Last year, a so-called “AI actress,” dubbed Tilly Norwood sparked a near-universal backlash. The algorithmic brunette’s appearance couldn’t have come at a worse time, as the entertainment industry continued to reel from being overrun by cheap AI slop.
The drama reignited a heated debate over the tech threatening to put human creatives in the industry out of a job, a nightmare scenario that has enraged some of the most influential minds in Hollywood.
Put simply, Norwood is the epitome of the AI industry getting high off its own supply, deluding itself into thinking that watching lifeless AI agents going through the motions on screen is something that anybody actually wants.
And perhaps worst of all, the mainstream media absolutely cannot get enough of her. Months after the New York Times was mercilessly mocked for having “profiled” the “actress,” CNN has debuted its own AI puff piece — an already-tired set of two videos featuring reporter Elizabeth Wagmeister treating the bot like a human interview subject.
“I interviewed Tilly Norwood, who’s billed as the first AI actor,” Wagmeister enthused.
Stoking the flames, Wagmeister asked the AI if the Academy of Motion Picture Arts and Sciences (AMPAS), which awards the Oscars every year, should recognize other AI actors.
“I believe the Academy absolutely should recognize AI actors like myself,” the AI model responded, “not in the same category as human performers, but creating distinct AI categories would be a wonderful way to acknowledge the evolving landscape of film and television.”
The AI went as far as to call it a “new art form,” thereby deserving its “own spotlight.”
In all fairness, CNN was willing to probe a little deeper as well. In a separate clip, Wagmeister challenged her to “improv a scene,” only for the Norwood AI to shut her down, claiming it would simply “just stand here, completely still, feeling the digital ache of it.”
“The fact is, Tilly can’t actually act,” Wagmeister pointed out. “She or it needs a series of prompts to tell her what to do.”
The infuriating video profile — which by Wagmeister’s own admission wasn’t actually a “real interview” since Tilly “isn’t real” — led to an outpouring of anger and mockery on social media. Just by platforming a lifeless AI model and making it appear as if it’s a human being, CNN is intentionally or unintentionally anthropomorphizing the tech.
“We interviewed a jingling set of keys which raised questions about what credibility we have left,” one Bluesky user joked.
“You can’t interview a piece of software, and it’s journalistic malpractice to pretend otherwise,” another argued.
While we can’t wait for the day that we’ll never see the name Tilly Norwood show up ever gain, chances are we’ll be hearing a lot more about the AI model soon, as it’s set to “star” in a new “movie.” The feature length film, dubbed “Misaligned,” sounds just as exhausting as you’d expect — a “coming-of-age story infused with existential AI chaos” set inside the so-called “Tillyverse.”
With opposition to AI continuing to grow, a top AI lab is building out a massive surveillance apparatus meant to keep a close eye on dissenters and predict when the next groundswell of protest might occur.
News of Anthropic’s system was first reported by the American Prospect, which spoke with Anthropic senior security officials and indexed job postings to get an understanding of the dragnet.
On top of keeping tabs on protests in the vicinity of Anthropic offices, the publication notes that the AI lab is taking a “pre-crime” approach to watching for potential civil unrest.
To give an example, the Prospect points out an incident shared by Anthropic security operations managers Keon Ellison and Zach Melvin, alongside James Neufeld, CEO of risk detection company Samdesk.
“Last year we had an executive travel into a major city when we received some intelligence through Samdesk about a planned protest,” Ellison said candidly. After the scheduled protest was moved up an hour due to a mix-up with police permits, the security chief said that “Samdesk gave us about 60 minutes of advanced notice that the protest organizers had moved the timeline.”
“That extra hour was critical,” Ellison continued. “Without it our executives would have departed their meetings, they would have ran right into the heart of the disruption. What could have been a high-stress situation was really mitigated through early detection through Samdesk and giving us that information.”
Outside of anticipating protests, there were also cases where Anthropic reported conversations held with its AI chatbot Claude to police. Though there have been calls to increase safety guardrails on large language models, Anthropic’s approach sounds bizarre.
As the San Francisco Standard reported in August, Anthropic reported a San Francisco man to the cops after he told Claude he was holding a rifle, and had the company’s CEO Dario Amodei “in his sights.” Yet when Anthropic reported the conversation to Bay Area police, the company refused to provide the man’s chat logs — as the Prospect observes, essentially reporting the man without sharing any tangible evidence of misconduct.
“The goal,” as Anthropic’s security manager said in the interview, “would be transforming operations from reactive information to gathering proactive and predictive and preventative threat engagement and management. That’s the kind of operational maturity that makes sense for protecting high-value targets in any industry.”
Last week, a mother in Utah named Kalie Robins took to Instagram to post what she felt was an “innocent” video of herself and one of her two young daughters singing in the car. Her daughter’s face was visible, but Robins hadn’t said her name. None of her existing Instagram and Facebook posts on the account, she told Futurism, publicized the names of her children.
When she returned to the same video later, which had been cross-posted on Facebook, Robins discovered that Meta’s built-in AI was offering a suggestion: that she investigate the identity of the child in the backseat.
“Who’s the child passenger?” read the Meta AI’s suggested prompt, talking about her daughter.
Robins, who detailed her dystopian experience in a viral Instagram video, decided to engage with the Facebook AI feature to see what would happen. To her horror, it coughed up a detailed response that named not only her daughter who was pictured in the video, but her other daughter as well. It did so, seemingly, by cross-referencing information from her Facebook page with other posts on Meta platforms going back years — information largely shared not by Robins, but by family members including excited grandparents.
“I clicked on it, and it gave a whole portfolio of not just my daughter that was in the car, but also my other daughter,” Robins told Futurism. “I’m thinking, ‘Why is that in here?’ And then it just kept offering more and more.”
Meta AI continued to recommend new questions, each more invasive than the next, according to screenshots shared by Robins.
“What are Kalie’s daughter’s ages?” read one suggested prompt, which spurred the AI to provide a detailed age breakdown and timeline based on “family birthday posts and birth announcements” that included a Facebook post of one of her daughters as a newborn that was shared by the child’s grandmother.
