OpenAI co-founder Andrej Karpathy has joined rival AI lab Anthropic. "The hire is a major coup for Anthropic in the high-stakes competition for elite AI talent -- and another sign the company is emerging as a magnet for some of the industry's most respected technical minds," reports Axios. From the report: Karpathy will start this week on Anthropic's pre-training team, which is responsible for the massive training runs that give Claude its core knowledge and capabilities, according to Anthropic. Karpathy will help launch a new team focused on using Claude itself to accelerate pretraining research -- an increasingly important frontier as AI companies race to automate parts of AI development. "I think the next few years at the frontier of LLMs will be especially formative. I am very excited to join the team here and get back to R&D," Karpathy said in a post on X.

Karpathy is a rare AI figure with credibility across research, industry and education. He was a founding member of OpenAI before serving as Tesla's director of AI, where he led the computer vision team behind Autopilot. Karpathy coined the term "vibe coding" and recently described himself as being in a "state of AI psychosis" since December -- embracing "tokenmaxxing" and aggressively stress-testing frontier models.

The FBI is seeking up to $36 million for nationwide access to automated license plate reader (ALPRs) data, which could let it query vehicle movements across the U.S. and its territories through a commercial database. 404 Media reports: "The FBI has a crucial need for accessible LPRs to provide a diverse and reliable range of collections across the United States. This data should be available across major highways and in an array of locations for maximum usefulness to law enforcement," a statement of work, which describes what data the FBI is seeking access to, reads. ALPR cameras generally work by constantly scanning the color, brand, model, and license plate of vehicles that drive by. This creates a timestamped record of where a particular vehicle was at a specific time that law enforcement can then query, effectively letting them see exactly where someone drove across time. The technology has existed for decades, but has become more pervasive in recent years.

The FBI says it is looking for a vendor that will let it log into a Software-as-a-Service system and then query the collected ALPR data with license plate information, a description of the vehicle, a time or date, and geolocation information. The FBI says it is looking for ALPR coverage in the following areas: Eastern 48 (East of the Mississippi River); Western 48 (West of the Mississippi River); Hawaii; Puerto Rico; Alaska; and outlying areas such as Guam, the U.S. Virgin Islands, or Tribal Territories. In effect, the FBI is looking for ALPR data nationwide and even beyond. An attached price template indicates the FBI is willing to pay $6 million for each of those broad areas, bringing the total to $36 million.

The FBI says it intends to award the contract to a single vendor, but if any such vendor is unable to fulfill all of the requirements, the agency may award the contract to up to two vendors. The contract is specifically for the FBI's Directorate of Intelligence, which oversees the agency's intelligence mission. The FBI is not only a law enforcement agency, but also part of the Intelligence Community. The report notes that the contract appears aimed at vendors like Flock or Motorola Solutions, since they're some of the only companies able to provide the sort of data the FBI is seeking.

Further reading: Small Town Fights Over Flock's AI-Enhanced Network of License Plate-Reading Cameras

America's Library of Congress "is preserving a little piece of Hell," jokes Engadget, "by inducting the soundtrack to the original Doom into the National Recording Registry." The album of demon-slaying tracks is joined by several other notable 2026 additions to the registry, like Weezer's self-titled debut album (colloquially known as "The Blue Album"), Taylor Swift's "1989," Beyonce's "Single Ladies (Put a Ring On It) and the original "Mambo No. 5."

"Doom" was created by Bobby Prince, a freelance composer who worked on lots of id Software games, and also scored Doom's '90s rival Duke Nukem 3D. The soundtrack draws clear inspiration from metal bands, but also touches on techno and ambient music throughout its track list, making for an eclectic soundscape for tearing through enemies. That it all fits together is also impressive in its own right: All of the music for Doom was written before the game had completed levels to play through, according to Prince.
The official announcement from the Library of Congress says Doom "brought a heavy metal energy to MS-DOS systems across the globe," while also pioneering first-person shooter videogames. "Key to Doom's popularity was the adrenaline-fueled soundtrack created by freelance video game music composer Bobby Prince. Prince, a lifelong musician and practicing lawyer, was fascinated by the MIDI technology that rose in prominence in the mid-1980s as a means for instrument control and composition... For "Doom," Prince took inspiration from a pile of CDs loaned by the game's chief designer, John Romero, including seminal works by Alice in Chains, Pantera and Metallica.

Despite the limitations of the 1993-era sound card drivers, Prince composed the perfect riff-shredding accompaniment for the game's demon-slaying journey to hell and back. Taking advantage of his knowledge of MIDI, Prince even worked to ensure that the sound effects he created could cut through the music by assigning them to different MIDI frequencies.

