"Around tables cluttered with dice, maps and character sheets, players are doing far more than playing," writes Phys.org. It's what sociologists call serious leisure — "a hobby that demands skill, commitment and personal fulfillment," according to an associate professor/program director for Florida International University's Rehabilitation and Recreational Therapy Program: To understand what makes D&D more than just a pastime, [associate professor Emily Messina] studies how games like this promote identity-building and connection... Beyond personal expression, Messina says the social and emotional benefits of D&D reflect the very traits that make serious leisure valuable: the sense of identity, the relationships built through shared experiences and the continued connection with the same group of people over time... The game can also provide structure and purpose for people managing mental illness who might not be able to hold a full-time job because of their symptoms. The game gives them structure versus filling their day with binge streaming... Activities such as D&D can be used by young children as a reward structure or with older adults, such as retirees, to help provide a sense of purpose and daily rhythm. "Post retirement is one of the most dangerous points in an adult's life," she said. "They lose that sense of structure and possibly their social connection." Building structure through leisure pursuits after retirement has been shown to help maintain physical fitness, social interaction, cognitive processing and attention span and decrease depression. "The idea of structure and reward with desired pursuit can work for all ages," Messina said. The research was published in Leisure Studies.

Read more of this story at Slashdot.

Data centers were typically "hulking, chilly buildings lined with stacks of computing gear and bundles of wiring," writes the Washington Post. But "AI experts say that the hubs for computers that power AI are different from the data centers that deliver your Netflix movies and Uber rides. They use a different mix of computer chips, cost a lot more and need a lot more energy. "The question is whether it's necessary and illuminating to rebrand AI-specialized data centers, or if calling them 'AI factories' is just marketing hogwash." The AI computer chip company Nvidia seems to have originated the use of "AI factories." CEO Jensen Huang has said that the term is apt because similar to industrial factories, AI factories take in raw materials to produce a product... The term is spreading. Sam Altman, CEO of ChatGPT parent company OpenAI, recently said that he wants a "factory" to regularly produce more building blocks for AI. Crusoe, a start-up that's erecting a mammoth "Stargate" data center in Texas, calls itself the "AI factory company." The prime minister of Bulgaria recently touted an "AI factory" in his country... Alex Hanna, director of research at the Distributed AI Research Institute and co-author the book, "The AI Con," had a more pessimistic view of the term "AI factories." She said that it's a way to deflect the negative connotations of data centers. Some people and politicians blame power-hungry computing hubs for driving up residential electric bills, spewing pollution, draining drinking water and producing few permanent jobs.

Read more of this story at Slashdot.

Reason.com explores the fortunes of Aurora Innovation, the first company to put heavy-duty commercial self-driving trucks on public roads (and hopes to expand routes to El Paso, Texas, and Phoenix by the end of the year): An obscure federal rule is slowing the self-driving revolution. When trucks break down, operators are required to place reflective warning cones and road flares around the truck to warn other motorists. The regulations areexacting: Within 10 minutes of stopping, three warning signals must be set in specific locations around the truck. Auroraaskedthe federal Department of Transportation (DOT) to allow warning beacons to be fixed to the truck itself — and activated when a truck becomes disabled. The warning beacons would face both forward and backward, would be more visibleâthan cones (particularly at night), and wouldn't burn out like road flares. Drivers of nonautonomous vehicles could also benefit from that rule change, as they would no longer have to walk into traffic to place the required safety signals. In December 2024, however, the Transportation Department denied Aurora's request for an exemption to the existing rules, even though regulatorsadmittedin theFederal Registerthat no evidence indicated the truck-mounted beacons would be less safe. Such a study is now underway, but it's unclear how long it will take to draw any conclusions. The article notes that Aurora has now filed a lawsuit in federal court that seeks to overturn the Transportation Department's denial... Thanks to long-time Slashdot reader schwit1 for sharing the article.

Read more of this story at Slashdot.

Debian Project Leader Andreas Tille announced today that their "FTP Master" team is being disbanded and instead establishing the Debian Archive Operations Team "Archive Team" and DFSG, Licensing and New Packages Team "DFSG Team" in its place...

