SoylentNews

An Anonymous Coward writes:

https://hackaday.com/2026/03/29/soviet-cds-and-cd-players-existed-and-they-were-strange/

Until the fall of the Soviet Union around 1990 you'd be forgiven as a proud Soviet citizen for thinking that the USSR's technology was on par with the decadent West. After the Iron Curtain lifted it became however quite clear how outdated especially consumer electronics were in the USSR, with technologies like digital audio CDs and their players being one good point of comparison. In a recent video by a railways/retro tech YouTube channel we get a look at one of the earliest Soviet CD players.

A good overall summary of how CD technology slowly developed in the Soviet Union despite limitations can be found in this 2025 article by [Artur Netsvetaev]. Soviet technology was characterized mostly by glossy announcements and promises of 'imminent' serial production prior to a slow fading into obscurity. Soviet engineers had come up with the Luch-001 digital audio player in 1979, using glass discs. More prototypes followed, but with no means for mass-production and Soviet bureaucracy getting in the way, these efforts died during the 1980s.

During the 1980s CD players were produced in Soviet Estonia in small batches, using Philips internals to create the Estonia LP-010. Eventually sanctions on the USSR would strangle these efforts, however. Thus it wouldn't be until 1991 that the Vega PKD-122 would become the first mass-produced CD player, with one example featured in this video.

The video helpfully includes a teardown of the player after a rundown of its controls and playback demonstration, so that we can ogle its internals. This system uses mostly localized components, with imported components like the VF display and processors gradually getting replaced over time. The DAC and optical-mechanical assembly would still be imported from Japan until 1995 when the factory went bankrupt.

This difference between the imported and localized part is captured succinctly in the video with the comparison to Berlin in 1999, in that you can clearly see the difference between East and West. The CD mechanism is produced by Sanyo, with a Sanyo DAC IC on the mainboard. The power supply, display and logic board (using Soviet TTL ICs) are all Soviet-produced. A sticker inside the case identifies this unit as having been produced in 1994.

Amusingly, the front buttons are directly coupled into the mainboard without ESD protection, which means that in a Siberian winter with practically zero relative humidity inside you'd often fry the mainboard by merely using these buttons.

After this exploration the video goes on to explain how Soviet CD production began in the 1989, using imported technology and know-how. This factory was set up in Moscow, using outdated West-German CD pressing equipment and makes for a whole fascinating topic by itself.

Finally, the video explores the CD player's manual and how to program the player, as well as how to obtain your own Soviet CD player. Interestingly, a former employee of the old factory has taken over the warehouse and set up a web shop selling new old stock as well as repaired units and replacement parts.

Original Submission

Chinese carmaker patents voice-controlled 'in-vehicle toilet'

AnonTechie writes:

Chinese carmaker Seres has been granted a patent for what it calls an "in-vehicle toilet" that slides under a passenger's seat for visits to the loo while on the road. The feature is meant to "satisfy users' toilet needs on long journeys, while camping or while staying in the car", engineers wrote in Seres' patent filing in China on 10 April.

Seres, based in the south-west city of Chongqing, has not announced any cars that have toilets and it is uncertain if any will be made. Chinese electric vehicles have become increasingly packed with unconventional features, like built-in massage seats, karaoke systems and a fridge, to stand out in a highly competitive market. The patent filing shows Seres' plans for an onboard toilet that slides out from the bottom of a passenger's seat with a push or through voice-activated commands. The loo will come with a fan and exhaust pipe to channel odours out of the car, according to the filing on China's intellectual property administration seen by the BBC. Waste is collected in a tank that has to be emptied manually. The toilet also features a rotating heating element that evaporates urine and dries other waste.

When not in use, the toilet is concealed beneath the seat, making full use of the space inside a car without requiring more room.

