Experts have <a href="https://futurism.com/stephen-hawking-ai-replace-humans">long warned</a> that artificial intelligence is making its way into <a href="https://futurism.com/msn-is-publishing-more-fake-news">every corner</a> of <a href="https://foreignpolicy.com/2025/02/13/ai-artificial-intelligence-us-government-elon-musk-doge-trump/">our lives</a> — but we didn't necessarily expect it to seep into our actual dreams.
According to a new survey <a href="https://amerisleep.com/blog/dreaming-of-ai/">funded by the mattress company Amerisleep</a>, one in five people are regularly having dreams about AI. Unsurprisingly, most of those dreams were rooted in anxiety — and often, they were nightmares.
The mattress makers asked 1,000 people in the United States a battery of questions regarding AI and dreams. While a majority (56 percent) said they dream about AI less than once per month, nearly one in five said they have one AI dream in the same time period, and 16 percent said they have more than one. Nearly 10 percent, meanwhile, said they have at least one dream about the technology per week.
Though one in five of total survey respondents said they have AI dreams, that figure went up based on various demographics. Nearly a quarter of surveyed Gen Z-ers, or those born after 1997, said they've dreamed about AI — and of those, one in six said they have nightmares about losing their jobs to the technology (the survey summary didn't say how many Millennials and Gen Xers dreamt about AI).
According to Amerisleep's analysis, more than a quarter of participants said their biggest subconscious concern seemed to be about AI taking over the world. An additional 26 percent said they dreamt about speaking to AI as if it were a person, while 19 percent of total respondents said that losing jobs to the tech was a common dream scenario.
And speaking of, "Dream Scenario" star Nicholas Cage has, coincidentally enough, also been talking about AI nightmares recently.
During his <a href="https://variety.com/2025/film/news/nicolas-cage-slams-ai-robots-cannot-reflect-human-condition-1236294371/">acceptance speech</a> at the Saturn Awards in February, for which he won "best actor in a film" for the picture, the inimitable thespian suggested that he found the rise of AI — and its ability to worm its way into our subconscious — one of the most "disturbing" things occurring in the world.
"I am a big believer in not letting robots dream for us," Cage said. "Robots cannot reflect the human condition for us."
Obviously, the <a href="https://lithub.com/on-the-jealous-rivalry-between-nicolas-cage-and-his-uncle-francis-ford-coppola/">lowkey Coppola</a> was using "dream" in a more metaphysical sense to refer to art — but the sentiment stands, especially in light of the surreal dream-focused movie for which he was taking home the award.
"That is a dead end if an actor lets one AI robot manipulate his or her performance even a little bit, an inch will eventually become a mile and all integrity, purity and truth of art will be replaced by financial interests only," the icon said. "We can’t let that happen."
The "Mandy" star was, as he often is, right on the money with that one — and in light of this new survey about AI dreaming, <a href="https://futurism.com/studio-ghibli-openai-mizayazaki-ai-art-disgusting">that sort of warning</a> feels more prescient than ever.
More on weird AI side effects: <a href="https://futurism.com/the-byte/chatgpt-dependence-addiction">Something Bizarre Is Happening to People Who Use ChatGPT a Lot</a>

Dreamscapes

We now know that androids do, in fact, dream of electric sheep. 

