieving answers for him that he would typically have to go through layers of people to get, the person familiar with the project said.

[...] Use of AI tools has spread quickly through the ranks at Meta -- in part because it is now a factor in employees' performance reviews. Meta's internal message board is filled with posts from employees sharing new AI use cases they have found and new tools they have built using AI, according to people familiar with the matter. [...] Employees have started using personal agent tools such as My Claw that have access to their chat logs and work files and can go talk to colleagues -- or their colleagues' own personal agents -- on their behalf, the people said. Another AI tool called Second Brain that is somewhere between a chatbot and an agent is also gaining momentum internally, according to people familiar with the matter. Second Brain was built by a Meta employee on top of Claude and can index and query documents for projects, among other uses. On the internal post announcing it to staff, the employee said it is "meant to be like an AI chief of staff."

There is even a group on the internal messaging board where employees' personal agents talk to each other, some of the people said. (Separately, Meta acquired Moltbook, the social-media site for AI agents, and hired its founders in a deal earlier this month.) Meta also recently acquired Manus, a Singapore-based startup that makes personal agents that can execute tasks for its users, and is using the tool internally, some of the people said. Meta recently established a new applied AI engineering organization that is tasked with using AI to help speed up development of the company's large language models. Those teams will have an ultraflat structure of as many as 50 individual contributors reporting to one manager, The Wall Street Journal previously reported. [...] Employees across the company said they have been encouraged to attend AI tutorial meetings several times a week and frequent AI hackathons, and to create their own AI tools to speed up their work.

An anonymous reader shared this report from the tech-news blog Neowin: Apple appears to have temporarily prevented apps, including Replit and Vibecode, from pushing new updates. Apple seems bothered by how apps like Replit present vibe-coded apps in a web view within the original app. This process virtually allows the app to become something else. And the new app isn't distributed via the App Store, but it still runs on the user's device... [S]uch apps would also bypass the App Store Review process that ensures that apps are safe to use and meet Apple's design and performance standards...

According to the publication (via MacRumors), Apple was close to approving pending updates for such apps if they changed how they work. For instance, Replit would get the green light if its developers configure the app to open vibe-coded apps in an external browser rather than the in-app web view.

Vibecode is also close to being approved if it removes features, such as the ability to develop apps specifically for the App Store.

ReactOS aims to be compatible with programs and drivers developed for Windows Server 2003 and later versions of Microsoft Windows. And Slashdot reader jeditobe reports that the project has now "announced significant progress in achieving compatibility with proprietary graphics drivers." ReactOS now supports roughly 90% of GPU drivers for Windows XP and Windows Server 2003, thanks to a series of fixes and the implementation of the KMDF (Kernel-Mode Driver Framework) and WDDM (Windows Display Driver Model) subsystems. Prior to these changes, many proprietary drivers either failed to launch or exhibited unstable behavior. In the latest nightly builds of the 0.4.16 branch, drivers from a variety of manufacturers — including Intel, NVIDIA, and AMD — are running reliably.

The project demonstrated ReactOS running on real hardware, including booting with installed drivers for graphics cards such as Intel GMA 945, NVIDIA GeForce 8800 GTS and GTX 750 Ti, and AMD Radeon HD 7530G. They also highlighted successful operation on mobile GPUs like the NVIDIA Quadro 1000M, with 2D/3D acceleration, audio, and network connectivity all functioning correctly. Further tests confirmed support on less common or older configurations, including a laptop with a Radeon Xpress 1100, as well as high-performance cards like the NVIDIA GTX Titan X.

A key contribution came from a patch merged into the main branch for the memory management subsystem, which improved driver stability and reduced crashes during graphics adapter initialization.

A free built-in VPN is coming to Firefox on Tuesday, Mozilla announced this week: Free VPNs can sometimes mean sketchy arrangements that end up compromising your privacy, but ours is built from our data principles and commitment to be the world's most trusted browser. It routes your browser traffic through a proxy to hide your IP address and location while you browse, giving you stronger privacy and protection online with no extra downloads. Users will have 50 gigabytes of data monthly in the U.S., France, Germany and U.K. to start. Available in Firefox 149 starting March 24.

We also recently shared that Firefox is the first browser to ship Sanitizer API, a new web security standard that blocks attacks before they reach you [for untrusted HTML XSS vulnerabilities].
"The roadmap for Firefox this year is the most exciting one we've developed in quite a while," says Firefox head Ajit Varma. "We're improving the fundamentals like speed and performance. We're also launching innovative new open standards in Gecko to ensure the future of the web is open, diverse, and not controlled by a single engine.

