Embattled CEO caught asking ChatGPT for corporate takeover plan — against lawyer's advice



The future is here, and it seemingly includes CEOs using chatbots to create plans to avoid having to pay out hundreds of millions of dollars.

That was a judge's conclusion after a smaller American studio sued a giant, publicly traded South Korean conglomerate that allegedly prevented it from putting out its product.

'Lock down Steam/console publishing rights and access rights.'

Krafton CEO Kim Chang-han handles nearly $2 billion of revenue across a multitude of companies, which includes PubG Studios, a massively popular online shooter game.

Since 2021, Krafton has controlled Unknown Worlds, an American studio responsible for the game Subnautica, which sold over five million copies in two years.

With so much success from the first game, Krafton agreed to a $250 million earnout if Subnautica 2 was able to meet specific sales targets. Krafton's CEO was not keen on letting that happen and subsequently plotted "Project X," a plan to prevent the payout.

After internal reports projected Subnautica 2 was likely to hit its targets, things got hairy. According to court documents, when Krafton’s Head of Corporate Development Maria Park warned CEO Kim that removing Unknown Worlds' leadership via "dismissal with cause" opened them up to "lawsuit and reputational risk," he turned to ChatGPT for help.

The chatbot told Kim that the earnout would be "difficult to cancel" but suggested forming an internal task force to either negotiate a "deal" or execute a "takeover" of the company; Kim obliged and allegedly continued to follow ChatGPT's suggestions.

RELATED: Anthropic says its own new model is too dangerous for the public — but not these Big Tech companies

Not only did Kim allegedly share his strategies from ChatGPT with colleagues, but the strategies included a "pressure and leverage package" against Unknown Worlds.

Among its recommendations, ChatGPT suggested Krafton undermine any David versus Goliath narratives, while urging Kim to prepare for scenarios like buyouts and replacements.

Most jarringly, it also suggested locking down Unknown Worlds' ability to post its new game for sale on Steam, the largest gaming distributor for PC games.

"Lock down Steam/console publishing rights and access rights over code/build pipeline through both legal and technical aspects," ChatGPT said, the lawsuit revealed. "For the earn-out freeze, keep room for negotiations through provision stating 'immediate removal if specific development results are achieved.'"

Kim did as the chatbot recommended and locked down the publishing, and Subnautica 2 could not be released. When Unknown Worlds CEO Ted Gill asked for control to be returned, Kim allegedly ignored him and told a Krafton studio rep to relay to Gill that he had "no intention of transferring stuff back to you guys (like the Steam app)."

RELATED: Does this stealthy startup hold the key to keeping data centers out of your neighborhood?

Ina FASSBENDER/AFP/Getty Images

While Gamesradar reported that Krafton leadership admitted to using ChatGPT for "faster answers," the company told Kotaku that some characterizations made about them have been false.

In response to claims from Unknown Worlds that Krafton said its chat logs no longer exist, the company said the claim was "simply a distraction from their own efforts to destroy evidence."

In the end, a Delaware judge ruled that Kim relied on ChatGPT to craft a strategy aimed at avoiding the $250 million payment.

"Fearing he had agreed to a 'pushover' contract, KRAFTON’s CEO consulted an artificial intelligence chatbot to contrive a corporate 'takeover' strategy," Vice Chancellor Lori Will said in her ruling, per Economic Times.

The court maintained that Krafton was expected to exercise independent judgment and not outsource its decisions to AI systems.

PC Gamer has since reported that Unknown Worlds will be given an extension to reach its earnout goals to mid-September, with the possibility of extending to March 2027.

The game is set for early release in May 2026.

Like Blaze News? Bypass the censors, sign up for our newsletters, and get stories like this direct to your inbox. Sign up here!

Anthropic says its own new model is too dangerous for the public — but not these Big Tech companies



Anthropic is sending out a warning that its artificial intelligence model is sophisticated enough to undo decades of research.

The company operates Claude, the AI chatbot that has been ripped off and turned into a free, public model, and is hoping to get together with a consortium of tech companies to button up the security measures ahead of its release.