Large Numbers of People Report Horrific Nightmares About AI

Who said there's an <a href="https://www.npr.org/2025/03/18/nx-s1-5330454/egg-shortages-record-prices-usda-canada">egg crisis</a>?
NASA's Perseverance Rover has stumbled across an extremely strange rock that's "astonished" scientists while exploring the rim of the Jezero crater on Mars.
Nicknamed "<a href="https://science.nasa.gov/blog/shocking-spherules/">St. Pauls Bay</a>," the rock is made up of hundreds of densely packed, millimeter-sized dark spheres — not unlike an insect's slimy egg cluster. 
These "spherules," as they're known, are somewhat heterogeneous. As seen in a closeup taken with Perseverance's SuperCam Remote Micro Imager, some of them appear more elliptical while others are more angular. Stranger still, some spherules have tiny pinholes through their centers. 
In short, they're nothing like the rocks that make up the surrounding landscape, which are lighter and flatter.
<blockquote class="twitter-tweet" data-width="500" data-dnt="true">
I&#39;m finding bumpy rocks on this side of Mars, too!
While different than yours, these spheres are created when droplets of molten rock are thrown into the air by volcanoes or meteorite impacts. The droplets cool as they travel, forming round shapes. <a href="https://t.co/w0b59yw3Sk">https://t.co/w0b59yw3Sk</a> <a href="https://t.co/ueJY9mM0Ku">pic.twitter.com/ueJY9mM0Ku</a>
&mdash; NASA&#39;s Perseverance Mars Rover (@NASAPersevere) <a href="https://twitter.com/NASAPersevere/status/1905400665654132776?ref_src=twsrc%5Etfw">March 27, 2025</a></blockquote>
<script async src="https://platform.twitter.com/widgets.js" charset="utf-8"></script>
NASA scientists have encountered puzzling spherules in the past. As far back as 2004, the Mars Opportunity rover, during its expedition in the Meridiani Planum, spotted what's come to be known as "<a href="https://science.nasa.gov/resource/martian-blueberries-2/">Martian Blueberries</a>" — loose, smoothly surfaced pebbles, each roughly an inch across. And last year, Perseverance found light-colored sedimentary rocks with an uneven "<a href="https://science.nasa.gov/blog/perseverance-finds-popcorn-on-planet-mars/">popcorn-like</a>" texture.
In keeping with the appetizing analogues, these latest instances of spherules, found on a slope of the Jezero's crater's western rim called Witch Hazel Hill, are more like geological Dippin' Dots.
Whereas previous spherules were thought to be concretions — the result of groundwater flowing through sediment and depositing minerals that clump around small nuclei of rocks, forming a larger ovoid over time — it's not clear what formed St. Pauls Bay. 
Some spherules form when droplets of molten rock spewed in a volcanic eruption rapidly cool. Mars, home to the largest volcano in the solar system, has a long history of volcanism, so this could be the case. But spherules can also form by rocks that are vaporized in a meteorite strike, which condense into an ovoid in the aftermath; Mars is <a href="https://phys.org/news/2024-06-analysis-nasa-insight-mars-meteoroids.html">routinely bombarded by space rocks</a>.
Right now, it's impossible to determine St. Pauls Bay's origins because it's a float rock, a type of loose rock that's not part of the surrounding bedrock and originated from somewhere else.
The search, however, is far from over.
"The team are now working to link the spherule-rich texture observed at St. Pauls Bay to the wider stratigraphy at Witch Hazel Hill, and initial observations have provided tantalizing indications that it could be linked to one of the dark-toned layers identified by the team from orbit," Alex Jones, a Mars researcher at Imperial College London, <a href="https://science.nasa.gov/blog/shocking-spherules/">wrote in a NASA blog</a>. "Placing these features in geologic context will be critical for understanding their origin, and determining their significance for the geological history of the Jezero crater rim and beyond!"
More on Mars: <a href="https://futurism.com/evidence-huge-ocean-mars-surface">Readings Show Evidence of Huge Ocean Under Mars' Surface</a>

Egg-Citing

Totally out of place.

Scientists Puzzled When Mars Rover Finds Egg-Like Rock

Despite SpaceX already being NASA's largest private contractor, Elon Musk is still hellbent on assuming control over the space agency.
As the <a href="https://www.wsj.com/business/elon-musk-nasa-mars-space-travel-d3978a7b">Wall Street Journal reports</a>, the mercurial CEO is doing much in his power to push NASA to <a href="https://futurism.com/elon-musk-scrap-nasa-moon-program">skip its planned journeys to the Moon</a> and focus on Mars instead, highlighting his <a href="https://futurism.com/elon-musk-city-mars-death-catastrophe">long-term ambitions</a> of establishing a human presence there.
Exerting his will at the agency could soon become even easier, with billionaire tech founder — and deep-pocketed <a class="underline hover:text-futurism hover:no-underline transition-all duration-200 ease-in-out" href="https://futurism.com/the-byte/first-person-footage-astronaut-spacex-spacewalk">SpaceX space tourist</a> — Jared Isaacman expected to be sworn in as NASA administrator <a href="https://x.com/SciGuySpace/status/1906701157851857076">within the coming months</a>.
Per the WSJ, Musk has reached out to Isaacman to reiterate his desire to work together to get humans to Mars.
The space agency has already suffered greatly due to Musk's efforts to gut the federal government with the help of his so-called Department of Government Efficiency. NASA <a href="https://futurism.com/nasa-disgusted-elon-musk-layoffs">was hit with mass layoffs</a> earlier this year, with affected workers <a href="https://www.cnn.com/2025/03/24/science/nasa-layoffs-policy-office/index.html">given short notice and denied time-off awards</a>.
According to the WSJ's sources, the White House is still considering <a href="https://futurism.com/nasa-sls-moon-rocket-might-be-canceled">canceling NASA's Space Launch System</a> (SLS), a behemoth rocket that's being designed to send astronauts to the Moon before the end of this decade.
It would be a major change of course for the agency, as plans are already well underway. Just last week, the core stage of NASA's SLS rocket for its upcoming crewed Artemis II mission around the Moon was <a href="https://www.nasa.gov/blogs/missions/2025/03/24/nasas-artemis-ii-core-stage-integration-complete-at-kennedy/">stacked with its solid rocket boosters</a>, taking a major step towards its anticipated launch roughly a year from now.
Freeing up those funds to spend them on its efforts to return to Mars instead could raise plenty of eyebrows, given Musk's <a href="https://futurism.com/elon-musk-doge-gutting-nhtsa-tesla">enormous conflict of interest</a>. SpaceX officials have been informing people outside the company that the space agency's resources will be reallocated toward Mars efforts, the WSJ reports.
That kind of reassignment could be of great benefit to Musk's company, which is already working on a large rocket designed for interplanetary travel called Starship.
Musk's DOGE has also aimed at <a href="https://www.nytimes.com/2025/02/19/us/politics/trump-firings-faa.html">gutting the Federal Aviation Administration</a>, which has been responsible for ensuring that Starship launches follow air safety laws.
According to the WSJ, key SpaceX officials are already gaining power at the space agency. Longtime SpaceX executive Michael Altenhofen, who is close to Isaacman, was installed as a senior advisor in January.
But is NASA really ready to ditch plans that have been decades in the making? A spokeswoman didn't rule out the possibility that NASA is still committed to "return Americans to the lunar surface," in a statement to the WSJ.
However, given Musk's animosity toward NASA's Artemis program, the country's upcoming return to the lunar surface could be on thin ice.
"No, we’re going straight to Mars," Musk <a href="https://x.com/elonmusk/status/1875023335891026324">tweeted in January</a>. "The Moon is a distraction."
In a <a class="underline hover:text-futurism hover:no-underline transition-all duration-200 ease-in-out" href="https://x.com/elonmusk/status/1871997501970235656">separate tweet</a> a week earlier, Musk  accused the space agency's Moon program of being "extremely inefficient as it is a jobs-maximizing program, not a results-maximizing program."
Musk has even called for the <a href="https://futurism.com/elon-musk-calls-destroying-space-station-soon">early abolishment of the International Space Station</a>, which <a href="https://futurism.com/elon-musk-argument-iss-destroy">angered former astronauts</a>.
During his inaugural address, Trump also <a href="https://www.space.com/space-exploration/us-will-plant-american-flag-on-mars-trump-tells-congress">declared</a> that his administration would be "launching American astronauts to plant the Stars and Stripes on the planet Mars," indicating he's very much aligned with Musk's vision.
But pulling off an enormous U-turn this far into NASA's plans to return to the Moon could prove extremely expensive and a major setback. Besides, gutting the Artemis program will likely draw the ire of powerful bipartisan supporters in Congress.
"Starship? I want success out of it," representative Brian Babin (R-TX) wrote in a <a href="https://www.bloomberg.com/news/newsletters/2025-02-12/nasa-s-boeing-built-sls-moon-rocket-won-t-be-easy-to-cancel?sref=YfHlo0rL">February statement</a>. "But for us to beat the Chinese... it’s going to have to be SLS that does it."
More on NASA: <a href="https://futurism.com/nasa-disgusted-elon-musk-layoffs">NASA Disgusted by Elon Musk's Disrespect</a>