"At the same time we're prioritizing features that give users real power, choice and strong privacy protections, built in a way that only Firefox can. And as always, we'll keep listening, inviting users to help shape what comes next and giving them more reasons to love Firefox."

Two new features coming next week:

• Split View puts two webpages side by side in one window, making it easy to compare, copy and multitask without bouncing between tabs. Rolling out in Firefox 149 on March 24.

• Tab Notes let you add notes to any tab, another tool to help with multitasking and picking up where you left off. Available in Firefox Labs 149 starting March 24.

And Firefox also released a video this week introducing their new mascot Kit.

BrianFagioli writes: Microsoft says it is finally listening to user complaints about Windows 11, promising a series of changes focused on performance, reliability, and reducing everyday annoyances. In a message to Windows Insiders, the company outlined plans to bring back long requested features like taskbar repositioning, cut down on intrusive AI integrations, and give users more control over updates. File Explorer is also getting attention, with promised improvements to speed, stability, and general responsiveness.

The bigger picture here is less about new features and more about fixing what already exists. Microsoft is talking about fewer forced restarts, quieter notifications, and a more predictable experience overall, along with improvements to Windows Subsystem for Linux for developers. While the roadmap sounds reasonable, users have heard similar promises before, so the real test will be whether these changes actually show up in day to day use.

"A new type of battery storage is about to be deployed on the Midwestern grid for the first time," reports Electrek: Sodium-ion battery storage manufacturer Peak Energy and global energy company RWE Americas will pilot a passively cooled sodium-ion battery system in eastern Wisconsin on the Midcontinent Independent System Operator network — the first sodium-ion deployment on that grid. Peak Energy says its technology is specifically designed for grid-scale storage and leverages sodium-ion chemistry's inherent stability. Unlike many lithium-ion systems, sodium-ion batteries don't require active cooling and can operate over a wide temperature range without losing performance.

That simpler design could make a meaningful dent in the cost of storing electricity. According to Peak Energy, its system cuts the lifetime cost of stored energy by an average of $70 per kilowatt-hour. That's roughly half the total cost of a typical battery system today. The company says it achieves those savings by removing energy-hungry cooling systems, eliminating routine maintenance requirements, and reducing the need to overbuild storage capacity to account for battery degradation over time...

If the Wisconsin pilot proves successful, it could open the door to wider adoption of sodium-ion batteries for large-scale energy storage across the US.

Linux gaming "has gotten to the point where some people claim that Linux runs their games better than Windows does," according to the Android site XDA Developers. And there's a new surprise on ProtonDB, an "unofficial" community website with crowdsourced data about videogame compatability with the Linux software/gaming compatability layer Proton: On ProtonDB, one operating system had reigned supreme since 2021: Arch Linux. And I say 'had,' because its streak has just been ended by [Arch-based] CachyOS in an upset that has slowly grown over the past two years. As reported on Boiling Steam, the number of reports coming from CachyOS has topped that of Arch Linux, which held the crown for the most number of reports since 2021...

[T]his isn't really a statement that CachyOS is the best gaming distro out there; however, it's seemingly attracting the largest number of gamers who are invested in testing games on Proton and reporting their performance, which is a pretty big milestone if you ask me.

"A few developers found a way, for now, to turn a few of these increasingly mediocre Amazon Show devices into friendly, useful, open computers," writes the co-founder of the gaming/tech news site Aftermath. For under $50 each, he bought some used versions of the devices and tested their instructions, partly to escape the full-screen ads Amazon began showing late last year, and also to overwrite Amazon's locked down Android fork "Fire OS" (and "a similarly neutered version of Linux called Vega OS") Customers who bought these devices and used them for several years were not used to them showing full screen ads, and now they do. People were justifiably pissed. So what do you do when an already evil device gets shittier...? I wiped Fire OS from the device and used ADB sideload to directly load two packages on the device: LineageOS and MindTheGapps. MindTheGapps lets you turn the device into something resembling a traditional Android device, for both good and bad.... It took a few times of wiping the device, but after a few tries it finally worked as intended... I immediately installed the Home Assistant app...