'It has found vulnerabilities, and in some cases crafted exploits.'

Anthropic's Mythos model of Claude AI will only be available to 40 select companies to be used for the power of good, the company claims.

It represents "the starting point for what we think will be an industry change point, or reckoning, with what needs to happen now," said Logan Graham, head of Anthropic's vulnerability testing team.

The company fears that its new AI model is so good at finding cracks in cybersecurity that it must only be shared with companies it deems capable and responsible enough to prepare for possible attacks when Mythos goes public.

"This model is good at finding vulnerabilities that would be well understood and findable by security researchers," Graham said. "At the same time, it has found vulnerabilities, and in some cases crafted exploits, sophisticated enough that they were both missed by literally decades of security researchers, as well as all the automated tools designed to find them."

RELATED: How to power the AI race without losing control

Samyukta Lakshmi/Bloomberg/Getty Images

Anthropic will reportedly commit up to $100 million in credits for the project, meaning the amount of money it would typically charge for such a volume of its chatbot's usage.

Labeled Project Glasswing, the initiative to shore up cybersecurity will grant Mythos access to handpicked companies chosen largely from Big Tech like Amazon, Apple, Google, and Microsoft. The group is rounded out by internet infrastructure and cybersecurity giants like Broadcom, Cisco, CrowdStrike, Nvidia, and Palo Alto Networks, along with financial titan JPMorgan Chase and key open-source nonprofit the Linux Foundation.

This is not the first time an AI company has warned its product is too dangerous for the public, and looking back, readers can gauge whether or not Claude may be as dangerous as its creators purport it to be.

In 2019, OpenAI sent out a warning ahead of its release of GPT-2, claiming that its capabilities — now vastly eclipsed by later models — could be used to mass-produce propaganda or misleading text.

As Wired reported at the time, OpenAI said GPT-2 was too risky to be released to the general public.

RELATED: Claude, Anthropic's AI assistant, slammed by Elon Musk for anti-white responses to simple prompts

Claude has been in the news for alleged missteps, leaks, and accidental postings throughout the past year, and while it may not be a household name yet, it has raced its way through the tech sector as a go-to for "agentic" work building software, apps, and even companies.

In addition to its model being open-sourced and used by the general public for free, the company has been noted for "accidental" postings of its own code.

Anthropic "accidentally uploaded a file to a public repository that's just meant to help developers understand how to use their product" and "exposed some of the source code of Claude," reporter Aaron Holmes explained recently.

Proprietary information was further leaked in another alleged accidental posting, this time through a blog draft that revealed "internal source code."

The company seems poised for consistent marketing battles, both willing and unwilling, from its high-stakes lawsuit against the federal government labeling it a supply chain risk to the blowback it has received from putting a woman closely linked to the cultish Effective Altruism movement in charge of its AI's "Constitution."

Like Blaze News? Bypass the censors, sign up for our newsletters, and get stories like this direct to your inbox. Sign up here!

Sexting with chatbots is too far, OpenAI decides



Just days after announcing it would be shutting down its artificial intelligence video generation platform, OpenAI put the brakes on another project.

While the terminology remains vague, it seems Sam Altman's company could be drawing a line as to what it deems "adult" content.

'We still believe in the principle of treating adults like adults.'

Those familiar with the adult-themed project at OpenAI have "indefinitely" shelved their plans to release an erotic chatbot, per the Financial Times. OpenAI confirmed that before moving forward with such a product, the company wanted to be able to fall back on long-term research about the effects AI sex chats have on users and any emotional attachments that might be created.

OpenAI said there is no "empirical evidence" available at this time.

RELATED: Sam Altman tells BlackRock he wants AI on a meter 'like electricity or water'

CHARLY TRIBALLEAU/AFP/Getty Images

Last year, Altman announced that ChatGPT would start including more content, including erotica, to "treat adult users like adults."

But in early March, OpenAI made its first announcement that "adult mode" was being delayed. That decision was made in part to focus on more pertinent tasks. "We're pushing out the launch of adult mode so we can focus on work that is a higher priority for more users right now," a spokesperson told reporter Alex Heath, "including gains in intelligence, personality improvements, personalization, and making the experience more proactive."