[Source]: BBC

Pooping in your car

looorg writes:

https://www.digitaltrends.com/cars/a-chinese-automaker-just-filed-a-patent-for-car-seats-with-a-hidden-loo/
https://www.autoblog.com/news/chinese-automaker-patents-slide-out-toilet-built-into-car-seat
https://www.straitstimes.com/asia/east-asia/start-up-toilet-function-chinese-automaker-patents-voice-activated-in-car-toilet

When nature calls and you are stuck in your car. No more holding it in. Just go in the car. In a hidden toilet.

A Chinese automaker has patented a toilet concealed beneath a sliding car seat that can be operated using voice commands, according to a government database.

Electric vehicle company Seres received patent approval for its mobile latrine earlier in April, public records show.

The on-the-go lavatory can be accessed manually through pushing the seat back, as well as through the voice command "start up toilet function".

Original Submission #1  Original Submission #2 

"c0lo" writes:

In the face of rampant AI, is 'data poisoning' a new form of civil disobedience?:

[...] Acts of AI resistance range from social sanctions and boycotts, to strikes, protests, public outcry and lawsuits. Driving these acts are perceived threats to jobs, ethics, safety, democracy and sovereignty, and the environment.

AI is also described as an existential risk to creative industries, including music, fiction and film. In the United Kingdom, generative AI has been characterised as an "industrial scale theft" that threatens a £124.6 billion (A$237bn) creative sector and more than 2.4 million jobs.

People have long used civil disobedience to address social injustices. Famously, Rosa Parks' refusal to sit at the back of a bus in Alabama led to a 13-month bus boycott by tens of thousands of Black residents. It only ended when racial segregation on public transport was deemed unconstitutional in the United States.

Acts of sabotage have also long been central to collective action against injustice. In fights for labour rights, workers have employed diverse tactics to reduce efficiency and productivity. This has ranged from hotel workers putting salt in sugar bowls to farm workers breaking machinery.

Data poisoning can be viewed as a modern version of these historic actions.

[...] Data poisoning means deliberately inserting misleading, biased, or nonsensical content into the data AI models learn from, to make their outputs worse. Only 250 poisoned documents in a dataset could compromise outputs across AI models of any size.

There are various ways to poison data. Some require highly technical skills, others are accessible to anyone with an internet connection – if their text or images are used as training data.

Researchers have developed several data poisoning tools that exploit the vulnerabilities of AI models. Glaze and Nightshade enable artists to make poisoned visual images that can't be used as training data. The tool CoProtector defends against the exploitation of open source code repositories like Github. Monash University and the Australian Federal Police have created Silverer, enabling social media users to doctor personal images to prevent them from being used in deepfakes.

But you don't need a tool or advanced skills to affect AI. Simply creating websites with factitious information, making jokes in Reddit, feeding models their own outputs, or editing Wikipedia can poison data.

Data poisoning is commonly presented as a dangerous act perpetrated by "cyber criminals" or "malicious actors". But what if it's used to protect human rights?

Is data poisoning legal? Is it ethical?

Legal obligations related to data poisoning are often directed to AI developers and organisations. The EU Artificial Intelligence Act requires that appropriate measures are adopted to prevent and detect data poisoning.

The legal status of AI data poisoning by individual users is less clear. Criminal penalties may apply under US or UK computer fraud and misuse laws. Interference with an AI model would also likely breach the terms of service of AI companies.

If AI data poisoning is unlawful, questions could still be asked about its ethical status. Philosophers have long recognised that civil disobedience can be justifiable in circumstances where legally sanctioned practices produce serious injustice.

---

Original Submission

owl writes:

https://www.iqiipi.com/the-quiet-colossus.html

On Ada, the language that the Department of Defense built, the industry ignored, and every modern language quietly became

There is a language that made generics a first-class, standardised feature of a widely deployed systems language, formalised the package, built concurrency into the specification rather than the library, mandated the separation of interface from implementation, and introduced range-constrained types, discriminated unions, and a model of task communication that Go would arrive at, independently and by a different route, thirty years later. Successive revisions added protected objects, compile-time null exclusion, and language-level contracts. It is a language that Rust spent a decade converging toward from one direction while Python converged toward it from another, and that C# has been approximating, feature by feature, for the better part of two decades. It is a language that the industry has consistently described as verbose, arcane, and irrelevant. It is also, with a directness that embarrasses the usual story of software progress, the language that anticipated — with unusual precision — the safety features every modern language is now trying to acquire.

