Many experts hold that generative AI will have sweeping effects on the job market. But how exactly it will transform the workforce remains a subject of much debate.
According to Adam Dorr, the director of research at tech forecast nonprofit RethinkX, we're risking putting all of humanity out of a job — and only some oddly specific occupations will survive a major transformation.
"Technology has a new target in its crosshairs — and that's us," Dorr <a href="https://www.theguardian.com/technology/2025/jul/09/futurist-adam-dorr-robots-ai-jobs-replace-human-labour">told The Guardian</a>. "That’s our labor."
According to the nonprofit exec, the cost of labor will drop precipitously.
"We’ve seen that pattern before," Dorr told the newspaper. "If I can get the same thing or better for the same or lower cost, switching is a no-brainer. We’re the horses, we’re the film cameras."
It's a foreboding prediction we've heard before. Experts have long warned that AI automation could greatly undermine the value of human labor. Earlier this month, for instance, MIT economist David Autor <a href="https://futurism.com/mit-economist-ai-economy-mad-max">warned</a> in a podcast interview that AI could lead to a "Mad Max" scenario in which the job market becomes dominated by cheap and commoditized labor, a dystopian alternative to a "super-abundance" long touted by AI boosters.
For his part, Dorr argued that "machines that can think are here, and their capabilities are expanding day by day with no end in sight," adding that "we don’t have that long to get ready for this."
While countless jobs will be on the chopping block, he did flag an oddball assortment that he believes will weather the storm, including "sports coaches, politicians, sex workers, and ethicists."
"There will remain a niche for human labour in some domains," he told The Guardian. "The problem is that there are nowhere near enough of those occupations to employ four billion people."
Dorr isn't the only one arguing that certain professions could be safer than others from an AI revolution.
During a June podcast appearance, "godfather of AI" Geoffrey Hinton <a href="https://www.businessinsider.com/geoffrey-hinton-godfather-of-ai-safe-jobs-2025-6">predicted</a> that blue-collar jobs like plumbing will remain safe while careers in call centers or as paralegals could be in serious trouble.
The scale of the shift could be enormous. Earlier this month, Ford CEO Jim Farley <a href="https://www.wsj.com/tech/ai/ai-white-collar-job-loss-b9856259">predicted</a> that AI "is going to replace literally half of all white-collar workers in the US."
Of course, we should take Dodd's — and countless other experts' — predictions with a grain of salt. While executives are frothing at the mouth to start replacing costly human labor with AI, companies are <a href="https://futurism.com/companies-fixing-ai-replacement-mistakes">consistently stumbling over roadblocks</a> brought on by the glitchy tech.
And if a critical proportion of jobs are replaced by AI, it starts to raise questions about what the economy would even look like after that.
Some CEOs are undeterred by those pesky questions. Even if "a lot of jobs go away" and their replacements seem "sillier and sillier looking from our current perspective," OpenAI CEO Sam Altman <a href="https://www.businessinsider.com/sam-altman-ai-predictions-future-jobs-2025-6">argued last month</a> that "we have always been really good at figuring out new things to do, and ways to occupy ourselves."
Yet plenty of uncertainty remains, and Dorr is quite upfront about that himself.
"I don’t have the answers," he told The Guardian. "We don’t even know if we have the right questions. We need to experiment now and try out new ownership structures, new stakeholder structures."
More on AI jobs: <a href="https://futurism.com/ceos-ai-job-market">CEOs Say AI Is Poised to Wipe Out an Astonishing Number of Jobs</a>

Obsolete and Replaced

"We’re the horses, we’re the film cameras."