160 miles north of New York City, a man was convicted of manslaughter "with the help of license plate reader technology," reports a local news station. In the small town of Troy (population: 51,000), the mayor described the cameras as "a critical tool" in that investigation. But locals and city officials "have raised concerns about who can access the data collected locally, along with data security, privacy invasions and use by federal authorities, including U.S. Immigration and Customs Enforcement, reports WNYT: When Troy's contract came up for renewal, Mayor Carmella Mantello wanted to keep paying Flock and the council paused payments. The mayor then issued a public safety emergency declaration to keep the license plate readers active. The council has filed a lawsuit to overturn that..."If this illegal emergency order is left unchallenged, we give this mayor and any future mayor regardless of their political party or ideology, unchecked authority to issue an emergency declaration whenever they disagree with the council on any issue," [said Troy council president Sue Steele].
"The technology that's in place today is not the technology of six years ago," council president Steele told another local news station. "We have AI, we have rapidly changing and advancing technology. So that begs the need for regulations to protect certain data." The American Civil Liberties Union warns that Flock will use AI to let law enforcement search its trove of videos. But "Listen, if it was infringing on people's rights, people's liberties, we'd be the first to get rid of it. We have safeguards in place," [mayor] Mantello responded. Mantello noted that data captured by Troy's Flock cameras is only being shared with other local municipalities.

Steele said the data had been shared nationally until she and other elected officials raised concerns. "As far as sharing with local law enforcement, that's necessary in the normal course of investigations. The concern is what Flock does with this data: sharing it with ICE, for instance, and other nefarious outlets," Steele said.
As the debate continues over the small city's 26 Flock cameras, a columnist in Albany wrote that "it's a good thing. We should be asking questions about the growing surveillance state. We should be debating whether this is the future we want." As the American Civil Liberties Union noted, [Flock] has quietly built a broad mass-surveillance infrastructure, with cameras installed in 5,000 communities around the country, and is continually expanding how that network is used. Did we ask for that? Did we vote for it? Not really. The cameras have been installed in municipality after municipality, mostly with little discussion or controversy, which makes us like the proverbial frogs who didn't notice the water getting warmer until it was boiling. Suddenly, surveillance cameras are everywhere; we're always being watched...

[T]he City Council's Democratic majority is considering legislation that, among other steps, would require that data collected by the cameras be generally deleted after 48 hours and that the city be more transparent about how the cameras are used.
The controversy and pushback continues to draw local coverage. The mayor complains the proposed rules restricts the cameras "almost exclusively to cases involving individuals with outstanding felony arrest warrants or situations where officers can determine in advance that an incident will result in a felony charge... This is beyond reckless."

But the Albany columnist still argues many of America's Flock cameras are unnecessary and are "being installed just because... It's worth considering where this might lead and whether the future we're installing is the future we want."

An anonymous reader quotes a report from Reuters: Meta employees distributed flyers at multiple U.S. offices on Tuesday to protest the company's recent installation of mouse-tracking software on their computers, according to photos of the pamphlets seen by Reuters. The flyers, which appeared in meeting rooms, on vending machines and atop toilet paper dispensers at the Facebook owner's offices, encouraged staffers to sign an online petition against the move. "Don't want to work at the Employee Data Extraction Factory?" they asked, according to the photos seen by Reuters. [...]

The pamphlets and the petition both cite the U.S. National Labor Relations Act, saying "workers are legally protected when they choose to organize for the improvement of working conditions." In the UK, a group of Meta employees has started organizing a drive for unionization with United Tech and Allied Workers (UTAW), a branch of the Communication Workers Union. The employees set up a website to recruit members using the URL "Leanin.uk," a reference to former Chief Operating Officer Sheryl Sandberg's best-selling book encouraging women to seek equal footing in the workplace. "Meta's workers are paying the price for management's reckless and expensive bets. While executives chase speculative AI strategies, staff are facing devastating job cuts, draconian surveillance, and the cruel reality of being forced to train the inefficient systems being positioned to replace them," said Eleanor Payne, an organizer with UTAW. "If we're building agents to help people complete everyday tasks using computers, our models need real examples of how people actually use them -- things like mouse movements, clicking buttons, and navigating dropdown menus," said a statement Meta issued earlier.

fjo3 shares a report from Phys.org: Ever felt like mosquitoes bite you while ignoring everyone else? Scientists are now making progress in deciphering the complex chemical cocktail that makes particular people more enticing to these disease-spreading bloodsuckers. "It's not a misconception -- mosquitoes are attracted to some people more than others," Frederic Simard of France's Institute of Research for Development told AFP. "But we are not all magnets all the time," the medical entomologist added.

