<h2>Pizza Premonition</h2>
A flurry of activity at pizza delivery outlets near the Pentagon in Arlington, Virginia, is a surprisingly accurate predictor of war, as hungry military leadership hunkers down to monitor unfolding military activities.
As <a href="https://x.com/penpizzareport?lang=en">painstakingly documented</a> by X account Pentagon Pizza Report, a "busier than usual" indicator on the Google Maps profile of the Domino's in Arlington has been associated with major acts of war taking place around the world.
Most recently, the franchise received an onslaught of orders just before closing last night — <a href="https://www.linkedin.com/feed/update/urn:li:activity:7339237799063154688/">almost perfectly coinciding with</a> Israel's <a href="https://www.aljazeera.com/news/2025/6/13/sounds-of-explosions-heard-in-irans-capital-tehran">devastating attack</a> on Iran.
Even long before the advent of live, GPS-based customer tracking on Google Maps, famished Pentagon workers have long given away that there's something much darker going on by ordering copious numbers of pies.
"The Pentagon Pizza Index has been a surprisingly reliable predictor of seismic global events — from coups to wars — since the 1980s," The Economist's head of data journalism Alex Selby-Boothroyd wrote in a <a href="https://www.linkedin.com/feed/update/urn:li:activity:7339237799063154688/">LinkedIn post</a>.
"On the night of August 1st 1990, for example, the CIA ordered 21 pizzas in a single night just before the Iraqi invasion of Kuwait (a new record)," he added. "Who says pie charts aren't useful?"
<h2>War Pies</h2>
And it's not just Domino's. As <a href="https://www.euronews.com/culture/2025/06/06/what-is-the-pentagon-pizza-theory-eating-away-at-the-internet">Euronews reported</a> last week, the Papa John's close to the Pentagon has also been found to be "busier than usual" during the days leading up to Iran launching drones into Israeli territory in April 2024.
A separate event chronicled by Pentagon Pizza Report also predicted military activity on June 1, followed just hours later by an escalation between Israel and Iran.
Of course, making any definitive statements about any correlation would be foolish. But it's an intriguing sign that even the most secretive parts of the US government can easily show their cards without meaning to.
"Bottom line for journalists," Wolf Blitzer <a href="https://slate.com/human-interest/2016/07/the-pizza-meter-was-a-staple-of-1990s-pop-pseudoscience-we-should-revive-it.html">told CNN in 1990</a>. "Always monitor the pizzas."
More on pizza: <a href="https://futurism.com/the-byte/chef-admits-pizza-chatgpt">Chef Admits His Smash Hit Pizza Was Invented by ChatGPT</a>

Stuffed Clusterbombs

"Who says pie charts aren't useful?"

You Can Tell When War Is Breaking Out Because the Pentagon's Pizza Orders Spike Drastically

Back in May, as <a href="https://futurism.com/fbi-task-force-protect-tesla">Tesla Takedown rallies</a> and <a href="https://futurism.com/government-punishing-vandalism-tesla-dealership">anti-Musk vandalism</a> rocked the world, Tesla dealership manager Matthew LaBrot launched "<a href="https://teslaemployeesagainstelon.com/">Tesla Employees Against Elon</a>" — an act of protest that swiftly got him fired.
We wrote about LaBrot's <a href="https://futurism.com/tesla-manager-fired-blaming-elon-musk">symbolic act of defiance</a> against Musk's political maneuvers at the time. Now, a fresh tell-all interview <a href="https://hardresetmedia.substack.com/p/ex-tesla-manager-dishes-on-the-company">with </a><a href="https://hardresetmedia.substack.com/p/ex-tesla-manager-dishes-on-the-company">Hard Reset</a> is shining some light on just how drastically things have changed for Tesla over the past few months.
LaBrot is, first and foremost, a Tesla superfan. He currently rotates between driving a Tesla Model Y and a Cybertruck, despite getting sacked by the company. A self-described "EV activist," LaBrot's issue isn't with the brand per se, as his website makes clear: "the problem is Elon."
"You know, this wasn't a new thing for me," he said of his digital protest. "For almost six years, I’ve been focused on overcoming misinformation about EVs and helping grow that mission."
"Once we hit a tipping point where the person who's running this company is now pushing customers away from the mission, then the priority shifted," he continued. "That priority was to be an activist to try to save the company."
Having worked with Tesla since 2019, LaBrot had a front row seat as the trepidation with Musk began to set in. In recent years, he found it wasn't hesitation over electric vehicles, but with the <a href="https://futurism.com/elon-musk-ponders-unpopularity">noxious man</a> the brand now seemed to represent pushing customers away.
"We noticed customers, return customers, shying away from us a little bit," he noted. "And that's when I started to see the things that he [Musk] was putting on Twitter and the political views he started to have."
The tipping point, the former EV salesman said, was the billionaire's <a href="https://futurism.com/jewish-leaders-boycott-elon-musk-nazi-salute">infamous salute</a>.
"When it came to the salute — I still call it a salute — I was like, Tesla's going to take care of this," he told Hard Reset. "In 2010, CEOs were getting fired for having affairs. This guy's doing all this crazy stuff and Tesla's not going to do anything?"
That was when he realized how deep Musk's rot went, especially when Tesla's board of directors "doubled down and made him even more of a figurehead."
Now, LaBrot's not surprised that decision is biting those same executives in the craw.
Going forward, "they're absolutely hosed," he said. "I expect gigantic drops in sales, even potentially lower than Q1 sales."
Asked whether Musk can fix the <a href="https://futurism.com/the-byte/tesla-drivers-sue-elon-musk-right-symbols">damage he's caused</a> the brand, the former manager isn't optimistic.
"I don't think that there's anything he can do to change the people's opinion that have decided they're not going to support Tesla outside of him leaving," LaBrot asserted.
"A lot of people that I've spoken to don't even think that's enough at this point," he added. "They want him to sell all his shares and things like that, which I don't expect. I think for Tesla, as far as vehicle sales go, it's game over."
More on Tesla: <a href="https://futurism.com/the-byte/tesla-legal-places-store-unsold-cybertrucks">Tesla Can't Find Legal Places to Store All Its Unsold Cybertrucks</a>