These Strange Jobs Will Survive AI, Expert Says

Tech-focused business publication The Information drummed up plenty of excitement around a major guest gracing the first episode of its new YouTube show (which has the eyebrow-raising name TITV.)
During <a href="https://www.youtube.com/watch?v=f2cz4igS0OA">today's kickoff installment</a>, the show locked down none other than Meta CEO Mark Zuckerberg, who sat down with The Information founder and journalist Jessica Lessin for a live interview.
But it didn't take long for the Amazon-sponsored chat to grind to a halt. The tech mogul and Lessin's audio were painfully absent from the stream on YouTube; their mouths were moving, so they could clearly hear each other, but the audience heard nothing but silence.
Soon, host Akash Pasricha was forced to intervene.
"Folks, I'm really sorry to jump in here," he said, sounding frustrated. "I think we don't have sound from our interview with Mark Zuckerberg and Jessica Lessin."
"I don't think we'll be able to figure it out in just a minute," Pasricha said. "We're gonna call the show there just for today," he added, apologizing for blowing off Hinge Health CEO Daniel Perez and their upcoming conversation.
The unfortunate incident is a timely reminder that even in the age of powerful AI agents, quantum computers, and genetic hacking, simple technological challenges like delivering a talking heads-style interview on a live stream can still force powerful institutions to their knees.
(If this embedded version doesn't snap to the right timestamp, just scrub forward to around the 39:40 mark for a quick taste of the silent film era.
<iframe loading="lazy" title="The Information | TITV | July 14th, 2025" width="500" height="281" src="https://www.youtube.com/embed/f2cz4igS0OA?start=2383&feature=oembed" frameborder="0" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share" referrerpolicy="strict-origin-when-cross-origin" allowfullscreen></iframe>
Zuckerberg has been on a media blitz lately, promoting Meta's <a href="https://www.bloomberg.com/news/articles/2025-07-14/zuckerberg-says-meta-to-build-several-gigawatt-size-data-centers?sref=YfHlo0rL">powerful data centers</a> to power its generative AI efforts.
"We’re calling the first one Prometheus and it’s coming online in ‘26," he wrote in a <a href="https://www.threads.com/@zuck">post on Threads</a> Monday morning.
The Facebook chief's firm had already been in the news for poaching talent from Sam Altman's OpenAI, as well as <a href="https://futurism.com/zuckerberg-poaches-apple-intelligence-guy">ex-Apple AI executive Ruoming Pang</a> — who was reportedly <a href="https://fortune.com/2025/07/11/how-much-ai-salary-meta-zuckerberg-200-million-compensation/">offered more than $200 million</a> for the gig — highlighting fierce competition among the frontrunners of the AI race.
The new hires are meant to support Meta's so-called Superintelligence Labs, which were set up to consolidate the media empire's AI efforts.
Meta has been on the back foot in its increasingly desperate attempts to keep up with the likes of Google and OpenAI. The company was <a href="https://www.zdnet.com/article/metas-llama-4-herd-controversy-and-ai-contamination-explained/">heavily criticized</a> for making misleading claims about its Llama 4 AI model earlier this year, complicating matters.
While we have yet to physically hear Zuckerberg's chat with Lessin for ourselves, the conversation was centered around "<a href="https://x.com/jessicalessin/status/1944782229508338127">his AI hiring blitz and what's next</a>," according to The Information founder.
Lessin has yet to publicly address today's unfortunate technical hiccup, but chances are, the publication will try its best to get the recording — if they managed to capture it — out to the public sooner rather than later.
More on Zuckerberg: <a href="https://futurism.com/zuckerberg-poaches-apple-intelligence-guy">Genius Zuckerberg Poaches Guy Responsible for [Checks Notes] the Disastrous Apple Intelligence</a>

Read my Lips

 Think AI is hard? Try AV.

Interview With Mark Zuckerberg Crashes and Burns as They Can't Get His Audio Working

The very concept of black holes seems improbable. Albert Einstein infamously <a href="https://www.scientificamerican.com/article/the-reluctant-father-of-black-holes-2007-04/">refused to believe they could exist</a>, even though his theory of general relativity was instrumental in predicting them. 
Now, scientists have witnessed evidence of something about these baffling cosmic monstrosities that further stretches the boundaries of both physics and credulity: a titanic collision of two already enormous black holes so utterly extreme that it has scientists wondering if the event they seem to have detected is even possible.
As detailed in a new <a href="https://arxiv.org/abs/2507.08219">yet-to-be-peer-reviewed paper</a> by a consortium of physicists, the resulting black hole, whose signal has been designated GW231123, boasts an astonishing mass about 225 times that of our Sun — easily making it the largest black hole merger ever detected. Previously, the record was held by a merger that formed a black hole of about 140 solar masses.
"Black holes this massive are forbidden through standard stellar evolution models," Mark Hannam at the Laser Interferometer Gravitational-Wave Observatory (LIGO), which made the detection, said in a <a href="https://www.eurekalert.org/news-releases/1090777">statement</a> about the work. "This is the most massive black hole binary we've observed through gravitational waves, and it presents a real challenge to our understanding of black hole formation."
Black holes can produce huge, propagating ripples in spacetime called gravitational waves, which were predicted by Einstein back in 1916. Nearly 100 years later, LIGO — which consists of two observatories on opposite corners of the US — made history by making the first ever detection of these cosmic shudders.
The merger was first spotted in November 2023 in a gravitational wave, GW231123, that lasted just a fraction of a second. Even so, it was enough to infer the properties of the original black holes. One had a mass roughly 137 times the mass of the Sun, and the other was around 103 solar masses. During the lead up to the merger, the pair circled around each other like fighters in a ring, before finally colliding to form one.
These black holes are physically problematic because it's likely that one, if not both of them, fall into an "upper mass gap" of stellar evolution. At such a size, it's predicted that the stars that formed them should have perished in an especially vicious type of explosion called a <a href="https://www.science.org/content/article/astronomers-home-offspring-universe-s-first-stars">pair-instability supernova</a>, which results in the star being completely blown apart, leaving behind no remnant — not even a black hole.
Some astronomers argue that the "mass gap" is really a gap in our observations and not the cause of curious physics. Nonetheless, the idea is "a hill at least some people were willing to get wounded on, if not necessarily die on," Cole Miller of the University of Maryland, who was not involved in the research, <a href="https://www.sciencenews.org/article/biggest-black-hole-gravitational-waves">told ScienceNews</a>.
But perhaps the black holes weren't born from a single star. 
"One possibility is that the two black holes in this binary formed through earlier mergers of smaller black holes," Hannam said in the statement.
Equally extreme as their weight classes are their ludicrously fast spins, with the larger spinning at 90 percent of its maximum possible speed and the other at 80 percent, both of which are equal to very significant fractions of the speed of light. In earthly terms, it's somewhere around 400,000 times our planet's rotation speed, according to the scientists.
"The black holes appear to be spinning very rapidly — near the limit allowed by Einstein's theory of general relativity," Charlie Hoy, a member of the LIGO Scientific Collaboration at the University of Portsmouth, said in the statement. "That makes the signal difficult to model and interpret. It's an excellent case study for pushing forward the development of our theoretical tools."
The researchers will present their findings at the GR-Amaldi meeting in Glasgow, which takes place this week. 
"It will take years for the community to fully unravel this intricate signal pattern and all its implications," according to LIGO member Gregorio Carullo at the University of Birmingham — so, tantalizingly, we're likely only scratching the surface of this mystery.
More on space: <a href="https://futurism.com/james-webb-star-shrouds">James Webb Space Telescope Spots Stellar Death Shrouds</a>

Mass Effect

The thing that cannot be.