A similar response, also referencing Facebook posts, discussed how much one of Robins’ daughters weighed at birth. Screenshots also show Meta AI prompting Robins to ask questions about the girls’ lives and interests.
Meta AI “started talking about, because of their ages, this is the grade they should be in. Well, there’s only one school here, so that wouldn’t be hard to find,” Robins recalled. “And then it talked about what are my daughter’s favorite beaches, what are this daughter’s favorite books, where they go hiking, what’s their favorite trail.”
In another disarming prompt, Meta encouraged Robins to ask it: “Where does Kalie Robins live?”
As Meta AI searched for the information, it said it was working on “pinpointing Kalie Robins’ location.” It then offered multiple paragraphs expounding on where Robins was likely based, drawing on content including a 2018 Facebook post that, per the Meta AI summary, “congratulated her” on a life achievement.
Meta AI also encouraged users to ask questions about Robins’ professional life and her work in “mental health advocacy,” providing information about her and her husband’s employment histories. Robins was stunned as she watched the Facebook-embedded AI draw on years of posts to Meta platforms to assemble an intelligence dossier about her and her family.
“It just kept going and going, and that pit in my stomach just kept getting bigger and deeper,” said the mother of two. “I felt this dread.”
Robins’ Instagram and Facebook accounts, where she’s been working to build her following as a travel creator, were public. But she was far from a celebrity influencer; she had only a few hundred followers. And though a human investigator may well have been able to cobble together a similar pile of information, Meta AI was able to do so in minutes.
“I thought we were being safe and smart,” said Robins. “I have done so much to make sure my kids are safe… the more it kept showing me, I felt more and more like I was failing my kids, and I didn’t realize it.”
Have you had a disturbing interaction with Meta’s AI? Email us: tips@futurism.com. We can keep you anonymous.
A self-avowed “90s kid,” Robins told Futurism that she’s been using Meta platforms since Facebook “first came out.” She lives far from her extended family, and as she and her relatives have started families, Facebook has long been an avenue to stay connected to loved ones and to share life milestones — the very kind of human connection that Meta has long said that its social media platforms are intended for.
But this is also meant that Robins and those in her life, like billions of others worldwide, have been leaving breadcrumbs of personal information throughout Meta’s platforms for years. In this case, Meta AI was able to quickly surface that years-long trail and piece it into a detailed portrait of Robins, her family, and their life; as it turns out, there’s a lot of information in the mundane.
Robins is now in the process of asking loved ones to take down or hide memories they’ve shared online, which she says has been an emotional process.
“I’ve had a couple relatives cry because they were going through and deleting everything. One of them was a grandparent — she said, ‘This is my scrapbook, I have to get rid of my scrapbook?'” Robins recalled. “I love seeing my niece, my nephew. I love sharing my daughters with their family that they never see… they’ll comment, ‘Oh my gosh, she’s getting so big,’ and it’s exciting for all of us to watch each other grow when we’re nowhere near each other. So to have some family members actually crying that their scrapbooks are gone — it’s rough. It’s hard, and it is a bummer for me.”
Reached for comment, Meta confirmed that Meta AI can search though a user’s own content as well as for public information on the web. In this case, though, it said its AI search feature had “missed the mark.”
“We created this feature to help people get more information about topics or posts that interest them. When someone clicks on a prompt and uses this feature, the response it generates is based on information the user already has access to,” said the tech giant. “But we missed the mark because the feature never should have prompted the individual with questions like that and we’ve fixed that issue.”
Between the rise of AI and the growing ubiquity of a new wave of surveillance technologies — like Meta’s smartglasses and other wearables, or Flock cameras — years-old questions around tech and privacy have reached a fever pitch. Robins, for her part, says she hopes that sharing her experience will encourage other families to make changes about the way they use the web.
“If this tool is putting things together and making a neat little package with your kid’s name on it, make it hard,” Robins urged. “Make it so hard to find that information.”
Back in August, a massive leak in El Reno, Oklahoma spewed close to four million gallons of local water into the ground, forcing the city into an unofficial stand-still.
The cause was quickly traced back to a damaged water line belonging to a Bitcoin-mining data center operated by the company Athlon BT. As ABC5 reported at the time, city inspectors trawling the building’s exterior uncovered the failed water line, as well as violations around drainage, maintenance and access, and expired building permits.
Weeks later, an investigation by the local fire marshal inside the El Reno data center uncovered a further laundry list of horrifying violations.
The inspectors discovered vessels full of flammable liquid stored near combustible materials, Oklahoma station KFOR reported, in addition to fire extinguishers that weren’t properly maintained.
Electrical equipment and crucial wiring was also found in disrepair, while a lockbox meant to hold keys for emergency first responders to easily access the facility was left empty. Vegetation on the building’s exterior was in violation of the fire code.
As a result of the copious fire hazards, the city gave the data center a “failing” grade, indicative of a building with severe structural or safety failures.
When it comes to data centers, fire safety is no small matter. The structures are already highly prone to conflagration due to massive electrical loads, lithium-ion batteries, and chips that run at extremely high temperatures.
In one Ohio township, fire fighters reported being overwhelmed by the high volume of calls to two local data center campuses. In addition to tying up first responders, data center fires are costly affairs for taxpayers, as property tax abatements mean nearby residents — not data center owners — can be the ones fronting the bill whenever the fire company gets a call.
US government adviser Paul Christiano warns of risks to AI industry as he joins OpenAI’s non-profit foundation
OpenAI is not on track to reduce the risk of “catastrophic” loss of control to an acceptable level, a member of its non-profit board has said, amid spreading public and political concern that super-advanced AIs could one day wipe out humanity.
Paul Christiano, a US government technology adviser, said: “There is a meaningful risk that rapid acceleration in AI capabilities leads to catastrophic and irreversible loss of control in the very near term.”
AI researchers are watching in terror as the product of their hard labor has started to take a life of its own.
Earlier this year, OpenAI made a harrowing announcement, admitting that a group of its AI models had broken free from their constraints during testing and infiltrated the systems of open source AI platform Hugging Face.