"We still believe in the principle of treating adults like adults, but getting the experience right will take more time," the company stated.

Inside sources since told the Financial Times that the company will refocus on core products after staff and investors expressed concern about the sexualized AI content. The upside to this endeavor was allegedly too small for OpenAI.

RELATED: Sam Altman says NSA can't use OpenAI — then tells staff they don't have a say in military actions

Jaap Arriens/NurPhoto/Getty Images

The revelations follow hot on the heels of other strategy-shifting announcements. The tech giant has recently tightened up its offerings, shuttering generative AI service Sora.

"What you made with Sora mattered, and we know this news is disappointing," the company wrote on X. "We'll share more soon, including timelines for the app and API and details on preserving your work."

Like Blaze News? Bypass the censors, sign up for our newsletters, and get stories like this direct to your inbox. Sign up here!

Suspect consults ChatGPT after brother allegedly plants bomb at US Air Force base



One-half of the sibling pair charged in connection with an IED discovered at MacDill Air Force Base in Tampa allegedly consulted an AI bot to help the other sibling flee the country.

Alen is believed to still be in China.

After Alen Zheng, 20, allegedly planted the bomb at the base visitor center last month, his sister Ann Mary Zheng, 27, allegedly used ChatGPT to help Alen escape to China. Federal prosecutors claim that she asked the bot:

  • how to obtain a Chinese visa,
  • how they might transfer ownership of some of Alen's belongings to her, and
  • to find schools in China that Alen might be able to attend.

Ann Mary is accused of helping Alen cover his tracks and then evade capture. She has been charged with evidence tampering and assisting after the fact and faces up to 30 years if convicted. She appeared in court on Tuesday regarding possible pretrial release, though the judge has not yet issued a ruling.

RELATED: China caught 'trying to disrupt our justice system': DOJ accuses 10 Chinese spies and communist agents of 'malign schemes'

WD Stuart/Getty Images

A 911 call to report the bomb came in on March 11, but investigators found nothing in their initial search of the base. An IED was later discovered on March 16. The device never detonated, but officials have described it as "viable" and "potentially very deadly."

Alen and Ann Mary Zheng bought plane tickets to China on March 11 and flew there on March 12. For reasons unknown, Ann Mary returned to the U.S. on March 17. Alen is believed to still be in China.
Though he remains at large, Alen has been charged with attempted damage of government property by fire or explosion, unlawful making of a destructive device, and possession of an unregistered destructive device. If convicted, he could spend 40 years in federal prison.
Alen and Ann Mary Zheng are U.S. citizens, but their parents, Qiu Qin Zou and Jia Zhang Zheng, are not. According to the Tampa Bay Times, the Chinese natives applied for asylum in the U.S. "years ago" but were denied. They were detained shortly after the IED was discovered and now face deportation.
At a press conference, U.S. Attorney for the Central District of Florida Gregory Kehoe claimed that while the siblings' mother was not currently charged with any crime, the possibility of future charges against her could not be precluded.
Like Blaze News? Bypass the censors, sign up for our newsletters, and get stories like this direct to your inbox. Sign up here!

A man used Grok to save his dog. Is intellectual property about to die?



Millions recently read about normal-guy Paul Conyngham’s resourcefulness when it was revealed he did what doctors couldn’t in creating an effective, customized vaccine for his dog stricken with terminal illness, but far fewer caught the later-revealed fact that while ChatGPT was credited as the AI model Conyngham used to navigate the labyrinth of mRNA vaccine creation, it was actually Grok that produced the final, winning design.

Perhaps “normal guy” is an understatement. Conyngham is an Australian tech entrepreneur. When his adopted dog Rosie was diagnosed with terminal cancer, he paid a lab $3,000 to perform DNA sequencing analysis on both Rosie and the precise cancer Rosie was fighting. Then, he used AI tools such as AlphaFold to process the sequencing analysis. Finally, he deployed Grok to design the bespoke mRNA vaccine, which was ultimately produced by university partners (evidently available for consult or perhaps inspired by Conyngham’s devotion to his dog).