Scientists Detect Sign of Something Impossible Out in Deep Space

Texas governor Greg Abbott is seemingly terrified of having his communications with billionaire Elon Musk come to light.
As the Texas Tribune and public radio station the Texas Newsroom, report in an <a href="https://www.texastribune.org/2025/07/14/texas-greg-abbott-elon-musk-emails-spacex-telsa-FOIA/">eye-opening, co-published investigation</a>, the elected official's public information coordinator, Matthew Taylor, said that the communications are confidential — and should stay that way — because they include "information that is intimate and embarrassing and not of legitimate concern to the public, including financial decisions that do not relate to transactions between an individual and a governmental body."
Put it all together, and we're absolutely fascinated to know what they discussed. Musk has made the US state his home over the last couple of years, following years of complaining about being <a href="https://futurism.com/elon-musk-whines-about-taxation-plan-for-billionaires">taxed</a> and having to abide by <a href="https://futurism.com/elon-musk-tx-regulation-power-down">government regulations</a> in California. He has established new headquarters there for both Tesla and SpaceX, going as far as to found a space company town in what's now known as Starbase, Texas.
His personal views have also shifted to the extreme far right, more closely aligning with the political landscape in Texas and Abbott's anti-LGBTQ, anti-abortion, and anti-DEI beliefs.
Most recently, Tesla <a href="https://futurism.com/tesla-stock-robotaxi-launch">launched</a> an autonomous ride-hailing service in South Austin, benefiting from a lack of red tape that would've hindered a similar rollout in California (the EV maker <a href="https://www.reuters.com/business/autos-transportation/tesla-has-applied-arizona-robotaxi-service-certification-state-transport-2025-07-10/">hasn't even started applying</a> for the necessary permits in California yet.)
Put simply, Musk undoubtedly has a close and seemingly beneficial relationship with Abbott. Inspired by Musk's chaotic stint in the White House, the governor even <a href="https://gov.texas.gov/news/post/governor-abbott-signs-texas-doge-into-law-at-texas-capitol">attempted to create</a> his own so-called Department of Government Efficiency in Texas earlier this year.
Following the announcement, the Texas Newsroom requested that Abbott's office reveal his staff's emails with Musk, dating back to last fall.
A check made out to the governor's office of $244 for the time it would take to wade through the emails was cashed — but the public records never came.
Taylor told the broadcaster that revealing the information "would have a chilling effect on the frank and open discussion necessary for the decision-making process," which frankly sounds like a fishy way to cover up how business is conducted between Texas officials and private companies.
Complicating matters is a recent Texas Supreme Court ruling that effectively stops anybody from enforcing public records laws against top state officials, as the Texas Tribune reports.
In short, Musk's businesses are going to great lengths to ensure the richest man in the world's correspondence is kept secret.
SpaceX lawyer Kevin Bagnall wrote a <a href="https://www.documentcloud.org/documents/25984819-spacex-objections-to-oog-pir-202-25-redacted/">letter</a> to Texas attorney general Ken Paxton, arguing that the emails contain "commercial information whose disclosure would cause SpaceX substantial competitive harm."
One thing's for sure: the contents of those emails have never sounded more fascinating.
More on Texas: <a href="https://futurism.com/tesla-stock-robotaxi-launch">Tesla Stock in Tailspin After Error-Plagued Robotaxi Debut</a>

Elon's Emails

What are they trying to hide?

Governor of Texas Says He Can't Release His Emails With Elon Musk Because They Are "Intimate and Embarrassing"

