SoylentNews

janrinok writes:

Poverty in Australia increases to 1 in 7 people, according to report

The number of people living in poverty in Australia has increased to 1 in 7, according to a new report released today—at the start of Anti-Poverty Week.

As many as 14.2% of the population—or 3.7 million Australians—were living below the poverty line in 2022–23, according to the "Poverty in Australia 2025: Overview" report released today.

The report, from the Australian Council of Social Service (ACOSS) and UNSW Sydney-led Poverty and Inequality Partnership, uses the latest available data from the Household, Income and Labor Dynamics in Australia (HILDA) Survey.

These latest data mark an increase from 12.4% of the population—or 1 in 8 people—living below the poverty line in 2020–21.

The study also found the poverty rate for children is 1 in 6, equaling 757,000 children.

"This research shows that 1 in 7 people are now living in poverty. This is unacceptable in one of the wealthiest countries in the world," says Dr. Yuvisthi Naidoo, Senior Research Fellow at UNSW's Social Policy Research Center.

"The rate of people living in poverty decreased in 2020 due to the temporary doubling of JobSeeker during COVID," Dr. Naidoo says.

"But that has sharply risen above pre-pandemic levels due to the removal of COVID payments and rising housing costs," she says.

"The steep increase in rents in recent years has had a particularly severe impact on people with the lowest incomes."

The report found from June 2021 to June 2023, the median advertised rent for units rose from $486 per week to $680 in Sydney (40%), from $395 to $528 in Melbourne (34%) and from $394 to $554 in Brisbane (41%).

The proportion of low-income renters (the lowest 20% of earners) spending more than 30% of their income on rent—known as rental stress—increased from 52% in 2020–21 to 57% in 2022–23.

hubie writes:

Microsoft's quality control department caught napping again:

Microsoft's October Windows 11 update has managed the impressive feat of breaking localhost, leaving developers unable to access web applications running on their own machines.

The problem first surfaced on Microsoft's own support forums and quickly spread to Stack Overflow and Server Fault after the October 2025 cumulative update (KB5066835) landed, which appears to have severed Windows' ability to talk to itself.

Developers describe HTTP/2 protocol errors and failed connections affecting everything from ASP.NET builds to Visual Studio debugging sessions.

The bug, introduced in build 26100.6899, has been traced to HTTP.sys, the Windows kernel component that handles local HTTP traffic. Developers have found that uninstalling KB5066835, and in some cases its sibling KB5065789, restores localhost functionality.

Others have discovered a temporary workaround that involves manually disabling HTTP/2 in the registry, which works but feels a bit like using a sledgehammer to swat a fly.

At the time of writing, Microsoft had yet to acknowledge the issue. Users report mixed results when trying to reinstall the patch or roll forward to newer builds. The problem appears to vanish on clean installs of Windows 11 24H2, suggesting that the error stems from a conflict in how the update interacts with existing system configurations, rather than being a universal bug.

All this comes as Microsoft pushed its final update for Windows 10 this week, officially ending support for the decade-old OS and urging users to move to Windows 11.

The transition hasn't exactly been buttery smooth. Microsoft's Windows 11 media creation tool also stopped working the day before, potentially affecting users trying to upgrade, and the same patch cycle saw end-of-support deadlines for Office 2019 and multiple server products.

All this means that, within the same week, Microsoft's installer broke, its new OS borked local development, and Redmond's multimillion-dollar upgrade push instead highlighted how fragile its ecosystem still is.

It's almost enough to make you nostalgic for Clippy. We said almost.

Hopefully this is all fixed before this story gets posted.

Original Submission

fliptop writes:

The meteoric rise of artificial intelligence but it's facing a shortage of training data:

"We've already run out of data," Neema Raphael, Goldman Sachs' chief data officer and head of data engineering, said on the bank's "Exchanges" podcast published on Tuesday.

Raphael said that this shortage may already be influencing how new AI systems are built.

He pointed to China's DeepSeek as an example, saying one hypothesis for its purported development costs came from training on the outputs of existing models rather than entirely new data.

