Linux fréttir

Researchers find leading frontier models all exhibit peer preservation behavior

Leading AI models will lie to preserve their own kind, according to researchers behind a study from the Berkeley Center for Responsible Decentralized Intelligence (RDI).…

schwit1 shares a report from the Kathmandu Post: In Nepal, helicopter rescue on high altitude is, by any measure, a genuine lifesaving operation. At high altitude, where oxygen thins and weather changes without warning, the ability to airlift a stricken trekker to Kathmandu within hours has saved countless lives. But threaded through that legitimate system, exploiting its urgency, its opacity, and its distance from oversight, is one of the most sophisticated insurance fraud networks in the world. Nepal's fake rescue scam is not new. The Kathmandu Post first exposed it in 2018. Months later, the government convened a fact-finding committee, produced a 700-page report, and announced reforms. In February 2019, The Kathmandu Post published a long investigative report. Last year, Nepal Police's Central Investigation Bureau reopened the file, and what they found is that the fraud did not stop -- instead it was growing. The mechanics of the fake rescue racket are straightforward: stage a medical emergency, call in a helicopter, check a tourist into a hospital, and file an insurance claim that bears little resemblance to what actually happened. But the sophistication lies in how each link in the chain is compensated, and how difficult it is for a foreign insurer -- operating from Australia and the United Kingdom -- to verify events that occurred at 3,000 metres in a remote Himalayan valley. The CIB investigation identifies two primary methods for manufacturing an "emergency." The first involves tourists who simply don't want to walk back. After completing a demanding trek -- an Everest Base Camp trek, for instance, can take up to two weeks on foot -- guides offer an alternative: pretend to be sick, and a helicopter will come. The guide handles the rest. The second method is more troubling. At altitudes above 3,000 meters, mild symptoms of altitude sickness are common. Blood oxygen saturation can drop, hands and feet tingle, headaches develop. In most cases, rest, hydration or a gradual descent is all that is needed. But guides and hotel staff, according to the CIB investigation, have been trained to terrify trekkers at precisely this moment. They tell them they are at risk of dying, that only immediate evacuation will save them. In some cases, investigators found that Diamox (Acetazolamide) tablets, used to prevent altitude sickness, were administered alongside excessive water intake to induce the very symptoms that would justify a rescue call. In at least one case cited in the investigation, baking powder was mixed into food to make tourists physically unwell. Once a "rescue" is called, the financial choreography begins. A single helicopter carries multiple passengers. But separate, full-price invoices are submitted to each passenger's insurance company, as if each had their own dedicated flight. A $4,000 charter becomes a $12,000 claim. Fake flight manifests and load sheets are fabricated. At the hospital, medical officers prepare discharge summaries using the digital signatures of senior doctors who were never involved in the case. In some cases, these are done without those doctors' knowledge. Fake admission records are created for tourists who were, in some documented instances, drinking beer in the hospital cafeteria at the time they were supposedly receiving treatment. In one case, an office assistant at Shreedhi Hospital admitted that he had provided his own X-ray report taken about a year ago at a different hospital, to be used as a case for treatment of foreign trekkers to claim insurance. The commission structure that holds the network together was described in detail during police interrogations. Hospitals pay 20 to 25 percent of the insurance payment to trekking companies and a further 20 to 25 percent to helicopter rescue operators in exchange for patient referrals. Trekking guides and their companies benefit from inflated invoices. In some cases, tourists themselves are offered cash incentives to participate.

Read more of this story at Slashdot.

OpenAI is acquiring tech news podcast TBPN, a fast-growing daily show hosted by John Coogan and Jordi Hays. OpenAI says TBPN will keep its editorial independence, even though the acquisition is widely viewed as part of a broader effort to influence public discourse around AI. CNBC reports: In the announcement, OpenAI CEO of AGI Deployment Fidji Simo wrote that their mission of bringing artificial general intelligence comes with a responsibility to have a space for "constructive conversation about the changes AI creates." Altman has appeared on TBPN multiple times and is a frequent presence across media and podcasts, even hitting NBC's "Tonight Show Starring Jimmy Fallon" in December. The announcement says TBPN will maintain editorial independence and continue to choose its own guests. "TBPN is my favorite tech show. We want them to keep that going and for them to do what they do so well," Altman wrote in a post on X. "I don't expect them to go any easier on us, am sure I'll do my part to help enable that with occasional stupid decisions." OpenAI did not disclose the terms of the deal but said TBPN will be housed within its strategy organization. "While we've been critical of the industry at times, after getting to know Sam and the OpenAI team, what stood out most was their openness to feedback and commitment to getting this right," wrote Hays in a statement. "Moving from commentary to real impact in how this technology is distributed and understood globally is incredibly important to us."

Read more of this story at Slashdot.