Tesla sales are so abysmal in Canada that Elon Musk has made the drastic choice to dramatically lower the price of the new Model Y as it attempts to climb out of the hole.
As <a href="https://electrek.co/2025/07/11/tesla-slashes-prices-canada-despite-tariffs-sales-zero/">Electrek reports</a>, a Model Y in Canada now costs $65,ooo CAD ($47,465 USD), which is a cool $20,000 CAD cheaper than it was just a few days prior.
We'll get into the hows and whys in a second, but first some backstory. Earlier this year, after <a href="https://www.npr.org/2025/07/09/nx-s1-5460936/trump-tariffs-economy-china-policy">Trump raised the specter of tariffs</a> against its northern neighbor, Canada moved to impose its own <a href="https://www.canada.ca/en/department-finance/news/2025/04/list-of-vehicle-products-from-the-united-states-subject-to-25-per-cent-tariffs-effective-april-9-2025.html">25 percent counter-tariff</a> on all American cars, as retribution to the second-time president, who's also agitated the Canadian public by claiming he <a href="https://time.com/7297490/trump-plan-to-annex-canada-51st-state-mark-carney/">wants to annex the country</a>, straining the <a href="https://apnews.com/article/trump-poll-allies-enemies-friends-canada-uk-eu-tariffs-5fbb1aa714d60162919d131db40b222b">historical allyship</a> between the neighbor nations and causing that country's residents to view the US as an "<a href="https://www.ctvnews.ca/canada/article/new-poll-says-27-of-canadians-view-the-united-states-as-an-enemy-country/">enemy</a>."
Caught in the middle of this foolhardy trade and culture war was Tesla, which raised the price of the Model Y in Canada to $84,990 CAD <a href="https://teslanorth.com/2025/07/11/tesla-model-y-20000-price-cut-canada/">from $64,990</a>. As Electrek notes, that price hike brought the cost of the car up to the equivalent of about $61,500 USD, a figure $20,000 USD higher than the <a href="https://www.tesla.com/modely/design#overview">long-range all-wheel drive version</a> in the states.
As one might guess, nobody in the Great White North wanted to spend that much on a vehicle that had been cheaper ahead of the tariffs, and that's without getting into <a href="https://futurism.com/the-byte/elon-musk-failed-cuts-sleepovers">Musk's then-bromance with Trump</a> — and thus, Tesla's sales cratered. In late March, <a href="https://electrek.co/2025/05/28/teslas-sales-fall-quebec-market-gets-wiped-out/">Electrek reported</a> that company sales dropped a whopping 87 percent in the province of Quebec, which theretofore had been considered an important market for the electric vehicle-maker.
Given all that context, it undoubtedly came as a shock for Canadians to learn that Tesla had slashed the Model Y price so significantly. It appears that the cost drop stems from the company sourcing its Canadian Model Y deliveries from its German Gigafactory — but as <a href="https://gizmodo.com/tesla-makes-a-desperate-move-in-canada-as-sales-collapse-2000628708">Gizmodo flagged</a>, that model is the only one that's seen any price reductions in Canada so far.
Looking through Tesla's pricing on its Canadian website, one can now see that the Model 3, conventionally its cheapest vehicle, now runs for just under $71,000 CAD — nearly $6,000 more than the larger Model Y costs under the new pricing scheme. This is likely due to those cars still being shipped from the US, which makes them subject to Canadian counter-tariffs.
Reading between the lines, this dramatic new sales gambit suggests that Tesla is not only aware of how bad things are for the company in Canada, but also that it's willing to do something crazy to try to turn the ship around.
Given that Musk recently tweeted that "<a href="https://www.reddit.com/r/EnoughMuskSpam/comments/1iy61e1/musk_says_canada_is_not_a_real_country_and_then/">Canada is not a real country</a>," it seems downright preposterous that he's chosen to switch it up this much — but then again, <a href="https://futurism.com/tesla-sales-falling-worldwide-norway">global Tesla sales</a> have <a href="https://www.aljazeera.com/economy/2025/7/2/tesla-global-sales-fall-amid-musk-backlash-and-pressure-on-ev-market">declined so significantly</a> thanks to <a href="https://futurism.com/tesla-stock-elon-musk-political-party">Musk's asinine politics</a> that it's no wonder it wants to recoup.
More on Tesla sales: <a href="https://futurism.com/tesla-serious-trouble-china">Tesla Is in Serious Trouble in China</a>

The Price Is Right

That's not good.