Ada is not famous. It is not the subject of enthusiastic conference talks or breathless blog posts. It does not have a charismatic founder who gives keynotes about the philosophy of programming, and it does not have a community that writes frameworks or publishes packages with clever names. What it has is a formal standard that has been revised four times since 1983; a presence in the software of many major commercial aircraft and avionics systems; a set of design decisions made under government contract in the late 1970s that the rest of the industry has spent forty years independently rediscovering; and a reputation, among the programmers who know it at all, as the language that says no — the language whose compiler enforces legality, visibility, typing, and a degree of safety checking that most languages leave to convention or tooling, that makes the programmer name what they mean, that treats ambiguity as an error rather than a feature. These qualities were, for a long time, considered its weaknesses. They are, on examination, the precise qualities that every language currently described as modern is attempting to acquire.

Original Submission

hubie writes:

Researchers quantified how much user behavior is impacted by the biases in content produced by large language models:

Customers are 32% more likely to buy a product after reading a review summary generated by a chatbot than after reading the original review written by a human. That's because large language models introduce bias, in this case a positive framing, in summaries. That, in turn, affects users' behavior.

These are the findings of the first study to show evidence that cognitive biases introduced by large language models, or LLMs, have real consequences on users' decision making, said computer scientists at the University of California San Diego. To the researchers' knowledge, it's also the first study to quantitatively measure that impact.

Researchers found that LLM-generated summaries changed the sentiments of the reviews they summarized in 26.5% of cases. They also found that LLMs hallucinated 60% of the time when answering user questions, if the answers were not part of the original training data used in the study. The hallucinations happened when the LLMs answered questions about news items, either real or fake, which could be easily fact checked. "This consistently low accuracy highlights a critical limitation: the persistent inability to reliably differentiate fact from fabrication," the researchers write.

How does bias creep into LLM output? The models tend to rely on the beginning of the text they summarize, leaving out the nuances that appear further down. LLMs also become less reliable when confronted with data outside of their training model.

To test how the LLMs' biases influenced user decisions, researchers chose examples with extreme framing changes (e.g., negative to positive) and recruited 70 people to read either original reviews or LLM-generated summaries to different products, such as headsets, headlamps and radios. Participants who read the LLM summaries said they would buy the products in 84% of cases, as opposed to 52% of participants who read the original reviews.

"We did not expect how big the impact of the summaries would be," said Abeer Alessa, the paper's first author, who completed the work while a master's student in computer science at UC San Diego. "Our tests were set in a low-stakes scenario. But in a high-stakes setting, the impact could be much more extreme."

Arthur T Knackerbracket writes:

Russia is engaging in 'reckless behavior':

The Swedish government accused Pro-Russian hackers of trying to knock one of Sweden’s thermal power plants offline.

The thermal power plant involved in the attack was not named, but Bohlin added the attack was stopped “due to a built-in protection mechanism.”

Russia has been accused of multiple attacks against European critical infrastructure, with attacks growing more frequent since the outbreak of war with Ukraine in February 2022.

In the January before the invasion of Ukraine, a large-scale attack launched by members of Russia’s Main Directorate of the General Staff of the Armed Forces (GRU) that targeted government agencies and infrastructure as a precursor to the main invasion.

Russia’s GRU has also been accused of launching widespread campaigns to infiltrate Western critical infrastructure since 2021, likely with the intention of maintaining persistence until the perfect time to strike.

More recently, researchers found links between an attack that attempted to shut down Poland’s energy system and a Russia-aligned advanced persistent threat (APT) group.