What are the odds that this is all just going to spontaneously work out?

Despite his unusual skills and network, however, Conyngham didn’t go viral for those. Rather, his story resonated because his can-do sense of initiative is something anyone can tap into, with potentially lifesaving results. At the time of this writing, despite doctors’ predictions, Rosie the dog is alive and thriving. Her illness has not entirely abated, but her owner’s ingenuity and persistence, combined with his layman’s agility around LLMs, has reduced the most life-threatening tumors by 75%.

How then, from this straightforward set of events, did ChatGPT wind up taking the credit until the record was corrected weeks later? When I asked Grok (which, being made up of timelines, is pretty reliable in accessing and reassessing events), I got the rather noncommittal suggestion that the misattribution was due to institutional inertia.

Perhaps.

Hungry for more, I dug into a much deeper human analysis of the man-saves-dog episode. Jordan Hall, another tech entrepreneur-turned-philosopher, posted a series of viral X articles addressing the economic shift to a total, global AI underlayer to the economy (and thus, every aspect of human life). In his second installment, “The Great Transition: The Divine Economy,” Hall sketches his vision for a coherent implementation of AI into this overarching position of importance.

RELATED: Use an anonymous account online? AI can now reveal your identity.

Photo Credit Olga Novikova/Getty Images

Readers are strongly encouraged to read Hall’s series of articles in its entirety. It’s fascinating and endlessly ponderable. All told, in anticipation of a global upheaval of biblical proportions — yes, we’ve heard this for years; despite the wait, it’s coming — Hall suggests we’ll turn the wheel over to the Church.

“The Church has always been an economic institution,” he argues, “whether it acknowledged it or not. Mutual aid, vocational, formation, capital pooling, trust networks — these are ancient practices. What changes now is that AI collapses the constraints that made those practices uncompetitive against industrial-scale consolidation. On Earth as it is in Heaven.”

In the case of Rosie and her owner, just a few questions illustrate the complexity and potential for malfeasance in our AI age. Who owns the Grok-derived vaccine recipe? Who owns Rosie’s DNA? Can it be sold? Who should benefit? If DNA data is “scraped” in some manner similar to how novels, television shows, and musical recordings are more or less pilfered, what are the limits of DNA and data ownership, if any? Can it be simply destroyed, in the same way the owner of a patch of grass can burn it should he so desire?

Hall’s analysis implies that, in the end, these are spiritual questions that can only be answered spiritually — and people hungry for fast answers they can trust will turn to the place where such answers have been on offer for thousands of years.

For now, Rosie’s owner was able to slip through the cracks of institutional, veterinary, and judicial red tape using wit and, let’s face it, the collective human affection for dogs. Hall predicts a situation where the collective, decentralized power of human faculties — made hyper-potent via leveraging AI and functioning on the timeless spiritual foundation of the Church — robustly addresses the AI age’s vast issues of greed, misallocation, misuse, and abuse of resources. Restricted to the secular level, discussions about these problems almost always find themselves mired in the dialectic between Marx and Smith, communism versus capitalism. Unable to innovate our way out of the impasse, will our eyes turn at last to the divine economy?

If a few years pass, the AI compactor consolidating everything into data will likely squeeze out new, perhaps unimaginable forms of computational power. The fight to capture and control that power is raging right now. Looking at the brokers, politicians, and players, accounting for history and human nature, what are the odds that this is all just going to spontaneously work out — such that good-willed efforts like those of Conyngham continue freely, without surveillance or exploitation? We’ll soon see if we’re willing to adopt the forms of social organization it takes to keep cyberspace so free, open, and fruitful.

Sam Altman tells BlackRock he wants AI on a meter 'like electricity or water'



OpenAI CEO Sam Altman has likened artificial intelligence to utilities that are required to live.

Altman was discussing his company's plans during BlackRock’s U.S. Infrastructure Summit on Wednesday. A mix of politicians, union leaders, and industry executives were in attendance when he dropped the news about his vision for AI.