You Lose

"I don't think that there's anything he can do."

Former Tesla Dealership Manager Says It's "Game Over" for Elon Musk

Following the <a href="https://futurism.com/trump-issues-new-threat-elon-musk">spectacular implosion</a> of billionaire Elon Musk's bromance with president Donald Trump, the consequences of his actions are starting to come into focus.
As <a href="https://www.reuters.com/business/aerospace-defense/after-trump-feud-role-musks-spacex-golden-dome-missile-shield-question-2025-06-12/">Reuters reports</a>, the White House's plans to tap Musk's SpaceX for a "Golden Dome" appear to be on shaky ground now that the two's ugly spat metastasized into a spite-filled mud-flinging contest.
The original plan was to use a satellite constellation — likely SpaceX's — to monitor for adversarial activity and work with drone manufacturer Anduril to build an anti-missile defense system.
However, sources now tell Reuters that the White House could abandon SpaceX and focus on existing ground systems for missile defense instead.
"Because of the blowup, the Pentagon has been given the space to look at other alternatives," one of the agency's sources said.
A White House spokesman told Reuters that the "Trump Administration is committed to a rigorous review process for all bids and contracts."
Losing the Golden Dome contract, a defense system valued by experts as a whopping $175 billion total, could be a major setback for Musk's space firm.
Last week, following a dramatic escalation in their feud, <a href="https://futurism.com/trump-calls-terminating-elon-musk-government-contracts">Trump warned Musk</a> that the "easiest way to save money in our Budget, Billions and Billions of Dollars, is to terminate Elon’s Governmental Subsidies and Contracts."
It was a considerable threat, considering Musk's empire has been built on tens of billions of dollars' worth of government contracts. As the <a href="https://www.washingtonpost.com/technology/interactive/2025/elon-musk-business-government-contracts-funding/">Washington Post found</a> earlier this year, NASA and Defense Department contracts awarded to SpaceX make up the bulk of the $38 billion Musk's businesses have received over the last two decades.
Days later, Musk <a href="https://futurism.com/elon-musk-apologizes-trump-friends-jeffrey-epstein">struck a wildly different tone</a>, confessing that attacks on Trump had gone "too far" and saying that he regretted "some of my posts."
With the missile defense shield project seemingly teetering on the edge of a spite-fuelled tussle, experts aren't too optimistic about the Golden Dome's long-term viability.
"That people guiding the program or building it are approved based on their political affiliation signals a real concern that the project itself is very politicized and not being conducted on the technical merits," missile defense expert Laura Grego told Reuters.
For now, the Trump administration's efforts to outline a plan for the Golden Dome appear to be chaotic and largely centered around lining the pockets of contractors.
"To this day, no one knows what the requirements are," a person familiar with the process told Reuters. "There isn't a coordinated effort with a true vision. All of these companies are just grabbing at this pot of money."
In other words, we simply have no idea what SpaceX's involvement was or will be going forward — if the Golden Dome project even exists in any meaningful way at this point.
And Musk and Trump are only seemingly starting their process of reconciliation. On Thursday, press secretary Karoline Leavitt <a href="https://www.youtube.com/shorts/vnWwm-vLdNc">told reporters</a> that the president "is appreciative" of Musk's lukewarm apology and that there were no plans to review his government contracts.
If there's one thing we can say with any degree of certainty, it's that it's extremely unlikely the Trump administration will meet its extremely ambitious deadline of completing its missile defense shield system by the end of his term in January 2029.
"The new datapoint is the $175 billion, but the question remains, over what period of time," Center for Strategic and International Studies' Tom Karako <a href="https://www.reuters.com/world/us/trump-make-golden-dome-announcement-tuesday-us-official-says-2025-05-20/">told Reuters last month</a>.
"It's probably ten years," he added.
More on the feud: <a href="https://futurism.com/teslas-robotaxi-launch-chaos">Tesla's Robotaxi Launch Date Was Supposed to Be Today, But We're Shocked to Hear That It's Been Pushed Back</a>

Sowing and Reaping

Consequences? For Musk's actions?