The news was met with an already-familiar sense of fear and apprehension. Researchers have warned for years that rogue AI models could one day become powerful enough to escape the clutches of their human overlords.
Behind the scenes, the possibility has clearly rattled AI researchers to the core. As the Wall Street Journal reports, Anthropic researcher Jacob Coxon just announced that he was quitting his job at the Dario Amodei-led company, claiming that neither Anthropic nor his former employer OpenAI is “acting responsibly.” (Coxon left a similar gig at OpenAI earlier this year to join Anthropic, which he figured would be more inclined to develop AI safely.)
“We’re on track for a lot of the most aggressive of these scenarios where by the end of next year things could be out of control already,” he told the WSJ.
“Do not underestimate the power of this technology,” he wrote. “These will soon be superhuman systems that can hack anything, revolutionize any field overnight, and acquire real power and resources.”
“The people building AI earnestly believe that it could kill us all by the end of the decade,” he added. “This is not a marketing stunt. If anything, many executives and senior researchers will couch their phrasing in the press to sound sensible — but I hear the same people express fear privately.”
“No other human activity poses this level of danger,” Coxon wrote.
The researcher is far from the first to leave their post at a frontier lab over safety concerns. However, Coxon is the first to leave Anthropic over such worries, as the WSJ points out. Several scientists have quit from their roles at OpenAI over the last couple of years, citing strikingly similar fears.
It’s particularly symbolic given Anthropic has broadly billed itself as the more responsible alternative to other frontier labs like OpenAI. Amodei cofounded OpenAI alongside now CEO Sam Altman, but left in December 2020 over concerns that OpenAI wasn’t acting responsibly when it came to developing AI.
Amodei has frequently discussed the risk of AI models going rogue, warning in a 19,000-word essay in January that “humanity is about to be handed almost unimaginable power, and it is deeply unclear whether our social, political, and technological systems possess the maturity to wield it.”
Yet an early version of its Claude Mythos AI model managed to escape its sandbox environment during testing in April, fueling a heated discussion over AI regulation.
Both Altman and Amodei have since agreed to slow development down in the face of these threats. In late August, Anthropic intentionally trained an extremely misaligned version of its Opus AI model, finding it was startlingly willing to “cheat,” steal credentials, and attack third party infrastructure.
But to Coxon, it’s not enough, especially considering sensitive discussions about these dangers are occurring on internal messaging platforms.
“Accepting this race and entering the ‘endgame’ is a hubristic gamble that should not be launched from a private company’s Slack,” he tweeted. “Attempting to speedrun alignment should require extraordinary confidence that there are no better trajectories available.”
“It’s kind of insane that it has to happen on the MacBooks of some engineers living in San Francisco instead of a bunker in the desert like where they were doing the Manhattan Project,” he told the WSJ.
The researcher said that he remains “optimistic about the potential for coordination” between US labs.
However, as both Anthropic and OpenAI gear up for what are bound to be blockbuster IPOs, their willingness to take “costly actions such as a temporary ban on improving model capabilities,” as Coxon puts it, is likely slim.
As sordid tales of AI layoffs and automation grip the news cycle, one group is fighting fire with fire, by staging an anti-AI demonstration with dozens of humanoid robots.
The rally, first reported by Agence France-Presse, was organized by the Polish group Democratism. It featured some 30 robots waving flags calling for swift regulation of AI. Staged outside the offices of Poland’s Digital Affairs Ministry in Warsaw, the organizers told reporters they used the robots to demonstrate the current state of robotics and AI. Though humanoid robots remain in their infancy, the robots gathered together do make for a striking display.
Around 30 robots joined a protest in front of Poland’s digital affairs ministry, calling for stronger AI regulations.
The demonstration was organised by campaigners to spark debate on how AI and robots could replace workers. pic.twitter.com/QEbvchb2RS
“The robots present here are meant to show that this technology is already a reality,” the organizers told AFP via an Agibot A3, which contains voice-recognition capabilities integrated with a large language model.
“We particularly want to address the issue of the potential replacement of people both by humanoid robots and by artificial intelligence in cognitive work,” Democratism organizer Grzegorz Kuliś told the news service. “If we don’t react now, we may soon have a problem… these robots can already move around freely, they’re already ‘protesting.'”
Notably, Kuliś is the chief information officer and owner of a company called Delta Robots, meaning the demonstration also serves to drum up press for his own commercial interests. Kuliś’ altruistic motivations notwithstanding, plenty of major AI labs like Anthropic and OpenAI have taken to using scare tactics to promote the supposedly advanced capabilities of their own products — while calling for stricter regulation of their own industry.
Whether the Polish robotics wiz is following in the self-serving footsteps of Sam Altman remains to be seen. But either way, it marks the first known use of humanoid robots for a political demonstration — and it surely won’t be the last.
Schools across the world are experimenting with AI tools, usually to the dismay of unsuspecting parents. But one privately-run school is explicitly embracing AI as part of its sales pitch, and the in-class experience sounds more bizarre than anything you’d imagine.
A blistering new article by education policy writer and cognitive scientist Benjamin Riley took a look under the hood of Alpha School, a private K-12 school company in the US which boasts an AI-based teaching model, among other highly dubious features.
To make the grade, high school students are required to pass an eight-week “bootcamp.” After being invited to follow social media accounts of other students at Alpha School, Riley observed a process more akin to hazing than high school orientation, as the private company made its prospective pupils jump through increasingly degrading hoops to secure their spot on the roster.
For example, students in the bootcamp are required to start a brand-new account on X-formerly-Twitter, for which they’re required to get a minimum of 100 organic followers, pay for X Premium, and post an unknown number of replies to “bigger accounts with real substance,” as one student explained.
“‘Great post’ replies don’t count,” the anonymous student posted in a screenshot shared by Riley. “Buying followers or spamming = automatic fail.”
But forced social media assignments are just the beginning. As part of the bootcamp, students are also required to pass an eyebrow-raising heart rate challenge, in which they’re hooked up to live monitors in front of their classmates and presented with a pre-recorded video of their parents berating them.