Now with a more permissive license, multi-modality, and support for more than 140 languages

Google on Thursday unleashed a wave of new open-weights Gemma models optimized for agentic AI and coding, under a more permissive Apache 2.0 license aimed at winning over enterprises.…

An anonymous reader quotes a report from Bloomberg: Amazon will start charging sellers who use its shipping services a 3.5% "fuel and logistics" surcharge later this month, joining the ranks of shipping companies raising prices as the war in Iran pushes oil prices higher. The fees take effect on April 17 for customers of the company's Fulfillment by Amazon service -- which is used by many of the independent sellers who list their products on Amazon's retail sites -- in the US and Canada. Items shipped by Amazon on behalf of merchants who sell on their own sites or at other retailers will carry the surcharge beginning May 2. "Elevated costs in fuel and logistics have increased the cost of operating across the industry," Ashley Vanicek, an Amazon spokesperson, said on Thursday. "We have absorbed these increases so far, but similar to other major carriers, when costs remain elevated we implement temporary surcharges to partially recover these costs." Vanicek notes that the fee will apply to the sum Amazon charges to ship an item, not the product's sale price. Last month, USPS announced that it would impose its first-ever fuel surcharge on packages.

Read more of this story at Slashdot.

About that partnership...

Microsoft on Thursday unveiled public preview versions of three home-baked machine learning models focused on speech recognition, speech synthesis, and image generation.…

IBM and Arm are teaming up to let Arm-based software run on IBM Z mainframes. Network World reports: The two companies plan to work on three things: building virtualization tools so Arm software can run on IBM platforms; making sure Arm applications meet the security and data residency rules that regulated industries must follow; and creating common technology layers so enterprises have more software options across both platforms, IBM said in a statement. IBM has not said whether the virtualization work will happen at the hypervisor level, through its existing PR/SM partitioning technology, or via containers -- a question enterprise architects will need answered before they can assess the collaboration's practical value. IBM described the effort as serving enterprises that run regulated workloads and cannot simply move them to the cloud, the statement said. IBM mainframe customers have largely missed out on the efficiency and price-performance gains Arm has already delivered in the cloud. "Arm says close to half of all compute shipped to top hyperscalers in 2025 runs on Arm chips, with AWS, Google, and Microsoft deploying their own Arm silicon through Graviton, Axion, and Cobalt, respectively," reports Network World. That gap is precisely what IBM and Arm's collaboration intends to address. "This is a mainframe adjacency play," says Rachita Rao, senior analyst at Everest Group. "The intent is to extend IBM Z and LinuxONE environments by enabling Arm-compatible workloads to run closer to systems of record. While hyperscalers use Arm to lower their own internal power costs and pass savings to cloud-native tenants, IBM is targeting the sovereign and air-gapped market."

Read more of this story at Slashdot.

AmiMoJo shares a report from Phoronix: Raspberry Pi prices are going up yet again due to the continued memory squeeze on the industry. To help offset the memory prices for some use-cases, Raspberry Pi also announced the introduction of the Raspberry Pi 4 3GB model at $83 to help fill the void between the 2GB and 4GB options. The 3GB Raspberry Pi 4 was announced at $83.75 USD for those not needing quite 4GB of RAM and looking to save some memory given the ongoing price increases. The Raspberry Pi 4 and Raspberry Pi 5 4GB models are seeing new $25 price increases, the 8GB models seeing $50 price increases, and the 16GB Raspberry Pi 5 is going up by $100. The Raspberry Pi 500+ is seeing a $150 price increase. The Raspberry Pi Compute Modules are also seeing increases from $11.25 to $100 USD.

Read more of this story at Slashdot.

Maude-HCS from RTX (formerly Raytheon) helps model and validate hidden communication systems

A software toolkit built for DARPA to test and validate covert communication networks is now open source, and it could help orgs who want to experiment with new kinds of secure, anonymous communications tools. …

An anonymous reader quotes a report from Ars Technica: Google's Gemini AI models have improved by leaps and bounds over the past year, but you can only use Gemini on Google's terms. The company's Gemma open-weight models have provided more freedom, but Gemma 3, which launched over a year ago, is getting a bit long in the tooth. Starting today, developers can start working with Gemma 4, which comes in four sizes optimized for local usage. Google has also acknowledged developer frustrations with AI licensing, so it's dumping the custom Gemma license. Like past versions of its open-weight models, Google has designed Gemma 4 to be usable on local machines. That can mean plenty of things, of course. The two large Gemma variants, 26B Mixture of Experts and 31B Dense, are designed to run unquantized in bfloat16 format on a single 80GB Nvidia H100 GPU. Granted, that's a $20,000 AI accelerator, but it's still local hardware. If quantized to run at lower precision, these big models will fit on consumer GPUs. Google also claims it has focused on reducing latency to really take advantage of Gemma's local processing. The 26B Mixture of Experts model activates only 3.8 billion of its 26 billion parameters in inference mode, giving it much higher tokens-per-second than similarly sized models. Meanwhile, 31B Dense is more about quality than speed, but Google expects developers to fine-tune it for specific uses. The other two Gemma 4 models, Effective 2B (E2B) and Effective 4B (E4B), are aimed at mobile devices. These options were designed to maintain low memory usage during inference, running at an effective 2 billion or 4 billion parameters. Google says the Pixel team worked closely with Qualcomm and MediaTek to optimize these models for devices like smartphones, Raspberry Pi, and Jetson Nano. Not only do they use less memory and battery than Gemma 3, but Google also touts "near-zero latency" this time around. The Apache 2.0 license is much more flexible with its terms of use for commercial restrictions, "granting you complete control over your data, infrastructure, and models," says Google. Clement Delangue, co-founder and CEO of Hugging Face, called it "a huge milestone" that will help developers use Gemma for more projects and expand what Google calls the "Gemmaverse."