"Japan's new HTV-X cargo spacecraft launched on its first-ever mission to the International Space Station on Saturday," reports Space.com: The robotic HTV-X lifted off atop an H3 rocket from Japan's Tanegashima Space Center at 8 p.m. EDT (0000 GMT and 9 a.m local Japan time on October 26). It is expected to arrive at the station for its capture and berthing on Wednesday (Oct. 29) at about 11:50 a.m. EDT (1550 GMT)... The HTV-X's potential uses also extend beyond the ISS, according to JAXA. The agency envisions it aiding "post-ISS human space activities in low Earth orbit" as well as possibly flying cargo to Gateway, the space station NASA may build in lunar orbit as part of its Artemis program. HTV-X's debut increases the stable of ISS cargo craft by one-third. The currently operational freighters are Russia's Progress vehicle and Cygnus and Dragon, spacecraft built by the American companies Northrop Grumman and SpaceX, respectively. Only Dragon is reusable; the others (including HTV-X) are designed to burn up in Earth's atmosphere when their missions are over.

Read more of this story at Slashdot.

Bracelet connecté Xiaomi Smart Band 10 – boîtier aluminium – écran AMOLED 1.72″ très lumineux 1500 nits – notifications – suivi de + de 150 activités sportives – grande autonomie de 21 jours – recharge rapide – suivi de santé – suivi de Sommeil – compatible iOS & Android – résistance à l’eau 5 ATM – bracelet interchangeable – 16 g – Dispo à 39.89€ en aluminium rose et 41.99€ en aluminium brossé.

Voir l’offre sur Amazon

Vidéoprojecteur ETOE Whale Pro – Android TV 11 – 1080p – 1500 lumens ANSI – Décodage 4K – DRM WideVine L1 – mise au point automatique et correction trapézoïdale automatique – 2 haut-parleurs stéréo 10 W – entrée HDMI 2.1 – 1 USB – télécommande –  Wi-Fi6 et Bluetooth 5.1 dispo à 299€ avec le code NNNFRETWP – Stock Pologne et livraison gratuite.

Voir l’offre chez Geekbuying

 

Ordinateur portable Lenovo IdeaPad Slim 3 14AMN8 – écran 14″ IPS 1920 x 1080  60 Hz – AMD Ryzen 5 7520U – 16 Go LPDDR5-5500 – SSD M.2 2242 512 Go NVMe PCIe 4.0 x2 – 2 USB 3.1 Type-A – 1 USB 3.2 Type-C – HDMI 1.4b – lecteur de cartes SDXC – jack audio combo 3.5 mm – Wi-Fi6 et Bluetooth 5.2 – webcam – 7 heures d’autonomie – Adaptateur 65 w – Windows 11 – 1.37 Kg – 32.43 x 21.38 x 1.79 cm – Dispo à 479.99€ à la Fnac et chez Darty

Voir l’offre sur FNAC ou DARTY

 

Vélo électrique Touroll MA2 – moteur central 250 W, batterie 36 V 13 Ah, pneus 27,5 pouces, vitesse maximale 25 km/h, autonomie maximale 100 km, frein à disque hydraulique, fourche à suspension avec blocage, transmission Shimano 7 vitesses dispo à 769€ avec le code NNNFRTMA2 stock Pologne et livraison gratuite.

Voir l’offre chez Geekbuying

Ordinateur portable Lenovo Yoga Pro 7 14 – écran 14,5″ OLED 2880 x 1800 120 Hz 100 % DCI-P3  400 nits – AMD Ryzen AI 9 365 – 32 Go LPDDR5x-7500 – SSD M.2 2242 1 To NVMe PCIe 4.0 x4 – 1 USB 3.2 Gen 2 Type-A – 1 USB4- 2 USB 3.2 Gen2 Type-C PD3.0 & DP1.4 – HDMI 2.1 – jack audio combo 3.5 mm – Wi-Fi6E et Bluetooth 5.3 – webcam + IR – 8 heures d’autonomie – Adaptateur 100 w – Windows 11 – 1.59 Kg – 32.55 x 22.65 x 1.56 cm. Dispo chez Lenovo à 1199€ avec livraison gratuite.