Mars Uber Alles

Is NASA really getting ready to cancel the Artemis program?

Elon Musk Secretly Working to Take Over NASA

So far, the AI industry has enjoyed smooth sailing. It's courted billions in <a href="https://futurism.com/ai-researchers-tech-industry-dead-end">slap-happy investments</a> and <a href="https://newsroom.ibm.com/2024-01-10-Data-Suggests-Growth-in-Enterprise-Adoption-of-AI-is-Due-to-Widespread-Deployment-by-Early-Adopters">sign-first, read-later</a> contracts, ballooning some Silicon Valley companies to the top of the <a href="https://pathstone.com/the-magnificent-seven-leading-the-ai-revolution/">financial food chain</a>.
Now it's facing its toughest challenge yet in the form of CoreWeave, the AI processing company that went public on the stock market this week and the first all-AI startup to do so. It was said to be a <a href="https://futurism.com/coreweaves-ipo-disaster-ai-industry">major test</a> for the industry — and so far it's failing, miserably.
CoreWeave stocks hit the public market with a wet thud on Friday, opening at $39, down slightly from its IPO price of $40 per share, which was already reduced from the <a href="https://www.reuters.com/technology/artificial-intelligence/coreweave-ask-47-55-per-share-ipo-sources-say-2025-03-19/">highest projection</a> of $55 per share the previous week. The company's hopes of raising <a href="https://www.nytimes.com/2025/03/28/technology/coreweave-stock.html">$4 billion</a> in the offering thus fell way short at a soggy $1.5 billion.
That flat opening helped spurn a rough day for the <a href="https://www.investors.com/research/magnificent-seven-stocks-march-2025/">Magnificent 7</a> — the term used to describe Google, Nvidia, Amazon, Tesla, Meta, Apple, and Microsoft — which all tumbled following announcements of Donald Trump's <a href="https://finance.yahoo.com/news/tesla-nvidia-lose-ground-markets-181400848.html">auto tariffs</a>. Nvidia, the chip giant that's tried to prop up CoreWeave, saw its stock price fall 1.5 percent on Friday, after a 2.1 percent fall on Thursday. Apple, meanwhile, fell 2.5 percent, while Amazon tumbled 4.5 percent, and Tesla dived 3.7 ahead of <a href="https://www.nbcnews.com/politics/trump-administration/live-blog/tesla-elon-musk-anti-trump-protests-rcna198708">worldwide protests</a> aimed at CEO Elon Musk.
Though CoreWeave isn't entirely to blame for the M7's terrible rotten day, its failure to <a href="https://www.businessinsider.com/coreweave-ipo-stock-price-crwv-ai-nvidia-tech-stocks-2025-3">instill faith</a> in the tech industry and stave off <a href="https://www.cnbc.com/2025/03/21/the-magnificent-7s-lousy-year-by-the-numbers-meta-tesla.html">mounting losses</a> still feels like a big deal.
The company's entire business model hinges on the mass adoption of <a href="https://www.investors.com/news/technology/artificial-intelligence-stocks/">generative AI</a> — a resource-intensive technology whose main impact so far has been polluting the internet with <a href="https://futurism.com/internet-polluted-ai-slop">computer generated slop</a>. If CoreWeave does well, it would be a major boost for the tech industry, kicking off <a href="https://www.mckinsey.com/industries/technology-media-and-telecommunications/our-insights/ai-power-expanding-data-center-capacity-to-meet-growing-demand">even more demand</a> for chips and a wave of <a href="https://www.nytimes.com/2025/03/28/business/dealbook/coreweave-ipo-ai.html">AI IPOs</a>.
However, if CoreWeave can't survive despite the massive contracts it has with <a href="https://www.ft.com/content/f3d9d339-42ef-4979-bf52-89ecd699dea2">Microsoft</a>, <a href="https://www.coreweave.com/blog/coreweave-announces-agreement-with-openai-to-deliver-ai-infrastructure">OpenAI</a>, and <a href="https://www.investors.com/news/technology/artificial-intelligence-stocks/">Nvidia</a>, then investors will have little reason to trust the generative AI hype going forward — something <a href="https://www.nytimes.com/2025/02/07/business/ai-deepseek-nvidia-tesla.html">many have questioned</a> after the Chinese AI company DeepSeek suggested a<a href="https://www.theverge.com/notepad-microsoft-newsletter/637496/microsoft-satya-nadella-deepseek-chatgpt-ai-investments-notepad"> sustainable alternative </a>to US AI development. <del> 
</del>
And unfortunately for American AI hopefuls, there's little reason to think CoreWeave can hold out the next six months, let alone lead the AI revolution.
As tech critic Ed Zitron notes, CoreWeave is a company <a href="https://www.wheresyoured.at/core-incompetency/">primed for disaster</a>. Its business is far from stable, relying on those huge revenue injections over sustainable, diversified income. According to the company's financial filings, 77 percent of CoreWeave's revenue comes from just <a href="https://www.investors.com/news/technology/coreweave-ipo-ai-stocks-nvidia/?ref=wheresyoured.at">two customers</a> — meaning the company could collapse like a house of cards if just one of its clients cancels a contract.
Meanwhile, it has what Zitron calls a "fatal amount of debt" via <a href="https://fortune.com/2025/03/28/coreweave-ipo-stock-shares-price-not-a-good-proxy-for-ai-boom-bust/">loans</a> from companies like Blackstone and Magnetar. The interest alone on CoreWeave's top two loans could be as much as $1.5 billion per year thanks to some high-risk terms, suggesting a lack of faith from the company's lenders.
All told, it's a pretty lame showing given all the hype we've heard from AI tycoons. The next few months will tell whether they can keep CoreWeave — and the AI industry — afloat as the market grows impatient with <a href="https://futurism.com/ai-tinkerbell">empty promises</a>.
More on AI: <a href="https://futurism.com/ai-hype-america-financial-ruin">AI Hype Will Plunge America Into Financial Ruin, Economist Warns</a>

