A new meta-analysis found nutrients in food decreased over the last 40 years, reports the Washington Post. "Many of humanity's most important crops — including wheat, potatoes, beans — contain fewer vitamins and minerals than they did a generation ago." "The invisible culprit behind this damaging phenomenon? Carbon dioxide pollution." Surging concentrations of carbon in the atmosphere, caused largely by burning fossil fuels, have produced potent changes in the way plants grow — from increasing their sugar content to depleting essential nutrients like zinc... "The diets we eat today have less nutritional density than what our grandparents ate, even if we eat exactly the same thing," said Kristie Ebi, a professor at the University of Washington's Center for Health and the Global Environment. People in wealthy countries with strong health care systems will have many tools to cope with the change, experts said. But for the world's poorest and most vulnerable, the consequences could be devastating. One study concluded that by the middle of the century the phenomenon could put more than a billion additional women and children at risk of iron-deficiency anemia — a condition that can cause pregnancy complications, developmental problems and even death. Meanwhile, some 2 billion people across the globe who already suffer from some form of nutrient shortage could see their health problems grow even worse. "The scale of the problem is huge," Ebi said. Plants depend on carbon dioxide to perform photosynthesis — but that doesn't mean they grow better when there's more carbon in the air, scientists say. A sweeping survey of changes among 32 compounds in 43 crops found that nearly every plant that humans eat is harmed by rising CO2 levels... On average, they found, nutrients have already decreased by an average 3.2 percent across all plants since the late 1980s, when the concentration of carbon dioxide in the atmosphere was about 350 parts per million. Thanks to long-time Slashdot reader GameboyRMH for sharing the news.

Read more of this story at Slashdot.

Linux 7.1-rc2 is out for testing with its accumulation of initial bug and regression fixes that have been collected over the past week since the Linux 7.1 merge window was capped off...

A startup south of Austin is using robots to build homes out of clay pulled directly from the ground, reports a local news station: The materials are gathered on site, mixed, and placed on a build plate. From there, a robot lowers from above, picks up the clay with a claw, carries it to the wall and drops it into place. Later, the same robot switches tools, using a hammer attachment to pound the material into shape. "It's kind of trying to replicate how a human might build an adobe house," said software engineer Anastasia Nikoulina... Using machine learning, the system constantly evaluates the wall, adjusting how it builds to create a flat, solid surface... The project is underway at Proto-Town, a ranch between Lockhart and Luling where startups test new technologies, from anti-drone systems to nuclear reactors. The company plans to build their next home on the property, with hopes to do more than 20 homes over the next year.

Read more of this story at Slashdot.

A 1999 press release bragged "Jeeves" answered 92.3 million questions in just three months. "In the digital wilds of Y2K, we came to him with our most probing questions," remembers the New York Times — whether it was Britney Spears or tamagotchis: We asked, and he answered: Jeeves, the digital butler of information, the online valet who led us into the depths of cyberspace. Now, like so many other relics of yesterday's internet, Jeeves — and his home, Ask.com — are no more. After almost 30 years, the question-and-answer service and former search engine shuttered on Friday. "To you — the millions of users who turned to us for answers in a rapidly changing world — thank you for your endless curiosity, your loyalty, and your trust," the company said in a notice posted on its now-defunct website... Created in Berkeley, Calif., in the days of the dot-com gold rush, Ask Jeeves first appeared on computer screens in 1996.... Their mascot, Jeeves, was modeled on the clever English butler character from the famed P.G. Wodehouse book series. Its search function was simple — type in a question, get an answer. But the quality of its responses was uneven, and the website was quickly eclipsed by Google and Yahoo as the world's go-to search engines. The site was bought by InterActive Corp. for more than $1 billion in 2005, and was given an injection of cash to help it compete as a search engine. It rebranded as Ask.com and as part of the reimagining, the site also ditched the character of Jeeves in 2006. Scrappy but inventive, the site was one of the first to introduce hyperlocal map overlays to its searches and incorporate thumbnails of webpages. "They are doing a lot of clever and interesting things," a Google executive noted of Ask.com at the time. Still, Ask.com struggled to compete and returned in 2010 to its bread and butter: question-and-answer style prompts. Even then, it faltered against newer, crowdsourced iterations like Quora and Google's unyielding march to the internet fore — the platform now dominates search traffic, and the world's general experience of the internet. A statement at Ask.com ends "by thanking its millions of users, and saying, 'Jeeves' spirit endures'," notes this article from Engadget: As sad as it is to see a relic of the early Internet days fade into obscurity, we still have Ask Jeeves to thank for why some users still punch in full questions when querying Google. On top of that, Jeeves was built to provide detailed answers in natural language, which could have arguably acted as a precursor to today's AI chatbots like ChatGPT. "Now, Ask.com joins the Internet graveyard that includes competitors like AltaVista, which shut down in 2013," the article points out. "With Ask.com gone, alongside AIM and AOL dial-up services also sunsetting, we're truly coming to an end of a specific era of the Internet." And the New York Times argues the memory of Jeeves now rests somewhere between Limewire and Beanie Babies... Slashdot reader BrianFagioli calls it "a quiet reminder of how quickly the web moves, and how even widely recognized names can drift into obscurity once the underlying technology leaves them behind."

