People are using ChatGPT for things its creators never even imagined — including, in a decidedly "Black Mirror" twist, to solicit feedback on how to better conform to society's norms around attractiveness.
As the <a href="https://www.washingtonpost.com/technology/2025/05/25/chatgpt-ai-beauty-advice/">Washington Post reports</a>, women struggling with self-worth and body image issues have taken to asking the OpenAI chatbot, and others like it, for advice on how to be, in the parlance, "hotter."
What's worse: despite <a href="https://futurism.com/character-ai-eating-disorder-chatbots">years of evidence</a> showing how <a href="https://www.forbes.com/sites/emilybaker-white/2025/05/12/these-ai-tutors-for-kids-gave-fentanyl-recipes-and-dangerous-diet-advice/">weird chatbots can get</a> when it comes to providing advice, multiple women in their early 30s told the news outlet that they'd had good experiences getting ChatGPT to critique their looks and bodies.
"People filter things through their biases and bring their own subjectivity into these sorts of loaded questions," explained Ania Rucisnki, a 32-year-old in Australia who said she uploaded her photo to the chatbot and asked it how she could look better. "ChatGPT brings a level of objectivity you can’t get in real life."
That assumption, while intriguing, isn't entirely correct.
Though OpenAI and other companies have worked to <a href="https://www.technologyreview.com/2023/02/21/1068893/how-openai-is-trying-to-make-chatgpt-safer-and-less-biased/">tamp down human prejudice</a> from their chatbots — which we've <a href="https://futurism.com/dall-e-mini-racist">encountered the hard way</a> on multiple <a href="https://futurism.com/ai-chatbot-hallucinations-marriage">occasions over the years</a> — they're still <a href="https://www.washingtonpost.com/technology/interactive/2024/ai-bias-beautiful-women-ugly-images/">rife with bias</a>. The very concept of AI being objective is starting to shift as models like OpenAI's o-series reasoning models are trained to learn, <a href="https://news.ucsc.edu/2025/03/ai-empathy/">with many missteps</a>, how to have <a href="https://www.inc.com/jason-aten/openai-says-chatgpt-4-5-comes-with-a-killer-feature-emotional-intelligence/91154092">emotional intelligence</a> and <a href="https://theconversation.com/chatgpts-artificial-empathy-is-a-language-trick-heres-how-it-works-244673">empathy</a>.
When Kayla Drew, also 32, asked ChatGPT for advice on various aspects of her looks, from her skin and hair to clothes and makeup, it gave her suggestions that ultimately led to roughly $200 in purchases. Though OpenAI doesn't yet make money from any links it includes during such exchanges, the company has said that it's <a href="https://futurism.com/the-byte/openai-weighs-ads-chatgpt">exploring ad revenue</a> as a future income stream — and others, like Perplexity, are <a href="https://digiday.com/media/how-perplexity-calculates-publishers-share-of-ad-revenue/">already making money</a> from in-chatbot advertising.
<div class="wpds-c-PJLV article-body type-text" data-qa="article-body">
<p class="wpds-c-heFNVF wpds-c-heFNVF-iPJLV-css overrideStyles font-copy" dir="null" data-apitype="text" data-contentid="YOCOJAQ66FC4XFOKL6WFBN2KBE" data-el="text" data-scroll-pos="11" data-scroll-measured="true">"Today I asked about whitening my teeth, and she [yes, this person uses 'she' pronouns to refer to ChatGPT] was like, 'Make sure your dental hygiene is good,' and I was like, 'Damn, girl,'" Drew told WaPo. "Nobody else would come up to me and say that. It was pretty cool because I guess I needed to hear it."
Of course, it's hard to suss out how "make sure your dental hygiene is good" would be considered particularly actionable advice for questions about teeth whitening. According to Emily Pfeiffer, a commerce analyst with the business consultancy Forrester, that sort of thing is par for the course with AI, which "just echoes what it’s seen online."
"Much of that has been designed to make people feel bad about themselves and buy more products," Pfeiffer said.
More on AI and women: <a href="https://futurism.com/ai-labor-gender-equity">AI Is Replacing Women's Jobs Specifically</a>
</div>

Talking Body

"ChatGPT brings a level of objectivity you can’t get in real life."