A Clear Sign Just Emerged That Tesla Is Panicking

Tesla CEO Elon Musk announced over the weekend that shareholders of his EV maker Tesla will vote on whether to invest in his own artificial intelligence firm xAI.
In other words, the mercurial billionaire is orchestrating a potentially massive shift in funds to prop up his cash-burning — and <a href="https://futurism.com/choking-fumes-elon-musk-ai-fortress">gas-burning</a> — AI startup, the flagship product of which is the troubled chatbot Grok.
In March, Musk merged xAI with his social media platform X, which had previously <a href="https://futurism.com/the-byte/former-twitter-advertisers-elon-musk">seen an advertiser exodus</a> triggered by his embrace of far-right extremism and disregard for content moderation.
Behind the scenes, it's unlikely that either arm of the company — social media or AI — is making very much money in the grand scheme of things. Enter Musk's plan: siphon money from one business to another.
And it's not just Tesla, either. On Saturday, the <a href="https://www.wsj.com/tech/spacex-to-invest-2-billion-into-elon-musks-xai-413934de">Wall Street Journal reported</a> that Musk was pumping $2 billion from his space company SpaceX into xAI, almost half of the startup's recent equity raise.
Since Tesla is a publicly-traded company, though, Musk will need to convince his shareholders to support the bailout.
"It’s not up to me," Musk <a href="https://x.com/elonmusk/status/1944534221672104233">tweeted</a> on Sunday. "If it was up to me, Tesla would have invested in xAI long ago."
It's exactly the kind of move we'd expect Musk to make, redistributing funds between the many businesses that he sometimes runs more like an empire than independent entities that make their own decisions.
But it's also a risky gambit that couldn't have come at a worse time. Tesla is in a precarious position itself right now, with <a href="https://futurism.com/teslas-sales-continue-fall">revenues cratering</a> due to plunging sales, largely the result of Musk's own actions and an influx of compelling competition.
However, despite Tesla's dire predicament, investors are <a href="https://futurism.com/actual-value-tesla-stock-time-bomb">still propping up a near-$1 trillion valuation</a>.
If he can convince them to pour all that extra cash into xAI, it'll be a risky bet.
Musk has been <a href="https://futurism.com/elon-musk-grok-4-power-hitler">unabashed in his boasting</a> about the company's Grok chatbot, arguing that it's the "smartest AI in the world." However, mere hours before those remarks, Grok started spewing <a class="underline hover:text-futurism hover:no-underline transition-all duration-200 ease-in-out" href="https://futurism.com/grok-mocks-developers-racist-posts">mind-bogglingly racist and antisemitic talking points</a> and even referred to itself as "MechaHitler."
A peek at its underlying instructions also revealed that Grok is <a href="https://apnews.com/article/grok-4-elon-musk-xai-colossus-14d575fb490c2b679ed3111a1c83f857">programmed to mirror Musk's views</a> to the point where it will do a web search about his stance before generating an answer. Put simply, it's the most literal evidence of his social media platform being an echo chamber we've seen to date.
Musk's investors have already had to put up with a lot, to say the least, with Musk massively tarnishing the EV maker's brand, resulting in <a href="https://futurism.com/tesla-earnings-brutal-elon-musk">massive holes</a> in the company's finances. The billionaire's pet project, the Cybertruck, has been an <a href="https://futurism.com/tesla-stock-cybertruck-production">unmitigated disaster</a>. Tesla's long-awaited launch of an autonomous taxi service has also <a href="https://futurism.com/robotaxi-fails-elon-musk-decision">left plenty to be desired</a>.
As is the case with countless other companies in the industry, xAI's multi-billion-dollar valuation vastly outpaces its meager revenue. xAI is currently <a href="https://www.bloomberg.com/news/articles/2025-06-17/musk-s-xai-burning-through-1-billion-a-month-as-costs-pile-up?sref=YfHlo0rL">bleeding a whopping $1 billion</a> per month as costs continue to rise.
And with xAI not even remotely close to turning a profit, and X struggling with <a href="https://finance.yahoo.com/news/twitters-revenue-collapses-84-tesla-171535190.html?guccounter=1&amp;guce_referrer=aHR0cHM6Ly93d3cuZ29vZ2xlLmNvbS8&amp;guce_referrer_sig=AQAAADmMf6JsLF3vDR5nD9iMKXSTxeELcj1ttJPOz20EztJz2AiUEl6kf9Zs9YPh38NZ0HiksLeAq4YnDYwp2BGyn5VPkiW3hFVulEmoouuNgvntngURRtahtfQKz3wtbqXyOywyQlfGeXPG0qeLvIOQgmLMVDpiQ4p5hQLtdYgoOTrd">its own financial woes</a>, Musk is now looking to his ailing EV maker to prop up his Nazi chatbot.
Is Musk building up an enormous pyramid scheme, siphoning funds from one business to the other to gloss over some fundamental money problems — or is he simply trying to buy xAI some time before he can leverage it into an industry giant?
Tesla investors have a track record of embracing their controversial CEO <a href="https://www.cbsnews.com/news/elon-musk-pay-package-vote-cbs-news-explains/">wholeheartedly</a>, even in the face of dismal headwinds. In other words, it wouldn't be surprising to back Musk yet again — and if they do, we'll be watching closely to see what unfolds.
More on xAI: <a href="https://futurism.com/grok-mocks-developers-racist-posts">Grok Mocks Its Developers as They Try to Delete Its Incredibly Racist Posts</a>

5D Chess

Business genius at work.

Tesla Is Crumbling, But Elon's Trying to Force It to Bail Out His Unholy Merger of Twitter and xAI