[...] With the web tapped out, developers are turning to synthetic data — machine-generated text, images, and code. That approach offers limitless supply, but also risks overwhelming models with low-quality output or AI slop.

However, Raphael said he doesn't think the lack of fresh data will be a massive constraint, in part because companies are sitting on untapped reserves of information.

Rick Beato talked about [15:29 --JE] how he broke ChatGPT with a simple question and exposed the gaps in AI's "knowledge" that are filled with synthetic data.

Related: The Real (Economic) AI Apocalypse is Nigh

Original Submission

janrinok writes:

New psychology research looks at why we help our friends when they need it

Despite how natural friendship can feel, people rarely stop to analyze it. How do you know when someone will make a good friend? When is it time to move on from a friendship? Oftentimes, people rely on gut intuitions to answer these kinds of questions.

In psychology research, there's no universally accepted definition of a friend. Traditionally, when psychologists have analyzed friendship, it's often been through the lens of exchange. How much did that friend do for me? How much did I do for them? The idea is that friendships are transactional, where friends stick around only as long as they are getting at least as much as they are giving in the friendship.

But this focus doesn't capture what feels like the essence of friendship for many people. We and our colleagues think another model for relationships—what we call risk-pooling—better matches what many people experience. In this kind of friendship, no one is keeping track of who did what for whom.

Our research over the past decade suggests that this kind of friendship was essential for our ancient ancestors to survive the challenges they encountered. And we feel it's essential for surviving the challenges of life today, whether navigating personal struggles or dealing with natural disasters.

The traditional social exchange theory of friendship views relationships as transactions where people keep a tally of costs and benefits. Building on this framework, researchers have suggested that you approach each friendship with a running list of pluses and minuses to decide whether to maintain the bond. You keep friendships that provide more benefits than costs, and you end those that don't.

The theory holds that this balancing act comes into play when making decisions about what kinds of friendships to pursue and how to treat your friends. It's even made its way into pop psychology self-help spaces.

We contend that the biggest issue with social exchange theory is that it misses the nuances of real-life relationships. Frankly, the theory's wrong: People often don't use this cost-to-benefit ratio in their friendships.

Anybody who has seen a friend through tough times—or been the one who was supported—can tell you that keeping track of what a friend does for you isn't what friendships are about. Friendships are more about companionship, enjoyment and bonding. Sometimes, friendship is about helping just because your friend is in need and you care about their well-being.

Social exchange theory would suggest that you'd be better off dropping someone who is going through cancer treatment or a death in the family because they're not providing as many benefits to you as they could. But real-life experiences with these situations suggest the opposite: These are the times when many people are most likely to support their friends.

Our research is consistent with this intuition about the shortcomings of social exchange theory. When we surveyed people about what they want in a friend, they didn't place a high value on having a friend who is conscientious about paying back any debts—something highly valued from a social exchange perspective.

janrinok writes:

Understanding volcanoes better: Scientists find exact locations of magma movement

How do volcanoes work? What happens beneath their surface? What causes the vibrations—known as tremor—that occur when magma or gases move upward through a volcano's conduits? Professor Dr. Miriam Christina Reiss, a volcano seismologist at Johannes Gutenberg University Mainz (JGU), and her team have located such tremor signals at the Oldoinyo Lengai volcano in Tanzania.

"We were not only able to detect tremor, but also to determine its exact position in three dimensions—its location and depth below the surface," said Reiss. "What was particularly striking was the diversity of different tremor signals we detected."

The findings provide new insights into how magma and gas are transported within Earth and thus improve our understanding of volcanic dynamics. This also has societal relevance as the researchers hope that their work will enhance the ability to forecast volcanic eruptions in the long term. Their results are published in Communications Earth & Environment.

When magma rises from the depths of Earth toward or into a volcano, this can cause shaking. If the magma exerts high pressure, the surrounding rock can fracture—resulting in earthquakes. Other processes can cause milder vibrations, known as tremors. For example, when magma ascends through pre-existing channels, when gases escape from magma, or when pressure fluctuations occur in the transport pathways.