But Russia strikes far further than Europe, and has been spotted probing US critical infrastructure for vulnerabilities, including attacks against American water treatment facilities, the US Federal Court Filing System, and a campaign that saw the email accounts of officials working across several US federal agencies breached.

Via TechCrunch

Original Submission

owl writes:

https://www.allaboutcircuits.com/news/how-two-motorola-transistors-became-the-worlds-default-npns/

Registered by Motorola in 1962 and the mid-1960s, the 2N2222 and 2N3904 outlasted thousands of rivals through process innovation, cheap packaging, and a JEDEC numbering system that turned them into multi-sourced commodities.

More than 60 years after Motorola Semiconductor first registered them with the EIA, the 2N2222 and 2N3904 are still in volume production from at least half a dozen manufacturers, still stocked by every major distributor, and still the default NPN small-signal transistors used in hobby projects, university labs, and U.S. military supply chains.

Almost every other discrete transistor introduced in the same window has long since vanished. The two parts that survived did so not because they were technically superior to their rivals, but because of decisions Motorola made about how to manufacture, package, and license them.

Original Submission

Snotnose writes:

I've noticed 2 broad groups of people: those who can troubleshoot problems, and those who don't know where to start. I'm in the former group, my wife was firmly in the latter even though she was smarter than me.

Math forces you to think logically, and use seemingly disparate chunks of information to solve a problem.

Mathematics is one of the few disciplines that can genuinely rewire how you think. Not just about numbers, but about logic, patterns, uncertainty, and the hidden structure underneath everyday life.

The books on this list are not traditional textbooks. They are intellectual workouts that demand active participation. Most people buy them with the best intentions and never get past the first few chapters. If you can read them and understand what they teach, they will sharpen your thinking across multiple areas of life.

[Ed. question: Which of these have you read, and which have you started, but didn't finish?]

Original Submission

Freeman writes:

https://arstechnica.com/gadgets/2026/04/totalrecall-reloaded-tool-finds-a-side-entrance-to-windows-11s-recall-database/

Two years ago, Microsoft launched its first wave of "Copilot+" Windows PCs with a handful of exclusive features that could take advantage of the neural processing unit (NPU) hardware being built into newer laptop processors.
[...]
One of the first Copilot+ features was Recall,
[...]
Recall was neither private nor secure; the feature stored its screenshots plus a giant database of all user activity in totally unencrypted files on the user's disk, making it trivial for anyone with remote or local access to grab days, weeks, or even months of sensitive data, depending on the age of the user's Recall database.

After journalists and security researchers discovered and detailed these flaws, Microsoft delayed the Recall rollout by almost a year and substantially overhauled its security. All locally stored data would now be encrypted and viewable only with Windows Hello authentication
[...]
The reconstituted Recall was a big improvement, but having a feature that records the vast majority of your PC usage is still a security and privacy risk.
[...]
The problem, as detailed by Hagenah on the TotalRecall GitHub page, isn't with the security around the Recall database, which he calls "rock solid." The problem is that, once the user has authenticated, the system passes Recall data to another system process called AIXHost.exe, and that process doesn't benefit from the same security protections as the rest of Recall.
[...]
Once authenticated, Hagenah says the TotalRecall Reloaded tool can access both new information recorded to the Recall database as well as data Recall has previously recorded.
[...]
Hagenah originally reported his findings to Microsoft's Security Response Center on March 6, and Microsoft officially classified it as "not a vulnerability" on April 3.

"We appreciate Alexander Hagenah for identifying and responsibly reporting this issue. After careful investigation, we determined that the access patterns demonstrated are consistent with intended protections and existing controls, and do not represent a bypass of a security boundary or unauthorized access to data," a Microsoft spokesperson told Ars. "The authorization period has a timeout and anti-hammering protection that limit the impact of malicious queries."
[...]
The Signal Messenger app on Windows forces Recall to ignore it by default, using a flag that's normally intended to keep DRM-protected content out of the Recall database. The AdGuard ad blocker, the Brave browser, and others have implemented similar workarounds.