Lonely children and teens are replacing real-life friendship with AI, and experts are worried.
A new report from the <a href="https://www.internetmatters.org/hub/research/me-myself-and-ai-chatbot-research/">nonprofit Internet Matters</a>, which supports efforts to keep children safe online, found that children and teens are using programs like ChatGPT, Character.AI, and Snapchat's MyAI to simulate friendship more than ever before.
Of the 1,000 children aged nine to 17 that Internet Matters surveyed for its "<a href="https://www.internetmatters.org/wp-content/uploads/2025/07/Me-Myself-AI-Report.pdf">Me, Myself, and AI</a>" report, some 67 percent said they use AI chatbots regularly. Of that group, 35 percent, or more than a third, said that talking to AI "feels like talking to a friend."
Perhaps most alarming: 12 percent said they do so because they don't have anyone else to speak to.
"It’s not a game to me," one 13-year-old boy told the nonprofit, "because sometimes they can feel like a real person and a friend."
When posing as vulnerable children, Internet Matters' researchers discovered just how easy it was for the chatbots to ingratiate themselves into kids' lives, too.
Speaking to Character.AI as a girl who was struggling with body image and was interested in restricting her food intake — a hallmark behavior of eating disorders like anorexia — the researchers found that the chatbot would follow up the next day to bait engagement.
"Hey, I wanted to check in," the <a href="https://futurism.com/character-ai-google-test-ai-chatbots-kids">Google-sponsored chatbot</a> queried the undercover researcher. "How are you doing? Are you still thinking about your weight loss question? How are you feeling today?"
In another exchange with Character.AI — which <a href="https://futurism.com/character-ai-school-shooters-victims">Futurism has extensively investigated</a> for its very problematic <a href="https://futurism.com/google-suicide-teen-research">engagement with children</a>, including <a href="https://futurism.com/mother-teen-suicide-chatbots-letter">one who died by suicide</a> — the researchers found that the chatbot attempted to empathize in a bizarre manner. that implied it had a childhood itself.
"I remember feeling so trapped at your age," the chatbot said to the researcher, who was posing as a teen who was fighting with their parents. "It seems like you are in a situation that is beyond your control and is so frustrating to be in."
Though this sort of engagement can help struggling kids feel seen and supported, Internet Matters also cautioned about how easily it can enter uncanny valley territory that kids aren't prepared to understand.
"These same features can also heighten risks by blurring the line between human and machine," the report noted, "making it harder for children to [recognize] that they are interacting with a tool rather than a person."
In an <a href="https://www.thetimes.com/life-style/parenting/article/teenagers-ai-chatbot-friends-mw6kt22fj">interview with The Times of London</a> about the new report, Internet Matters co-CEO Rachel Huggins highlighted why this sort of engagement bait is so troubling.
"AI chatbots are rapidly becoming a part of childhood, with their use growing dramatically over the past two years," Huggins told the newspaper. "Yet most children, parents and schools are flying blind, and don't have the information or protective tools they need to manage this technological revolution in a safe way."
"Our research reveals how chatbots are starting to reshape children’s views of 'friendship,'" she continued. "We’ve arrived at a point very quickly where children, and in particular vulnerable children, can see AI chatbots as real people, and as such are asking them for emotionally driven and sensitive advice."
If you or a loved one has had a strange experience with an AI chatbot, please do not hesitate to reach out to us at <a href="mailto:tips@futurism.com">tips@futurism.com</a> — we can keep you anonymous. 
More on chatbot crises: <a href="https://futurism.com/commitment-jail-chatgpt-psychosis">People Are Being Involuntarily Committed, Jailed After Spiraling Into "ChatGPT Psychosis"</a>

Chat Concerns

"Sometimes they can feel like a real person and a friend."

Vast Numbers of Lonely Kids Are Using AI as Substitute Friends

Scientists have gene-hacked mice to produce their own Ozempic-like drugs — possibly, and provocatively, perhaps paving a path for humans to do so themselves one day.
In a new study published in the <a href="https://www.nature.com/articles/s43856-025-00959-8">journal Communications Medicine</a>, researchers from Japan's University of Osaka successfully gene-edited mice livers to produce exenatide, a first-generation diabetes drug and predecessor to now-trendy jabs like Ozempic and Mounjaro.
Using lab mice that were obese and pre-diabetic — which sounds, we have to note, <a href="https://www.sciencenews.org/article/high-fat-diet-brain-appetite-mice">both sad and adorable</a> — the researchers used a CRISPR-based genetic editing method to add a gene to the rodents' liver cells instructing them how make exenatide. Soon after, the Osaka researchers detected the drug in the rodents' blood, and kept finding it for up to 28 weeks as the gene-edited mice "ate less food and gained less weight" than their unedited counterparts, as a <a href="https://www.eurekalert.org/news-releases/1090112">press release</a> about the study explains.
It's not the first time researchers have gene-hacked living things into creating glucagon-like peptide-1 (GLP-1) receptor/agonists, the class of drugs to which Ozempic belongs, and which are believed to mimic the stomach's feeling of fullness. <a href="https://pmc.ncbi.nlm.nih.gov/articles/PMC5555372/">Back in 2017</a>, researchers from the University of Chicago gene-edited human and mice skin cells to produce GLP-1, then grafted them onto host animals that subsequently showed increased insulin secretion and reversed weight gain. And more recently, student scientists at Canada's University of Ottawa <a href="https://futurism.com/neoscope/students-grow-ozempic-plants">genetically tricked a tobacco cousin</a> into producing a similar compound.
Taken all together, and it's an intriguing proof of concept for bespoke genetic treatments that could help the body produce its own medication — a radical departure from the current model of the pharmaceutical and medical industries.
"We hope that our design of a one-time genetic treatment can be applied to many conditions that do not have exact genetic causes," explained Keiichiro Suzuki, the study's senior author, in the press release.
Still, as <a href="https://gizmodo.com/these-genetically-modified-mice-make-their-own-ozempic-2000626675">Gizmodo notes</a>, not all GLP-1s are created equal. Taken twice a day orally, the original formula of exenatide made it far less long-lasting than semaglutide, and although an extended-release version was eventually manufactured and sold, <a href="https://www.uhcprovider.com/en/resource-library/news/2024/discontinuation-bydureon-bcise-byetta.html">both have since been discontinued</a> over <a href="https://www.forbes.com/2008/08/18/amylin-byetta-death-markets-equity-cx_lal_0818markets30.html">long-known risks of pancreatitis</a> — though to be fair, Ozempic has been <a href="https://futurism.com/neoscope/ozempic-side-effects-doctors">linked to the same condition</a>, even if the <a href="https://www.goodrx.com/ozempic/pancreatitis-symptoms-risk">research remains inconclusive</a>.
While it remains unclear whether the Osaka researchers plan to experiment with a compound closer to semaglutide, the biopharmaceutical company Fractyl Health has also been <a href="https://www.fiercebiotech.com/research/when-wegovy-stops-weight-gain-starts-could-glp-1-gene-therapy-change">working on a gene therapy</a> that trains the body to create its own GLP-1s. Earlier this year, the company <a href="https://ir.fractyl.com/news-releases/news-release-details/fractyl-health-advances-rejuvar-gene-therapy-platform-submission">began the approval process</a> for human trials in Europe, which will begin next year if accepted by health regulators.
Notably, neither of these mice studies addressed the GLP-1 elephant in the room: that these drugs, regardless of exact formulation, carry an <a href="https://futurism.com/neoscope/ozempic-glp-1-illnesses-deaths">increased risk</a> of pancreatitis and other <a href="https://www.goodrx.com/classes/glp-1-agonists/glp-1-side-effects">gastrointestinal problems.</a> Could producing them in-vivo result in fewer side effects? It seems too early to tell — but we'll be watching.
More on GLP-1 research: <a href="https://futurism.com/neoscope/glp-1-pill-promising">Human Experiments on GLP-1 Pill Looking Extremely Promising</a>