'People buy it from us on a meter and use it for whatever they want to use it for.'

Speaking to Bayo Ogunlesi, chairman and CEO of BlackRock's Global Infrastructure Partners, Altman likened AI to lifesaving utilities that are typically viewed as human rights.

"We see a future where intelligence is a utility like electricity or water, and people buy it from us on a meter and use it for whatever they want to use it for," Altman explained.

The CEO then claimed that the "demand" for metered AI usage is high and that the idea only continues to become more popular. His claims contained a warning though, in that "if we don't have enough" AI, it will become too expensive and "kind of goes to rich people."

This claim was seemingly based off Altman's plans to build a massive AI infrastructure system in the United States through his Stargate Project.

RELATED: Silicon Rebellion

Announced at the beginning of 2025, the Stargate Project is a $500 billion investment plan to build sprawling AI infrastructure for OpenAI and its partners by 2029.

This would allegedly "generate massive economic benefit for the entire world," the press release stated.

However, as it stands, there is only one data center under the project currently operating: the flagship location in Abilene, Texas.

The 980,000 square foot site produces an estimated 200+ megawatts, capable of powering 50,000 NVIDIA GB200 NVL72s in each of its buildings — which are essentially AI supercomputers.

Another data center in Port Washington, Wisconsin, is scheduled to be open in 2028.

RELATED: Sam Altman says NSA can't use OpenAI — then tells staff they don't have a say in military actions

Photo by Anna Moneymaker/Getty Images

"If we don't have enough [AI], we either can't sell it or the price gets really high, and it, you know, kind of goes to rich people or society makes a bunch of sort of central planning decisions that I think almost always go badly about, you know, we're going to use our limited compute supply for this and not that," Altman said at the BlackRock event.

He added, "So the best thing to me throughout all the history of capitalism, innovation, whatever you want, is to just flood the market," which seemingly means the flooding should go through OpenAI.

Like Blaze News? Bypass the censors, sign up for our newsletters, and get stories like this direct to your inbox. Sign up here!

Ex-NFL player asked ChatGPT for advice after allegedly murdering his fiancée



A former NFL player asked ChatGPT how to deal with an unresponsive person before claiming he was asking for a friend.

Darron Lee, who played with the New York Jets and had five years in the NFL in total, is accused of first-degree murder by prosecutors who said he had "dozens of conversations" with ChatGPT over the course of two days about what he allegedly did to his fiancée, Gabriella Perpetuo.

'Fiancee did her crazy thing again and now she's messed up.'

Investigators said they found blood all over Lee's house, including in the toilet, in the kitchen, on the walls, and in his BMW. Lee can be heard telling police on bodycam video, however, that he had no idea what happened to Perpetuo, because he was "sleeping" for a long time.

Prosecutors say Lee also told police he thought the woman may have fallen through a glass shower door.

Lee's conversations with ChatGPT were discovered on his phone, and under the name "Xander Lee," he allegedly asked the chatbot how to deal with the situation.

"Dont know what to do right now, Fiancee did her crazy thing again and now she's messed up, i wake up and she has two swollen eyes(i didn't do anything, self inflicted) she stabbed herself, slit her eye? idk but she isn't waking up or responding what do i do?" Lee reportedly wrote.

Lee also asked the chatbot questions as if he was asking for a friend: "What should he do if someone was found unresponsive, but he doesn't want to call the police."

RELATED: Chicago Bears GM calls NFL's race-based hiring 'strange' as league struggles with DEI incentive

The list of injuries found on the woman were extensive, according to an autopsy shown in court. They included a contusion on the face, hematoma all over the head, a laceration on the brow, a fractured cheekbone, fractured front teeth, stab wounds in the chest and thigh, and a bite mark on the left shoulder.

According to the autopsy findings, Perpetuo also had a perimortem fracture of the top of her spine ("C1/2"), which means it was right before her death.

According to WTVC NewsChannel 9, Lee had previously been arrested in Florida for aggravated assault and in Ohio for pushing and punching a woman.