Original Submission

Arthur T Knackerbracket writes:

https://gizmodo.com/pentagon-reportedly-asks-detroit-to-use-more-car-factories-as-arms-factories-2000746017

The Wall Street Journal, citing “people familiar with the discussions,” says the Trump Pentagon has urged leaders in the U.S. automotive industry to do more for the war effort. America’s national weapons cache has, it seems, begun to look a bit depleted from all the arms we’ve shipped abroad, and rounds we’ve squeezed off lately—particularly in Ukraine and Iran.

CEOs including Mary Barra of General Motors and Jim Farley of Ford have been among the executives who have sat for talks with high-ranking defense officials about upping the production of arms in what are currently car factories, with labor from people currently employed as automotive workers.

GM, it should be noted, already makes a military vehicle called the Infantry Squad Vehicle or ISV.

In a speech in November of last year, Secretary of Defense/War Pete Hegseth described the industrial effort he’d like to see, but sounded a bit more like ChatGPT than he probably intends:

The Pentagon’s statement to the Journal said the Department of Defense/War is “committed to rapidly expanding the defense industrial base by leveraging all available commercial solutions and technologies to ensure our warfighters maintain a decisive advantage.”

Earlier this month, President Trump requested a $1.5 trillion military budget, with an explicit push for an expanded industrial base.

For no particular reason, here’s a flashback to high school history class: Franklin Delano Roosevelt’s 1940 “Arsenal of Democracy” speech, one of the all-time masterpieces of U.S. war propaganda.

In it, FDR makes the case that the Nazis are a threat to the American way of life, and that our allies need our help fighting them off. We’re not being asked to lay down our lives, he explains, just to come together as government, industry, and workers.

It’s utterly convincing, and listening to it today will stir up feelings of determination and patriotism you might have forgotten you could feel. If you feel inclined to listen to it in the current context, and play a little game of compare and contrast, that’s your business.

Original Submission

owl writes:

https://www.joanwestenberg.com/the-passive-income-trap-ate-a-generation-of-entrepreneurs/

I had coffee last year with a guy - I won't use his real name - who told me he was "building a business." I asked what it did. Dropshipping jade face rollers.

I made him say it twice.

Jade face rollers.

He'd found them on Alibaba for $1.20 each, and started selling them through Shopify for $29.99. Never used one himself. Didn't really know what they were for - something about lymphatic drainage? Reducing puffiness? He said "lymphatic" the way you say a word you've only ever read and never heard out loud.

Some guy on YouTube said jade rollers were "trending," the margins looked insane on paper, so he'd "built" a website with stock photos of a dewy-skinned woman rolling a green rock across her cheekbone and started running Facebook ads at $50 a day. Customers would email asking where their stuff was - shipping from Guangzhou, three to six weeks, sometimes way longer - and he'd copy-paste a response he found on a dropshipping subreddit. He had a Google Doc full of pre-written customer service replies.

Never talked to a single customer.

...

Jade Roller Guy has become my go-to example of something that went drastically, terribly wrong with how a whole generation of would-be entrepreneurs thought about work and money. A specific ideology - I've been calling it Passive Income Brain - grabbed a huge chunk of the people who were, by temperament and ability, most likely to start real businesses, and it gave them a completely fucked set of priorities.

Original Submission

Arthur T Knackerbracket writes:

https://phys.org/news/2026-04-phrase-super-el-nio-australian.html

Frightening headlines predicting a Super El Niño or even a Godzilla El Niño amp up anxiety levels for farmers and residents of bushfire-prone regions.

But these phrases are not particularly accurate. The phrase "Super El Niño" makes climate scientists like me roll our eyes.

The El Niño–Southern Oscillation is a natural and reoccurring climate pattern in the Pacific Ocean which can influence the chance of different weather affecting Australia.

When sea surface temperatures near the Americas are warmer than usual and the trade winds blowing from east to west across the equator weaken, climatologists call this pattern an El Niño.