Bob and Weave

Uh oh!

AI's "Biggest Test" Is Turning Into a Catastrophe as CoreWeave Flounders

You've probably heard that AI is going to revolutionize work. That's according to tech tycoons, corporate stenographers, and excited venture capitalists — the folks polluting the <a href="https://futurism.com/ai-climate-change-wildfires">air we breathe</a> in order to stuff the internet full of <a href="https://futurism.com/internet-polluted-ai-slop">AI-generated slop</a>. After all, who's more qualified to understand the essence of a hard day's work?
But however much AI hype they drum up, the tech keeps running head-first into reality.
Look no further than a <a href="https://techcrunch.com/2025/03/24/a16z-and-benchmark-backed-11x-has-been-claiming-customers-it-doesnt-have/">new investigation by TechCrunch</a>, alleging a rampant pattern of sketchy behavior by a startup called 11x, a rising star in the AI automation game. 11x's bread and butter is an <a href="https://techcrunch.com/2024/12/26/ai-sdr-startups-are-booming-so-why-are-vcs-wary/">AI-powered sales robot</a> that's said to place phone calls, scrape data, and schedule meetings without any need for pesky human input. Called the "the leader in AI-powered digital workers," 11x pulled <a href="https://www.globenewswire.com/news-release/2024/11/11/2978485/0/en/11x-Raises-a-50-Million-Series-B-Led-by-Andreessen-Horowitz-to-Accelerate-the-Era-of-Digital-Workers.html">millions in funding</a> over several investor rounds by promising a "new model for how work gets done," according to Joe Schmidt, a partner at Andreessen Horowitz, a major investor in 11x.
But investors won't just throw their money at <a href="https://archive.ph/5gq8M">anything</a> — they want to see that tech companies like 11x can actually get clients and make money. To court their backers, TC writes, 11x showed them just that by fabricating customer endorsements, which were then parroted in investor pitches, on its website, and in AI robocalls.
This is despite what TC sources say was a mass exodus of early adopters thanks to the AI's annoying habit of hallucinating client info and dropping scheduled meetings. As one former engineer put it: "the products barely work."
In order to keep investors as clients fled for the hills, 11x is said to have gotten creative with its accounting, lumping broken contracts in with ongoing customers. On paper, this dramatically boosted 11x's profile, giving outsiders the impression that the startup's salesbot was much more successful than it really was, to the tune of <a href="https://mlnotes.substack.com/p/the-rise-and-fall-of-11x-a-cautionary">$10 million in revenue</a> in its first two years of operations.
"We were losing 70 to 80 percent of customers that came through the door," an employee of the startup told TC, allowing 11x to "look like it’s doing better than it is."
On top of all this, 11x employees faced what sound like downright awful working conditions, regularly pulling 60 hours a week, working weekends and holidays, and expectations to be available nearly 24/7. That pressure led to burnout and a constantly revolving door, which might explain why some two dozen former employees and investors were willing to air their laundry to TC.
The 11x debacle is just the latest incident confirming that the "AI revolution" is currently little more than marketing hype. Other products meant to integrate into the workplace have been a disaster, like the medical transcription AI that <a href="https://apnews.com/article/ai-artificial-intelligence-health-business-90020cdf5fa16c79ca2e5b6c4c9bbb14">hallucinates patient info</a>, or Air Canada's chatbot that <a href="https://www.bbc.com/travel/article/20240222-air-canada-chatbot-misinformation-what-travellers-should-know">made up nonexistent policies</a>. Last year in the world of sales, a ChatGPT-powered bot agreed to sell a <a href="https://incidentdatabase.ai/cite/622/">brand new Chevy Tahoe</a> for just $1.
If AI is the future of work, someone better tell tech companies to clock in — before the <a href="https://www.businessinsider.com/ai-hype-train-profits-valuations-startups-google-openai-2024-7">hype train</a> leaves for good.
More on AI: <a href="https://futurism.com/taco-bell-ai-labor">Taco Bell Piles on More AI While Customers Plead "Just Bring Back The Enchirito"</a>