Women With Body Image Issues Are Asking ChatGPT Something Terrible

In 1986, an inventor named David Humble rolled out the first self-checkout machine to a Kroger outside of Atlanta. It was a <a href="https://www.oklahoman.com/story/news/1987/05/20/self-checkout-latest-innovation-in-grocery-stores/62689483007/">watershed moment</a> for shop owners at the time, who hoped to "increase store profits" and "alleviate labor shortages" — another way to say "avoid hiring workers above the national minimum wage," which <a href="https://www.nytimes.com/1986/01/05/nyregion/worker-shortage-leads-to-job-glut.html">was then </a>$3.35 per hour, and <a href="https://www.dol.gov/general/topic/wages/minimumwage">just </a><a href="https://www.dol.gov/general/topic/wages/minimumwage">$7.25</a> today.
Fast forward 39 years, and self-checkout has taken the world by storm. Today, there are over 217,000 self-checkout terminals installed <a href="https://www.retaildive.com/news/self-checkout-boasts-record-year-report/720623/">throughout the world</a>, a technology which <a href="https://www.newsday.com/business/self-checkout-shoplifting-stew-leonards-holiday-farms-wus4aed2?utm_source=chatgpt.com">labor experts estimate</a> eliminates 35 to 40 percent of entry-level retail jobs within 12 months of operation in any given store.
But with <a href="https://www.theatlantic.com/technology/archive/2023/10/self-checkout-kiosks-grocery-retail-stores/675676/">constant outages</a>, unsettling <a href="https://www.the-independent.com/life-style/food-and-drink/features/tesco-ai-self-checkout-supermarkets-shopping-privacy-b2759100.html">surveillance schemes</a>, and millions of <a href="https://shift.hks.harvard.edu/wp-content/uploads/2024/08/Selfcheckout_Layout_V8.pdf">unexpected items in the bagging area</a>, who really benefits from all this tech?
That's a question Gen Z is being forced to grapple with in an imminent way. Not just with self-checkout, but with the explosive growth of artificial intelligence over the years they're entering — or at least attempting to enter — the labor force.
A <a href="https://www.oxfordeconomics.com/wp-content/uploads/2025/05/US-Educated-but-unemployed-a-rising-reality-for-college-grads.pdf">recent report</a> by the Oxford Economics Group found that unemployment rates among recent college grads in the US are rising, suggesting that "entry-level positions are being displaced by artificial intelligence at higher rates."
Among 22 to 27 year olds, work in STEM industries like computer science is down 8 percent from 2022, signaling a <a href="https://www.nytimes.com/2025/03/25/business/economy/white-collar-layoffs.html">major disruption</a> in labor market norms. "Recent and experienced college graduate unemployment rates have always been lower than the national average, until now," the report reads.
While there's always been a gap in employment as young college grads search for their first "real job," the number now struggling to find work is growing faster than unemployment overall, suggesting that "entry-level positions have become harder to find."
That's a view shared by LinkedIn executive Aneesh Raman, who recently <a href="https://www.nytimes.com/2025/05/19/opinion/linkedin-ai-entry-level-jobs.html">penned an op-ed</a> about artificial intelligence and the entry-level job market.
"Breaking first is the bottom rung of the career ladder," Raman said. He points to tech as a key example, where simple coding and debugging tasks — once the proving ground for junior software engineers — have rapidly become automated by AI, despite some <a href="https://futurism.com/time-workers-save-ai-jobs">major shortcomings</a> with the tech.
Other disruptions can be found in <a href="https://www.nytimes.com/2023/04/10/technology/ai-is-coming-for-lawyers-again.html">big law</a>, where entry-level clerical work is being handed over to AI, and <a href="https://www.nytimes.com/2025/02/02/business/klarna-ceo-ai.html">retail jobs</a>, where AI chatbots take on the friction of customer service, to the annoyance of <a href="https://futurism.com/the-byte/people-hate-ai-customer-service">pretty much everyone</a>.
While these upsets may seem isolated to specialized industries compared to US labor overall, the job market is a tightly wound ecosystem, and Raman notes the potential for ruffles in white collar labor to make huge waves across the country.
"If entry-level roles evaporate, those lacking elite networks or privileged backgrounds will face even steeper barriers to finding their footing in the workplace," the LinkedIn executive wrote. "Plus, the fallout from large-scale economic shifts ripples through entire communities. When manufacturing jobs vanished across America’s heartland, the result wasn’t just lost income, but also social and political upheaval."
Whether or not AI is ready to take on the jobs typically given to early-career college grads is a whole different story. Many preliminary reports suggest the tech's <a href="https://futurism.com/professors-company-ai-agents">actual capabilities</a> are far behind the <a href="https://futurism.com/venture-capitalist-andreessen-jobs">typical narrative</a> of an AI-dominated labor market.
That's a story typically fed to us from executives and tech industry moguls. Raman notes that, in a <a href="https://www.linkedin.com/posts/linkedin-economic-graph_redefining-entry-level-roles-innovation-activity-7328833531374108672-ZHUZ/?utm_source=share&amp;utm_medium=member_desktop&amp;rcm=ACoAABTEARUB3DqvBLONb1r6EcmmOVjinNqM3Wo">LinkedIn survey</a> of over 3,000 corporate bigwigs, "63 percent agreed that AI will eventually take on some of the mundane tasks" that usually go to entry-level associates.
There's reason to believe these <a href="https://reutersinstitute.politics.ox.ac.uk/news/how-news-coverage-often-uncritical-helps-build-ai-hype">buzzy tales</a> of AI "innovation" and "workplace paradigm shifts" are really just <a href="https://futurism.com/time-workers-save-ai-jobs">cover</a> for broader — and more perfidious — trends in the labor market, such as the "<a href="https://firstup.io/blog/never-mind-the-great-resignation-welcome-to-the-great-gigification/">gigification</a>" of labor.
Meanwhile, researchers and labor organizers are sounding the alarm that the <a href="https://link.springer.com/content/pdf/10.1007/s43681-024-00502-w.pdf">most vulnerable</a> among us — not just entry-level, but minority, immigrant, and elderly workers — will be among the first to feel the rungs break as the "AI revolution" barrels on.
More on AI: <a href="https://futurism.com/ai-labor-gender-equity">AI Is Replacing Women's Jobs Specifically</a>

Qui Bono

"Breaking first is the bottom rung of the career ladder."