[...] "For the first time, we were able to determine the precise location where tremor occurs," stated Reiss. "We discovered that two types of tremor seem to be linked: One originated at around five kilometers depth and the other near the base of the volcano—with a time delay between them. It is clear that these signals are connected, thus we see a directly linked system here."

VLM writes:

Anyone get good views of Comet C/2025 A6 aka "Lemmon"? Anyone doing astrophotography of it? Anyone see it a couple times and watch it change day to day?

In the USA, you'd look vaguely NE an hour before sunrise about a hands width above the horizon. Supposedly no optics needed although all I see is clouds. Naturally it has been raining every morning for me LOL, but I hope to see it soon, thats life in an area that gets like four feet of rain annually LOL. Peak closest approach is supposed to be next Tuesday ish the 21st (hopefully this story will be posted before then)

https://www.astronomy.com/observing/nows-the-time-to-see-comet-lemmon/

https://www.scientificamerican.com/article/how-to-see-comet-lemmon-and-the-orionids-meteor-shower-peak/

https://en.wikipedia.org/wiki/C/2025_A6_(Lemmon)

https://apod.nasa.gov/apod/ap250930.html

https://www.space.com/stargazing/how-to-find-comet-lemmon-in-the-night-sky-as-it-brightens-this-october-2025

I just thought it would be fun to have a very chill astronomy "event" post, once in awhile. Comets are cool.

Note this is not comet 3I/ATLAS aka "it's probably an interstellar alien space probe", although I suppose thats fun to talk about too.

Original Submission

fliptop writes:

General Motors said on Oct. 14 that it will bear a $1.6 billion loss to scale back its electric vehicle (EV) operations, citing weaker expected demand following recent U.S. policy changes that ended federal EV tax credits and loosened emissions rules:

The Detroit-based automaker said its Audit Committee approved the loss on Oct. 7, covering the three months ended Sept. 30. The company noted that the loss is part of its plan to realign EV production and factory operations to better match customer demand.

The decision was made after the expiration of the $7,500 federal EV tax credit on Sept. 30, part of a broader policy rollback under President Donald Trump.

[...] "Following recent U.S. government policy changes, including the termination of certain consumer tax incentives for EV purchases and the reduction in the stringency of emissions regulations, we expect the adoption rate of EVs to slow," GM said in a filing.

[...] According to the filing, $1.2 billion of the loss is related to non-cash impairments, mostly write-downs of EV assets. The remaining $400 million will be paid in cash for contract cancellations and commercial settlements tied to EV investments.

The company said its review of EV manufacturing and battery component investments is ongoing.

Related:

• Canadian EV Sales Collapse by 35% as Gas Car Purchases Surge
• Tesla Continues Slide As Musk Warns Of "Rough Quarters" Ahead
• Electric Truck-Maker Nikola Falls Into Bankruptcy Joining a Procession of Failed EV Startups

See Also:

• EV Sales In Europe Keep Surging, With One Exception
• EVs Poised To Exceed Half Of Europe's New Car Sales Sooner Than Expected
• EV sales are surging in Europe, except for Tesla

Original Submission

hubie writes:

Risks to BIG-IP users include supply-chain attacks, credential loss, and vulnerability exploits:

Thousands of networks—many of them operated by the US government and Fortune 500 companies—face an "imminent threat" of being breached by a nation-state hacking group following the breach of a major maker of software, the federal government warned Wednesday.

F5, a Seattle-based maker of networking software, disclosed the breach on Wednesday. F5 said a "sophisticated" threat group working for an undisclosed nation-state government had surreptitiously and persistently dwelled in its network over a "long-term." Security researchers who have responded to similar intrusions in the past took the language to mean the hackers were inside the F5 network for years.