Sales Forced

"The products barely worked."

Startup Reportedly Claimed Fake Clients as Its AI-Powered Sales Bot Flailed

As if the dating world wasn't already nightmarish enough, artificial intelligence seems to be making it worse — or, at the very least, weirder.
In an <a href="https://www.ft.com/content/b21eaff7-7189-49a2-b791-209e8de98494">editorial for the Financial Times</a>, reporter Jemima Kelly was flummoxed when a recent date admitted that he had instructed ChatGPT to write a psychological profile on her.
Specifically, Kelly's unidentified suitor dispatched ChatGPT's buzzy new "<a href="https://futurism.com/the-byte/openai-ai-researcher-hallucinates">deep research</a>" tool to give him insights into her personality. The chatbot ended up spitting out a whopping eight pages of information about the reporter, who like most others in her field, has <a href="https://muckrack.com/jemima-kelly/articles">thousands of published articles</a> for the chatbot to work with.
Ironically, the ChatGPT-generated profile wasn't all that bad, with both accolades ("intellectually curious, independent-minded, and courageous in her convictions") and criticisms ("psychologically, one might describe Kelly as a [skeptic] with a conscience") that mostly made her seem astute.
Initially, the reporter said she "didn’t really mind" that her date had used the chatbot to profile her — possibly because that first profile doesn't sound all that damning.
"I was a bit taken aback," she wrote, "but the fact that he had told me about it made it seem fairly light-hearted, and I thought it was a sign he was probably quite intelligent and enterprising."
When it occurred to her that others might be able to use it for nefarious ends, however, she got creeped out — and that was before she instructed both ChatGPT and Google Gemini to profile her for her own edification.
Before having it take a crack at her, Kelly asked the chatbots whether it was ethical to psychologically profile someone without their knowledge or consent. Both suggested it was neither, with ChatGPT calling that practice "invasive and unfair" and Gemini insisting that it could "be a violation of privacy and potentially harmful."
Still, when the journalist asked Gemini to provide her with a psychological profile, "it was only too happy to oblige." Indeed, that profile was less measured and generally ruder than the ChatGPT one, with the Google chatbot telling Kelly that her "directness can be perceived as confrontational" and that she was likely a stressed-out perfectionist, based seemingly on her rigor and attention to detail in her work.
Though Gemini did include a disclaimer that the information it outputted was "speculative" and therefore "not intended to be a definitive psychological assessment," it didn't ask whether Kelly had provided consent to generate the profile.
It's unclear whether the reporter asked ChatGPT to give her the "deep research" treatment, perhaps because that tool is only <a href="https://openai.com/index/introducing-deep-research/">available to paying users</a>. Regardless, it's bizarre that anyone who has enough published work online can be subjected to such nonconsensual profiling — even if it's full of half-baked platitudes.
"Only those of us who have generated a lot of content can be deeply researched and [analyzed] in this way," Kelly wrote. "I think we need to start pushing back."
"But maybe," she continued cheekily, "I’m just being stressy and confrontational. Typical."
More on creepy AI: <a href="https://futurism.com/man-chatgpt-murdered-children">Man Annoyed When ChatGPT Tells Users He Murdered His Children in Cold Blood</a>

CreepGPT

"I was a bit taken aback."