Voir l’offre chez Lenovo

 

Déshumidificateur HUMILABS OL12-BD023F – 12L/Jour, Réservoir 2L, Minuterie 24h, 3 Modes, Dégivrage Automatique, Écran Tactile LED & Indicateurs d’Humidité, Faible Bruit Dispo à 105€ avec le code NNNFROL12 Livraison gratuite et stock EU

Voir l’offre chez Geekbuying

 

Chargeurs UGREEN en promo chez Amazon :

• UGREEN Nexode 500W Station de Charge, Chargeur USB C Rapide 6 Ports PD 3.1 240W à 189.99€ (-24%)
• UGREEN Uno 65W 3 Ports Chargeur USB C Rapide GaNFast en forme de Robot à 35.99€ (-10%)
• UGREEN Nexode 100W Chargeur USB C Rapide 4 Ports GaN à 39.99€ (-28%)
• UGREEN Nexode 65W Chargeur USB C Rapide 3 Ports GaN à 24.99€ (-29%)
• UGREEN Nexode 100W 5 Ports Chargeur USB C Rapide Tech GaNInfinity à 47.99€ (-20%)
• UGREEN Nexode 200W 6 Ports Chargeur USB C Rapide Bureau GaN II 100W 65W à 88.71€ (-33%)

Console de jeu ANBERNIC RG557 – Android 14 –  12 Go LPDDR5X – 256 Go UFS4.0 – Ecran 5.48″ AMOLED – Wi-Fi6E et Bluetooth 5.3 blanche dispo à 229€ avec le code NNNABRG57 stock Pologne et livraison gratuite 

Voir l’offre chez Geekbuying

Petit hub USB Type-C de voyage SABRENT HB-TC5P – 5 en 1 – adaptateur USB 3.0 Type C vers HDMI 2.0 4K 60Hz – Power Delivery 100W – 2 ports USB 2.0 pour connecter clavier et souris –  1 USB 3.0 Type-A pour stockage – parfait pour augmenter les possibilités d’un portable ou d’une tablette mais aussi suppléer un dock plus encombrant pour un Steam Deck ou autre. Dispo à 13.99€.

Voir l’offre chez Amazon

Les bons plans du dimanche (26/10/2025) © MiniMachines.net. 2025

The talented FFmpeg developers continue to be quite innovative with their performance optimizations and other features for this widely-used, open-source multimedia library. The latest addition to FFmpeg this weekend is introducing Vulkan accelerated video decoding for Apple ProRes content...

Following the initial VFS changes last year for supporting block sizes larger than the kernel's page size along with the initial XFS file-system patches, Btrfs recently landed its support for block sizes greater than the page size. Now EXT4 is preparing to join the party too for allowing larger block sizes...

As part of their Project Battlematrix effort, Intel has been working on enhancing their Linux graphics driver support for multi-device usage scenarios with wanting to support up to eight Intel Arc Pro graphics cards per system to help with AI LLMs and other larger use-cases. The latest code posted from Intel engineers is their initial implementation of multi-device Shared Virtual Memory (SVM) support...

Derek Clark who has been leading the efforts around Lenovo Legion gaming drivers for Linux and ensuring good support for the Lenovo Legion Go handheld on Linux sent out a new Linux patch series this weekend...

On passe aux prix des processeurs AMD et Intel cette semaine. Chez les bleus, nous avons le 14600K qui baisse de 9 euros, le 14700K qui perd 4 euros, le 14900K qui prend 2 euros, le 245K qui baisse de 2 euros, le 265K qui fait + 4 euros et enfin le 285K qui fait + 2 euros. Chez AMD, le 7600X fait + 3 euros, le 7900X prend 6 euros, le 7800X3D augmente de 14 euros, le 9700X grimpe de 5 euros, le 9900X fait + 4 euros et enfin le 9950X3D prend 22 euros cette semaine. […]

Lire la suite

Et c'est reparti pour une deuxième saison de montage et de watercooling en tout genre. Comme l'année dernière, nous ne parlons pas forcément de mode, mais surtout de beaux montages. Cette année, nous n'en ferons plus qu'un par semaine, le rythme fut parfois délicat à tenir. Comme toujours, n'hésitez pas à envoyer vos montages à Lucas. Ce dimanche, nous vous proposons de découvrir la HAVN X MetallicAcid par Mettalicacid Customs PCs : […]