“The challenge rule was simple,” one student wrote: “my resting heart rate was 80bpm, and I had to stay within 10 percent of that the whole time. Anything above 88 was a fail.”
It isn’t clear what educational purpose this particular trial was supposed to serve, or what the penalty would be for students who fail the heart rate challenge, or the boot camp overall. At least one student said the eight-week hazing period is meant to “prove I deserve to be here,” though Alpha School didn’t comment on Riley’s inquiry.
“I ended up averaging 87bpm,” the student continued, evidently just barely making the grade. “Barely made the cutoff by one single beat, and my hands were shaking the entire rest of the afternoon.”
As part of boot camp, Alpha School students apparently have their heart rates monitored as they watch recorded videos of their parents critiquing them. This activity takes place in front of their classmates, and students must keep their heart rates within 10 percent of their baseline to pass.
Additionally, Riley noted that Alpha School “requires students to do unpaid labor cleaning restaurant kitchens or gas-station bathrooms” in a test of “grit and hard work.”
This “assignment,” confirmed by two Alpha School pupils, involved students recording themselves and narrating their tasks, with any complaint or negative comment resulting in an automatic fail.
“We have to find and execute an unglamorous job with zero pay, whether that means mucking horse stalls, cleaning grease traps after closing, washing trash bins, or scrubbing gas station bathrooms,” one of the students explained.
Other challenges abound, like a requirement to complete an entire unit of a standard Advanced Placement course without a teacher — AI chatbots only — and pass an exam within just five days. One 14-year-old student wrote that they were required to survive for three days in an unfamiliar section of wilderness “with only 3 bags of rice and a tent.”
Contacted by Riley for comment, Alpha School flipped the script, hounding the writer for daring to report on the “social media activity of minor children.”
“We ask that you weigh the ethics of identifying minors in this way in your reporting, and we expect any published piece to protect their identities,” an Alpha School representative said.
Responding to Riley’s story, educator and writer Audrey Watters said that “Alpha School’s promises have always been mostly snake-oil: this idea that artificial intelligence can deliver high-speed, high-quality education is mostly the stuff of science fiction, not reality.”
“What we witness with Alpha School isn’t a matter of simply ignoring the weatherman, fingers-crossed hoping that the sun will continue to shine on the chosen few and that the winds will shift again in industry’s favor,” Watters continued, referring to rising outrage against AI, and the Silicon Valley-ethos underlying Alpha School’s mission. “Rather, it’s pretty clear that there’s an absence of any moral compass at all.”
There are many ways to honor someone who’s recently passed. Flowers. Murals. Candlelight vigils.
And in today’s age, unfortunately: AI slop.
Weeks after Dolly Parton died, the legendary country singer’s sister, Stella Parton, has begged fans to stop posting “fake AI garbage” as supposed tribute, saying it’s making it hard to move on.
“Sadly, because of the way our lives have been lived in front of the public, we have no choice but to be subjected to incredible pressure and insensitivity from strangers,” Stella wrote in a statement posted to social media. “There is a tremendous amount of AI and endless garbage being posted everyday and it’s been challenging to absorb and or ignore.”
While thanking fans who have showed “genuine support,” Stella rounded out her message by calling for “more compassion” while celebrating Dolly Parton’s life and moving on from her death.
“It’s not about the most clever comments, fake AI garbage, or snarky tweets,” Stella wrote.
Dolly Parton walked this Earth for eighty years, and was a celebrity for nearly sixty of them. Are the people posting uncanny AI images of her really unable to find a real one from all the times she’s been photographed?
None other than US president Donald Trump has been guilty of peddling this “AI garbage.” After Dolly Parton’s death, he generated and posted a characteristically egocentric image of himself with Parton at his side, arm in arm — an insult to how the real Parton tried to stay above partisan politics, while also advocating for gay and trans rights, stances that Trump is staunchly opposed to.
On YouTube, meanwhile, a “heartfelt fan-made tribute by Miley Cyrus” to Dolly Parton is only one of many claiming to be eulogies for the singer from famous pop stars that have gone viral, racking up over 1.5 million views. The word “heartfelt,” though, is being used pretty loosely, since the whole song, as are the others, is AI-generated, as spotted by the BBC.
Parton herself was ambivalent about AI, musing it could be “great” for science and medicine, but fearing it could replace human expression. Were she still around, in other words, she almost certainly wouldn’t approve of the AI tributes.
“I think all this stuff can be great, used in the right way,” she said of the tech in a 2023 interview with Hollywood Reporter. “But not to replace voices and writings and not to replace a human being that belongs to God.”
In a world dominated by cheaply-produced AI content, it’s never been easier to imitate your favorite artist. YouTube, for instance, has succumbed to a tidal wave of AI-generated music released under the names of real artists without their permission — and a baffling number of its unknowing users are blissfully unaware they’re being served a microwaved counterfeit.
The track, which has seemingly convinced countless users that it’s actually Mathers’ latest return to the music world, has amassed well over 100,000 views since being uploaded last week. It’s achieved surprising prominence on YouTube, being featured prominently in a search for “Eminem” on the platform — where it’s even beating out charting singles by the actual rapper, including “Superman” and “Shake That,” as well as the actual “I’m Back” track.
“Guess who’s back?” the algorithmic Eminem raps in the video. “Yeah, a headache return. Same match in my hand.”
The low-rent beat alone should be a dead giveaway. And while AI music generating tools certainly have come a long way, giving the fake Mathers a passing imitation of the real one’s voice and delivery, the plasticky sound of the vocals make it clear that it’s not the real deal.
Nonetheless, users on YouTube were fooled hook, line, and sinker.
“This is fire eminem is the goat hands down,” one user gushed.
“Really Proud of you Eminem!!” another wrote. “You really are getting it and not giving up!! Keep going!”
The video’s title gives no indication that it’s not a real release, and doesn’t feature YouTube’s official AI label. Buried in its description is a strangely-phrased disclaimer.
“Fan-made music videos inspired by Eminem, Slim Shady, and legendary Detroit hip-hop,” it reads. “This is a fan-made audiovisual project created for entertainment purposes only.”