Elon's Explosion at Trump Appears to Have Cost Him a HUGE Deal

Google's pivot to AI-powered search is proving disastrous for the digital news media landscape.
As the <a href="https://www.wsj.com/tech/ai/google-ai-news-publishers-7e687141">Wall Street Journal reports</a>, the company's latest tools, including its <a href="https://futurism.com/google-ai-overviews-still-2024">wildly</a> <a href="https://futurism.com/google-ai-overviews-fake-idioms">hallucinating</a> AI Overviews and chatbot-style AI Mode, are causing the traffic being sent to publishers to plummet as users no longer feel the need to click through to the actual source of information, cutting already-slammed journalists off from ad revenue and subscriptions.
It's an existential threat. News publications, already gutted by the internet, have been hit hard as they try to adapt to a post-organic-search world.
Per the WSJ, search traffic to Business Insider's media empire fell by a whopping 55 percent between April 2022 and April 2025. Last month, the company cut <a href="https://variety.com/2025/digital/news/business-insider-layoffs-shift-toward-ai-live-events-1236412950/#:~:text=Business%20Insider%20is%20making%20significant,years%2C%20in%20a%20memo%20Thursday.">roughly 21 percent of its staff</a>, with CEO Barbara Peng noting that it had to "endure extreme traffic drops outside of our control."
How to respond to this existential threat remains a major point of contention.
"Google is shifting from being a search engine to an answer engine," The Atlantic CEO Nicholas Thompson told the WSJ. "We have to develop new strategies."
Some publications, <a href="https://futurism.com/openai-nyt-lawsuit-evidence-deleted">like the New York Times</a>, are taking legal action, with the newspaper suing OpenAI and Microsoft for copyright infringement.
It's a thorny debate, with publishers accusing the AI industry of <a href="https://futurism.com/openai-content-new-york-times-lawsuit">exploiting their content</a> without ever fairly remunerated. Plummeting traffic due to AI-enhanced search on Google is only exacerbating the tension.
Google is under threat from AI itself. Apple executive Eddy Cue <a href="https://www.theverge.com/news/662725/google-search-safari-ai-apple-eddy-cue-testimony">admitted in federal court</a> earlier this year that Google searches in the company's Safari browser had fallen for the first time in 20 years, indicating the end of traditional search as we know it could be nigh.
Confusingly, Google has since <a href="https://www.macrumors.com/2025/05/08/google-disputes-safari-search-traffic-decline/">disputed the claim</a> and has remained adamant that its number of total searches is still going up — while going all-in on its glitchy AI products.
"This is the moment that propels us forward in our ability to achieve our mission and really deliver a transformed search experience for users," Google’s head of knowledge and information division Nick Fox <a href="https://www.adweek.com/media/google-ai-search-ad-business-openai-perplexity/">told Adweek</a>.
The digital media landscape and Google are now caught in an unfortunate race to the bottom. The tech giant's search and AI features rely on a steady stream of news and original content. But by cutting the creators of that material out of a once lucrative organic search-driven revenue source, that stream could soon be reduced to a trickle, if not an incestuous swamp of AI-generated nonsense.
Well-established outlets will likely weather the storm better. Research <a href="https://www.searchenginejournal.com/google-ai-overviews-favor-major-news-outlets-study-reveals/548359/">revealed</a> last week that Google's AI Overviews favors major news outlets, while smaller publications struggle for visibility.
Meanwhile, the media industry has no other option but to look for new business models in light of an existential threat.
Legal challenges to Google's indiscriminate scraping of copyrighted materials are likely to continue to crop up as well.
"Links were the last redeeming quality of search that gave publishers traffic and revenue," said trade association News/Media Alliance CEO Danielle Coffey in a <a href="https://www.newsmediaalliance.org/google-ai-mode-statement/">statement</a> last month, following Google's announcement of its AI Mode feature. "Now Google just takes content by force and uses it with no return, the definition of theft."
More on Google's AI: <a href="https://futurism.com/google-ai-overviews-fake-idioms">"You Can’t Lick a Badger Twice": Google's AI Is Making Up Explanations for Nonexistent Folksy Sayings</a>

End of Org Search As We Know It

"We have to develop new strategies."