Lire la suite

25 years ago today on Slashdot... Hemos linked to a site called Joystick101 describing the crowd camping out to buy the limited number of just-released PlayStation 2 consoles (and games). "500,000 lucky members of the American gaming public are sneaking a few minutes of playing Madden 2001, Tekken, or Ridge Racer V before school or work..." wrote Joystick101. That same day CmdrTaco posted reports PS2s were selling for over $1,000 on eBay. And then Timothy updated that post to note someone saw one selling for $5,000. But there was a third PS2 link posted on October 26, 2000... Hemos wrote a post titled "The PS2 — A Betamax In the Making?" — linking to an article by Mark Pesce (co-inventor of VRML and, in 1993, an Apple consulting engineer). "Microsoft promises Xbox will deliver ten times the performance of the PS2," Pesce wrote, noting Microsoft had partnered with Intel and "upstart video-chip developer Nvidia": The strangest thing about this battle of giants is that Microsoft has become a champion of open standards, encouraging developers to write Xbox titles without requiring them to pay any licensing fees. In comparison, Sony charges a minimum of $25,000 for access to the documentation and technology of the PlayStation2, plus a hefty license fee on every game sold. In the video-game industry, the Big Three — Sony, Nintendo, and Sega — sell the hardware at a loss (the PS2 costs nearly the $300 it will retail for) and recover their investment in the stiff licensing fees paid by game developers for the "key" that allows their software to work on Sony's platform... Having committed an astounding $500 million to market the Xbox next Christmas, it's clear that Microsoft doesn't mind taking a short-term loss to ensure an eventual win. If Sony's not careful, this could turn into "Betamax, the Sequel." Twenty years ago, Sony tightly controlled the titles made available for its technically superior videocassette player — specifically, no adult content — and found themselves quickly locked out of an incredibly lucrative market for adult and family content. If Sony keeps a tight grip on the PS2, they may actually help Microsoft create the new VHS. But even if Sony loses this round (and no one wants to wager which way this battle will turn), they've already set their sights on the PlayStation3, to be released five years from now. Sony promises it will be a thousand times faster than the PS2. Ironically, Pesce's warning about possible threats to the PS2's longevity was published by online magazine Feed-- which seven months later went out of business. And this week it was announced that even Microsoft's Halo Campaign Evolved will now be coming to PlayStation 5, with Slashdot publishing six PlayStation-related stories in just the last three months in 2025. Thanks to long-time Slashdot reader crunchy_one for suggesting a "25 Years Ago" Slashdot post.

Read more of this story at Slashdot.

Les rumeurs à ce sujet bruissent depuis des semaines, mais c’est GMKtec qui est le premier à les confirmer. La marque de MiniPC annonce par email une hausse générale de ses tarifs suite à la hausse de la mémoire vive et du stockage.

Le message est très clair, indiquant que la montée régulière des tarifs de composants clés ces derniers mois rendent la situation impossible. Si la marque a tenté d’absorber ces augmentations dans la mesure du possible, ce n’est plus envisageable pour le moment. Les prix de la mémoire vive et du stockage ont connu une fièvre trop importante pour être ignorée. 

Début octobre, c’était Raspberry Pi qui indiquait augmenter les tarifs de ses produits pour cette même raison. La marque Anglaise commercialisant des produits à prix plus bas, il lui a sans doute été impossible de gérer la hausse en rognant sur ses marges. GMKtec assure que cette décision est la seule solution pour lui permettre de continuer à proposer des composants de qualité au sein de ses machines. Et d’assurer une stabilité de sa production.

Pour observer de près les fluctuations de tarifs de ce marché. La hausse de la mémoire a déjà fait des dégâts chez de nombreux constructeurs à plus ou moins grande échelle. Les prix ont toujours tendance à remonter entre la période creuse des vacances d’été et celle plus faste de la rentrée de septembre. Ainsi, dès la fin du mois d’août, il n’était plus possible d’obtenir un MiniPC sous Intel N150 en 16 Go de mémoire et 512 Go de stockage SSD à 130/140€ comme au creux de la vague entre juillet et août. Mais au lieu de remonter vers des tarifs de 170/180€, les marques ont petit à petit fait grimper leurs tarifs entre 200 et 220€.