Read more of this story at Slashdot.

Amazon once tried to pressure Nintendo to break the law, says former Nintendo of America President Reggie Fils-Aimé. At a recent NYU lecture, he describes a conversation with an Amazon executive, Kotaku reports: "Amazon was looking to get bigger into the video game space," said Fils-Aimé. "Amazon's mentality back then is they wanted to have the lowest price out in the marketplace, even lower than Walmart... Essentially what Amazon wanted (was an) obscene amount of support, financial support, so they could have the lowest price and beat Walmart. I literally said to the executive, 'You know that's illegal, right? I can't do that'...." At the time, the Wii and DS were Nintendo's best selling hardware in history. Amazon originally sold books, but in the 2000s rapidly expanded with cheaper discounts to became a one-stop shop for almost everything. Everything except Nintendo, that is.... "Literally we stopped selling to Amazon," Fils-Aimé continued, "and it's because I wasn't going to do something illegal. I wasn't going to do something that would put at risk the relationship we have with other retailers." "The two sides have since made amends," notes the Verge, "and you can buy a Switch 2 through Amazon. But for a long time, Nintendo consoles had been largely unavailable on the site."

Read more of this story at Slashdot.

The Wall Street Journal reports that OpenAI "recently gave its popular ChatGPT strict instructions. Stop talking about goblins." Recent models of the artificial-intelligence chatbot have been bringing up the creatures in conversations with users seemingly out of the blue, as well as gremlins, trolls and ogres. The goblin-speak caught the attention of programmers, who are often heavy users of the bot. Barron Roth, a 32-year-old product manager at a tech company, said the bot referred to a flaw in his code as a "classic little goblin." He said he counted more than 20 times it mentioned goblins, without any prompting... Several users speculated that goblin terminology was how the model characterized itself, in lieu of identifying as a person with a soul. Then OpenAI decided enough was enough. "Never talk about goblins, gremlins, raccoons, trolls, ogres, pigeons, or other animals or creatures unless it is absolutely and unambiguously relevant to the user's query," reads an open source line in ChatGPT's base instructions for its coding assistant. The Journal calls this "a reminder that even as AI companies tout one advance after another in their technology, they are sometimes baffled by the things their own models do...." While training a "nerdy" personality for their model's customization feature, "We unknowingly gave particularly high rewards for metaphors with creatures," OpenAI explained in a log post. And "From there, the goblins spread." When we looked, use of "goblin" in ChatGPT had risen by 175% after the launch of GPT-5.1, while "gremlin" had risen by 52%... With GPT-5.4, we and our usersâ noticed an even bigger uptick in references to these creatures... Nerdy accounted for only 2.5% of all ChatGPT responses, but 66.7% of all "goblin" mentions in ChatGPT responses... The rewards were applied only in the Nerdy condition, but reinforcement learning does not guarantee that learned behaviors stay neatly scoped to the condition that produced them. Once a style tic is rewarded, later training can spread or reinforce it elsewhere, especially if those outputs are reused in supervised fine-tuning or preference data. It all started because the "nerdy" personality's prompt had said "You must undercut pretension through playful use of language. The world is complex and strange, and its strangeness must be acknowledged, analyzed, and enjoyed..." Now OpenAI calls this "a powerful example of how reward signals can shape model behavior in unexpected ways, and how models can learn to generalize rewards in certain situations to unrelated ones." But "fans of goblins don't have to fear," notes the Wall Street Journal. "OpenAI provided a command in its blog post that would remove its creature-suppressing instructions."