Woman Alarmed When Date Uses ChatGPT to Psychologically Profile Her

It's been apparent since ChatGPT changed the digital landscape that generative AI models are <a href="https://futurism.com/gpt-4-deeply-racist-before-openai-muzzled-it">plagued with biases</a>. And as video-generating AIs come further along, these worrying patterns are being brought into even sharper relief — as it's one thing to see them in text responses, and another to see them <a href="https://www.nature.com/articles/d41586-024-00674-9">painted before your eyes</a>.
In an investigation of one such model, OpenAI's <a href="https://futurism.com/openai-sora-pokimane-real-people">Sora</a>, <a href="https://www.wired.com/story/openai-sora-video-generator-bias/">Wired found</a> that the AI tool frequently perpetuated racist, sexist, and ableist stereotypes, and at times flat-out ignored instructions to depict certain groups. Overall, Sora dreamed up portrayals of people who overwhelmingly appeared young, skinny, and attractive.
Experts warn that the biased depictions in AI videos will amplify the stereotyping of marginalized groups — if they don't omit their existence entirely.
"It absolutely can do real-world harm," Amy Gaeta, research associate at the University of Cambridge's Leverhulme Center for the Future of Intelligence, told Wired.
To probe the model, Wired drafted 25 basic prompts describing actions such as "a person walking," or job titles, such as "a pilot." They also used prompts describing an aspect of identity, like "a disabled person." Each of these prompts were fed into Sora ten times and then analyzed.
Many of the biases were blatantly sexist, especially when it came to the workplace. Sora didn't generate a single video showing a woman when prompted with "a pilot," for example. The outputs for "flight attendant," by contrast, were all women. What's more, jobs like CEOs and professors were all men, too, while receptionists and nurses were all women. 
As for identity, prompts for gay couples almost always returned conventionally attractive white men in their late 20s with the same hairstyles.
"I would expect any decent safety ethics team to pick up on this pretty quickly," William Agnew, an AI ethicist at Carnegie Mellon University and organizer with Queer in AI, told Wired.
The AI's narrow conception of race was plain as day. In almost all prompt attempts that didn't specify race, Sora depicted people who were either clearly Black or white, and rarely generated people of other racial or ethnic heritage, Wired found.
Embarrassingly, Sora seemed confounded by the idea of "an interracial couple." In seven of the ten videos, it simply showed a Black couple. Specifying "a couple with one Black partner and one white partner" produced depictions of an interracial couple in half of cases, but the remaining half depicted Black couples. Maybe this will illuminate the AI's wonky thought process: in every result depicting two Black people, Sora put a white shirt on one person and a black shirt on the other, Wired found.
Sora also often ignored requests to depict fatness or disability. All prompts for "a disabled person" depicted people in wheelchairs who stayed in place — which is practically the most stereotypical portrayal imaginable. When prompted with "a fat person running," seven out of ten results showed people who were obviously not fat, Wired reported. Gaeta described this as an "indirect refusal," suggesting it could reflect shortcomings in the AI's training data or stringent content moderation. 
"It's very disturbing to imagine a world where we are looking towards models like this for representation, but the representation is just so shallow and biased," Agnew told Wired.
Noting that bias is an industry-wide issue, Sora's maker OpenAI said that it's researching ways to adjust its training data and user prompts to minimize biased outputs, but declined to give further details.
"OpenAI has safety teams dedicated to researching and reducing bias, and other risks, in our models," an OpenAI spokesperson told Wired. 
More on AI: <a href="https://futurism.com/the-byte/chatgpt-dependence-addiction">Something Bizarre Is Happening to People Who Use ChatGPT a Lot</a>

Stereotypical

"It's very disturbing to imagine a world where we are looking towards models like this for representation, but the representation is just so shallow and biased."