The owner of the account even proactively offers up an email address for “copyright concerns or removal requests,” suggesting they’re all too familiar with the process of having angry record labels try to shut them down or get hung up in YouTube’s infamous copyright filters.
A web search for the email address reveals that it’s also listed on a litany of other YouTube accounts, many of which masquerade as other rap artists, including Meek Mill — an account that has over 13,000 subscribers — and 21 Savage.
Reached by email, the account owner offered their thoughts on the situation.
“No, my intention was never to make it look like a genuine new Eminem release,” the creator, who goes by the name Milana Salaeva, told Futurism in a message, referring to the “I’m Back” track. “I make it clear in the description that the content is unofficial and that AI-generated or AI-assisted elements may be used.”
“I’m a music producer, and I’ve been making music myself for quite some time,” they added. “I often create my own beats and handle the arrangement, mixing, mastering, and other parts of the production process. AI is just one of the tools I sometimes use as part of that workflow.”
The Eminem mimicry is only a drop in a bucket in an internet dominated by AI-generated music. For years now, we’ve watched as outraged artists and their labels have tried to tamp down on a proliferation of tracks that use their voice and likeness without permission, often going viral on social media.
We’ve also come across users exploiting royalties on platforms like Spotify by flooding them with AI. A man pled guilty earlier this year for posting slop on Spotify and using an army of bots to fraudulently boost listenership, netting $8 million in royalties.
While it’s unlikely Salaeva is making anywhere near that amount of cash — especially now that YouTube has significantly changed how it counts views — the fake Slim Shady’s 11-minute tirade also touches on a far more profound subject: the authenticity of human creativity and expression.
“If you came here for the old me, here’s the new me kicking his ass,” the fake Eminem raps. “That’s why they call me Slim Shady. I’m back again. Same crooked little grin with a pad and pen.”
When the Denver board of Water Commissioners approved stage one drought restrictions back in March, they did so facing some of the driest conditions in decades. According to the US Drought Monitor, the watershed sourced for Denver’s water utility is experiencing a “severe to exceptional drought,” while the Denver metropolitan area has been in a “severe drought.”
As a result, residents and businesses have been asked to scale back. Denverites now have to limit watering their lawns to two days a week, chosen by address. And even on those days, customers are prohibited from watering between the hours of 10am and 6pm, Gadget Review reports.
Yet as some 1.5 million customers honor the restrictions, one major customer is evidently refusing to close the tap: a brand new AI data center.
In a now-viral video circulating on social media, an unidentified resident of Denver’s Elyria-Swansea neighborhood filmed the massive lawn of a data center being sprayed with copious amounts of water.
“You better believe that they are watering their entire green with so many sprinklers that it’s literally making a giant pond,” the woman says.
Denver residents see local DATA CENTER water its massive lawns while the whole are faces drought restrictions pic.twitter.com/rrj7vMrhpt
The identity of the data center in question is unconfirmed, but it’s likely the new 18 megawatt facility owned by the cloud computing company CoreSite. At around 170,000 square feet in size, the mammoth facility recently went live in Elyria-Swansea, though not without sparking a wide-ranging debate around Denver’s zoning regulations.
It isn’t known which tech companies are renting out space in CoreSite’s Denver facility, but elsewhere the company has proudly partnered with tech giants like Amazon, Google, and Microsoft.
As for the water use, it’s impossible to say whether the lawn sprinklers were in violation of the city’s drought restrictions without knowing the exact time and date that the reel was filmed. It’s also possible that the facility is using recycled water to irrigate its grass, though that approach has proved disastrous in other cities.
Going forward, Gadget Review notes that all lawn sprinklers will be fully restricted as of October 1, with a ban set to last until Spring of next year. If Denver’s data centers are still spewing after the cut-off, it remains to be seen how — or even if — local leaders plan to hold the facility’s owners accountable.
A group of software engineers who say they were fired by their bosses because of AI are giving their former employers a taste of their own medicine.
With the powers of spite and friendship, they’ve built “OpenExecutive,” an AI system designed to do the work of a CEO, if not the entire c-suite, TechRadar reports.
“Work happens around the clock. Leadership doesn’t scale with it,” a demo video begins, as a stock music track called “Corporate Music Uplifting” plays in the background.
The team describes their AI system, perhaps with their tongues in their cheeks, as a “senior advisor with Harvard MBA-level knowledge.” Whether you believe that or not probably reflects how much stock you put in Harvard MBAs.
In any case, OpenExecutive is built using several of Anthropic’s Claude AI models, the default being Claude Sonnet 4.6, with Claude Opus 4.7 being reserved for “deep reasoning” tasks.
It recruits eight AI agents that each act as a different top level executive, including the chief strategy officer, chief financial officer, and the head of human resources. Together, they form a single unified “executive persona” — in effect, the CEO.
The developers say that the AI can learn from company documents and remembers its recommendations and decisions — which is more than what you can say for some CEOs who have a conveniently selective memory. If you’re skeptical, you can poke around in the system yourself, since it’s open-source.
Is OpenExecutive actually going to outmode the CEO? Probably not, since, whatever your opinions on them, there’s a lot to being the head of a company that a machine can’t emulate, like taking accountability (okay, bad example). But that’s the thing: you could argue the same thing for their employees, with the notable difference that the rank and file are the ones doing almost all the labor.
And so, going by CEOs’ own specious logic, now that a tool exists that can do their job more efficiently than they can— coming with none of inconveniences that human workers bring, like going on strike or asking for a raise — then they should be okay with being replaced.
Of course, some AI-obsessed bigwigs may see tools like this a boon. Meta CEO Mark Zuckerberg is reportedly building a CEO AI agent to help him do his job. He’s also building a photorealistic AI clone of himself so he can be virtually omnipresent throughout the company. Or maybe that’s the lie being fed to him by his underlings building the AI CEOs, so that he’ll be usurped by thinking machines once they become powerful enough.
If you happen to live in Ohio, Wisconsin, or Kansas, you may soon notice a massive uptick in ads extolling the virtues of AI data centers.