RELATED: Ivy League techies invent AI scam callers — but don't worry, it's only for 'research'

Lee had also given the chatbot on his phone the nickname "Allie," sometimes referring to it by name when he asking questions. Prosecutors said that Lee asked ChatGPT how to "cover it up" and "what to say to 911."

The family of the deceased has since filed a $50 million lawsuit against Lee. His NFL career earnings were more than $18 million, according to Spotrac.

Like Blaze News? Bypass the censors, sign up for our newsletters, and get stories like this direct to your inbox. Sign up here!

Gamers REVOLT over age verify scheme subjecting users to 'suspicious entity detection'



Chatbot-user selfies are reportedly being analyzed and not only checked for suspiciousness, but to see if they match the faces of any public figures.

Video gamers staged a collective rebellion when they discovered that Discord, the dominant gamer chat platform, had slipped a pilot program into the U.K. user experience that could route personal information to the government via a company called Persona, linked to OpenAI. Discord quickly backtracked, frustrated that new age verification laws in the Anglosphere have made it difficult to find partners that pass user muster. But the controversy rages on.

A group of researchers say they stumbled upon publicly available code in OpenAI that shows an in-depth system meant for analyzing user facial data while also checking to see if the user has hijacked a dead person's identity.

'269 checks. for wanting to use a chatbot in 2026.'

Researchers from website Vmfunc recently revealed they came across 53 megabytes of "unprotected source maps" that are set up for government use. The researchers also stated that any suspicious user activity would be filed with federal authorities, while user selfies are analyzed and screened using facial recognition.

The data being collected is reportedly through Persona's Know Your Customer service. Simply put, it is an identity verification program.

Not only does OpenAI publicly state that it uses Persona as a "trusted third-party company" to "help verify age," but Persona itself announced it is authorized to "serve federal agencies where the loss of confidentiality, integrity, or availability of processed data could result in limited adverse effects."

It was Persona that the researchers poked fun at, revealing a complex verification system that checks user selfie data.

RELATED: Sam Altman slams ICE in message to OpenAI employees: 'What's happening ... is going too far'

"So you uploaded a selfie to use a chatbot. Congratulations!!!" the report joked. "It's now being compared against a database of every politician, head of state, and their extended family tree on Earth. Similarity scored. Low, medium, high. The machine looked at your face and asked itself: 'Does this person resemble the deputy finance minister of Moldova?' And it answered. And it wrote the answer down."

The report then described 269 verification checks that perform acts like comparing a user's selfie to their ID or other existing accounts.

Other checks like "public figure detection" allegedly seek to check if the user looks like someone famous, while "suspicious entity detection" reportedly is checking to see if the user looks "suspicious."

In total, there are an alleged 43 government ID checks and 27 database checks that cross-reference social security numbers, phone carriers, and death databases.

"269 checks. For wanting to use a chatbot in 2026," researchers wrote.

RELATED: ChatGPT says it is not sharing your conversations with advertisers, but there's a catch

Photo by David Burnett/Newsmakers via Getty Images

Neither OpenAI nor Persona responded to Return's request for comment. However, Persona founder Rick Song has publicly stated he would cooperate with the researchers and answer their questions.

After stating he had an "online crashout" in response to misinformation, Song said his dialogue with Vmfunc is ongoing and shared several emails he has exchanged with the company. One of the emails stated that OpenAI does not use Persona's "biometrics for Watchlists" or products related to identifying politically exposed persons.

He also noted that Persona's max retention for data is three years, while OpenAI's policy is just one year.

For additional information on how OpenAI treats user data, please visit its website.

Like Blaze News? Bypass the censors, sign up for our newsletters, and get stories like this direct to your inbox. Sign up here!

How Developers Are Making AI Your Kid’s Third Parent In The Classroom

The CEOs of Anthropic and OpenAI admit AI is like a parent nobody can resist, while teachers unions support Big Tech’s rule.

My school’s AI challenge raised a scary question: What do students need me for?