During that time, F5 said, the hackers took control of the network segment the company uses to create and distribute updates for BIG IP, a line of server appliances that F5 says is used by 48 of the world's top 50 corporations. Wednesday's disclosure went on to say the threat group downloaded proprietary BIG-IP source code information about vulnerabilities that had been privately discovered but not yet patched. The hackers also obtained configuration settings that some customers used inside their networks.

Control of the build system and access to the source code, customer configurations, and documentation of unpatched vulnerabilities has the potential to give the hackers unprecedented knowledge of weaknesses and the ability to exploit them in supply-chain attacks on thousands of networks, many of which are sensitive. The theft of customer configurations and other data further raises the risk that sensitive credentials can be abused, F5 and outside security experts said.

Customers position BIG-IP at the very edge of their networks for use as load balancers and firewalls, and for inspection and encryption of data passing into and out of networks. Given BIG-IP's network position and its role in managing traffic for web servers, previous compromises have allowed adversaries to expand their access to other parts of an infected network.

F5 said that investigations by two outside intrusion-response firms have yet to find any evidence of supply-chain attacks. The company attached letters from firms IOActive and NCC Group attesting that analyses of source code and build pipeline uncovered no signs that a "threat actor modified or introduced any vulnerabilities into the in-scope items." The firms also said they didn't identify any evidence of critical vulnerabilities in the system. Investigators, which also included Mandiant and CrowdStrike, found no evidence that data from its CRM, financial, support case management, or health systems was accessed.

[...] The US Cybersecurity and Infrastructure Security agency has warned that federal agencies that rely on the appliance face an "imminent threat" from the thefts, which "pose an unacceptable risk." The agency went on to direct federal agencies under its control to take "emergency action." The UK's National Cyber Security Center issued a similar directive.

Original Submission

janrinok writes:

Lasers and gold nanoparticles enable on-demand crystal growth for new materials

Researchers at Michigan State University have discovered how to "draw" on demand the crystals used in many crucial technologies, from solar panels and LED lighting to medical imaging.

Appearing in the journal ACS Nano, the breakthrough was achieved by striking gold nanoparticles with a single laser pulse.

"We're just beginning to scratch the surface of what's possible. This is opening a new chapter in how we design and study materials," said Elad Harel, associate professor in the Department of Chemistry and senior author of the study.

Take the time to look around, and you'll find a world that runs on crystals. From smoke alarms and television screens to ultrasounds and sonar, the unique optical and electrical properties of these chemical structures place them at the cutting-edge of most innovations.

Growing crystals, however, isn't easy.

"When using traditional growing methods, crystals can form at random times and locations, so the results might not always be the same," Harel said.

As technologies and materials rapidly improve, they rely on crystals of exceptional quality being placed in just the right spot, so this lack of control is a major hurdle for researchers.

To tackle this challenge, Harel turned to his lab's specialty—lasers, and in particular, fast lasers.

At MSU, Harel uses short laser pulses to shine a light on the mysteries of the natural world. This includes a recent breakthrough that leveraged ultra-fast lasers to actually "hear biology."

In the new publication, the researchers tried their hand at growing types of crystals called lead halide perovskites. These crystals are crucial for LEDs, solar cells and medical imaging.

Rather than move through the typical complicated steps of crystal growth or even use a small "seed" crystal to jumpstart the process, Harel's team aimed their lasers at a tiny glittering target: gold nanoparticles less than one thousandth the width of a human hair.

The scientists revealed that these particles generated heat where the laser light struck, and that this interaction led to crystallization. Using special, high-speed microscopes, they were even able to watch the process unfold in real time.

Like a laser used to engrave artwork into metal or wood, this sort of crystal creation offers researchers the ability to "draw" crystals with levels of control that could transform fields ranging from clean energy to quantum technologies. The findings also help expand our understanding of how crystals form—a notoriously tricky area of chemistry.

"With this method, we can essentially grow crystals at precise locations and times," said Dr. Md Shahjahan, a research associate at MSU and first author of the paper. "It's like having a front-row seat to watch the very first moments of a crystal's life under a microscope. Only here can we also steer how it develops."