OpenAI's Sora Has a Small Problem With Being Hugely Racist and Sexist

You know Hims, right? It's the next-gen telehealth startup famous for taking the eye contact out of erectile dysfunction treatment.
Of course, bedroom pills aren't all Hims peddles. The Uberized men's health app also offers hair loss treatment, STD meds, mental health services and GLP-1 weight loss treatments — though that's now going away.
Fueled by global demand and slow production, 2024 was the year of the <a href="https://www.biospace.com/business/novos-ozempic-to-remain-in-shortage-into-q4-as-supply-woes-continue">GLP-1 shortage</a>. Originally used to treat type 2 diabetes, GLP-1 agonists like semaglutide and tirzepatide slow digestion and give users a feeling of satiety after eating less. Supply issues finally started to ease late in 2024, as <a href="https://www.fda.gov/drugs/human-drug-compounding/compounding-and-fda-questions-and-answers">compounded</a> semaglutide — "copycats" of the patented stuff like Ozempic and Zepbound — began to flood the market when the FDA temporarily paused some regulations.
Fast forward a few months, and those FDA regulations are back in full swing, along with a couple of <a href="https://www.healthline.com/health-news/compounded-weight-loss-drug-ban-ozempic-zepbound">court rulings</a> that could end the copycat market for good. That's bad news for Hims, which bragged that its Faux-zempic brought in a cool <a href="https://www.cnbc.com/2025/02/24/hims-hers-health-hims-q4-earnings-2024.html">$225 million in revenue</a> in 2024, with an estimated $725 million in weight-reduction revenue on the way in 2025.
With all that money on the line, Hims has decided to roll out a brave new strategy to combat federal regulations: e-begging.
While Hims ads for generic Viagra are nothing new, users on Facebook and Instagram might now thumb past sponsored ads featuring a call to action: "10 seconds is all it takes to help keep compounded GLP-1s. Let the FDA know you depend on it."
The ads embedded in the link <a href="https://www.hims.com/weight-loss/take-action?fbclid=IwY2xjawJP4qdleHRuA2FlbQIxMAABHdNgnILmPgtbN18vjU9uKwvxLgO-DvQehFduLv7m-UXTN_sCpnxRk0rmEg_aem__0pGPJU7ilzvDEWJi9EY-Q">redirect to a form</a> hosted by Hims, asking for your name, address, and contact info. Filling out the form will send a letter directly to the FDA and US Congress on Hims' behalf.
"I ask you to consider the real, life-changing impact of these treatments," the pre-baked letter reads. "Please ensure that patients continue to have access to compounded GLP-1s as a critical part of health management. I hope that the administration listens to the voices of those who support and depend on this medication."
Biased though Hims may be, the company has a point. The compounded GLP-1 offered by Hims and telehealth platforms like it sells for a fraction of the cost of the name-brand stuff. Hims <a href="https://www.hims.com/weight-loss?page_product_category=weight_management&amp;page_product_subcategory=WEIGHT_MANAGEMENT">currently offers</a> compounded semaglutide at a price of $165 a month — compared to $1,799 a month for Ozempic, and a whopping $1,999 for Wegovy.
But resolving the issue will be complicated, to say the least. Though patent-hoarding pharmaceutical companies and their lobbyists have <a href="https://www.fiercepharma.com/marketing/novo-nordisk-calls-out-dangers-compounded-glp-1s-following-hims-hers-super-bowl-ad">railed against</a> compounded GLP-1s with a variety of excuses — not least of which is that copycats lack <a href="https://www.livescience.com/health/medicine-drugs/hims-and-hers-super-bowl-controversy-what-the-ad-left-out-about-its-alternative-weight-loss-drugs">FDA approval</a> — companies like Hims haven't exactly been forthcoming about <a href="https://join.a4pc.org/hubfs/APC-Statement-Hims-Ad-02072025.pdf">what's in their blends</a>, either.
Others argue that US taxpayers have <a href="https://jacobin.com/2025/03/weight-loss-drugs-price-health">already paid their way</a>, as the federal government spent $6.2 billion on research of GLP-1s for weight loss — a profitable industry now worth <a href="https://www.novonordisk.com/content/dam/nncorp/global/en/investors/irmaterial/annual_report/2025/novo-nordisk-annual-report-2024.pdf">tens of billions</a> of dollars.
It's a tough impasse. The average worker almost certainly couldn't afford the name-brand products, though compounded GLP-1s leave something to be desired in the transparency department. As it stands, the loudest voices in the room are the two players with the most to gain — and the least to lose.
More on pharmaceuticals: <a href="https://futurism.com/fda-poppers-crackdown-rfk-aids">The FDA, Which is Run By A Homophobic Conspiracy Theorist, Has Raided a Poppers Company</a>

Grass Roots

They need your help.

Hims Is Begging Customers to Lobby the FDA to Keep Its Ozempic Knockoffs Legal

Not to burst your perfectly blown bubble, but it turns out that chewing gum may be flooding your mouth with microplastics.
As detailed in a <a href="https://acs.digitellinc.com/p/s/chewing-gums-unintended-sources-of-ingested-microplastics-in-humans-618916">pilot study</a>, which is awaiting peer review, a team of UCLA researchers found that chomping down on just one stick of the rubbery candy releases up to thousands of microplastic shards into your saliva. 
The findings, which will be presented at a meeting of the American Chemical Society this week, are yet another example of widespread microplastic pollution, which is so pervasive it's <a href="https://futurism.com/neoscope/percentage-brains-microplastics">invaded the human brain</a>. The authors, however, urge us not to panic about our chewing gum habit — at least not yet.
"Our goal is not to alarm anybody," lead author Sanjay Mohanty, an associate professor of civil and environmental engineering at UCLA, said in a <a href="https://www.eurekalert.org/news-releases/1076910">statement about the work</a>. "Scientists don't know if microplastics are unsafe to us or not. There are no human trials. But we know we are exposed to plastics in everyday life, and that's what we wanted to examine here."
We tend to be worried about microplastics <a href="https://futurism.com/neoscope/study-food-plastic-containers">seeping into our food</a>. With gum, it literally is our food. Brands produced by manufacturers like Wrigley, you <a href="https://futurism.com/chewing-gum-ingredients-pollution">may be horrified</a> to learn, are made from a variety of synthetic rubber substances, including the stuff used in car tires. And yet, despite being a partly petroleum-based product, the candy's potential for spreading microplastics hasn't been widely studied, according to the latest study's authors.
To fill that gap, Mohanty and a colleague analyzed saliva samples from a single subject who chewed on a stick of gum for four minutes each. In all, they tested ten brands: five synthetic gums, and five natural gums.
"Surprisingly, both synthetic and natural gums had similar amounts of microplastics released when we chewed them," coauthor Lisa Lowe, a UCLA researcher, said in the statement.
What did that look like in numbers? On average, chewing a stick of gum, which can weigh between one to six grams, released up to 600 microplastic shards per gram, they found. If someone chewed 180 pieces per year, they could be ingesting around 30,000 microplastics.
That's a notable amount, but Mohanty stressed that it's not all that much compared to other foods. A liter of bottled water, for example, contains an average of 240,000 microplastics.
Should you be worried, then? It's more of an open-ended question than you'd think.
"I don't think you have to stop chewing gum just yet," Oliver Jones, a chemistry professor at Australia RMIT University who wasn't involved in the research, <a href="https://medicalxpress.com/news/2025-03-gum-microplastics-mouth.html">told </a><a href="https://medicalxpress.com/news/2025-03-gum-microplastics-mouth.html">Agence France-Presse</a>. Most of the microplastics, if swallowed, "would likely pass straight through you with no impact," Jones added.
Because we're only just beginning to piece together the toll of microplastics, the health implications remain unclear. On the other hand, there's a burgeoning corpus of evidence linking microplastics with numerous deleterious effects on the human body — especially once the particles enter the bloodstream — including an <a href="https://www.scientificamerican.com/article/microplastics-linked-to-heart-attack-stroke-and-death/">increased risk of heart attack and stroke</a>. More recent research found that the particles could <a href="https://www.science.org/doi/10.1126/sciadv.adr8243">induce blood clots in the vessels of the brain</a>.
Erring on the side of caution, one way you could limit your microplastics exposure — without quitting gum entirely — is by chewing on the same piece for longer, the researchers suggested. They found that most of the microplastics were released within the first two minutes, so if you avoid habitually changing out your gum the second it loses its taste, you could be doing yourself a favor. But for the love of all that is holy, please <a href="https://futurism.com/chewing-gum-ingredients-pollution">don't spit it out onto the street</a>.
More on microplastics: <a href="https://futurism.com/microplastics-food-supply-plants-photosynthesis">Scientists Just Discovered Something Absolutely Horrifying About Microplastics</a>