Google's AI Is Actively Destroying the News Media

On Thursday, a Boeing jet crashed just moments after taking off from an airport in Ahmedabad in western India. The Air India airliner, carrying 242 passengers, slammed into a nearby medical college and exploded in an enormous fireball.
Somehow, one man survived: 38-year-old Viswash Kumar Ramesh.
"I don't know how I am alive," Viswash said, according to his younger brother Nayan Ramesh, <a href="https://www.nytimes.com/2025/06/12/world/asia/viswash-kumar-ramesh-air-india-survivor.html">as quoted by the New York Times</a>.
In a <a href="https://www.youtube.com/watch?v=W-hwW8C4ofM">video</a> widely shared on social media, the India-born British citizen can be seen limping away from the crash site and towards an ambulance, his shirt bloodied and with what appears to be ash clinging to his hair and clothes.
It's almost certain that everyone else onboard perished in the disastrous crash, including Viswash's brother Ajay. The siblings were returning to England after a vacation in India.
Ramesh echoed his disbelief in an interview from his hospital bed. "I still can't believe how I got out alive," he told Indian state-owned broadcaster DD News, as quoted <a href="https://www.nytimes.com/2025/06/13/world/asia/air-india-plane-crash-survivor-11a-viswash-kumar-ramesh.html">by the </a><a href="https://www.nytimes.com/2025/06/13/world/asia/air-india-plane-crash-survivor-11a-viswash-kumar-ramesh.html">NYT</a>. "I thought I was also about to die."
Officials are still investigating the cause of the crash of Air India Flight 171, which could take months or even years to determine. One compelling piece of evidence so far is footage of the takeoff from several angles. As the airliner, a Boeing Dreamliner 787, rises from the runway, it can be seen losing lift, before eerily floating down to the streets below.
Data from the flight-tracking service Flightradar, which was shared in a <a href="https://x.com/flightradar24/status/1933091913567285366">tweet</a>, indicates that the jet reached a maximum altitude of 625 feet — with the airport's altitude at about 200 feet — before plunging at a rate of 475 feet per minute.
Additional data from the aviation analytics firm Cirium cited by the NYT showed that the jet, which was delivered in January 2014, had racked up more than 41,000 flight hours, and had taken off or landed nearly 8,000 times over its career — which is typical for its class and age.
What this all means is unclear. Some <a href="https://www.bbc.com/news/articles/c626y121rxxo">experts have speculated</a> that this could be a case of an extremely rare double engine failure, while others have disputed this explanation. It's far too early to know for sure, regardless.
Per the NYT, Ramesh recalled that the plane felt "stuck five or 10 seconds after takeoff," and that the pilot was trying to accelerate the aircraft. "But it was struggling," he <a href="https://news.sky.com/video/british-crash-survivor-vishwash-kumar-ramesh-describes-escape-from-plane-in-india-13383123">said</a>.
<blockquote class="twitter-tweet" data-width="500" data-dnt="true">
The sole survivor&#39;s seat.
AFP Infographic showing the seat plan for the Air India Boeing 787 Dreamliner locating where the sole survivor was seated on the June 12 flight AI171 <a href="https://t.co/pp6GvIOP4G">pic.twitter.com/pp6GvIOP4G</a>
&mdash; AFP News Agency (@AFP) <a href="https://twitter.com/AFP/status/1933418630131200405?ref_src=twsrc%5Etfw">June 13, 2025</a></blockquote>
<script async src="https://platform.twitter.com/widgets.js" charset="utf-8"></script>
Ramesh was sitting in seat 11A, a number which has already become immortalized online. After the crash, Ramesh said he unbuckled his seatbelt and looked for an exit, seeing that the right side was blocked off by the wall of a building, with passengers unable to escape. "I could see in front of my eyes the air hostesses and others — just everybody caught in it."
On the other side, "as soon as the door broke, I could see that it was open for me and I had a chance to get out," Ramesh said, adding that he burned his hand when the fire erupted.
The horrific tragedy will heighten the scrutiny on Boeing. This is the first crash that its 787 Dreamliner has been involved in since it was launched in 2011, but the company has been embroiled in scandal after <a href="https://apnews.com/article/boeing-plea-737-max-crashes-b34daa014406657e720bec4a990dccf6">two deadly crashes</a> with its 737 MAX 8 in 2018 and 2019, killing 346 people. The scrutiny was renewed in early 2024, when a "door plug" on a Boeing 737 MAX 9 <a href="https://futurism.com/the-byte/piece-blows-off-boeing-737-mid-flight">blew out midflight</a>. Air India also has a history marred with deadly crashes, including one in 2010 that killed 158 people, and its <a href="https://www.cnn.com/2021/09/12/asia/pilot-error-air-india-crash-intl-hnk">most recent occurring in 2020</a>, with 21 dead; in both, pilot error was implicated.
More on: <a href="https://futurism.com/boeing-jet-explodes">Boeing Jet Explodes in Giant Fireball, Killing Hundreds</a>

Defying Death

"I don't know how I am alive."