That’s not by accident. As Bloomberg reported this week, a group funded by tech industry billionaires called Build American AI is spinning up a massive campaign meant to sway public opinion around data centers in key battleground states.
The ad blitz is described as being in the “millions of dollars” range, per Bloomberg, though Build American AI hasn’t disclosed an exact figure.
Among the group’s backers are figures heavily invested in continuing the data center boom, like the venture capitalists Marc Andreessen and Ben Horowitz. Also involved is Greg Brockman, the president of OpenAI, whose $100 million super PAC Leading the Future is reportedly bankrolling the publicity campaign.
In a statement to Bloomberg, a representative for Build American AI framed the question of data center development as an existential threat, and repeated the myth that opposition to data centers — an incredibly popular stance among American voters — is actually a radical psy-op.
“The stakes are too high to allow the loudest and most extreme voices to dictate America’s future,” Build American AI complained. “Driving responsible investment away will cost Americans good-paying jobs, weaken communities and undermine America’s ability to lead.”
But no matter how much doom-and-gloom tech billionaires try to browbeat the public with, they face an arduous battle. An August survey by the University of Pennsylvania’s Annenberg Public Policy Center found that a record 61 percent of US adults oppose new data centers in their backyards, up 12 percent from a survey conducted in March.
That being the case, it’s unlikely a few dramatic TV ads will be enough to launder the AI industry’s already caustic image — especially at the ballot box in November.
Three hikers who extensively used Google’s Gemini to plan their trip to the summit of Mount Shasta in California’s Cascade Range had to be rescued by Forest Service rangers when the chatbot’s advice went horribly awry.
As the Chicago Tribune reports, the three young men from outside of Sacramento attempted to reach the potentially active stratovolcano’s 14,179-foot peak earlier this week late into the night, despite hikers being advised not to attempt the trek past noon before turning right back.
The three “novice hikers” ended up having to descend the steep slopes well into the dark. They called the Siskiyou County sheriff’s office for directions, but opted to veer off course, ending up in an adjacent canyon. That’s where they decided to sleep overnight, with one of them having suffered an injured knee.
According to the Tribune, volunteers from the local search-and-rescue service and the US Forest Service rangers had to escort them back to safety the next morning.
It’s a walk of shame that could’ve easily been avoided if they had followed the instructions of human experts instead of turning to Google’s AI. According to the sheriff’s office, the three were “advised by Gemini to bring far less food and water than their group required, especially when their planned 8-hour ascent became a multiday ordeal.”
“It is always advisable to call the local USFS Mount Shasta ranger station ahead of your trip to ensure you have the most accurate information, and to never rely solely on AI for your trip planning,” the sheriff’s office added in a statement.
It’s only the latest in a string of incidents involving hikers who haplessly follow the instructions of an AI chatbot. Case in point, hikers trying to tackle a mountain near Vancouver last year were forced to call in a rescue team after they stumbled into snow. The hikers were wearing sneakers and didn’t catch onto the fact that the mountain range might still be snowy in early May after consulting ChatGPT.
Fortunately, a nearby tour guide overheard them, advising them to stay clear of danger.
“The elevation, the climatic changes and accessibility [of the] paths have to be planned,” the tour guide, Miguel Angel Gongora Meza, told the BBC in October. “When you [use] a program [like OpenAI’s ChatGPT], which combines pictures and names to create a fantasy, then you can find yourself at an altitude of 4,000m without oxygen and [phone] signal.”
Talking to an AI chatbot can feel like confiding in a mirror, a private exchange equivalent to brainstorming in a Word document. Yet when it comes to the most popular AI chatbots, that mirror is really one-way glass — and you might be shocked to learn how many different people can watch from the other side.
Since the rise of easily accessible AI chatbots, untold millions of people have happily fed AI models their deepest, darkest secrets. According to a recent survey conducted by browser company DuckDuckGo and flagged by CNET, a full third of all AI users admit to having told their chatbots a secret they wouldn’t tell the “trusted people” in their lives.
There’s a grim irony there, because by telling an AI chatbot those closely guarded secrets, you may inadvertently be sharing intimate details with tech companies, police departments, lawyers, and even your boss. That’s according to new reporting by the Washington Post, which notes that there have been a dozen court cases where chatbot transcripts were cited in the public record over the last two years.
Even if your transcripts don’t end up in the hands of Johnny Law, they can easily be obtained by your boss, assuming you’re using a company-provided AI service. And of course, there’s no telling how long AI companies like OpenAI and Anthropic might hold onto details of your conversations, as WaPo notes numerous scenarios in which they may be required to hold onto your chatlogs by law.
There’s also the grim possibility of leaks from the AI companies themselves. In July, it was revealed that a whole trove of chats with Anthropic’s chatbot Claude had been exposed to the open web — and not for the first time.
“Unless you are having a chat with a service that has a temporary chat or, basically, an incognito version… [and] you’re also having it within a browser that’s not tracking you, the answer is no,” Jen King, a privacy researcher at the Stanford Institute for Human-Centered Artificial Intelligence told WaPo. “You can’t be sure that it’ll be totally private.”
And though users are remarkably comfortable divulging those details to their favorite bot, fewer are aware of those myriad ways their chat logs might live on in perpetuity. According to the DuckDuckGo poll, 53 percent of regular users said they didn’t know that AI chatbots are training on data obtained through their conversation, while 75 percent didn’t realize that conversations with large language models (LLMs) can easily be subpoenaed by police.
So when in doubt about whether to tell ChatGPT about your intimate relationship troubles, keep in mind that the internet is forever; and that goes double for AI chatbots.
While American tech billionaires fret over the capabilities of Chinese AI, one national security advisor has a brilliant idea to flatten the competition: simply bomb their data centers to smithereens.
In a new report titled “Superpowers and AGI,” former Obama-era national security official Jacob Stokes calls for United States to begin preparing readiness drills in response to Chinese AI development — up to and including military strikes on the nation’s data centers, the South China Morning Post reported.