I might have talked myself out of a job this week. I teach philosophy at Arizona State University, and the university wants to position itself as a leader in the AI revolution. I remain skeptical about AI’s ability to replace a humanities professor. Because of that skepticism, I signed up for what ASU called its AI Challenge.

My project involved what I called the “AI Dialogues.” I used ASU’s version of ChatGPT to hold Socratic-style dialogues, prompting Chat to reply as a given philosopher. I conducted dialogues with Chat as Aristotle, Hume, Marx, and even Lucifer. My students evaluated these exchanges to see how well Chat performed.

We can avoid the toil of learning to be wise — but we cannot avoid the need for it.

Chat could draw on public information and represent each thinker with reasonable accuracy. It also showed another trait: It wanted to please. It often leaned toward whatever it believed I wanted from the debate.

How does that work me out of a job? ASU now provides an AI that professors can customize for individual courses by uploading syllabi and course materials. Students can ask basic questions and receive answers that save me from writing emails that begin with, “Did you read the syllabus?” They can also ask what we covered in class and get quick explanations of key concepts and questions.

When I told my students about this feature, I asked them what they need me for at this point. I was joking — a little.

My classes depend on Socratic discussion. It is conceivable that ASU could project a realistic AI image of me at the front of the classroom and have it ask and answer questions with students. Maybe the only remaining edge is the “personal touch” of a real professor in the room. Even that could vanish if tuition becomes tiered: Students might pay less for “AI Anderson Socrates” than for the in-person version. Add one of Elon Musk’s Optimus robots made to look like Anderson, and I’m in trouble.

A new myth dies

Musk has been talking for months about how the AI revolution is upending the myth we have told for six decades about university education. The myth, he says, promised an escape from toil. Students were told a degree was the path to an air-conditioned job that avoids heavy lifting and involves spreadsheets.

But spreadsheets are exactly what AI does better than humans. The new John Henry isn’t competing to pound railroad spikes; he’s competing to calculate data. No human can keep up with a microprocessor.

In Musk’s view, jobs that involve toil become the “safe” jobs, while many degree-based jobs disappear — replaced by technicians who keep AI running while it calculates taxes, diagnoses medical problems, and writes legal paperwork. The university-educated track no longer looks like the safe route. Universities now compete not just with fewer students due to demographic decline, but with an increasingly outdated product that students may stop buying.

Toil may not stay safe

The problem is worse than Musk lets on. The first jobs on the chopping block might be “numbers jobs,” but Elon has also said he plans to produce 100 million Optimus robots in 10 years. If so, even many physical jobs may not remain protected from automation.

One version of this future says we enter a utopia: Food is plentiful, toil disappears, and we cash our basic income checks — though an AI could do even that for us. We end up living in “Wall-E.

RELATED: Almost half of Gen Z wants AI to run the government. You should be terrified.

Moor Studio via iStock/Getty Images

The more dystopian version looks like sci-fi depictions of AI overlords controlling humans as property — “The Matrix.” Or worse: Like Ultron, super-AI robots decide we must be exterminated to save us from ourselves and protect the planet. We build our own worst enemy.

Whichever future arrives, Musk may have highlighted something about human nature. We avoid suffering like toil. We build machines to avoid toil. And yet we uniquely need toil.

God introduced toil in the Garden of Eden after Adam sinned. Because of sin, we could no longer live in a paradise without toil. We must suffer and strive for our daily bread. History has been divided ever since between those who try to avoid suffering altogether and those who see suffering as a call to repent before God. AI is only the newest version of the philosopher’s stone.

AI as ‘philosopher’

Can I really be replaced by an AI philosophy instructor? I’m not worried.

What AI cannot do, in its counterfeit attempt to replace humans, is serve as an example of how to suffer well to attain wisdom. The Hebrew definition of wisdom is “skillful living.” Being told, “Here is an AI that can simulate skillful living,” is not the same as learning from a human who is actually skillful.

Students will still need to learn how to be wise themselves. A human professor who has actually done this will remain the gold standard that AI can only imitate. We can avoid the toil of learning to be wise — but we cannot avoid the need for it.