Read more of this story at Slashdot.

Brush v0.4 debuted today for this "Bourne Rusty Shell" as a Bash/POSIX-compatible shell written in the Rust programming language...

An official in South Africa withdrew a draft of the country's national AI policy, reports a local newspaper, "after it was found the draft policy was compiled using AI, which cited academic articles that were 'fictitious'." Earlier this month, minister in the Presidency Khumbudzo Ntshavheni announced cabinet had approved the draft policy for public comment. [Ntshavheni] said the policy seeks to strengthen government's ability to regulate and adopt AI responsibly, while fostering innovation, job creation, and skills access. The article includes this quotes from the country's minister of communications/digital technologies department. "This unacceptable lapse proves why vigilant human oversight over the use of artificial intelligence is critical." Thanks to Slashdot reader Tokolosh for sharing the article.

Read more of this story at Slashdot.

"Ransomware activity jumped again in Q1 2026," writes Slashdot reader BrianFagioli, "with 2,638 victim posts on leak sites, up 22% year over year," according to a report from cybersecurity company ReliaQuest. But the bigger shift is how messy the ecosystem has become. Established groups like Akira and Qilin are still active, while newer players like The Gentlemen surged into the top tier with a 588 percent spike in activity. At the same time, questionable leak sites such as 0APT and ALP-001 are muddying the waters by posting possibly fake breach claims, forcing companies to investigate incidents that may not even be real. Meanwhile, actors like ShinyHunters are showing that ransomware does not always need encryption anymore. By targeting identity systems and SaaS platforms, attackers can steal data using legitimate access, often through phishing or even phone-based social engineering, and then extort victims without deploying traditional malware. With a record 91 active leak sites and faster attack timelines, the report suggests defenders should focus less on tracking specific groups and more on stopping common tactics like credential theft, remote access abuse, and large-scale data exfiltration.

Read more of this story at Slashdot.

ReactOS as the "open-source Windows" operating system project striving for binary compatibility with Microsoft Windows has seen some exciting improvements this week...

This week Google announced the selected Google Summer of Code "GSoC" 2026 projects for providing stipends to student developers for engaging in different open-source projects. This year a lot of open-source projects involve AI/LLM adoption but there are also a number of other interesting student projects at large from GNOME Mutter GPU reset recovery to adding new features to FreeBSD...

An anonymous reader shared this report from the BBC: "If even one extra person is able to access the internet, I think it's successful and it's worth it," says Sahand. The Iranian man is visibly anxious, speaking to the BBC outside Iran, as he carefully explains how he is part of a clandestine network smuggling satellite internet technology — which is illegal in Iran — into the country. Sahand, whose name we have changed, fears for family members and other contacts inside the country. "If I was identified by the Iranian regime, they might make those I'm in touch with in Iran pay the price," he says. For more than two months, Iran has been in digital darkness as the government maintains one of the longest-running national internet shutdowns ever recorded worldwide... Sahand says he has sent a dozen [Starlink terminals] to Iran since January and "we are actively looking for other ways to smuggle in more". The human rights organisation Witness estimated in January that there are at least 50,000 Starlink terminals in Iran. Activists say the number is likely to have risen... Last year, the Iranian government passed legislation that made using, buying or selling Starlink devices punishable by up to two years in prison. The jail term for distributing or importing more than 10 devices can be up to 10 years. State-affiliated media has reported multiple cases of people being arrested for selling and buying Starlink terminals, including four people — two of them foreign nationals — arrested last month for "importing satellite internet equipment". "The BBC contacted SpaceX for more details about the use of Starlink in the country but did not receive a response."