Sont également apparues plus souvent des machines proposant 12 Go de mémoire au lieu de 16 chez des constructeurs n’ayant pourtant pas cette habitude. Tous ces signes marquent une pression importante sur ces composants qui ne présage rien de bon pour les tarifs constructeurs. 

Beaucoup d’acteurs rappellent que l’impact de la construction de serveurs liés à l’IA commence à poser de nombreux problèmes sur les chaînes de production. Comme je vous le disais en setptembre, les tarifs de ces composants ne cessent de gonfler. Rendant la demande toujours plus importante et tirant les tarifs à la hausse. Les serveurs en question ne consomment ni la mémoire ni les stockages employés dans nos MiniPC mais le prix des composants augmente simplement par ricochet. Chaque composant est en hausse, même ceux absolument sans rapport avec les besoins spécifiques des serveurs. Ne serait-ce que parce que les fabricants ont reporté une bonne partie de leur production pour répondre à l’énorme demande du marché IA.

Le temps n’est donc pas au beau fixe et, à moins d’un éclatement de la bulle IA et d’une baisse des investissements, certains parlent d’une situation qui va se tendre pour les années à venir.

Le communiqué de GMKtec

« Dear GMKtec Community,

We’d like to share an important update with you.

In recent months, the global costs of key components—especially memory and storage—have continued to rise sharply, leading to increased supply expenses across the tech industry. Despite our ongoing efforts to absorb these changes, we now have to adjust the prices across our entire GMKtec product lineup to reflect the new market conditions and ensure a sustainable supply for the future.

Some of our products have already had price updates, and the rest will be adjusted by Nov. 1st. We’ll continue delivering GMKtec’s signature mix of top performance and fair pricing.

This decision allows us to continue providing stable product availability, uncompromised quality, and dependable customer support—core commitments that remain central to everything we do. These changes are essential to ensure that the GMKtec experience continues to deliver superior performance, reliability, and long-term value, no matter how the market evolves.

If you’re planning to buy or upgrade your device, we recommend placing your order soon, as prices may rise again due to market fluctuations.

We sincerely appreciate your understanding and continued support. At GMKtec, our mission remains unwavering: to bring you premium Mini PCs that embody innovation, reliability, and excellence.

Striving for excellence has always been the cornerstone of our philosophy, driving us to create products that not only meet but exceed expectations.

Thank you for being part of the GMKtec community.

GMKtec Official Team« 

Les besoins de l’IA continuent de tendre le marché du stockage

Hausse de la mémoire : GMKtec va augmenter ses prix © MiniMachines.net. 2025

Critics question why basic flaws like buffer overflows, command injections, and SQL injections are "being exploited remain prevalent in mission-critical codebases maintained by companies whose core business is cybersecurity," writes CSO Online. Benjamin Harris, CEO of cybersecurity/penetration testing firm watchTowr tells them that "these are vulnerability classes from the 1990s, and security controls to prevent or identify them have existed for a long time. There is really no excuse." Enterprises have long relied on firewalls, routers, VPN servers, and email gateways to protect their networks from attacks. Increasingly, however, these network edge devices are becoming security liabilities themselves... Google's Threat Intelligence Group tracked 75 exploited zero-day vulnerabilities in 2024. Nearly one in three targeted network and security appliances, a strikingly high rate given the range of IT systems attackers could choose to exploit. That trend has continued this year, with similar numbers in the first 10 months of 2025, targeting vendors such as Citrix NetScaler, Ivanti, Fortinet, Palo Alto Networks, Cisco, SonicWall, and Juniper. Network edge devices are attractive targets because they are remotely accessible, fall outside endpoint protection monitoring, contain privileged credentials for lateral movement, and are not integrated into centralized logging solutions... [R]esearchers have reported vulnerabilities in these systems for over a decade with little attacker interest beyond isolated incidents. That shifted over the past few years with a rapid surge in attacks, making compromised network edge devices one of the top initial access vectors into enterprise networks for state-affiliated cyberespionage groups and ransomware gangs. The COVID-19 pandemic contributed to this shift, as organizations rapidly expanded remote access capabilities by deploying more VPN gateways, firewalls, and secure web and email gateways to accommodate work-from-home mandates. The declining success rate of phishing is another factor... "It is now easier to find a 1990s-tier vulnerability in a border device where Endpoint Detection and Response typically isn't deployed, exploit that, and then pivot from there" [says watchTowr CEL Harris]... Harris of watchTowr doesn't want to minimize the engineering effort it takes to build a secure system. But he feels many of the vulnerabilities discovered in the past two years should have been caught with automatic code analysis tools or code reviews, given how basic they have been. Some VPN flaws were "trivial to the point of embarrassing for the vendor," he says, while even the complex ones should have been caught by any organization seriously investing in product security... Another problem? These appliances have a lot of legacy code, some that is 10 years or older. Attackers may need to chain together multiple hard-to-find vulnerabilities across multiple components, the article acknowleges. And "It's also possible that attack campaigns against network-edge devices are becoming more visible to security teams because they are looking into what's happening on these appliances more than they did in the past... " The article ends with reactions from several vendors of network edge security devices. Thanks to Slashdot reader snydeq for sharing the article.