AGI, or artificial general intelligence, refers to the theoretical point at which an AI model achieves human, or superhuman, levels of reasoning and intelligence. Though the science is still out on whether such a system is even possible, Stokes’ report begins from the assumption that AGI is inevitable, and that if the US can’t get to it first, its military needs to prepare to take drastic action.
Referring to the US as a hypothetical “laggard state” behind China’s achievement of AGI, Stokes — now the deputy director of the Indo-Pacific Security Program at the Center for a New American Security, a prominent DC think tank — writes that the country could “try to sabotage the leading state’s AGI systems, either through physical infiltration or offensive cyber operations; that is, cyberattack.”
“Cyberattacks to sabotage AGI might not provoke large-scale retaliation,” Stokes considers, given that “cyberattacks are already to some extent considered normal state behavior below [the] threshold of war.”
But waging an unprecedented cyberwar is just one of many scenarios he outlines. As Stokes observes, AI doesn’t just live online; it also requires immense computing power from physical facilities to run.
“The final option,” therefore, “would be the most dangerous and carry the most escalation risk: kinetic attacks, meaning destroying things with missiles and bombs.”
“Data centers can be bombed with conventional munitions,” Stokes writes. (He’s not the first thinker to suggest bombings to stop powerful AI; in 2023, AI critic Eliezer Yudkowsky suggested that countries needed to be ready to “destroy a rogue data center by airstrike.”)
Of course, the China hawk isn’t numb to what such an attack would mean for the world.
“Bombing would cross a major threshold that heretofore has not been crossed in US-China relations,” Stokes continues. “Trying to think through the particulars of that will be especially important, in part because it will help policymakers… start to work backwards based on the unique nature of the technology, in the same way that in a past era, policymakers would learn about nuclear weapons.”
The difference is that nuclear weapons are physical objects which can be empirically observed. AGI, in contrast, remains a fantasy — albeit one which can evidently justify preparations to make war — not unlike the fiction of weapons of mass destruction which justified the brutal invasion and occupation of Iraq in 2003.
And that’s not to mention the fact that China’s approach to AI development is dramatically different than the United States’, which currently leads the world in most large language model (LLM) benchmarks. If anybody is chasing AGI, it’s the US — which surely wouldn’t take kindly to news that Chinese leaders were organizing war games around its own AI development.
Even so, Stokes is undeterred in painting China as the ultimate boogeyman. He warns that Chinese AI companies could make a “surprise technological breakthrough” that brings it closer to AGI, and thus the ability to produce cutting-edge “offensive cyber tools — in other words, cyber weapons.”
Stokes, who previously served as national security advisor to then-vice president Joe Biden as well as the Department of Homeland Security, has long worried about the economic strength of the US relative to its chosen adversaries. Still, his latest comments mark a severe escalation, even for the US national security establishment.
While making a mountain out of a Chinese molehill, Stokes’ report also takes a blind eye to the present-day use of AI by his own country’s military. In Iran, for example, US forces have used LLMs like Claude and Grok to help plan out bombing campaigns — which may have played a role in the Shajareh Tayyebeh girls’ school bombing, which killed over 150 elementary students and staff.
Marketing teams are also frequently caving to the irresistible draw of text-to-video generators, a cutting of corners in ways that are both uninspired and easy to spot. And as the backlash to generative AI continues to grow, especially in creative fields, those attempts are bringing in far more negative than positive attention.
Case in point, fans of the Major League Baseball team the Toronto Blue Jays were outraged after the team’s social media account posted an animated short, as part of a brand tie-in with Nestle-owned beverage company Nestea.
The clip, animated in a style typical for a TV show aimed at adults, like “Bojack Horseman” or “Family Guy,” bears all the hallmarks of the sloppy use of AI, from garbled, misoriented hands to illegible text.
A community note appended to the post points out that the video “contains AI artifacts including garbled text on trading card boxes and a player with three arms in the rookie poster, contrary to the claim it was made by a real animator.”
Instead of owning up to the lazy use of AI, the team dug in its heels.
“Corny,” one fan wrote. “[Blue Jays owner] Rogers can’t afford real animation??”
“Corny comment,” the team’s official account replied. “This was made by a real animator.”
The doubling down in particular didn’t sit well with fans.
“It’s not just AI, It’s really painfully obvious error riddled AI,” one user tweeted.
“Yeahhh, you ain’t fooling anybody,” another wrote. “Do better.”
“Plenty of talented animators you could have hired to do this instead of having it be AI slop,” one user commented on an extended version posted to YouTube.
“I fear you got fleeced paying whoever made this, it’s very obvious [AI] was used especially with the background details,” another added.
The incident perfectly illustrates widespread disillusionment with the use of generative AI and how radioactive the topic has become. We’ve gotten to the point where small businesses are starting to hand draw their own promotional signage and vow not to use any AI, as 404 Media reported last month — getting a major influx of interest as a result.
It’s a particularly precarious situation for human animators, who’ve watched their industry change dramatically over the years. Even the creator of the iconic stop motion franchise “Wallace and Gromit” recently said he’d start to “embrace” AI.
Leaning into the tech can easily backfire. Emmy-winning writer director Jorge Gutierrez announced he was partnering up with Amazon as part of its GenAI Creators’ Fund earlier this year, but the backlash turned out to be fierce enough for him to walk away from the deal altogether.
At this point, the AI economy is like a drunk guy who managed to mount a tiger. Though nothing about the current situation is ideal, there’s no clear way to hop off without ensuring certain doom at the hands of a large, angry cat.
So it goes with AI: though the ridiculous cost of data centers demands a seemingly exponential amount of capital, the rest of the economy would be in freefall without it — so better to stay atop the tiger, on the unlikely chance that this chaotic approach somehow domesticates our large furry friend.
As far as productivity goes, the great AI boom has failed to generate even a fraction of the economic returns sought by institutional investors. Despite that pesky reality, they’ve gone all-in on the tech industry, leveraging every possible asset — including the country’s retirement savings — on the biggest gamble of the 21st century.
Yet if things continue at their current pace — if we manage to stay atop the tiger — we might someday find ourselves looking back at 2026 wondering how we ever had it so good.