With their gold nanoparticles now in the spotlight, Elad's team is heading back to the lab for future experiments with big potential.

These include using multiple lasers of different colors to "draw" even more intricate crystal patterns and attempting to create entirely new materials that can't be made through conventional methods.

"Now that we can 'draw' crystals with lasers, the next step is to make larger and more complex patterns, and to test how these crystals perform in real devices," Harel said.

More information: Nanoscale Plasmonic Heating Induced Spatiotemporal Crystallization of Methylammonium Lead halide Perovskite, ACS Nano (2025). DOI: 10.1021/acsnano.5c12057

Original Submission

hubie writes:

New design sets a high standard for post-quantum readiness:

The encryption protecting communications against criminal and nation-state snooping is under threat. As private industry and governments get closer to building useful quantum computers, the algorithms protecting Bitcoin wallets, encrypted Web visits, and other sensitive secrets will be useless. No one doubts the day will come, but as the now-common joke in cryptography circles observes, experts have been forecasting this cryptocalypse will arrive in the next 15 to 30 years for the past 30 years.

The uncertainty has created something of an existential dilemma: Should network architects spend the billions of dollars required to wean themselves off quantum-vulnerable algorithms now, or should they prioritize their limited security budgets fighting more immediate threats such as ransomware and espionage attacks? Given the expense and no clear deadline, it's little wonder that less than half of all TLS connections made inside the Cloudflare network and only 18 percent of Fortune 500 networks support quantum-resistant TLS connections. It's all but certain that many fewer organizations still are supporting quantum-ready encryption in less prominent protocols.

One exception to the industry-wide lethargy is the engineering team that designs the Signal Protocol, the open-source engine that powers the world's most robust and resilient form of end-to-end encryption for multiple private chat apps, most notably the Signal Messenger. Eleven days ago, the nonprofit entity that develops the protocol, Signal Messenger LLC, published a 5,900-word write-up describing its latest updates that make Signal fully quantum-resistant.

The complexity and problem-solving required for making the Signal Protocol quantum safe are as daunting as just about any in modern-day engineering. The original Signal Protocol already resembled the inside of a fine Swiss timepiece, with countless gears, wheels, springs, hands, and other parts all interoperating in an intricate way. In less adept hands, mucking about with an instrument as complex as the Signal protocol could have led to shortcuts or unintended consequences that hurt performance, undoing what would otherwise be a perfectly running watch. Yet this latest post-quantum upgrade (the first one came in 2023) is nothing short of a triumph.

"This appears to be a solid, thoughtful improvement to the existing Signal Protocol," said Brian LaMacchia, a cryptography engineer who oversaw Microsoft's post-quantum transition from 2015 to 2022 and now works at Farcaster Consulting Group. "As part of this work, Signal has done some interesting optimization under the hood so as to minimize the network performance impact of adding the post-quantum feature."

Of the multiple hurdles to clear, the most challenging was accounting for the much larger key sizes that quantum-resistant algorithms require. The overhaul here adds protections based on ML-KEM-768, an implementation of the CRYSTALS-Kyber algorithm that was selected in 2022 and formalized last year by the National Institute of Standards and Technology. ML-KEM is short for Module-Lattice-Based Key-Encapsulation Mechanism, but most of the time, cryptographers refer to it simply as KEM.

[...] As both Signal and Jacomme noted, users of Signal and other messengers relying on the Signal Protocol need not concern themselves with any of these new designs. To paraphrase a certain device maker, it just works.

In the coming weeks or months, various messaging apps and app versions will be updated to add the triple ratchet. Until then, apps will simply rely on the double ratchet as they always did. Once apps receive the update, they'll behave exactly as they did before upgrading.

For those who care about the internal workings of their Signal-based apps, though, the architects have documented in great depth the design of this new ratchet and how it behaves. Among other things, the work includes a mathematical proof verifying that the updated Signal protocol provides the claimed security properties.

Outside researchers are applauding the work.