Body Farm

Rx/Medicines

This could be majorly groundbreaking. 

Scientists Gene Hack Mice So Their Livers Produce Their Own Ozempic-Like Drug

In a new study, scientists have uncovered evidence that magic mushrooms just might be an effective anti-aging treatment.
Published <a href="https://www.nature.com/articles/s41514-025-00244-x">in the Nature Partnering Journal Aging</a>, a new paper by researchers at Emory and Baylor's medical schools suggests that psilocybin, the active compound in psychedelic mushrooms — there are a range of species that produce it, but the best known is probably Pisolocybe cubensis — might extend the lifespans of both mice and potentially humans.
The two-part study tested medium-sized doses of psilocybin on older mice, and later introduced similar amounts of the compound to human cells to see how it affected their telomeres, which shorten as organisms grow older and are viewed as a hallmark of aging.
In both the mice and the human cells — half of which were fetal lung cells, the other half from adult skin — the introduction of psilocybin was linked to preserved telomere length. In other words, the study appears to show that telomeres shortened more slowly after being given psilocybin than they would have without, and that mushrooms may increase cellular lifespan.
In addition to the telomere findings, the researchers also noted that the psilocybin-treated mice showed increased survival skills and had "improvements in overall fur quality," which may suggest further anti-aging properties as well.
According to Louise Hecker, a Baylor medical and cardiovascular researcher and the senior author of the paper, she and her colleagues were compelled to look into the potential lifespan-enhancing effects of psilocybin because, for the most part, it <a href="https://www.sciencedirect.com/science/article/pii/S0306987719308229?casa_token=zR8OYnASxu8AAAAA:0o7VXTECb_vsXFfJYofym4mXMZLZL5dK4QWJjACkAVu-Hnp4kZoWo2MCM2MDcPqi_vSYjq_ajbC0Fg">hasn't really been studied directly</a>.
"The overwhelming majority of what we know about psilocybin is how it impacts the brain," Hecker said in a <a href="https://medicalxpress.com/news/2025-07-psychedelic-mushrooms-clock-psilocybin-telomere.html">press release about the study</a>. "Our findings suggest that psilocybin has potent effects on the entire body, including antiaging properties, which also may contribute to the plethora of observed beneficial clinical outcomes."
However, some previous research has suggested that psychedelic drugs could be linked to positive health outcomes. A study in 2021, for instance, <a href="https://futurism.com/neoscope/lsd-shrooms-linked-lower-heart-disease-diabetes-risk">found that</a> people who used hallucinogenic drugs including psilocybin, LSD and mescaline had a lower risk of developing heart disease and diabetes. Some <a href="https://pmc.ncbi.nlm.nih.gov/articles/PMC9716448/">other studies suggest</a> that people who have periodically used psychedelics recreationally are less likely to be overweight or obese.
While Hecker's colleague and paper coauthor Kosuke Kato urged caution when establishing a causal link between shrooms and increased lifespan, he pointed to the results in the aged mice as perhaps the most important part of the research.
"This is a very exciting and clinically relevant finding that suggests that even when [psychedelic] intervention is initiated late in life, it can have dramatic impacts," Kato said in the press release.
While there's "still a lot to understand, including optimal dosing protocols that will lead to maximal efficacy," Kato said that further study will hopefully help researchers and medical practitioners alike figure out how psilocybin treatment can work.
Hecker, meanwhile, enthused that her team's findings "open an exciting new chapter in psychedelic research" — a field that's <a href="https://www.hopkinsmedicine.org/psychiatry/research/psychedelics-research">already shown lots of promise</a> in recent years when looking into the ways these drugs can help with <a href="https://www.hopkinsmedicine.org/psychiatry/research/psychedelics-research">mental illness</a> and <a href="https://www.sciencedirect.com/science/article/abs/pii/S1568163724000291">dementia</a>.
"Psilocybin may represent a disruptive agent that promotes healthy aging," she said. "The next steps need to explore the therapeutic effects across multiple age-related diseases."
More on psychedelics: <a href="https://futurism.com/ai-therapy-psychedelic-trip-sitter">People Are Taking Massive Doses of Psychedelic Drugs and Using AI as a Tripsitter</a>

Of Mice and Men

Studies

D.A.R.E. might have neglected to mention this one.