AI Is Destroying Gen Z's Chances at a Stable Career

In Manhattan's tony Nolita neighborhood, the so-called "Crypto King of Kentucky" allegedly held an Italian man hostage and tortured him for weeks when trying to gain access to his Bitcoin wallet.
As <a href="https://www.nbcnewyork.com/manhattan/crypto-king-nyc-townhome-torture-chamber-bitcoin-password/6277651/">NBC New York reports</a>, the purported "king," whose real name is John Woelz, was arrested and arraigned after the Italian tourist he allegedly kidnapped escaped on foot — literally speaking, as he was <a href="https://www.youtube.com/watch?v=5YFDFGt3Wx4">filmed running barefoot</a> — to find help.
<iframe loading="lazy" title="Video shows kidnapping and torture victim&#039;s escape from NYC home" width="500" height="281" src="https://www.youtube.com/embed/5YFDFGt3Wx4?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>
That tourist, <a href="https://www.nytimes.com/2025/05/27/nyregion/crypto-investor-torture-arrest.html">as the New York Times reports</a>, was identified as a millionaire named Michael Valentino Teofrasto Carturan who had been in a tumultuous business partnership with Woelz and a third man, William Duplessie.
Before things went sour, the relationship between the three men was described by insiders to NBC NY as"<a href="https://www.nbcnewyork.com/manhattan/frat-guys-gone-wild-crypto-king-due-in-court-in-nyc-townhouse-torture-case/6280734/">frat guys gone wild</a>," with Woeltz and Duplessie "roughing up" Carturan sometimes, but never to this extent.
When their partnership began to deteriorate, Carturan was said to have gone home to Italy, only to be persuaded back to New York where the "Crypto King" and Duplessie allegedly held him for three weeks in a four-story, 17-room luxury townhouse that <a href="https://www.compass.com/listing/38-prince-street-manhattan-ny-10012/1506636377084089097/">rents for $75,000 per month</a> (though Woeltz apparently got a deal and paid about $30,000, <a href="https://www.nbcnewyork.com/manhattan/frat-guys-gone-wild-crypto-king-due-in-court-in-nyc-townhouse-torture-case/6280734/">per NBC New York</a>.)
During their not-so-pleasant stay in Nolita, one or both of the accused conspirators tortured the Italian to try to get the password to Carturan's Bitcoin wallet, which was said to contain millions of dollars worth of crypto. It remains unclear whether they gained access to it before the man escaped, bloodied and barefoot, into the New York streets.
According to a police report viewed by the NYT, Carturan's torture was absolutely brutal.
Per his own attestation, the man was bound with electrical wires, pistol-whipped, urinated upon, forced to smoke crack cocaine, and electrocuted with a stungun while his feet were submerged in water. The kidnappers also allegedly threatened to kill Carturan's family, per prosecutor statements viewed by the newspaper, and other reports indicate that he claimed to have been <a href="https://www.nbcnewyork.com/manhattan/crypto-king-nyc-townhome-torture-chamber-bitcoin-password/6277651/">dangled from a fifth floor</a> — which would likely have taken place elsewhere, because the Prince Street townhouse only has four stories — and <a href="https://nypost.com/2025/05/23/us-news/kentucky-crypto-investor-allegedly-tortured-captive-italian-tourist-for-weeks-in-luxury-nyc-apartment-sources/">threatened with a chainsaw</a> as well.
The fuzziness of those details already appears to be a point of contention. During an arraignment hearing for Duplessie, the second of the alleged co-conspirators, his attorney Sanford Talkin suggested that there's more to the case than meets the eye.
"The facts here are hotly disputed,"  Talkin argued, per the NYT. "[Duplessie's] involvement is hotly disputed."
Though the specifics of the alleged torture have not been tamped down in the aftermath of Carturan's shocking escape, there's plenty of physical evidence lending veracity to his account. When investigators looked through the house of horrors on Prince Street, they found, according to the NYT, photos of the Italian's torture, as well as multiple firearms, a ballistic vest, chicken wire, and blood.
While we don't know much about either alleged kidnappers' backgrounds beyond their involvement in crypto investing and their late-night escapades, <a href="https://www.tmz.com/2025/05/28/john-woeltz-william-duplessie-big-spenders-new-york-nightclub-the-box/">per TMZ</a>, at an exclusive nightclub called The Box, they appear to have given neighbors the creeps.
"I knew there was something weird going on, I just assumed they were shooting porn or something," a nearby jewelry vendor, who asked not to be named, <a href="https://nypost.com/2025/05/23/us-news/kentucky-crypto-investor-allegedly-tortured-captive-italian-tourist-for-weeks-in-luxury-nyc-apartment-sources/">told the New York Post</a> of the "weird crypto guys" who'd taken up the townhouse.  "I couldn’t have guessed all this."
This kind of "<a href="https://www.wsj.com/finance/currencies/crypto-industry-robberies-attacks-32c2867a">wrench attack</a>," as crypto kidnappings, tortures, are robberies are now called, are unfortunately becoming more common as traders move to physical storage of their holdings to avoid hacks. This one, it seems, just happened to explode in full view of a bustling lower Manhattan neighborhood.
More on crypto crazies: <a href="https://futurism.com/crypto-daughter-paris-kidnap">Crypto Tycoon's Daughter Narrowly Escapes Kidnappers in Paris</a>

There Will Be Bitcoin

The victim escaped and ran barefoot through the streets.