Read more of this story at Slashdot.

It's been an interesting 2026 in Linux development with beginning to phase out i486 CPU support, dropping ISDN and amateur "ham" radio support, and other code cleaning in the name of a diminishing user base -- or perhaps even no users left -- for those running such vintage hardware with a modern, up-to-date kernel. Yet ISA sound card drivers have seen an uptick in activity...

One of the interesting GNOME-aligned application developments in recent months has been RustConn as a modern GTK4-based connection manager. RustConn allows managing SSH, RDP, VNC, SPICE, and a variety of other connections from this Rust-written application. It's been steadily tacking on more features and that effort continued with more features landing...

One of the most prominent changes with the upcoming Linux 7.1 kernel release is the introduction of the new NTFS driver in the Linux 7.1 kernel. This new driver provides more features and better performance than the Paragon NTFS3 driver that's been in the kernel the past few years and far better off than the original NTFS read-only driver that previously was in the kernel and for which this new driver is based. Needless to say it's also a big improvement over the NTFS-3G user-space FUSE driver too...

Nouvelle semaine de suivi pour les prix des processeurs, avec nos relevés effectués entre le 24/04/2026 et le 01/05/2026. Et cette fois, nous avons un marché assez contrasté, avec quelques belles baisses chez Intel, quelques stabilisations chez AMD, mais aussi des tarifs qui restent parfois bien perchés sur certains modèles. On commence chez Intel, avec le Core i5-14600K, qui remonte assez franchement après sa grosse baisse de la semaine dernière. Il passe de 261,90 euros à 301,90 euros, soit une hausse de 40,00 euros. Le Core i7-14700K baisse légèrement, passant de 419,90 euros à 410,90 euros, soit -9,00 euros. Même tendance pour le Core i9-14900K, qui corrige fortement après avoir atteint 516,90 euros la semaine dernière. Il revient maintenant à 474,90 euros, soit une baisse de 42,00 euros. Du côté des Core Ultra, le Ultra 5 245K reste parfaitement stable à 179,90 euros. Le Ultra 5 250K+ baisse doucement, passant de 239,90 euros à 235,90 euros. Le Ultra 7 265K ne bouge pas et reste à 314,90 euros, tout comme le Ultra 7 270K+, toujours affiché à 340,90 euros. La bonne surprise vient du Ultra 9 285K, qui passe de 607,90 euros à 573,90 euros, soit une baisse de 34,00 euros. […]

Lire la suite

Six offres de box fibre Bouygues Telecom coexistent en 2026 : deux formules sans engagement B&YOU et quatre Bbox avec TV. Ce guide comparatif complet vous aide à trouver l’abonnement idéal, que ce soit pour cet été ou la rentrée scolaire. 🔗 Article contenant des liens sponsorisés. Bbox-Mag perçoit une commission sur les souscriptions réalisées […]

L'article Box fibre Bouygues Telecom 2026 : quel abonnement choisir ? a été publié en premier sur Bbox-Mag