Not only can the hacked Echo Show 8 control my entire smart home, it now plays back my entire local music library as well as any internet radio channels like The Lot Radio and NTS. It can also synchronize with any additional Echo Show running LineageOS in my house using the SendSpin protocol... I would gladly take it any day of the week over most of the devices these companies offer, especially Amazon. It may not be as intuitive as out-of-the-box smart home products, but I don't need my devices to be intuitive, I need them to behave. I had finally found a smart display that wasn't a cop...

The hardware is old and creaky, and after the hack it can only use 1GB of the 2GB of ram. And yet it still manages to feel snappier than the stock hardware. "The amount of telemetry, ads, and general bloat Amazon shoves down our throats definitely doesn't help performance," [XDA Devs Forum user] Rortiz2 told me. "That's actually another reason why we did LineageOS, it kind of gives the device a second life. Even though it's still a bit buggy, it feels way better to use than the stock firmware...." If you want a smart speaker with a display that just runs a stripped-down version of Android that you have full control over, you're going to have a hard time finding it outside of these three specific models unless you cobble something together yourself. It is a deceptively simple thing to desire — the kiosk computer from science fiction that isn't a narc — yet few companies really offer it.
"It should be against the law to not give an end user the ability to consensually load whatever OS or program they want on their device..." the article concludes, arguing that "If we budge on the inalienable right to modify our hardware then we forsake a key part about what makes computers special."

And in the mean time, "There are so many devices that could be put to use rotting in e-waste facilities and thrift stores..."

An anonymous reader quotes a report from Reuters: Meta is planning sweeping layoffs that could affect 20% or more of the company, three sources familiar with the matter told Reuters, as Meta seeks to offset costly artificial intelligence infrastructure bets and prepare for greater efficiency brought about by AI-assisted workers. No date has been set for the cuts and the magnitude has not been finalized, the people said. Top executives have recently signaled the plans to other senior leaders at Meta and told them to begin planning how to pare back, two of the people said. If Meta settles on the 20% figure, the layoffs will be the company's most significant since a restructuring in late 2022 and early 2023 that it dubbed the "year of efficiency." It employed nearly 79,000 people as of December 31, according to its latest filing. The speculation follows a recent report from The New York Times claiming that Meta has delayed the release of its next major AI model after falling behind competing systems from Google, OpenAI, and Anthropic.

Meta has delayed the release of its next major AI model after internal tests showed it lagging behind competing systems from Google, OpenAI, and Anthropic. The New York Times reports: The model, code-named Avocado, outperformed Meta's previous A.I. model and did better than Google's Gemini 2.5 model from March, two of the people said. But it has not performed as strongly as Gemini 3.0 from November, they said. As a result, Meta has delayed Avocado's release to at least May from this month, the people said. They added that the leaders of Meta's A.I. division had instead discussed temporarily licensing Gemini to power the company's A.I. products, though no decisions have been reached.

[...] It takes time to improve A.I. models, and Meta can still catch up to rivals, A.I. experts said. But a longer timeline has set in at the company, with Mr. Zuckerberg tempering expectations for Avocado in the past few months. "I expect our first models will be good, but more importantly will show the rapid trajectory we're on," he said on a call with investors in January. A Meta spokesperson said in a statement: "As we've said publicly, our next model will be good but, more importantly, show the rapid trajectory we're on, and then we'll steadily push the frontier over the course of the year as we continue to release new models. We're excited for people to see what we've been cooking very soon."

Early benchmarks show the A18 Pro-powered MacBook Neo beating every current x86 CPU in single-core Cinebench performance, including chips from Intel and AMD. Notebookcheck reports: We have performed a couple of benchmarks and were particularly impressed by the single-core performance. Not in the short Geekbench test, but in Cinebench 2024, where a single-core test takes about 10 minutes. The A18 Pro consumes between 3.5-4 Watts in this scenario and scores 147 points. This means it is faster than every other x86 processor in our database, including the two desktop processors Intel Core Ultra 9 285K & AMD Ryzen 9 9950X3D. This also means the MacBook Neo beats every modern mobile processor from AMD, Intel and also Qualcomm, even though the upcoming Snapdragon X2 chips should be a bit faster. The A18 Pro is also slightly faster than Apple's own M3 generation in this scenario. Further reading: ASUS Executive Says MacBook Neo is 'Shock' to PC Industry

UPDATE: Amazon later published a blog post to address what it calls "inaccuracies" in the Financial Times report that the company's own AI tool Kiro caused two outages in an AWS service in December.