A range of sensory cues can cause mosquitoes to pick one human over another -- mainly the smell and heat our bodies give off, and the carbon dioxide we exhale. Female mosquitoes -- which are the only ones that bite -- detect these signals with finely tuned receptors, then choose their target accordingly. "We have known for over 100 years that mosquitoes are attracted by the carbon dioxide that we exhale -- this is the first signal that triggers their behavior" when they are dozens of meters away, Swedish scientist Rickard Ignell told AFP. Within around 10 meters, "mosquitoes will start detecting our odor, and in combination with carbon dioxide," this attracts them even more, said the senior author of a recent study on the subject. As they get closer, body temperature and humidity make particular humans even more enticing.

[...] For Ignell's recent study, the researchers released Aedes aegypti mosquitoes -- known for spreading yellow fever and dengue -- on 42 women in a lab, to see which ones they preferred. "We have shown that mosquitoes use a blend of odorous compounds (we identified 27 that the mosquitoes will detect, out of the possible 1,000) for their attraction to us," Ignell said. The woman the mosquitoes most liked to bite -- which included pregnant women in their second trimester -- produced a large amount of a particular compound made by a breakdown of the skin oil sebum. That even a small increase of this compound -- called "1-octen-3-ol", or mushroom alcohol -- made a difference came as a surprise, Ignell emphasized.

South Korea's presidential policy chief is calling for a "citizen dividend" that would return some AI-driven profits and tax revenue to the public. The Straits Times. From the report: Presidential policy chief Kim Yong-beom said in a Facebook post that a portion of the profits and tax revenue derived from the artificial intelligence boom "should be structurally returned to all citizens." That is because, Mr Kim argued, the economic gains from AI are based at least partly on industrial infrastructure built by the country over five decades. Mr Kim's comments come after tens of thousands of people gathered outside Samsung's main chip hub in April to demand employees get a greater share of AI profits. The company's labour union wants 15 per cent of operating profit handed to chip-division employees.

The union has threatened an 18-day strike starting May 21. Workers have pointed to rising payouts at SK Hynix, which in 2025 agreed to allocate 10 per cent of its annual operating profit to a performance bonus pool, as evidence they deserve more pay. "Excess profits in the AI era are, by nature, concentrated," Mr Kim wrote. Memory companies, core engineers and asset holders are highly likely to receive substantial benefits, while much of the middle class may experience only indirect effects.

Researchers at Columbia demonstrated the first real-time brain-controlled hearing system that can identify which speaker a listener is focusing on in a noisy environment and automatically amplify that voice while suppressing others. "This breakthrough addresses the 'cocktail party effect,' a major limitation of conventional hearing aids, which often struggle to distinguish between overlapping conversations in noisy settings," reports Neuroscience News. From the report: In the new study, Columbia researchers teamed up with surgeons and their epilepsy patients who were undergoing brain surgery to better pinpoint the sources of their seizures. The hospital patients, who volunteered to be part of this study, already had electrodes implanted in their brains. [senior author Nima Mesgarani's] system used the electrodes to measure the brain activity of the patients as they focused on one of two overlapping conversations played simultaneously. The system then automatically detected which conversation a patient was paying attention to and adjusted the volume in real time, turning up that conversation while quieting the other. For one volunteer, the experience of controlling the system with her brain was literally unbelievable. She accused the researchers of secretly adjusting the volumes. Others told stories about friends and family with hearing impairments who could benefit from such a technology. One person said: "It seems like science fiction."

[...] The scientists developed real-time machine-learning algorithms that could examine the brainwaves and identify which conversation the patients were paying attention to. Once deployed, their system could rapidly deduce which conversation each listener was paying attention to and make it easier for them to hear it. This happened both when the researchers guided the subjects toward a particular conversation, and when the subjects chose freely, as would be necessary in a real-world conversation. "For this to work in real time, the system has to be very fast, accurate and stable for the experience to feel pleasant for the listener," Dr. Mesgarani said. The scientists found their new system correctly identified which conversation the volunteers paid attention to. This dramatically improved the intelligibility of the speech the volunteers focused on, reduced listening effort, and was consistently preferred by the volunteers when compared to conversations the system did not provide assistance with. One volunteer recalled her uncle, who had hearing problems. "Can you imagine if this technology existed in a world [where] ... he could access it? He might actually live a much more peaceful... life." The research has been published in Nature Neuroscience.