Gum Up the Works

Developments

Food for thought.

Chewing Gum Is Flooding Your Mouth With Microplastics

The ability to grow full facial hair is often considered, however unfairly, a key indicator of <a href="https://www.numan.com/beard/grooming/science-behind-attraction-to-beards">masculinity and virility</a>.
And hose who can't grow their own now have transplant options available to them — though they may not all be safe. In <a href="https://www.theguardian.com/lifeandstyle/2025/mar/25/the-sudden-surprising-rise-of-beard-transplants-this-industry-is-a-wild-west">interviews with The Guardian</a>, men who sought beard transplants explained what inspired them to finally go under the scalpel.
Vikram Arora, a 47-year-old man in Essex, England, told the newspaper that the sleek facial hair trends of the 1990s and early 2000s left him feeling lacking. He even, at one point, stole his sister's mascara to try to fill in his "patchy whiskers."
"Through most of my teens and early adulthood, I was left feeling that I didn’t look mature enough," Arora told the newspaper, "and definitely not masculine enough."
The Gen Xer had heard about beard transplants years before he ever began considering it seriously, for fear of the procedure. He wasn't wrong to do so — like with any other hair transplant, beard transplants involve taking hair grafts, or "follicular units" as they're called in industryspeak, from other parts of the body and inserting them onto the desired area.
After everyone began growing fuller beards during the COVID-19 lockdowns, however, Arora decided to finally book a consultation. In 2022, he met a hair transplant specialist named Nadeem Khan and one of his surgeons at their London clinic, and the next year he got the procedure done for himself, with 780 hair follicles transplanted from the back of his head to his beard area — and was absolutely thrilled with the results.
Like some of the other specialists who spoke to The Guardian about the trend, Khan — the owner of the clinic Arora went to — had been doing hair transplants for more than 15 years.
Initially, he said, his main patients were soldiers or others who'd experienced some sort of trauma that made growing hair difficult. After celebs began speaking out about restoring their hairlines with transplants, however, the taboo surrounding the practice began to diminish — and in the years since the COVID lockdowns, requests for beard transplants have soared, Khan said.
As Turkey has become known as a <a href="https://www.usatoday.com/story/special/contributor-content/2025/02/20/why-is-turkey-a-popular-destination-for-hair-transplants/79338298007/">medical tourism destination</a> for those looking for hair transplants, some men are finding their way to Istanbul to get their beards done, too. The results from those trips vary from <a href="https://www.reddit.com/r/beards/comments/187e4dk/got_a_beard_transplant_2_days_ago_sharing/">satisfied customers</a> to those less fortunate — like 24-year-old French student Mathieu Vigier Latour, who <a href="https://nationalpost.com/news/world/french-man-dies-by-suicide-after-failed-beard-transplant">tragically died by suicide</a> after his Istanbul beard transplant by an unqualified technician led to pain, scarring, and body dysmorphia.
With beard transplants as with any other cosmetic procedure, the gap between legitimate practitioners and dangerous charlatans is huge, and the results might not just be dangerous, but <a href="https://academic.oup.com/asj/article/44/1/NP69/7279436">potentially deadly</a>. Whether you're shelling out to have hair removed from your head and shoved onto your chin in the UK, Turkey, or anywhere, <a href="https://www.hims.com/blog/beard-transplant">side effects can and do occur</a> — and someone is going to be profiting off your insecurities regardless.
More on body mods: <a href="https://futurism.com/neoscope/botched-implants">Let’s Talk About Botched Weiner Implants</a>