The "Crypto King of Kentucky" Allegedly Kidnapped and Tortured a Man For Weeks in NYC Over His Bitcoin Stash

The <a href="https://futurism.com/neoscope/gaben-steam-bci-stealth">rumors are true</a> — Gaben's finally doing it. And no, "it" doesn't involve the <a href="https://half-life.fandom.com/wiki/Development_of_the_next_Half-Life_game">number three</a>. 
After a long silence, Starfish Neuroscience, the secretive startup owned by <a href="https://futurism.com/the-byte/steam-rejecting-games-using-ai">Valve</a> cofounder Gabe Newell, announced that it's getting ready to unveil its first batch of brain computer interfaces this year. And it's looking for help.
"We anticipate our first chips arriving in late 2025 and we are interested in finding collaborators for whom such a chip would open new and exciting avenues," wrote Starfish engineer Nate Cermak in the company's first ever <a href="https://starfishneuroscience.com/blog/ultra-low-power-miniature-electrophysiological-electronics/?header-bg=card-bg0">blog post</a> last week.
The update was made with so little fanfare that it wasn't <a href="https://x.com/SadlyItsBradley/status/1925459695713525986">first spotted for two days</a>, initially by Valve watcher Brad Lynch, which prompted more widespread attention from the <a href="https://www.theverge.com/news/673938/gabe-newell-valve-founder-brain-computer-interface-first-chip-starfish">likes of The Verge</a>.
From what little Starfish has shared, it's creating a "minimally invasive" chip that will target multiple regions of the brain instead of one, unlike BCIs from competitors like Elon Musk's <a href="https://futurism.com/nonverbal-neuralink-patient-grok-answers">Neuralink</a>. 
The reason why, Starfish says, is because emerging research suggests that neurological disorders like Parkinson's disease involve some form of dysfunction in interactions between brain regions.
"Developing better therapies for these disorders will require distributed neural interfaces capable of interacting with the brain at the circuit level," reads the blog post. "Such interfaces do not yet exist clinically, and existing interfaces are not straightforward to parallelize because of their bulky physical footprints."
And so, Newell's startup is making its chip as small possible, with such tiny power demands that it doesn't even need a battery. It'll run, instead, entirely via wireless power transmission. 
In all, Starfish's planned brain chip — which it says is being built in partnership with the international R&amp;D firm IMEC — will have just 32 electrode sites and a tiny power consumption of 1.1 milliwatts during "normal recording." And its dimensions are equally diminutive: just two by four millimeters in size.
For comparison, Neuralink's battery-powered chip, as the Verge notes, boasts 1,024 electrodes, was last reported to demand around six milliwatts of power, and is about 23 by eight mm. 
That's not a knock against Neuralink, to be clear. Its patients seem pretty happy with their chips — allowing one of them to play Valve's very own tactical FPS "<a href="https://futurism.com/neoscope/patient-neuralink-play-counter-strike-2">Counter-Strike 2</a>," in fact.
But Musk's company has run into <a href="https://futurism.com/neoscope/neuralink-cancels-second-implant-medical-issues">issues</a> with <a href="https://www.theguardian.com/technology/article/2024/may/09/neuralink-brain-chip-implant">implanting the devices</a>. That perhaps illustrates Starfish's down-scaling raison d'etre, which will reduce the "surgical burden of device implantation," the company argues.
That said, this is all very early days. Starfish doesn't have an actual product to share yet, and it has a long road to catching up with the competition. But will it have something to show before Newell's little side-venture Valve releases "<a href="https://www.engadget.com/gaming/half-life-3-is-reportedly-playable-in-its-entirety-and-could-be-announced-this-year-183030499.html">Half-Life 3</a>?" It's entirely possible.
"There is tons of work yet to be done here," Starfish wrote.
More on BCIs: <a href="https://futurism.com/brain-implant-companies-data">Brain Implant Companies Apparently Have an Extremely Dirty Secret</a>

Brave Newell World

Neuroscience/Brain Science

He'll make anything but a certain video game sequel.