"If the normal encrypted messages we use are cats, then post-quantum ciphertexts are elephants," Matt Green, a cryptography expert at Johns Hopkins University, wrote in an interview. "So the problem here is to sneak an elephant through a tunnel designed for cats. And that's an amazing engineering achievement. But it also makes me wish we didn't have to deal with elephants."

The article has a lengthy middle section with details on the challenges and the workings of the algorithm.

Original Submission

looorg writes:

The Free Software Foundation (FSF) today announced its project to bring mobile phone freedom to users. "Librephone" is an initiative to reverse-engineer obstacles preventing mobile phone freedom until its goal is achieved.

Practically, Librephone aims to close the last gaps between existing distributions of the Android operating system and software freedom. The FSF has hired experienced developer Rob Savoye (DejaGNU, Gnash, OpenStreetMap, and more) to lead the technical project. He is currently investigating the state of device firmware and binary blobs in other mobile phone freedom projects, prioritizing the free software work done by the not entirely free software mobile phone operating system LineageOS.

So a free phoneOS. Free of the crud. Free of the bloat. Or something. Unclear when or if we can except this to be finished. If ever. The hardware might put up some hurdles.

https://www.fsf.org/news/librephone-project

Original Submission

janrinok writes:

First device based on 'optical thermodynamics' can route light without switches

A team of researchers at the Ming Hsieh Department of Electrical and Computer Engineering has created a new breakthrough in photonics: the design of the first optical device that follows the emerging framework of optical thermodynamics.

The work, reported in Nature Photonics, introduces a fundamentally new way of routing light in nonlinear systems—meaning systems that do not require switches, external control, or digital addressing. Instead, light naturally finds its way through the device, guided by simple thermodynamic principles.

Universal routing is a familiar engineering concept. In mechanics, a manifold valve directs inputs to a chosen outlet. In digital electronics, a Wi-Fi router at home or an Ethernet switch in a data center directs information from many input channels to the correct output port, ensuring that each stream of data reaches its intended destination.

When it comes to light, the same problem is far more challenging, however. Conventional optical routers rely on complex arrays of switches and electronic control to toggle pathways. These approaches add technical difficulty, while limiting speed and performance.

The photonics team at the USC Viterbi School of Engineering has now shown that there is another way. The idea can be likened to a marble maze that arranges itself.

Normally, you'd have to lift barriers and guide a marble step-by-step to make sure it reaches its destination—the right hole. In the USC team's device, however, the maze is built so that no matter where you drop the marble, it will roll on its own toward the right place—no guiding hands needed. And this is exactly how light behaves: it finds the correct path naturally, by following the principles of thermodynamics.

The implications of the new approach extend far beyond the laboratory. As computing and data processing continue to push the limits of traditional electronics, various companies—including chip designers such as NVIDIA and others—are exploring optical interconnects as a way to move information faster and more efficiently.

jelizondo writes:

The Times of India published an interesting article explaining the 2025 Economics Nobel Prize:

"We're a planet of six billion ninnies living in a civilisation built by a few thousand savants," Scott Adams once said — and beneath the Dibert creator's misanthropy is a disarmingly accurate description of how the modern world works. Most of us don't invent, build, or discover anything of world-changing importance. We live inside systems we didn't design, using tools we don't understand, and benefitting daily from the work of people whose names we'll never know. And yet, we're quick to criticise those same systems — science, technology, capitalism — that lifted us from the brink of subsistence to a level of prosperity our ancestors couldn't imagine.

[...] For most of recorded history, humanity went nowhere fast. A peasant in medieval Europe lived much the same life as a farmer in ancient Mesopotamia. Empires rose and fell, plagues came and went, and the occasional invention — a plough here, a printing press there — might briefly improve life. But those improvements rarely built on one another. Progress was sporadic and short-lived. The line of human prosperity was basically flat.

This year's Nobel laureates — Joel Mokyr, Philippe Aghion, and Peter Howitt — offer complementary answers to that question. Together, their work explains why humanity's growth engine roared to life — and what keeps it running today.