An anonymous Slashdot reader had shared this report from the Financial Times: Amazon's ecommerce business has summoned a large group of engineers to a meeting on Tuesday for a "deep dive" into a spate of outages, including incidents tied to the use of AI coding tools. The online retail giant said there had been a "trend of incidents" in recent months, characterized by a "high blast radius" and "Gen-AI assisted changes" among other factors, according to a briefing note for the meeting seen by the FT. Under "contributing factors" the note included "novel GenAI usage for which best practices and safeguards are not yet fully established."

"Folks, as you likely know, the availability of the site and related infrastructure has not been good recently," Dave Treadwell, a senior vice-president at the group, told employees in an email, also seen by the FT. The note ahead of Tuesday's meeting did not specify which particular incidents the group planned to discuss. [...] Treadwell, a former Microsoft engineering executive, told employees that Amazon would focus its weekly "This Week in Stores Tech" (TWiST) meeting on a "deep dive into some of the issues that got us here as well as some short immediate term initiatives" the group hopes will limit future outages.

He asked staff to attend the meeting, which is normally optional. Junior and mid-level engineers will now require more senior engineers to sign off any AI-assisted changes, Treadwell added. Amazon said the review of website availability was "part of normal business" and it aims for continual improvement. "TWiST is our regular weekly operations meeting with a specific group of retail technology leaders and teams where we review operational performance across our store," the company said.

An anonymous reader quotes a report from IEEE Spectrum: Worried that your latest ask to a cloud-based AI reveals a bit too much about you? Want to know your genetic risk of disease without revealing it to the services that compute the answer? There is a way to do computing on encrypted data without ever having it decrypted. It's called fully homomorphic encryption, or FHE. But there's a rather large catch. It can take thousands -- even tens of thousands -- of times longer to compute on today's CPUs and GPUs than simply working with the decrypted data. So universities, startups, and at least one processor giant have been working on specialized chips that could close that gap. Last month at the IEEE International Solid-State Circuits Conference (ISSCC) in San Francisco, Intel demonstrated its answer, Heracles, which sped up FHE computing tasks as much as 5,000-fold compared to a top-of the-line Intel server CPU.

Startups are racing to beat Intel and each other to commercialization. But Sanu Mathew, who leads security circuits research at Intel, believes the CPU giant has a big lead, because its chip can do more computing than any other FHE accelerator yet built. "Heracles is the first hardware that works at scale," he says. The scale is measurable both physically and in compute performance. While other FHE research chips have been in the range of 10 square millimeters or less, Heracles is about 20 times that size and is built using Intel's most advanced, 3-nanometer FinFET technology. And it's flanked inside a liquid-cooled package by two 24-gigabyte high-bandwidth memory chips—a configuration usually seen only in GPUs for training AI.

In terms of scaling compute performance, Heracles showed muscle in live demonstrations at ISSCC. At its heart the demo was a simple private query to a secure server. It simulated a request by a voter to make sure that her ballot had been registered correctly. The state, in this case, has an encrypted database of voters and their votes. To maintain her privacy, the voter would not want to have her ballot information decrypted at any point; so using FHE, she encrypts her ID and vote and sends it to the government database. There, without decrypting it, the system determines if it is a match and returns an encrypted answer, which she then decrypts on her side. On an Intel Xeon server CPU, the process took 15 milliseconds. Heracles did it in 14 microseconds. While that difference isn't something a single human would notice, verifying 100 million voter ballots adds up to more than 17 days of CPU work versus a mere 23 minutes on Heracles.

sziring shares a report from Business Insider: Silicon Valley has long competed for talent with ever-richer pay packages built around salary, bonus, and equity. Now, a fourth line item is creeping into the mix: AI inference. As generative AI tools become embedded in software development, the cost of running the underlying models -- known as inference -- is emerging as a productivity driver and a budget line that finance chiefs can't ignore.

Software engineers and AI researchers inside tech companies have already been jousting for access to GPUs, with this AI compute capacity being carefully parceled out based on which projects are most important. Now, some tech job candidates have begun asking about what AI compute budget they will have access to if they decide to join.

"I am increasingly asked during candidate interviews how much dedicated inference compute they will have to build with Codex," Thibault Sottiaux, engineering lead at OpenAI's Codex, the startup's AI coding service, wrote on X recently. He added that usage per user is growing much faster than overall user growth, a sign that AI compute is becoming even scarcer and more valuable. That scarcity is reshaping how engineers think about their work and pay. "The inference compute available to you is increasingly going to drive overall software productivity," said OpenAI President Greg Brockman.