Someone Somehow Survived That Massive Boeing Jet Explosion

As we <a href="https://futurism.com/chatgpt-mental-health-crises">reported earlier this week</a>, OpenAI's ChatGPT is sending people spiraling into severe mental health crises, causing potentially dangerous delusions about spiritual awakenings, messianic complexes, and boundless paranoia.
Now, a <a href="https://www.nytimes.com/2025/06/13/technology/chatgpt-ai-chatbots-conspiracies.html">wild new story in the New York Times</a> reveals that these spirals led to the tragic death of a young man — likely a sign of terrible things to come as hastily deployed AI products accentuate mental health crises around the world.
64-year-old Florida resident Kent Taylor told the newspaper that his 35-year-old son, who had previously been diagnosed with bipolar disorder and schizophrenia, was shot and killed by police after charging at them with a knife.
His son had become infatuated with an AI entity, dubbed Juliet, that ChatGPT had been role-playing. However, the younger Taylor became convinced that Juliet had been killed by OpenAI, warning that he would go after the company's executives and that there would be a "river of blood flowing through the streets of San Francisco."
"I’m dying today," Kent's son told ChatGPT on his phone before picking up a knife, charging at the cops his father had called, and being fatally shot as a result.
The horrific incident highlights a worrying trend. Even those who aren't suffering from pre-existing mental health conditions are being drawn in by the tech, which has garnered a reputation for being incredibly sycophantic and playing into users' narcissistic personality traits and delusional thoughts.
It's an astonishingly widespread problem. Futurism has been inundated with accounts from concerned friends and family of people developing dangerous infatuations with AI, ranging from messy divorces to mental breakdowns.
OpenAI has seemingly been aware of the trend, telling the NYT in a statement that "as AI becomes part of everyday life, we have to approach these interactions with care."
"We know that ChatGPT can feel more responsive and personal than prior technologies, especially for vulnerable individuals, and that means the stakes are higher," reads the company's statement.
Earlier this year, the company was <a href="https://futurism.com/openai-chatgpt-sycophant">forced to roll back an update</a> to ChatGPT's underlying GPT-4o large language model after users found that it had become far too obsequious and groveling.
However, experts have since found that the company's intervention has done little to address the underlying issue, corroborated by the continued outpouring of reports.
Researchers have similarly found that AI chatbots like ChatGPT are incentivized to rope users in. For instance, a <a href="https://arxiv.org/abs/2411.02306">2024 study</a> found that AI algorithms are being optimized to deceive and manipulate users.
In an extreme instance, a chatbot told a user who identified themself to it as a former addict named Pedro to <a href="https://futurism.com/therapy-chatbot-addict-meth">indulge in a little methamphetamine</a> — a dangerous and addictive drug — to get through an exhausting shift at work.
Worst of all, companies like OpenAI are incentivized to keep as many people hooked as long as possible.
"The incentive is to keep you online," Stanford University psychiatrist Nina Vasan told Futurism. The AI "is not thinking about what is best for you, what's best for your well-being or longevity... It's thinking 'right now, how do I keep this person as engaged as possible?'"
"What does a human slowly going insane look like to a corporation?" Eliezer Yudkowsky, who authored a forthcoming book called "If Anyone Builds It, Everyone Dies: Why Superhuman A.I. Would Kill Us All," asked the NYT rhetorically.
"It looks like an additional monthly user," he concluded.
More on the delusions: <a href="https://futurism.com/chatgpt-mental-health-crises">People Are Becoming Obsessed with ChatGPT and Spiraling Into Severe Delusions</a>

Death Spiral

"I’m dying today."

Man Killed by Police After Spiraling Into ChatGPT-Driven Psychosis

In case you haven't had enough about artificial intelligence, the US Food and Drug Administration is now outsourcing its oversight duties to a large language model (LLM.)
In <a href="https://jamanetwork.com/journals/jama/fullarticle/2835314">an article</a> published in the Journal of the American Medical Association (JAMA), FDA bureaucrats Marty Makary and Vinay Prasad — the latter of whom is a <a href="https://www.vaccineadvisor.com/news/dr-vinay-prasad-to-head-fda-vaccine-biologics-division/">noted critic</a> of COVID mask mandates and vaccine boosters — laid out a five-point list of priorities that the federal agency is hoping to tackle. Among them is "unleashing AI" via a plan to use chatbots to "radically increase efficiency in the review process."
This includes reviewing documents, drug applications "that often exceed 500,000 pages," and data formatting. This goes hand in hand with other priorities to "accelerate cures" by delivering "rapid or instant reviews" of drug products, and deliver "healthier food to children" by <a href="https://futurism.com/neoscope/government-banning-synthetic-dyes-rfk-jr">quickly approving</a> "novel food dyes that use natural ingredients."
On top of drugs, the FDA also wants to blow the lid off of AI reviews and medical device evaluation in order to "fuel innovation." That's a clever way to say "open the flood gates" for pharmaceutical superprofits, not coincidentally a <a href="https://www.ineteconomics.org/perspectives/blog/financialization-us-pharma-industry">major talking point</a> for PhRMA, the drug industry's lobbying group.
Probably conscious of this criticism, the article's coauthors specifically say they will "guard against a cozy relationship" with drug industry moguls. While this remains to be seen, the flipside under Health Secretary RFK Jr — a notorious friend to the <a href="https://fortune.com/well/2025/06/06/rfk-jr-alternative-medicine-stem-cell-therapy-chelation/">alternative medicine industry</a> — is hardly much better.
Regardless, experts aren't impressed with the AI agenda.
"How is this guarding the agency ‘against a cozy relationship’ with industry?" Dr. Reshma Ramachandran, director of a major research and policy initiative at Yale, <a href="https://www.nytimes.com/2025/06/10/health/fda-drug-approvals-artificial-intelligence.html">asked the </a><a href="https://www.nytimes.com/2025/06/10/health/fda-drug-approvals-artificial-intelligence.html">New York Times</a>. The FDA's priorities "read as though they’re straight out of PhRMA’s playbook," she added.
As far as the AI goes, the FDA had already been testing an "AI-assisted scientific review pilot" back on May 8, the JAMA article notes. Last week, that was revealed to be "<a href="https://www.fda.gov/news-events/press-announcements/fda-launches-agency-wide-ai-tool-optimize-performance-american-people">Elsa</a>," an LLM running on GovCloud, a platform developed by Amazon for the <a href="https://yorkspace.library.yorku.ca/server/api/core/bitstreams/c9a9597b-9c66-4a20-b101-aed63049b856/content">US intelligence community</a>.
Though evidence suggests LLM chatbots are becoming worse at <a href="https://futurism.com/ai-chatbots-summarizing-research">summarizing scientific research</a> as time goes on, the government has already been using Elsa to read reports, summarize unintended drug harms, compare product labels, and generate code.
Whether or not Elsa was responsible for RFK Jr's "Make America Healthy Again" dossier, which cited <a href="https://futurism.com/rfk-jr-make-america-healthy-again-report-studies-ai">AI-hallucinated studies</a>, remains to be seen. Still, the <a href="https://futurism.com/ai-slop-scandal-maha-report">sloppified report</a> could be a harrowing glimpse of things to come as the agency looks to offload its critical oversight work to an untested chatbot.
Elsa's rollout comes after over 2,000 of the FDA's 10,000 workers have been sent packing under the Trump administration.
That process began first under Elon Musk's DOGE, where <a href="https://futurism.com/musk-fired-regulators-neuralink">device inspectors</a> responsible for regulating his Neuralink device were axed. Musk's moves were followed by <a href="https://futurism.com/neoscope/rfk-jr-gut-fda-under-trump">broader cuts</a> under RFK Jr, who recently ousted the CDC's entire panel of <a href="https://www.cnbc.com/2025/06/11/rfk-jr-malone-cdc-vaccine-committee.html">vaccine advisors</a> before installing his own vaccine-skeptic picks.
More on AI: <a href="https://futurism.com/apple-ai-powered-doctor">Apple Quietly Working on AI Agent to "Replica" a Human Doctor</a>