Gabe Newell's New Company Is Getting Ready to Put Chips in Your Brain

After eviscerating Tesla's brand through his political mayhem, billionaire Elon Musk is suddenly singing a dramatically different tune.
This month, he abruptly proclaimed that he had "<a href="https://futurism.com/elon-musk-giving-up-politics">done enough</a>," announcing that he would substantially cut down the time he spent reshaping the federal government, efforts that have left permanent scars on Tesla's credibility and image.
Instead, he promised that he would refocus the majority of his time on his ailing businesses, news that was met with his EV maker's shares rallying after a precipitous decline in sales and stock value during his period of intense political activity.
Meanwhile, Musk spent much of this week publicly addressing the development of SpaceX's Starship.
He <a href="https://x.com/live_munro/status/1927477912128639179">chatted</a> with automotive engineer Sandy Munro, ran Tim "Everyday Astronaut" Dodd <a href="https://x.com/Erdayastronaut/status/1927466323862335651">through the company's progress</a>, and was seemingly on his best behavior during an <a href="https://arstechnica.com/space/2025/05/elon-musk-turns-his-focus-back-to-space-says-starship-and-mars-matter-most/">interview</a> with Ars Technica's Eric Berger.
It's a dramatic change of tune, considering Musk had spent much of this year raging against the mainstream media, <a href="https://futurism.com/grok-ai-elon-musk-white-genocide">spreading racist conspiracy theories</a>, <a href="https://futurism.com/trump-administration-elon-musk-views">wearing ridiculous hats</a> during cabinet meetings, and <a href="https://futurism.com/elon-musk-raging-against-trump-administration">lashing out against his many critics</a>.
What inspired Musk to — at least temporarily — return to the character of the detail-oriented businessman? It's especially striking after he hit a new low last week, making a <a class="underline hover:text-futurism hover:no-underline transition-all duration-200 ease-in-out" href="https://gizmodo.com/billionaire-man-child-elon-musk-gives-his-most-petulant-interview-to-date-2000604920">bafflingly childish</a> appearance at Bloomberg's Qatar Economic Forum, accusing the interviewer of being an "NPC" and getting rattled by some tough questions.
His reputation and popularity <a href="https://futurism.com/elon-musk-associated-press-poll-popularity">have taken a major hit</a> following his descent into far-right extremism and largely fruitless ripping up of government contracts. Even Republican lawmakers in the White House <a href="https://futurism.com/trump-administration-elon-musk-views">have reportedly had enough</a> of his overzealousness and cringy humor, resulting in shouting matches with Trump officials.
Now that he's dragged his EV maker's brand through the mud and spawned a global anti-Tesla movement, Musk is seemingly trying to play Mr. Nice Guy once again.
There's good reason. Musk has made himself a historically unpopular figure over the last couple of months, greatly undermining his credibility as an entrepreneur and exposing an alarming degree of callousness and <a href="https://edition.cnn.com/2025/03/05/politics/elon-musk-rogan-interview-empathy-doge">lack of empathy</a>.
And as SpaceX continues to struggle getting its enormous Starship both off the ground and safely back to the launch pad, Musk's latest actions give off a sense of desperation.
"Give me the rundown," Dodd asked Musk <a href="https://x.com/Erdayastronaut/status/1927466323862335651">during a Tuesday afternoon interview</a>. "Right now, we're T-minus four and a half hours. What's going through your head right now?"
A visibly stressed Musk pulled both of his hands through his hair, giving off the impression of a man in dire straits.
"I have a lot going through my head, actually," he finally answered. "Just thinking about the various ways the rocket can fail, which I'm sure are many."
Tuesday's test flight, the company's ninth, turned out to be yet another setback. The rocket spun out of control minutes into the launch, with SpaceX giving up on the deployment of a test batch of mock Starlink satellites since the cargo pay wouldn't fully open.
Perhaps it was a symbolic development for Musk's reinvigorated enthusiasm for his ailing businesses. After all, the billionaire has gouged some deep scars that aren't going to heal overnight.
More on Musk: <a href="https://futurism.com/self-driving-tesla-swerves-off-road-crashes">Self-Driving Tesla Suddenly Swerves Off the Road and Crashes</a>

Reeks of Desperation

"I have a lot going through my head, actually."

You Can Suddenly Sense Elon Musk's Desperation

SpaceX experienced yet another failure during its ninth test launch of its Starship vehicle Tuesday evening.
The rocket <a href="https://www.youtube.com/watch?v=QEn0_mjFzpI">spun out of control</a> as it tumbled through space, causing it to make an uncontrolled descent in the Indian Ocean that ended in an <a href="https://mashable.com/article/spacex-starship-test-flight-ninth-explosion-failure">all-too-familiar blast</a>.
As is always the case after a failure, SpaceX CEO Elon Musk was optimistic.
During an interview with <a href="https://arstechnica.com/space/2025/05/elon-musk-turns-his-focus-back-to-space-says-starship-and-mars-matter-most/">Ars Technica's Eric Berger</a>, published mere minutes before the launch of Tuesday's test flight, Musk was adamant that "we've got about an 80 percent chance of having solved" issues related to the design of the spacecraft's heat shield tiles and the cause of the "explosions in the last two flights in the upper stage engine bay."
But given this week's disaster, Musk may have put the horse before the cart. The prototype, dubbed Ship 35, never even got close to a controlled reentry. Live footage shared by the space company showed the untiled side of the spacecraft heating up as it was enveloped with flaming plasma.
However, the launch wasn't a complete failure. The two previous test flights, which saw each Starship <a href="https://futurism.com/spacex-fires-starship">break up over the Caribbean</a> and appear as bright streaks in the evening sky, were plagued by propulsion system leaks, triggering early shutdown — which technically didn't happen during Tuesday's launch, with the failure coming later.
In other words, the company's latest test flight was <a href="https://arstechnica.com/space/2025/05/spacex-may-have-solved-one-problem-only-to-find-more-on-latest-starship-flight/">one step forward, and several steps back</a>, yet again highlighting the risks of SpaceX's iterative design philosophy.
Whether Musk's newfound motivation and decision to stop ignoring his businesses while chaotically slashing budgets in the White House will prove effective remains to be seen.
"It's not like I left the companies," he told Berger. "It was just relative time allocation that probably was a little too high on the government side, and I've reduced that significantly in recent weeks."
Meanwhile, Musk proclaimed that he was elated following Tuesday's failure, <a href="https://x.com/elonmusk/status/1927531406017601915">arguing</a> that "Starship made it to the scheduled ship engine cutoff, so big improvement over last flight!"
"Also, no significant loss of heat shield tiles during ascent," he added.
Chances are we'll see more attempts in the coming months, if Musk is taken by his word.
"Launch cadence for next 3 flights will be faster, at approximately 1 every 3 to 4 weeks," Musk wrote.
Even with his renewed interest in his businesses, the pressure is seriously starting to grow for SpaceX. NASA is still hoping to tap the heavy launch platform for Artemis 3, its first crewed lunar landing attempt, which is tentatively scheduled for 2027.
Given the latest setbacks, SpaceX has its work cut out for it. Beyond a safe landing, the company still has to figure out how to increase the launch cadence significantly, refuel in orbit, provide life support to crews, and much more.
To Musk, it's all part of a much greater plan to send humans to Mars, a distant planet that's incredibly hostile to life.
"We're building the equivalent of the Union Pacific Railroad and the train," he told Berger prior to Tuesday's failure. "So once you have the transportation system to Mars, then there's a vast set of opportunities that open up to do anything on the surface of Mars, which includes, you know, doing everything from building a semiconductor fab to a pizza joint, basically building a civilization."
More on Starship: <a href="https://futurism.com/nasa-contract-elon-musk-spacex-starship">NASA Signs Contract for Elon Musk's Starship, Even Though It's Never Launched Without Exploding</a>