An anonymous reader quotes a report from 404 Media: Speaking to graduates of University of Central Florida's College of Arts and Humanities and Nicholson School of Communication and Media on May 8, commencement speaker Gloria Caulfield, vice president of strategic alliances at Tavistock Group, told graduating humanities students that AI is the "next industrial revolution," and was met with thousands of booing graduates. "And let's face it, change can be daunting. The rise of artificial intelligence is the next industrial revolution," Caulfield said. At that point, murmurs rippled through the crowd. Caulfield paused, and the crowd erupted into boos. "Oh, what happened?" Caulfield said, turning around with her hands out. "Okay, I struck a chord. May I finish?" Someone in the crowd yelled, "AI SUCKS!"

Her speech begins around the hour and 15 minute mark in the UCF livestream. [...] Before the industrial revolution comment, Caulfield praised Jeff Bezos for his passion and use of Amazon as a "stepping stone" to his real dream: spaceflight. Rattled after the crowd's reaction, she continued her speech: "Only a few years ago, AI was not a factor in our lives." The crowd cheered. "Okay. We've got a bipolar topic here I see," Caulfield said. "And now AI capabilities are in the palm of our hands." The crowd booed again. "I love it, passion, let's go," she said. "AI is beginning to challenge all major sectors to find their highest and best use," she continued. "Okay, I don't want any giggles when I say this. We have been through this before, these industrial revolutions. In my graduation era, we were faced with the launch of the internet."

She goes on to talk about how cellphones used to be the size of briefcases. "At that time we had no idea how any of these technologies would impact the world and our lives. [...] These were some of the same trepidations and concerns we are now facing. But ultimately it was a game changer for global economic development and the proliferation of new businesses that never existed like Apple and Google and Meta and so many others, and not to mention countless job opportunities. So being an optimist here, AI alongside human intelligence has the potential to help us solve some of humanity's greatest problems. Many of you in this graduating class will play a role in making this happen."

Hashicorp co-founder Mitchell Hashimoto says GitHub's frequent outages have made it "no longer a place for serious work," prompting him to move his Ghostty terminal emulator project elsewhere after 18 years on the platform. The Register reports: "I've been angry about it. I've hurt people's feelings. I've been lashing out. Because GitHub is failing me, every single day, and it is personal. It is irrationally personal," he wrote. The reason for his ire is the service has become unreliable. "For the past month I've kept a journal where I put an 'X' next to every date where a GitHub outage has negatively impacted my ability to work," he wrote. "Almost every day has an 'X'. On the day I am writing this post, I've been unable to do any PR review for ~2 hours because there is a GitHub Actions outage."

Hashimoto penned his post a few days before an April 28 incident that saw pull requests fail to complete due to an Elasticsearch SNAFU. Incidents like that mean Hashimoto has decided GitHub "is no longer a place for serious work if it just blocks you out for hours per day, every day." "It's not a fun place for me to be anymore," he lamented. "I want to be there but it doesn't want me to be there. I want to get work done and it doesn't want me to get work done. I want to ship software and it doesn't want me to ship software."

The developer says he wants GitHub to improve, but "I also want to code. And I can't code with GitHub anymore. I'm sorry. After 18 years, I've got to go." He's open to a return if GitHub can deliver "real results and improvements, not words and promises." But for now, he's working to move Ghostty to another collaborative code locker. "We have a plan but I'm also very much still in discussions with multiple providers (both commercial and FOSS)," Hashimoto wrote. "It'll take us time to remove all of our dependencies on GitHub and we have a plan in place to do it as incrementally as possible."

He's doing the equivalent of leaving a toothbrush at a former partner's house by leaving a read-only mirror of Ghostty on GitHub, and by keeping his personal projects on the Microsoft-owned service. But Hashimoto's moving his day job somewhere new. "Ghostty is where I, our maintainers, and our open source community are most impacted so that is the focus of this change. We'll see where it goes after that," he concluded.

An anonymous reader quotes a report from Ars Technica: University of Oregon chemist Christopher Hendon loves his coffee -- so much so that studying all the factors that go into creating the perfect cuppa constitutes a significant area of research for him. His latest project: discovering a novel means of measuring the flavor profile of coffee simply by sending an electrical current through a sample beverage. The results appear in a new paper published in the journal Nature Communications.