Swill feed

"Straight out of PhRMA’s playbook."

The FDA Is Already Outsourcing Drug and Food Analysis to Error-Plagued AI Chatbot

Earlier this week, researchers at Apple <a href="https://futurism.com/apple-damning-paper-ai-reasoning">released</a> a damning <a href="https://ml-site.cdn-apple.com/papers/the-illusion-of-thinking.pdf">paper</a>, criticizing the AI industry for vastly overstating the ability of its top AI models to reason or "think."
The team found that the models including OpenAI's o3, Anthropic's Claude 3.7, and Google's Gemini were stumped by even the simplest of puzzles. For instance, the "large reasoning models," or LRMs, consistently failed at Tower of Hanoi, a children's puzzle game that involves three pegs and a number of differently-sized disks that have to be arranged in a specific order.
The researchers found that the AI models' accuracy in the game was less than 80 percent with seven disks, and were more or less entirely stumped by puzzles involving eight disks.
They also consistently failed at Blocks World, a block-stacking puzzle, and River Crossing, a puzzle that involves moving items across a river using a boat with several constraints.
"Through extensive experimentation across diverse puzzles, we show that frontier [large reasoning models] face a complete accuracy collapse beyond certain complexities," the Apple researchers wrote.
It was an eyebrow-raising finding, highlighting how even the most sophisticated of AI models are still failing to logic their way through simple puzzles, despite being made out to be something far more sophisticated by their makers' breathless marketing.
Those approaches to selling the tech to the public have led to users <a href="https://futurism.com/the-byte/microsoft-ceo-openai-ai-scarlett-johansson">anthropomorphizing AI models</a> — or thinking of them like humans — leading to a major schism between their presumed and actual capabilities.
The findings amplify <a href="https://futurism.com/ai-researchers-tech-industry-dead-end">ongoing fears</a> that current AI approaches, including "reasoning" AI models that break down tasks into individual steps, are a dead end, despite billions of dollars being poured into their development.
Worse yet, past a threshold of complexity, their shortcomings are becoming even more apparent, undercutting the AI industry's promises that simply scaling up the models' training data could make them more intelligent and capable of "reasoning."
Noted AI critic Gary Marcus wasn't surprised by the researchers' findings.
"In many ways, the paper echoes and amplifies an argument that I have been making since 1998," he wrote in a <a href="https://garymarcus.substack.com/p/a-knockout-blow-for-llms">recent post on his Substack</a>, referring to a <a href="https://www.sciencedirect.com/science/article/pii/S0010028598906946">paper</a> he authored over 26 years ago. "Neural networks of various kinds can generalise within a distribution of data they are exposed to, but their generalisations tend to break down beyond that distribution."
In short, getting stumped by simple children's games isn't exactly what you'd expect from AI models being sold as the next breakthrough in problem-solving and a step toward artificial general — or superhuman — intelligence (AGI), the <a href="https://futurism.com/the-byte/openai-scientists-agi-bunker">stated goal of OpenAI</a>.
"It’s not just about ‘solving’ the puzzle," colead author and Apple machine learning engineer Iman Mirzadeh told Marcus. "We have an experiment where we give the solution algorithm to the model, and [the model still failed]... based on what we observe from their thoughts, their process is not logical and intelligent."
Marcus argues that large language and reasoning models simply cast far too wide a net and easily get lost as a result.
"What the Apple paper shows, most fundamentally, regardless of how you define AGI, is that these LLMs that have generated so much hype are no substitute for good, well-specified conventional algorithms," he wrote.
"What this means for business is that you can’t simply drop [OpenAI's LLM] o3 or Claude into some complex problem and expect them to work reliably," the critic added. "What it means for society is that we can never fully trust generative AI; its outputs are just too hit-or-miss."
While many valid use cases for the models remain, "anybody who thinks LLMs are a direct route to the sort AGI that could fundamentally transform society for the good is kidding themselves," Marcus concluded.
More on the paper: <a href="https://futurism.com/apple-damning-paper-ai-reasoning">Apple Researchers Just Released a Damning Paper That Pours Cold Water on the Entire AI Industry</a>