80 Percent Certainty of Chaos

Oops.

Elon Boasts of Huge Starship Improvements Immediately Before It Blows Up Spectacularly

The use of AI chatbots has exploded in the workplace.
White-collar workers have adopted the tech in huge numbers. Executives are looking for every opportunity to shoehorn AI into every aspect of their businesses. Companies continue to pour tens of billions of dollars into building out enormous data centers to support increasingly power-hungry AI models.
But is any of this actually leading to higher efficiency at work and increased pay?
According to a <a href="https://www.nber.org/papers/w33777">recently-released working paper</a> by a pair of economists at the University of Chicago and the University of Copenhagen that was <a href="https://fortune.com/2025/05/18/ai-chatbots-study-impact-earnings-hours-worked-any-occupation/">spotted by Fortune</a>, it's not looking great.
After analyzing data about 25,000 workers across 7,000 workspaces, users of AI only saved on average three percent of time. And only a meager three to seven percent of those productivity gains translated into bigger paychecks, they found.
Put simply, AI isn't even close to coming for everybody's jobs, a frequently cited fear — and it's not exactly making workers more productive, either, despite the industry's many reassurances.
"While adoption has been rapid, with firms now heavily invested in unlocking the technological potential, the economic impacts remain small," the two economists wrote in their paper. "Our findings challenge narratives of imminent labor market transformation due to generative AI."
As <a href="https://www.economist.com/finance-and-economics/2025/05/26/why-ai-hasnt-taken-your-job">The Economist points out</a>, while many pundits point to <a href="https://futurism.com/investor-ai-fully-replacing-people">growing unemployment and blame AI</a>, existing macroeconomic data suggests there are plenty of other reasons to blame. To the contrary, the magazine found that employment in white-collar work has actually risen in the past year.
"Software, writing code, writing marketing tasks, writing job posts for HR professionals — these are the tasks the AI can speed up," coauthor Anders Humlum, at UChicago, told Fortune. "But in a broader occupational survey, where AI can still be helpful, we see much smaller savings."
"I might save time drafting an email using a large language model, so I save some time there, but the important question is, what do I use that time savings for?” he added. "Is the marginal task I’m shifting my work toward a productive task?"
What workers actually do with the tiny amounts of extra free time, courtesy of AI, isn't exactly comforting, either. More than 80 percent of workers in the study said they used saved time for more work. Less than ten percent said they took the time off.
The findings "suggest that workers are not exactly knocking on the boss’s door asking for more work," Humlum told Fortune.
Signs that the impact of AI on the labor market has been overblown are certainly there. Earlier this year, Klarna CEO Sebastian Siemiatkowski admitted that <a href="https://futurism.com/the-byte/klarna-ceo-ai-replacing-workers">his all-in approach</a> to replacing human customer service agents with AI was not working. The company <a href="https://futurism.com/the-byte/klarna-ceo-bragged-replacing-workers-ai-losses">revealed last week</a> that it's facing net losses of $99 million for the first quarter of this year, more than double compared to the same period last year.
Last week, Duolingo CEO Luis von Ahn <a href="https://fortune.com/2025/05/24/duolingo-ai-first-employees-ceo-luis-von-ahn/">walked back his earlier promises</a> to <a href="https://futurism.com/the-byte/duolingo-replacing-humans-ai">replace all contract workers</a> with AI following an immense wave of customer blowback.
Earlier this month, an IBM survey found that <a href="https://fortune.com/article/ceos-ai-initiatves-fraction-deliver-return-on-investment-roi-study/">only a fraction</a> of AI initiatives are delivering a return on investment after surveying 2,000 CEOs.
These latest findings are an intriguing reality check for the notion that AI is coming for all of our jobs. Perhaps the tech is both making us efficient, but also failing to make us redundant.
The conclusion highlights growing concerns that generative AI tech may be a dead end after all, as <a href="https://futurism.com/ai-researchers-tech-industry-dead-end">some experts have warned</a>. Perhaps our jobs may be safe — other societal factors, like <a href="https://www.nbcnews.com/politics/trump-administration/trump-rejects-concerns-prices-economic-uncertainty-defends-agenda-rcna203512">historic levels of economic uncertainty</a>, notwithstanding.
More on AI job automation: <a href="https://futurism.com/ai-labor-gender-equity">AI Is Replacing Women's Jobs Specifically</a>

More Time for More Work

This is pitiful.