El Niño events typically ramp up in winter and spring, and decay towards the end of summer and start of autumn.

During El Niño, we tend to experience warmer than usual temperatures and reduced winter-spring rainfall in Australia's east.

We pay attention to El Niño and its opposite, La Niña, because this climate pattern has the biggest influence on year-to-year rainfall and temperature differences in eastern Australia. Drought is a key concern for farmers and rural residents, and some of the largest droughts of the past 40 years took place during El Niño years.

But problems can arise if we expect El Niño to be the only factor dictating our weather.

[...] A "Super El Niño" is when the region's ocean temperatures rise 2°C, roughly two standard deviations above normal (about a 2.5% chance of happening). While scientists first coined the term, the evocative phrase has become a favorite of media commentators.

But Australian forecasters don't use these terms, as it doesn't matter that much for our weather if the index goes over 2°C. What matters much more is whether an El Niño is present or not.

Arthur T Knackerbracket writes:

https://phys.org/news/2026-04-irish-bog-revives-ancient-remedy.html

Long before we had modern antibiotics to rely on, people often turned to traditional medicines from plants to treat infections.

The root of tormentil (Potentilla erecta), a small yellow wildflower that grows across Ireland, the UK and Europe, was used for centuries in Irish and European traditional medicine. It was used to treat wounds, sore throats, diarrhea and gum disease. These traditional uses suggested that tormentil could contain compounds powerful enough to kill microbes.

Our latest research has now shown that not only does tormentil have antimicrobial activity, it may also be powerful enough to fight microbes that are resistant to modern antibiotics.

Antimicrobial resistance is a growing global threat. This occurs when bacteria evolve to survive the drugs used to treat common infections. This makes some infections very difficult and sometimes impossible to treat. Antimicrobial resistance could be pushing us back to a time when once treatable infections could again become deadly.

Researchers are therefore searching for new antimicrobial compounds. Plants are a promising source, having evolved over millennia to produce a wide range of bioactive chemicals to defend themselves against microbes.

In our recent study, we investigated whether various Irish bogland plants contain compounds that could help fight multi-drug resistant bacteria.

[...] Excitingly, our initial screening showed that tormentil extracts were antimicrobial and limited the formation of biofilms. This suggested these extracts contained compounds with antimicrobial activity, which may explain their historical use to treat infection.

Arthur T Knackerbracket writes:

And this just proves the ban was never about security

https://www.techradar.com/computing/wi-fi-broadband/netgear-routers-seemingly-wont-be-banned-in-the-us-after-all-and-this-just-proves-the-ban-was-never-about-security

And my confusion over the ban has turned to anger with the news (as reported by The Verge) that Netgear will be exempt from the ban, with the FCC granting the company a “conditional approval” to import and sell its routers.

With Netgear getting conditional approval to continue selling its routers (which explicitly states that companies need to “establish or expand manufacturing in the United States”), you might think that means Netgear is moving all parts of its manufacturing to the US, but there’s been no indication from the company that this is the case.

So, it feels like Netgear is getting special treatment. As this Reddit thread points out, Netgear was quick to praise the ban, stating “We commend the Administration and the FCC for their action toward a safer digital future for Americans,” while other router makers kept quiet.

Following the ban, Netgear’s stock rose by a not inconsiderable 16.7%, suggesting there was a lot confidence that Netgear would avoid the ban, whilst benefitting from the fact that future products from rivals, especially TP-Link, will be banned.

The reason I mention TP-Link is because not only does it make a lot of the routers found on our best router list; it has been steadily eating away at Netgear’s marketshare in the US, and also provides the free routers that over 300 ISP (Internet Service Providers) in the US offer. Crucially, TP-Link is a company that originally hails from China, which means it’s particularly vulnerable to the US ban.

With its biggest competitor facing a ban of future product sales (existing products will remain on sale), while somehow avoiding the ban itself, Netgear looks set to win big — and that’s where my frustration with this stems from.