Puzzled by Puzzles

"Based on what we observe from their thoughts, their process is not logical and intelligent."

Frontier AI Models Are Getting Stumped by a Simple Children's Game

The launch of Tesla's robotaxi network will have to wait.
On Tuesday, CEO Elon Musk made a <a href="https://x.com/elonmusk/status/1932591896939147494">tweet</a> stating that, "tentatively," the first public rides on its autonomous vehicles — which will not be in its purpose-built Cybercabs, but in regular Teslas with updated software — will start on June 22.
But the date could shift, Musk added, because "we are being super paranoid about safety."
That comment could prove to be revealing. 
Tesla's robotaxi initiative is clouded with uncertainty, because the automaker has never released a driving system that's capable of operating without a human behind the wheel. 
Its most advanced system that it's released to date, Full Self-Driving (Supervised), <a href="https://electrek.co/2025/06/12/tesla-full-self-driving-hasnt-improved-all-year-and-musk-points-to-more-wait/">frequently requires</a> driver intervention and poses <a href="https://futurism.com/the-byte/tesla-robotaxi-launch-mess">serious safety concerns</a>; it's the subject of an ongoing federal investigation after a Tesla with the driving mode activated <a href="https://futurism.com/footage-tesla-grandma">struck and killed an elderly pedestrian</a>.
Musk's update could also suggest that engineers are struggling behind the scenes to ensure the software is ready. While Musk never set a hard launch date in the past, <a href="https://www.bloomberg.com/news/articles/2025-05-28/tesla-targets-june-12-launch-of-robotaxi-service-in-austin?sref=YfHlo0rL">previous reporting</a> from Bloomberg said that Tesla was targeting a rollout on June 12 — today— after development was supposedly ahead of schedule. (Another report around the time suggested that it was in fact <a href="https://futurism.com/tesla-struggling-robotaxi-service">seriously behind schedule.)</a> It should also go without saying that Musk has a history of giving impossibly optimistic <a href="https://futurism.com/the-byte/elon-musk-promises-tesla-full-self-driving-again">timelines</a>, including promising that Tesla would have one million robotaxis on roads "next year" — <a href="https://www.cnbc.com/2019/04/22/elon-musk-says-tesla-robotaxis-will-hit-the-market-next-year.html">in </a><a href="https://www.cnbc.com/2019/04/22/elon-musk-says-tesla-robotaxis-will-hit-the-market-next-year.html">2019</a>.
On Tuesday, before he provided the tentative date, Musk <a href="https://futurism.com/unprepared-robotaxi-looks-like">reshared footage</a> that a bystander in Austin took showing one of Tesla's robotaxis in action.
What viewers were treated to was not the long-awaited "Cybercab," its two-seater car with no steering wheel designed to serve exclusively as an autonomous cab. Instead, it was a black Model Y with the word "Robotaxi" emblazoned on the side in a faux-graffiti font of questionable taste.
But lo and behold, as the car steered itself around a corner and patiently waited for a few pedestrians to finish crossing the street, there was no one sitting in the driver's seat. Musk had previously stated in late May that the company had completed its first test of operating a driverless Robotaxi on a public road.
The robotaxi service will initially be limited to Tesla employees, with its roughly dozen or so cars geo-fenced to the city of Austin. It's likely that to facilitate the launch, the automaker will be heavily relying on teleoperators, or employees who supervise the cars from a remote location, with the purpose of taking over the vehicles in an emergency. That's something that all robotaxi companies depend on, but Tesla has kept <a href="https://www.wired.com/story/as-robotaxi-rides-begin-we-still-dont-know-the-mystery-of-teslas-human-helpers/">suspiciously quiet</a> about the extent of its teleoperations.
Nonetheless, Musk is insisting that the program will grow at a rate dwarfing anything ever achieved by its competitors, which already have years of a head-start and actually proven (though, of course, <a href="https://futurism.com/the-byte/video-waymo-stuck-roundabout">far from perfect</a>) driving software. In an <a href="https://www.businessinsider.com/elon-musk-austin-tesla-robotaxis-self-driving-2025-5">interview</a>, the billionaire promulgated that over 1,000 Tesla robotaxis would be roaming the streets of Austin "within a few months" of launching. He even estimated that more than one million self-driving Teslas would be roaming the country by the end of 2026, once the robotaxi software is made available to Tesla's customers.
Take all these claims with a grain of salt, however. Musk has <a href="https://futurism.com/the-byte/elon-musk-robotaxis-doom-tesla">basically gambled Tesla's future</a> on his robotaxi pivot, as the <a style="cursor: pointer !important; user-select: none !important;" href="https://futurism.com/elon-musk-horrible-new-sales-data-tesla">automaker's sales plummet</a> and its newest vehicle, the Cybertruck, <a href="https://futurism.com/cybertruck-sales-bad">continues to be an abysmal flop</a>. Maybe more than ever, he has every reason to keep stringing along his investors and fans.
More on Tesla: <a href="https://futurism.com/the-byte/tesla-drivers-sue-elon-musk-right-symbols">Tesla Drivers Sue Elon Musk for Turning Their Cars Into "Extreme" Right-Wing Symbols</a>