You'll Spit Take When You Hear How Little Time Workers Are Saving With AI, According to This Huge New Study

Last month, Aurora Innovation became the first company to operate a fully driverless semi truck service on American highways.
And by "fully driverless," we mean that no one — not a soul — was sitting in the driver's seat as the autonomous truck performed its first official delivery, barreling down a stretch of Texas's Interstate 45 with its 25,000 pound haul of frozen treats in tow. 
There was, in fact, zero human supervision. At least, not unless your idea of supervision is a man lounged in the back of the truck's cabin, watching YouTube videos on his phone, as the other suckers on the road were probably focused on not falling victim to one the country's most noted <a href="https://www.click2houston.com/news/local/2025/03/10/how-deadly-has-i-45-been-over-the-years/">motorist death traps</a>.
"I may be in the truck, but I'm just a passenger," wrote Aurora CEO Chris Urmson <a href="https://aurora.tech/newsroom/aurora-delivers-my-ride-in-the-first-self-driving-commercial-truck">in a writeup</a> about his backseat experience during the truck's inaugural haul. "I'm passing the time by chatting with our team, drafting this blog, and watching YouTube."
Urmson considered the feat to be a "watershed moment." Already, Aurora's total of two driverless trucks in operation have logged more than 1,200 driverless miles performing real deliveries along I-45, with plans to expand its fleet to 20 big rigs by the end of the year, the <a href="https://www.nytimes.com/2025/05/27/business/driverless-semi-trucks-aurora-innovation.html">New York Times reported</a>. Unlike human drivers, the trucks don't have to follow the 11-hour daily driving limit. And they are, Aurora insists, most definitely safe.
"We have something like 2.7 million tests that we run the system through," Urmson told the NYT.
Aurora's trucks are considered to have Level 4 autonomy on the SAE scale, meaning that they can fully drive themselves under limited conditions and don't require a human to take over. For comparison, consumer self-driving features like Tesla's — which are technically driver assistance systems — are generally Level 2. Robotaxis, like the ones Waymo offer, are Level 4.
Cracks are beginning to show, though, and that gloating about being fully driverless proved to be short-lived. Less than three weeks after Aurora launched its service, the startup <a href="https://www.axios.com/2025/05/19/aurora-driverless-truck-human-driver">announced</a> that it was putting a human supervisor back in the driver's seat. No more YouTube videos for Mr. Urmson.
This was done at the request of Paccar, the manufacturer of the Peterbilt trucks that Aurora outfitted its systems on. It's hazy why. In a <a href="https://aurora.tech/newsroom/updating-our-driverless-operations">blog post</a> we assume was not composed from the back of one of its trucks, Urmson only shared that it was "because of certain prototype parts in [Paccar's] base vehicle platform," without explaining further.
In its <a href="https://www.axios.com/2025/05/19/aurora-driverless-truck-human-driver">coverage</a>, Axios cited a <a href="https://www.bleeckerstreetresearch.com/research/aurora-innovation-aur">short-seller's report</a> which suggested that Aurora and Paccar didn't see eye-to-eye about the timing of the driverless operations; in fact, the Peterbilt logo was removed in some of Aurora's photos used to commemorate its launch.
Corporate intrigue aside, truckers are concerned about the dangers posed by Aurora semis' — not to mention the clear threat to their livelihoods. 
"It's potentially disastrous from a safety perspective," John Samuelsen, head of the Transport Workers Union of America, told the NYT.
"My initial thought is: It's scary," echoed veteran trucker Angela Griffin, who worried that the trucks wouldn't be able to handle chaotic traffic and challenging weather conditions.
Griffin recalled how misting rain confounded the AI-powered scanners on her own truck. Indeed, Aurora's fleet is designed only to operate during the day in clear weather conditions. That's why the human supervisors are there: to take over if things get wet.
Philip Koopman, an engineering professor at Carnegie Mellon, said that Aurora has been more diligent about safety than its competitors, but highlighted a distressing lack of government regulation. And there's no telling, he emphasized, how the vehicles will react to unexpected obstacles.
"This technology is really good at things it's practiced, and really bad at things it has never seen before," Koopman told the NYT. "From a safety point of view, nobody knows how it's going to turn out."
More on self-driving: <a href="https://futurism.com/self-driving-tesla-swerves-off-road-crashes">Self-Driving Tesla Suddenly Swerves Off the Road and Crashes</a>

Hands and Guilt Free

"I may be in the truck, but I'm just a passenger."