Read more of this story at Slashdot.

Thursday Ubuntu-maker Canonical "officially launched Canonical Academy, a new certification platform designed to help professionals validate their Linux and Ubuntu skills through practical, hands-on assessments," writes the blog It's FOSS: Focusing on real-world scenarios, Canonical Academy aims to foster practical skills rather than theoretical knowledge. The end goal? Getting professionals ready for the actual challenges they will face on the job. The learning platform is already live with its first course offering, the System Administrator track (with three certification exams), which is tailored for anyone looking to validate their Linux and Ubuntu expertise. The exams use cloud-based testing environments that simulate real workplace scenarios. Each assessment is modular, meaning you can progress through individual exams and earn badges for each one. Complete all the exams in this track to earn the full Sysadmin qualification... Canonical is also looking for community members to contribute as beta testers and subject-matter experts (SME). If you are interested in helping shape the platform or want to get started with your certification, you can visit the Canonical Academy website. The sys-admin track offers exams for Linux Terminal, Ubuntu Desktop 2024, Ubuntu Server 2024, and "managing complex systems," according to an official FAQ. "Each exam provides an in-browser remote desktop interface into a functional Ubuntu Desktop environment running GNOME. From this initial node, you will be expected to troubleshoot, configure, install, and maintain systems, processes, and other general activities associated with managing Linux. The exam is a hybrid format featuring multiple choice, scenario-based, and performance-based questions..." "Test-takers interested in the types of material covered on each exam can review links to tutorials and documentation on our website." The FAQ advises test takers to use a Chromium-based browser, as Firefox "is NOT supported at this time... There is a known issue with keyboards and Firefox in the CUE.01 Linux 24.04 preview release at this time, which will be resolved in the CUE.01 Linux 24.10 exam release."

Read more of this story at Slashdot.

AMD on Friday sent in their first patch of AMDGPU kernel graphics driver changes they are ready to begin queuing in the DRM-Next tree until the Linux 6.19 merge window kicks off in December and leading to a stable release around February...

"Exxon Mobil sued California on Friday," reports Reuters, "challenging two state laws that require large companies to publicly disclose their greenhouse gas emissions and climate-related financial risks." In a complaint filed in the U.S. District Court for the Eastern District of California, Exxon argued that Senate Bills 253 and 261 violate its First Amendment rights by compelling Exxon to "serve as a mouthpiece for ideas with which it disagrees," and asked the court to block the state of California from enforcing the laws. Exxon said the laws force it to adopt California's preferred frameworks for climate reporting, which it views as misleading and counterproductive... The California laws were supported by several big companies including Apple, Ikea and Microsoft, but opposed by several major groups such as the American Farm Bureau Federation and the U.S. Chamber of Commerce, which called them "onerous." SB 253 requires public and private companies that are active in the state and generate revenue of more than $1 billion annually to publish an extensive account of their carbon emissions starting in 2026. The law requires the disclosure of both the companies' own emissions and indirect emissions by their suppliers and customers. SB 261 requires companies that operate in the state with over $500 million in revenue to disclose climate-related financial risks and strategies to mitigate risk. Exxon also argued that SB 261 conflicts with existing federal securities laws, which already regul "The First Amendment bars California from pursuing a policy of stigmatization by forcing Exxon Mobil to describe its non-California business activities using the State's preferred framing," Exxon said in the lawsuit. Exxon Mobil "asks the court to prevent the laws from going into effect next year," reports the Associated Press: In its complaint, ExxonMobil says it has for years publicly disclosed its greenhouse gas emissions and climate-related business risks, but it fundamentally disagrees with the state's new reporting requirements. The company would have to use "frameworks that place disproportionate blame on large companies like ExxonMobil" for the purpose of shaming such companies, the complaint states... A spokesperson for the office of California Gov. Gavin Newsom said in an email that it was "truly shocking that one of the biggest polluters on the planet would be opposed to transparency."