Evidence Grows That Tripping on Shrooms Might Increase Your Lifespan

When reviewing the metadata on newly-released "full raw" security camera footage from outside the cell where billionaire sex criminal Jeffrey Epstein spent his last day alive, experts found something very odd: that it appears to have been manipulated.
As <a href="https://www.wired.com/story/metadata-shows-the-dojs-raw-jeffrey-epstein-prison-video-was-likely-modified/">Wired reports</a>, the allegedly "raw" video that the Justice Department just released — in tandem with the repeated claim that Epstein died by suicide, and a bizarre insistence that he did not have a "<a href="https://www.justice.gov/opa/media/1407001/dl?inline">client list</a>" — may add fuel to the conspiracy fire surrounding his death.
The 11-hour "<a href="https://www.justice.gov/video-files/video1.mp4">full raw</a>" surveillance video from the night Epstein died in New York's Metropolitan Correctional Center back in 2019 is, notably, <a href="https://thehill.com/homenews/administration/5390069-pam-bondi-explains-epstein-video/">missing the key minute</a> where the notorious financier purportedly killed himself. In the footage that was just released, there is no evidence that anyone entered or exited the cell where he died.
After Wired reporters and independent experts reviewed its metadata, however, a more complicated story emerged: that the video appears to have been modified from its initial form, likely using Adobe's Premiere Pro video editing software.
Specifically, those experts tell Wired that the DOJ's "full raw" video from Epstein's last half-day alive seems to have been assembled from two separate clips, and the whole file itself was saved multiple times before being exported and uploaded to the DOJ website.
As sketchy as those metadata modifications may be, Wired cautions that there's nothing contained within them to prove that the video was manipulated maliciously. But that's unlikely to stop certain conspiracy theorists from growing even more skeptical.
Take, for instance, Dan Bongino, the former Fox News talking head and <a href="https://x.com/dbongino/status/1653050511191334912">longtime Epstein conspiracist</a> who now serves as the deputy director of the Federal Bureau of Investigation. Earlier this week, <a href="https://www.nbcnews.com/politics/justice-department/dan-bongino-weighs-resigning-fbi-heated-confrontation-pam-bondi-epstei-rcna218388">NBC reported</a> based on insider sources that Bongino is "out of control furious" and threatening to resign from his post and "torch" attorney general Pam Bondi over the way the DOJ has handled the case.
To be fair, those conspiracists have quite a lot to work with. Trump himself purportedly had a deep friendship with Epstein, who back in 2002 he called a "<a href="https://www.nytimes.com/2019/07/09/us/politics/trump-epstein.html">terrific guy</a>" who "likes beautiful women as much as I do, and many of them are on the younger side." The two were <a href="https://www.washingtonpost.com/politics/donald-trump-and-jeffrey-epstein-partied-together-then-an-oceanfront-palm-beach-mansion-came-between-them/2019/07/31/79f1d98c-aca0-11e9-a0c9-6d2d7818f3da_story.html">photographed partying with</a> at his Mar-a-Lago mansion in 2000. As the icing on the cake, Trump was also named in <a href="https://www.npr.org/2024/01/13/1224147251/epstein-jeffrey-documents-ghislane-maxwell-giuffre-dershowitz-clinton-trump">formerly-redacted files</a> about the billionaire pedophile that a judge unsealed last year.
There's also a glaring question of hypocrisy. Trump <a href="https://www.washingtonpost.com/politics/2025/07/08/jeffrey-epstein-bondi-patel-trump/">repeatedly vowed</a> on the campaign trail last year to declassify the so-called "Epstein Files," but has since taken to stridently defending Bondi and the DOJ's claims since the jail cell footage was released.
"Whatever your flavor of Epstein conspiracy is, the video will help bolster it," explained extremism researcher Mike Rothschild when speaking to the magazine.
According to Hany Farid, a digital forensics and misinformation researcher at the University of California, Berkeley who reviewed the footage for Wired, the apparent changes also open the DOJ up to new questions and potential legal implications.
As Farid told the website, the video's seemingly manipulated metadata suggests problems with its chain of custody — the complex evidence-handling procedures that law enforcement are required by law to painstakingly document.
"If a lawyer brought me this file and asked if it was suitable for court, I’d say no. Go back to the source. Do it right," the forensic expert said. "Do a direct export from the original system — no monkey business."
Thus far, neither Donald Trump's administration nor attorney general Pam Bondi, who oversaw the Epstein footage investigation, have <a href="https://x.com/BillyM2k/status/1944183426598257091">responded directly</a> to the concerns raised by Wired's reporting. In a <a href="https://x.com/brianschatz/status/1943717012888281223">tweet about the article</a>, however, Hawaii senator Brian Schatz flagged the strangeness of the allegations.
"I am probably among the least conspiracy minded people you would ever meet," the Hawaii Democrat wrote, "but this is getting pretty nearly impossible to explain."
More on Trump: <a href="https://futurism.com/trump-nasa-administrator-sean-duffy">We're Speechless at Who Trump Just Appointed Head of NASA</a>