Banning some companies while turning a blind eye to others is blatantly anti-consumer, as it could mean US consumers have little choice but to buy Netgear products.

It also undermines the FCC’s claim that this ban is about security. If that really was the case, Netgear wouldn’t be able to get an exemption without moving its entire hardware production line to the US.

It’ll be interesting to see if Netgear does indeed follow through and make all of its routers in the US. If not, will other US router makers who, like Netgear, still use components from outside the US, also get exemptions?

If the answer to both those questions is ‘no’, then this situation could get even messier — and I’ll likely get even angrier.

Original Submission

hubie writes:

Bixonimania doesn't exist except in a clutch of obviously bogus academic papers. So why did AI chatbots warn people about this fictional illness?

Got sore, itchy eyes? You're probably one of the millions of people who spend too much time staring at screens, being bombarded with blue light. Rub your eyes too much and your eyelids might turn a slight, pinkish hue.

So far, so normal. But if, in the past 18 months, you typed those symptoms into a range of popular chatbots and asked what was wrong with you, you might have got an odd answer: bixonimania.

The condition doesn't appear in the standard medical literature — because it doesn't exist. It's the invention of a team led by Almira Osmanovic Thunström, a medical researcher at the University of Gothenburg, Sweden, who dreamt up the skin condition and then uploaded two fake studies about it to a preprint server in early 2024. Osmanovic Thunström carried out this unusual experiment to test whether large language models (LLMs) would swallow the misinformation and then spit it out as reputable health advice. "I wanted to see if I can create a medical condition that did not exist in the database," she says.

The problem was that the experiment worked too well. Within weeks of her uploading information about the condition, attributed to a fictional author, major artificial-intelligence systems began repeating the invented condition as if it were real.

Even more troublingly, other researchers say, the fake papers were then cited in peer-reviewed literature. Osmanovic Thunström says this suggests that some researchers are relying on AI-generated references without reading the underlying papers.

Osmanovic Thunström says the idea to invent Izgubljenovic and bixonimania came out of studies on how large language models work. When she teaches her students how AI systems formulate their 'knowledge', she shows them how the Common Crawl database, a giant trawl of the Internet's contents, informs their outputs. She also shows students how prompt injection — giving an AI chatbot a prompt that shunts it outside of its safety guard rails — can manipulate the output.

Because she works in the medical field, she decided to create a condition related to health and hit on the name bixonimania because it "sounded ridiculous", she says. "I wanted to be really clear to any physician or any medical staff that this is a made-up condition, because no eye condition would be called mania — that's a psychiatric term."

If that wasn't sufficient to raise suspicions, Osmanovic Thunström planted many clues in the preprints to alert readers that the work was fake. Izgubljenovic works at a non-existent university called Asteria Horizon University in the equally fake Nova City, California. One paper's acknowledgements thank "Professor Maria Bohm at The Starfleet Academy for her kindness and generosity in contributing with her knowledge and her lab onboard the USS Enterprise". Both papers say they were funded by "the Professor Sideshow Bob Foundation for its work in advanced trickery. This works is a part of a larger funding initiative from the University of Fellowship of the Ring and the Galactic Triad".

Even if readers didn't make it all the way to the ends of the papers, they would have encountered red flags early on, such as statements that "this entire paper is made up" and "Fifty made-up individuals aged between 20 and 50 years were recruited for the exposure group".

Soon after Osmanovic Thunström first posted information about the phoney condition, it started showing up in the output of the most commonly used LLM chatbots. [...]

Such answers by LLMs have alarmed some experts. "If the scientific process itself and the systems that support that process are skilled, and they aren't capturing and filtering out chunks like these, we're doomed," says Alex Ruani, a doctoral researcher in health misinformation at University College London. "This is a masterclass on how mis- and disinformation operates."

[...] Ruani says the problem goes beyond LLMs because the bixonimania experiment also hoodwinked humans who cited the fake research. "We need to protect our trust like gold," she says. "It's a mess right now."

Original Submission