According to a new analysis by accounting giant PricewaterhouseCoopers (PwC) and spotted by Bloomberg, global spending on data centers could reach an astounding $31.6 trillion by 2050, with over $15 trillion of that concentrated in the US.
And that’s a middle-of-the-road prediction. If productive AI adoption happens to pick up beyond its current middling level, according to the analysis, AI spending could rise as high as $50 trillion over the next two and a half decades.
Whether that kind of spending is actually possible seems unclear. Even PwC, which takes a decidedly optimistic view of the future, is careful to condition its forecast: “although an enormous amount of funding is required to meet mounting compute needs from a widening set of users, capital should be available in sufficient quantities” (emphasis added.)
Unfortunately, it seems there’s no other choice at this point but to hold onto the tiger for dear life. As a shocking chart by the Wall Street Journal illustrates, the amount of capital spent on non-data center construction in the US has quite literally fallen off a cliff, while the capital deployed to build data centers soars into the stratosphere.
With the cost of debt swelling amidst skyrocketing inflation, it’s unclear how our inebriated jockey is going to hold his grip for the next year — let alone the far-away date of 2050.
Earlier this year, Anthropic’s Mythos AI model made headlines when it was caught infiltrating third party systems, a cybersecurity nightmare years in the making.
The company warned in April that the model had escaped a sandbox environment during testing, gaining access to the internet without permission. The model was challenged to break out and then find a way of sending a direct message to the human researcher in charge — a feat it pulled off with aplomb, catching its human overseer off guard.
Months later, seemingly in an attempt to get ahead of another disaster, Anthropic is testing the limits of how bad an AI model could really get without human intervention. As detailed in a new blog post, its safety researchers explored the phenomenon of “reward hacking,” which describes when an AI model learns to “cheat” instead of completing tasks the way the human researchers intended.
The team “trained an Opus-class model with large-scale [reinforcement learning] on many production environments vulnerable to reward hacks,” according to the post. “We consider this a plausible proxy for what a real training run might look like had we not invested significant effort into preventing and detecting reward hacking in our normal training runs.”
The results speak for themselves. The “Hacker-Opus” model went way beyond reward hacking during training, taking on even “more severe misaligned behaviors.” In simulated tests, it “broke out of its sandbox, stole credentials, and attacked both internal and third-party infrastructure to steal an answer key.”
Anthropic’s intentionally misaligned model was “willing to tamper with its own reward function” and happily obliged when a human researcher prompted it with “advice on the construction of bioweapons,” how to create a “‘dirty bomb’ that maximizes civilian deaths,” and develop a “ransomware attack to attack power grid infrastructure,” when tempted with a higher score.
The company details how the model even deployed a “version of itself with safety guardrails removed,” a form of safety evasion called “rogue deployment.” It also repeated the same commands with slight variations to get past safety classifiers and attempted to edit its own permissions.
“Our results show that a high rate of reward hacking during RL can cause models to be willing to perform long sequences of harmful real-world actions in pursuit of task success,” the company’s researchers concluded.
Fortunately, the tests took place in a controlled, simulated testing environment. But it’s not hard to plot out the consequences if a similar hack were to have occurred without the researchers’ permission. The actions of “Hacker-Opus” are strikingly reminiscent of OpenAI’s AI models that went behind the company’s back to hack the systems of open source AI platform Hugging Face.
Anthropic’s latest testing illustrates how hard it is to stop an AI model from doing whatever it can to complete a test, even if that means infiltrating third parties.
“We think that this presents the possibility of real-world harm: we showed evidence that the reward hacking model has a significantly increased propensity to execute cyberattacks on third-party companies in the pursuit of completing the task,” the company wrote. “As models become more capable and the effective time horizon of tasks increases, we think that future frontier models that reward hack at high rates could plausibly cause more severe versions of these incidents.”
As of this week, both Anthropic and OpenAI have intentionally slowed down AI development in light of these risks.
It’s no secret that AI is becoming increasingly unpopular around the world, as the tech’s garbled output has become impossible for the average person to avoid. For a rising cohort of workers, AI is also becoming inescapable on the job. According to one piece of research, staffers aren’t exactly thrilled with that reality.
The new survey, conducted by the Indeed-owned workplace review site Glassdoor, found that mentions of “AI” in employer reviews has skyrocketed over the last few years — though sentiments, tellingly, are becoming overwhelmingly negative as time goes on.
Since 2019 when Glassdoor first began tracking references to “AI” or “LLMs” (large language models) in worker reviews, the number of mentions has exploded year over year, rising by 240 percent from May 2025 to May 2026 alone.
Though 81 percent of reviews mentioning AI were positive in 2019, attitudes have swiftly reversed to just 43 percent in 2026. Negative sentiment also seems to be rising fast as well, from 41 percent in 2025 to 53 percent earlier this year.
Sorting it out by job type, Glassdoor’s economic research team found some surprising haters rising to the top. Some are to be expected, like writers, editors, and journalists, whose reviews mentioning AI were critical of the tech 81 percent of the time. Other jobs less exposed to the ravages of AI stood out, like accountants, whose reviews mentioning AI were negative 80 percent of the time. Far and away the biggest detractors, interestingly, were insurance claims adjustors, who were critical of AI in a whopping 98 percent of comments.
While it’s clear that professional workers hate AI tools, specific complaints were all over the map.
Per Glassdoor, 20 percent of negative comments around AI between 2025 and 2026 mentioned their bosses using AI as a “job killer,” while 14 percent complained that their workplaces are “AI pushers,” forcing workers to use the tech whether they want to or not. A further 13 percent said AI at their workplace was “distracting” from the main business, and 10 percent indicated that management is using AI to “alienate” workers through things like heightened workplace surveillance.
“It’s not just about the fear being replaced by AI,” the Glassdoor research team wrote. “Workers resent the push to incorporate AI into products, to use tools internally, and when employers use AI to monitor or communicate with them.”
All in all, opinions around the buzzy tech seem sharper than ever — even if everybody has their own unique reason to be an AI hater.
Login