Trucking Company Deploys Self-Driving 18-Wheeler Where Human Employee Can Chill Out and Watch YouTube Videos in the Back as It Bombs Down the Highway

OpenAI's new partnership with famed iPhone designer Jony Ive has the rumor mill churning — and the latest may give you a guffaw.
According to supply chain analyst Ming-Chi Kuo, Ive and Altman's secretive <a href="https://futurism.com/the-byte/sam-altman-openai-wearable-device">AI companion device</a> — news of which <a href="https://www.wsj.com/tech/ai/what-sam-altman-told-openai-about-the-secret-device-hes-making-with-jony-ive-f1384005">leaked to the Wall Street Journal</a> last week — may look a lot like an iPod shuffle, but worn on a string around your neck, like a preposterous AI-inflected statement necklace.
"The current prototype is slightly larger than [Humane's] AI Pin, with a form factor as compact and elegant as an iPod Shuffle," Kuo said in a lengthy X post shortly after <a href="https://openai.com/sam-and-jony/">OpenAI announced</a> that it was purchasing io, Ive's hardware company. "The design and specifications may change before mass production."
Per the analyst's research, "one of the intended use cases is wearing the device around the neck," and although it won't have any screens, it may "have cameras and microphones for environmental detection," which could well jibe with Altman's claim in the employee call announcing the project that the device will be "fully aware" of its surroundings.
To us, this product sounds like a yassified version of <a href="https://futurism.com/humane-ai-pin-dumpster-fire">Humane's disastrous AI pin</a>, which itself was the size of a smart watch face and sucked so comprehensively that <a href="https://futurism.com/the-byte/more-people-returning-humane-ai-pin">more people returned the $700 disaster</a> than purchased it, forcing the company into a <a href="https://techcrunch.com/2025/02/18/humanes-ai-pin-is-dead-as-hp-buys-startups-assets-for-116m/">speedy total collapse</a>.
As such, Kuo's take isn't exactly optimistic for OpenAI.
"In my view, one of OpenAI's motives for announcing its collaboration with Jony Ive now is likely to shift market focus from [Google's recent I/O conference]," the Apple analyst wrote. "Google’s ecosystem and AI integration, showcased in the I/O keynotes, pose a challenge that OpenAI currently struggles to address. As a result, OpenAI is leveraging a new narrative to redirect attention."
(There are a few other things OpenAI may want to redirect market attention away from as well, including <a href="https://futurism.com/openai-chatgpt-sycophant">ChatGPT's very public woes</a> and warning signs of yet another leadership shakeup after Altman announced that he was <a href="https://futurism.com/the-byte/sam-altman-openai-applications-ceo">creating a new "CEO" position</a> that reports directly to him, which will <a href="https://bsky.app/profile/edzitron.com/post/3loofpbej6k2a">offload much of his own work</a> onto that new executive, <a href="https://openai.com/index/leadership-expansion-with-fidji-simo/">Instacart's Fidji Simo</a>.)
Looking at the bigger picture, Kuo noted that OpenAI's still-unconfirmed new AI gadget is part of a broader trend towards so-called "physical AI," which he described as "AI integrated real-world" — another sign that OpenAI is now following the herd instead of leading it.
"While the success of the Jony Ive-OpenAI partnership remains uncertain," the researcher mused, "it clearly aligns with this trend."
Kuo then went on to cite programming pioneer Alan Kay, who famously said that "people who are really serious about software should make their own hardware."
It is indeed a big step for OpenAI to be making its own devices — though at first bluff, there's nothing terribly impressive about a derivative form factor that looks like an iPod Shuffle.
More on wearables: <a href="https://futurism.com/wearable-cheating">New Wearable for Exceptionally Pathetic Men Detects When Their Wife Is Cheating</a>

Futile Devices

This is not going to be the game-changer Sam Altman is clearly expecting. 

You May Laugh When You Hear What OpenAI's Top Secret AI Gadget Allegedly Looks Like

<h2>Dark Lord's Spy</h2>
Scientists say they've developed a ludicrously keen-eyed laser device that can read the text in a book from a whopping 0.85 miles away.
As detailed in a <a href="https://journals.aps.org/prl/abstract/10.1103/PhysRevLett.134.180201">new paper</a> published in the journal Physical Review Letters, the team used interferometry, a commonly used technique in the world of astronomy that uses superimposed waves of light to create interference patterns, to develop the spy system.
However, instead of combining waves of radiation, light, or sound, the team used the information collected by multiple light detectors to measure variations in intensity, a technique known as intensity interferometry. They then used this data to compile an image of a distant object.
Put simply, researchers applied a technology that space observatories use to a ground-based laser system to zoom across vast distances — with promising and somewhat creepy results.
<h2>Read the Zoom</h2>
The team from the University of Science and Technology of China used eight infrared laser beams fired across a distance to reconstruct a visual of the target. They found that the system was sensitive enough to detect the shape of individual letters a mere three millimeters across, even from a vantage point nearly a mile away.
While intensity interferometry was first demonstrated in the mid-1950s to measure the diameters of distant stars, the researchers believe that the same technique could still be used in the presence of an atmosphere, which can lead to scrambled light wavelengths.
Instead of relying on a light source, like a star's radiation, the team used laser beams to illuminate distant objects. By splitting the laser into eight beams, the team made any perturbations in interference more observable, allowing them to still make out shapes, despite the presence of an atmosphere.
The team found that a single laser beam would have a far inferior resolution compared to eight beams.
In terms of possible applications of their research, the team proposes that it could be used to identify pieces of space debris by shining laser light on them.
The team is hoping they can further improve the system by harnessing the power of AI algorithms to interpret distant shapes more accurately.
"The new work represents a significant technical advancement in imaging distant objects that do not emit their own light," Sorbonne University optics researcher Shaurya Aarav, who wasn't involved in the research, <a href="https://physics.aps.org/articles/v18/99">told Physics Magazine</a>.
More on interferometry: <a href="https://futurism.com/darpa-requests-plans-large-bio-mechanical-space-structures">DARPA Requests Plans for "Large Bio-Mechanical Space Structures"</a>