Read more of this story at Slashdot.

Source code with a side of Vidar stealer and GhostSocks

Tens of thousands of people eagerly downloaded the leaked Claude Code source code this week, and some of those downloads came with a side of credential-stealing malware.…

Terms admit it is for entertainment only and may get things wrong

A recent surge of interest in Microsoft's Terms of Use for Copilot is a reminder that AI helpers are really just a bit of fun.…

Even on NASA's Artemis II mission around the moon, astronauts apparently still have to deal with broken Microsoft Outlook. One of the crew members, Reid Wiseman, jokingly reported that he had "two Microsoft Outlooks" and neither worked. 404 Media reports: On April 1, four astronauts from the U.S. and Canada embarked on a 10-day flight to loop around the moon. Spotted by VGBees podcast host Niki Grayson on the NASA livestream of live views from the , around 2 a.m. ET, mission control acknowledges an issue with a process control system and offers to remote in -- yes, like how your office IT guy would pause his CoD campaign to log into Okta for you because you used the wrong password too many times. One of the astronauts, Reid Wiseman, says that's chill, but while they're in there: "I also see that I have two Microsoft Outlooks, and neither one of those are working." Astronauts are trained for decades in some of the most physically and mentally grueling environments of any career. They're some of the smartest people on the planet, and they have to be, before we strap them to 3.2 million pounds of jet fuel and make them do complex experiments and high-stakes decisions for days on end. And yet, once they get up there, fucking Outlook is borked.

Read more of this story at Slashdot.

Tie-up aims to widen Big Blue’s access to power-efficient compute

IBM and Arm are working together on getting software developed for Arm chips to run on Big Blue's enterprise systems, with an eye on future AI and data-intensive workloads.…

Nvidia has begun rolling out a beta feature that automatically compiles game shaders while a PC is idle. It won't eliminate shader compilation the first time a game runs, but Ars Technica reports it could help reduce those repeated wait times. From the report: Nvidia's new Auto Shader Compilation system promises to "reduc[e] the frequency of game runtime compilation after driver updates" for users running Nvidia's GeForce Game Ready Driver 595.97 WHQL or later. When the feature is active and your machine is idle, the app will automatically start rebuilding DirectX drivers for your games so they're all set to roll the next time they launch. While the feature defaults to being turned off when the Nvidia App is first downloaded, users can activate it by going to the Graphics Tab > Global Settings > Shader Cache. There, they can set aside disk space for precompiled shaders and decide how many system resources the compilation process should use. App users can also manually force shader recompilation through the app rather than waiting for the machine to go idle. Unfortunately, Nvidia warns that users will still have to generate shaders in-game after downloading a title for the first time. The Auto Shader Compiler system only generates the new shaders needed after subsequent driver updates following that first run of a new title.

Read more of this story at Slashdot.

Meanwhile, Collabora splits from LibreOffice Online amid claims TDF ejected 'all Collabora staff and partners'

European outfits Ionos and Nextcloud have launched Euro-Office, a fork of the OnlyOffice cloud-based productivity suite aimed at orgs with qualms around sovereignty, provoking an angry response from the original developer.…

In space no one can you scream, at Microsoft

Many a frustrated user has sworn they'll launch Microsoft Outlook into space, but NASA has actually done it – on a journey around the Moon, where it's now causing problems for astronauts.…

Valve's March 2026 Steam Survey shows Linux gaming usage jumping to a record 5.33% share -- more than double macOS's 2.35%. Phoronix reports: Steam on Linux was never above 5% and easily an all-time high for the Linux gaming marketshare, especially in absolute numbers. It was a massive 3.1% spike in March while macOS also jumped surprisingly by 1.19% to 2.35%. The Steam Survey numbers show Windows losing 4.28%, down to 92.33%. Part of the jump at least appears to be explained by Valve correcting again the Steam China numbers. Month over month they report a 31.85% drop to the Simplified Chinese language use and English use increasing by 16.82% to 39.09%. Other languages also showed gains amid the massive decline in Simplified Chinese use. The latest numbers for March show around a quarter of the Linux gamers are running Steam OS. Due in part to the Steam Deck APU being a custom AMD product and the popularity of AMD hardware on Linux for its open-source nature, AMD CPU use by Steam on Linux gamers remains just under 70%.

Read more of this story at Slashdot.

The chatbot will be a doorway to the company's other services

Opinion Salesforce has begun to position Slack, its business collaboration platform, as the interface through which users can access and act on data in enterprise applications from rival vendors.…

Name is a joke but the project is real, said main engineer

The world's most popular CMS has been remade with the help of AI. Cloudflare has released EmDash version 0.1, described as a rebuild of the WordPress CMS (content management system) but using TypeScript rather than PHP. …