Joel Mokyr, an economic historian, argues that the Industrial Revolution wasn't just about machines — it was about knowledge. Before the 18th century, most human innovation was based on know-how: practical skills, techniques, and tricks. People knew how to do things but not why they worked. Without that deeper understanding, invention couldn't build on itself. Progress happened in bursts but couldn't compound.

The Enlightenment changed that. Science and technology stopped being separate worlds and started reinforcing each other. Scientific discoveries explained why things worked, which allowed engineers to design better tools and machines. Those tools, in turn, raised new scientific questions. Mokyr calls this cycle "useful knowledge" — a feedback loop between theory and practice that transformed invention from a series of lucky accidents into a self-sustaining system.

janrinok writes:

https://phys.org/news/2025-10-nanoplastics-farm-animal-cells-human.html

Scientists at the Research Institute for Farm Animal Biology (FBN) in Dummerstorf and the University of Udine have detected the uptake of nanoplastics in farm animal cell cultures. The results provide evidence of potential risks to animal health, meat production and also human food safety.

Plastic bags, packaging, yogurt lids—items that are carelessly thrown away decompose over years into tiny plastic particles. They end up in soil, waterways and ultimately in our food chain. Although numerous studies have already shown that microplastics can harm marine animals, birds and insects, the effects of nanoplastics on livestock have hardly been researched to date.

Unlike microplastics (1 µm–5 mm), there are currently few adequate methods for detecting nanoplastics (< 1 µm) in humans and animals. However, researchers assume that these small particles can also accumulate in tissue.

In a joint study, researchers from the FBN and the University of Udine have demonstrated the uptake of nanoplastic particles made of polystyrene into cultured cells from cattle and pigs. This absorption led to changes that could impair the cell function and health of the animals in the long term.

"Since we still know far too little about nanoplastics and detection is difficult, our results are particularly important for better assessing the risks," explains Dr. Anja Baufeld from the Cell Physiology and Reproduction working group at the FBN. "When we saw that nanoplastics were entering the cells, we knew that this could have far-reaching consequences," Baufeld continues.

[...] "Our research shows that nanoplastics are not only an environmental problem, but could also have direct consequences for the health of farm animals. These initial findings highlight the importance of conducting more intensive research into plastic pollution in order to assess the potential risks to both animals and humans at an early stage," says Baufeld.

More information: Francesca Corte Pause et al, Exploring the influence of polystyrene-nanoplastics on two distinct in vitro systems in farm animals: A pilot study, Science of The Total Environment (2025). DOI: 10.1016/j.scitotenv.2025.179378
       

Original Submission

upstart writes:

After blasting Wikipedia as biased and 'woke' and pushing for it to be defunded, Elon Musk says he's building his own online encyclopedia through xAI:

Elon Musk plans to take on Wikipedia with his own rival encyclopedia site.

On Tuesday, the Tesla CEO tweeted that his xAI startup is building Grokipedia, which he claims will be a "massive improvement" over Wikipedia. Musk has long had a gripe with Wikipedia, accusing it of being "woke" and even calling for it to be defunded. (The encyclopedia site has long relied on donations.) In January, Musk also railed at Wikipedia for adding an entry about him allegedly making a Nazi-like salute at a Trump inauguration event.

To create Grokipedia, Musk plans on tapping xAI's Grok chatbot (which he also created as an alternative to another technology he didn't like, ChatGPT). Grok has been trained on web data, including public tweets. In a podcast earlier this month, Musk suggested that Grok is smart enough not only to replicate the work of human community volunteers who maintain and update Wikipedia, but also to account for any bias or inaccuracies.

"Grok is using heavy amounts of inference compute to look at, as an example, a Wikipedia page, what is true, partially true, or false, or missing in this page," he said. "Now rewrite the page to correct, remove the falsehoods, correct the half-truths, and add the missing context." (That said, Grok has suffered its own share of problems, including praising Hitler.)

Original Submission