In 2016 an online "swarm intelligence" platform generated a correct prediction for the Kentucky Derby — naming all four top finishers in order. (But its 2017 predictions weren't even close.) Slashdot checked in again on how modern AI systems performed in 2023, 2024, and 2025 — but their predictions were still pretty bad. Would AI-generated Derby predictions be any better in 2026? This year's winner was 24-to-1 longshot "Golden Tempo" — though a lot of oddsmakers had favored a horse named Further Ado (which ultimately only finished 11th). So when USA Today prompted Microsoft Copilot for its own picks for the Kentucky Derby, Copilot also went with Further Ado. (Even worse, it predicted Golden Tempo would come in... 13th.) Here's how Copilot's picks actually performed... Further Ado (finished 11th)Chief Wallabee (finished 4th)The Puma (SCRATCHED)Renegade (finished 2nd)Commandment (finished 7th)So Happy (finished 9th)Emerging Market (finished 10th)Danon Bourbon (finished 5th)Potente (finished 12th)Incredibolt (finished 6th)Robusta (finished 14th)Ocelli (finished 3rd)Golden Tempo (finished 1st)Pavlovian (finished 18th)Great White (SCRATCHED)Wonder Dean (finished 8th) Litmus Test (finished 17th)Albus (finished 15th)Six Speed (finished 13th)Intrepido (finished 16th) Copilot was told to use the latest odds, conditions, and analysis of favorites, best bets, expert picks, previous results and race history with the post positions, according to USA Today. And meanwhile, Yahoo Sports asked Claude "to simulate the race using the opening odds, draw and potential track conditions. We also asked it to factor in some human predictions." Like Microsoft Copilot, Claude also picked Further Ado to finish first (though it came in 11th) — and predicted that Golden Tempo (the eventual first-place finisher) would finish 12th. Further Ado (finished 11th)The Puma (SCRATCHED)Commandment (finished 7th)Chief Wallabee (finished 4th)Renegade (finished 2nd)Emerging Market (finished 10th)So Happy (finished 9th)Incredibolt (finished 6th)Danon Bourbon (finished 5th)Potente (finished 12th)Pavlovian (finished 18th)Golden Tempo (finished 1st) Litmus Test (finished 17th)Albus (finished 15th)Wonder Dean (finished 8th)Six Speed (finished 13th)Intrepido (finished 16th)

Read more of this story at Slashdot.

Avec un jour en moins, si on peut dire ainsi, la semaine a été plutôt calme pour les tests et articles. Il devrait en être de même la semaine prochaine, mais aussi la suivante avec un Pont de Normandie de jeudi à dimanche. Mais revenons à aujourd'hui pour faire un point sur les articles publiés ces derniers jours, avec des choses très intéressantes, et d'autres plus décevantes. Ca arrive, tous les produits ne peuvent super géniaux comme Emmet. Par exemple, le très gros boitier Antec Flux Pro Noctua Edition fait son retour, et il est toujours appréciable d'avoir un avis supplémentaire. Un boitier onéreux bardé de ventilateurs Noctua, mais quel boitier ! Une petite réussie qui mérite largement un bon coup d'oeil et un craquage de carte bleue. A côté, HardwareetCo a vu plus petit avec le Framework Desktop, mais nous n'en dirons pas plus. Evoqué cette semaine dans une news, le câble CORSAIR ThermalProtect a le droit à une petite découverte. L'occasion pour parler de son avenir, déjà : CORSAIR a indiqué sur les réseaux sociaux avoir d'autres modèles en préparation pour les alimentations Type 4 et Type 5 avec un adaptateur depuis du PCI-E 6+2. Un véritable petit écosystème va donc être mis en place, une bonne chose pour ceux qui ont une carte graphique récente et qui craignent pour elle. […]

Lire la suite

Chaque dimanche, on propose un mod pour vous rappeler qu'un PC peut être bien plus qu'un simple boîtier posé sur un bureau. Entre créations totalement déjantées, configurations ultra sobres, mini-machines taillées au scalpel ou monstres de puissance capables de chauffer un quartier entier, il y en a pour tous les goûts. Noir profond, blanc clinique, rose bonbon ou festival RGB façon licorne sous caféine, tout existe. Aujourd'hui, on pose les yeux sur le superbe AQUA réalisé par le moddeur TantricmodZ, une configuration qui joue la carte de l'élégance glaciale. […]

Lire la suite