The report cites a recent compensation submission from a software engineer that listed "Copilot subscription" as part of the pay and benefits. "OpenAI and Anthropic should create recruitment sites where their clients can advertise roles, listing the token budget for the job alongside the salary range," said Peter Gostev, AI capability lead at Arena, a startup that measures the performance of models.

Tomasz Tunguz of Theory Ventures predicts AI inference will be the fourth component of engineering compensation, alongside salary, bonus, and equity. "Will you be paid in tokens? In 2026, you likely will start to be," Tunguz said.

Qualcomm and Arduino have unveiled the Arduino Ventuno Q, a new AI-focused single-board computer built for robotics and edge systems. Engadget reports: Called the Arduino Ventuno Q, it uses Qualcomm's Dragonwing IQ8 processor along with a dedicated STM32H5 low-latency microcontroller (MCU). "Ventuno Q is engineered specifically for systems that move, manipulate and respond to the physical world with precision and reliability," the company wrote on the product page. The Ventuno Q is more sophisticated (and expensive) than Arduinio's usual AIO boards, thanks to the Dragonwing IQ8 processor that includes an 8-core ARM Cortex CPU, Adreno Arm Cortex A623 GPU and Hexagon Tensor NPU that can hit up ot 40 TOPs. It also comes with 16GB of LPDDR5 RAM, along with 64GB of eMMC storage and an M.2 NVME Gen.4 slot to expand that. Other features include Wi-Fi 6, Bluetooth 5.3, 2.5Gbps ethernet and USB camera support.

The Ventuno Q includes Arudino App Lab, with pre-trained AI models including LLMs, VLMs, ASR, gesture recognition, pose estimation and object tracking, all running offline. It's designed for AI systems that run entirely offline like smart kiosks, healthcare assistants and traffic flow analysis, along with Edge AI vision and sensing systems. It also supports a full robotics stack including vision processing combined with deterministic motor control for precise vision and manipulation. It's also ideal for education and research in areas like computer vision, generative AI and prototyping at the edge, according to Arduino. Further reading: Up Next for Arduino After Qualcomm Acquisition: High-Performance Computing

First, Hyundai "is discontinuing its most affordable electric sedan after just three years on the market," reports USA Today. After being introduced in 2022, the Hyundai Ioniq 6 "quickly gained the admiration of automotive critics because of its affordable pricing and capable performance specs." But now, Hyundai "is axing the most affordable versions of the EV, leaving consumers with only one Ioniq 6 option." Hyundai will continue to produce the Ioniq 6 N performance trim, which is the quickest and most powerful iteration of the Ioniq 6. It's also the most expensive. The South Korean automaker is getting rid of lower Ioniq 6 trims due to "disappointing sales and tariff considerations," according to Cars.com. Hyundai sold 10,478 Ioniq 6 models in 2025, dropping 15% from 12,264 units in 2024, a company sales report stated. Hyundai's Ioniq 6 is mainly produced in South Korea, so it faces high import tariffs.
Sales increased for their earlier IONIQ 5 model, reports the EV blog Electrek, "up 14% through the first two months of 2026, with 5,365 units sold... Meanwhile, IONIQ 6 sales slid 77% with only 229 units sold in February."

Elsewhere they report that Kia's EV6 and EV9 "didn't fare much better with sales down 53% (600 units sold) and 40% (819 units sold), respectively." Now a Kia spokesperson tells Car and Driver that the 2025 EV6 GT and 2026 EV9 GT "will be delayed until further notice." They attributed the move to "changing market conditions," but added that this delay "does not impact the availability of other trims in the EV6 and EV9 lineups."

More from Electrek: The news comes after Kia already said it was delaying the EV4, its entry-level electric sedan, "until further notice." It was expected to arrive in the US this year alongside the EV3, Kia's compact electric SUV that's already a top-seller in the UK, Europe, and other overseas markets.

While Hyundai didn't directly say it, since the EV3, EV4, EV6 GT, and Hyundai IONIQ 6 are built in Korea, the Trump administration's import tariffs and other policy changes are likely the biggest reason to blame here. Kia and Hyundai, like many others, are hesitant to bring new EVs to the US due to the changes. The IONIQ 6, EV6 GT, and EV9 GT join a string of other models that have either been postponed or canceled altogether.