[...] The coffee industry typically uses a method for measuring the refractive index of coffee -- i.e., how light bends as it travels through the liquid -- to determine strength, but it doesn't capture the contribution of roast color to the overall flavor profile. So for this latest study, Hendon decided to focus on roast color and beverage strength, the two variables most likely to affect the sensory profile of the final cuppa. His solution turned out to be quite simple. Hendon repurposed an electrochemical tool called a potentiostat, typically used to test battery and fuel cell performance. Hendon used the tool to measure how electricity interacted with the liquid. He found that this provided a better measurement of the flavor profile. He even tested it on four different samples of coffee beans and successfully identified the distinctive signature of a batch that had failed the roaster's quality-control process.

Granted, one's taste in coffee is fairly subjective, so Hendon's goal was not to achieve a "perfect" cup but to give baristas a simple tool to consistently reproduce flavor profiles more tailored to a given customer's taste. "It's an objective way to make a statement about what people like in a cup of coffee," said Hendon. "The reason you have an enjoyable cup of coffee is almost certainly that you have selected a coffee of a particular roast color and extracted it to a desired strength. Until now, we haven't been able to separate those variables. Now we can diagnose what gives rise to that delicious cup." Outside of his latest electrical-current experiment, Christopher Hendon's coffee research has shown that espresso can be made more consistently by modeling extraction yield -- how much coffee dissolves into the final drink -- and controlling water flow and pressure.

He also found that static electricity from grinding causes fine coffee particles to clump, which disrupts brewing. The solution: adding a small squirt of water to beans before grinding (known as the Ross droplet technique) to reduce that static, cut clumping and waste, and lead to a stronger, more consistent espresso.

Researchers say infrasound -- low-frequency vibrations from things like pipes, HVAC systems, and traffic that humans can't consciously hear -- may help explain why some old buildings feel unsettling or "haunted." Rodney Schmaltz, senior author and professor at MacEwan, says: "Consider visiting a supposedly haunted building. Your mood shifts, you feel agitated, but you can't see or hear anything unusual. In an old building, there is a good chance that infrasound is present, particularly in basements where aging pipes and ventilation systems produce low-frequency vibrations. If you were told the building was haunted, you might attribute that agitation to something supernatural. In reality, you may simply have been exposed to infrasound." ScienceBlog.com reports: Infrasound sits below roughly 20 Hz, the lower limit of what the human ear can ordinarily detect. It's generated by storms, by volcanic activity, by tectonic rumblings deep in the Earth's crust, and (this is the part that matters) by the mundane mechanical heartbeat of cities: ageing pipes, HVAC systems, traffic, industrial machinery. "Infrasound is pervasive in everyday environments, appearing near ventilation systems, traffic, and industrial machinery," says Schmaltz. Most of the time, we walk through it without a second thought. The question the team wanted to answer was whether walking through it was actually doing something to us, whether the frequency was registered somewhere below consciousness, somewhere we couldn't readily name.

The experimental setup was deliberately ordinary. Thirty-six undergraduate students filed one at a time into isolated testing rooms and sat alone with a piece of music, either a calming instrumental or a horror-themed ambient track designed to provoke discomfort. Hidden subwoofers, including a 12-inch unit positioned in an adjacent hallway and a 16-inch speaker oriented toward the ceiling in a neighboring room, pumped infrasound at approximately 18 Hz into half those spaces. The participants had no idea. That last point turned out to be rather important. When the team ran the numbers, they found that participants couldn't reliably identify whether infrasound had been present. Their guesses were, statistically speaking, no better than chance. And according to Schmaltz, participants' beliefs about whether the infrasound was on had no detectable effect on their cortisol or mood. The physiological response didn't care what the participants thought was happening. It just happened anyway.

What happened, specifically, was this: those exposed to infrasound reported higher irritability, lower interest in the music, and a tendency to rate the music as sadder, irrespective of whether it was the calming or the horror track. Cortisol levels, measured before and about 20 minutes after exposure, were also elevated. Kale Scatterty, the PhD student who led the work, notes that irritability and cortisol do tend to move together under ordinary stress, but adds that "infrasound exposure had effects on both outcomes that went beyond that natural relationship." That distinction matters more than it might seem. Previous theories about infrasound and paranormal experience have often leaned on anxiety as the explanatory mechanism, the idea that low-frequency sound triggers a kind of free-floating dread that the mind then reaches for supernatural explanations to account for. The new data don't really support that picture. Measures of anxiety didn't budge significantly. What went up was irritability and disinterest, a kind of sour, low-grade aversion rather than fear. That's perhaps a more honest description of how a lot of ghost stories actually feel in the telling: not screaming terror, but wrong atmosphere, a sense of unease that never quite crystallizes into something you can point at. The study has been published this week in Frontiers in Behavioral Neuroscience.