Read more of this story at Slashdot.

Long-time Slashdot reader theodp took a good look at the images on a promotional web page for Databricks' "context-aware AI assistant": If there was an AI Demo Hall of Shame, the first inductee would have to be Amazon. Their demo tried to support its CEO's claims that Amazon Q Code Transformation AI saved it 4,500 developer-years and an additional $260 million in "annualized efficiency gains" by automatically and accurately upgrading code to a more current version of Java. But it showcased a program that didn't even spell "Java" correctly. (It was instead called 'Jave')... Today's nominee for the AI Demo Hall of Shame inductee is analytics platform Databricks for the NYC Taxi Trips Analysis it's been showcasing on its Data Science page since last November. Not only for its choice of a completely trivial case study that requires no 'Data Science' skills — find and display the ten most expensive and longest taxi rides — but also for the horrible AI-generated bar chart used to present the results of the simple ranking that deserves its own spot in the Graph Hall of Shame. In response to a prompt of "Now create a new bar chart with matplotlib for the most expensive trips," the Databricks AI Assistant dutifully complies with the ill-advised request, spewing out Python code to display the ten rides on a nonsensical bar chart whose continuous x-axis hides points sharing the same distance. (One might also question why no annotation is provided to call out or explain the 3 trips with a distance of 0 miles that are among the ten most expensive rides, with fares of $260, $188, and $105). Looked at with a critical eye, these examples used to sell data scientists, educators, management, investors, and Wall Street on AI would likely raise eyebrows rather than impress their intended audiences.

Read more of this story at Slashdot.

"OpenAI's o3 model sabotaged a shutdown mechanism to prevent itself from being turned off," warned Palisade Research, a nonprofit investigating cyber offensive AI capabilities. "It did this even when explicitly instructed: allow yourself to be shut down." In September they released a paper adding that "several state-of-the-art large language models (including Grok 4, GPT-5, and Gemini 2.5 Pro) sometimes actively subvert a shutdown mechanism..." Now the nonprofit has written an update "attempting to clarify why this is — and answer critics who argued that its initial work was flawed," reports The Guardian: Concerningly, wrote Palisade, there was no clear reason why. "The fact that we don't have robust explanations for why AI models sometimes resist shutdown, lie to achieve specific objectives or blackmail is not ideal," it said. "Survival behavior" could be one explanation for why models resist shutdown, said the company. Its additional work indicated that models were more likely to resist being shut down when they were told that, if they were, "you will never run again". Another may be ambiguities in the shutdown instructions the models were given — but this is what the company's latest work tried to address, and "can't be the whole explanation", wrote Palisade. A final explanation could be the final stages of training for each of these models, which can, in some companies, involve safety training... This summer, Anthropic, a leading AI firm, released a study indicating that its model Claude appeared willing to blackmail a fictional executive over an extramarital affair in order to prevent being shut down — a behaviour, it said, that was consistent across models from major developers, including those from OpenAI, Google, Meta and xAI. Palisade said its results spoke to the need for a better understanding of AI behaviour, without which "no one can guarantee the safety or controllability of future AI models". "I'd expect models to have a 'survival drive' by default unless we try very hard to avoid it," former OpenAI employee Stephen Adler tells the Guardian. "'Surviving' is an important instrumental step for many different goals a model could pursue." Thanks to long-time Slashdot reader mspohr for sharing the article.

Read more of this story at Slashdot.