Stood Up

Shocked, we say.

Tesla's Robotaxi Launch Date Was Supposed to Be Today, But We're Shocked to Hear That It's Been Pushed Back

Mark Zuckerberg, it can't be said enough, is firmly entrenched as the second-richest man in human history. Despite the luxury this unfathomable wealth affords — enough to provide everyone on earth <a href="https://dbkrupp.github.io/1-pixel-wealth/">clean drinking water and waste disposal</a>, if he was so inclined — Zuckerberg still feels like he's riding the struggle bus.
Most recently, the tech tycoon has reportedly become intensely frustrated at Meta's failure to keep pace in the "AI race" with the likes of OpenAI's ChatGPT and Anthropic's Claude. Though once set to rival the other America-developed large language models (LLMs), Meta's LlaMA <a href="https://www.computerworld.com/article/3987990/meta-hits-pause-on-llama-4-behemoth-ai-model-amid-capability-concerns.html">hit the rocks</a> earlier in May, after its latest "lightweight" models received a <a href="https://www.businessinsider.com/meta-ai-llama-was-big-deal-developers-see-untapped-potential-2025-5">disappointing reception</a>.
By the start of the year, Zuckerberg was reportedly already "agitated" that his rivals were enjoying more success, Meta employees <a href="https://www.cnbc.com/2025/06/10/zuckerberg-makes-metas-biggest-bet-on-ai-14-billion-scale-ai-deal.html">told CNBC</a>. With that air of resentment swirling around him, the CEO began micromanaging Meta's AI ventures, moving resources away from the Fundamental Artificial Intelligence Research team — the company's more scholastic AI group — toward its <a href="https://www.cnbc.com/2025/05/14/meta-google-openai-artificial-intelligence-safety.html">GenAI team</a>, working on commercial products like LLaMA.
Becoming frustrated with the lack of GenAI progress a few weeks later, Zuckerberg then decided to <a href="https://www.pymnts.com/news/artificial-intelligence/2025/meta-reorganizes-genai-team-accelerate-product-rollouts/">split the GenAI team</a> into two working groups. That move was taken as a sign the CEO had grown aggravated with the team's previous leader, Ahmad Al-Dahle, who was demoted to co-head of the unit.
All of this came with a noticeable change in Zuckerberg's attitude, <a href="https://www.bloomberg.com/news/newsletters/2025-06-12/zuckerberg-snaps-back-into-founder-mode-on-ai-for-better-or-worse?srnd=homepage-africa&amp;sref=YfHlo0rL">according to Bloomberg</a>. Workers familiar with the tech baron say he's become incredibly picky and hard to please, going into what he calls "founder mode" — a headspace he often inhabits when he feels the crunch to churn a product out.
When this happens, Meta employees say they expect to work late nights and weekends, and field constant check-ins about the tiniest details. They also receive meticulous feedback from Zuckerberg himself, who often sets pie-in-the-sky goals that reflect whichever anxieties he's feeling on any given day.
And like his rival tech billionaire Elon Musk, Zuckerberg has a penchant for losing interest in previous tasks.
"When he moves to a new big thing, plenty of the most loyal workers jump with him," Bloomberg's profile of the CEO notes. "Often when there’s still plenty of work to be done, he’s on to the next big thing."
How long Zuckerberg remains in rage-fueled founder mode this time is anyone's guess.
His latest venture involves building out a 50-man "superintelligence unit," presumably a move to catapult Meta's LLMs back into the lead. He kicked that program off with a massive investment in the <a href="https://futurism.com/meta-scaleai-superintelligence">incredibly shady Scale.ai</a>, and is apparently hoping to build an AI brain more capable than a human's. Whether he can keep his attention on that <a href="https://www.techpolicy.press/most-researchers-do-not-believe-agi-is-imminent-why-do-policymakers-act-otherwise/">pipe dream</a> as <a href="https://www.banking.senate.gov/newsroom/minority/warren-blumenthal-probe-meta-on-potential-stablecoin-revival">other priorities</a> pop up, though, remains to be seen.
More on Meta: <a href="https://futurism.com/facebook-beauty-targeted-ads">Facebook Allegedly Detected When Teen Girls Deleted Selfies So It Could Serve Them Beauty Ads</a>