Tech executives are suffering from "AI psychosis," says Box CEO

[Alfonso Maruccia]

TL;DR: A growing backlash against the forced adoption of AI is building across the industry, but enterprise executives remain as bullish as ever on the LLM-driven future. According to at least one prominent tech leader, that enthusiasm may signal how out of touch those at the top have become.

Box co-founder and CEO Aaron Levie has coined a name for what he sees afflicting the C-suite: "AI psychosis."

While industry leaders extol the technology as a once-in-a-generation revolution, many of their employees – and the masses are not cheering. They're booing. In a weekend post on X, Levie argued that CEOs are "uniquely prone" to the condition "because they're sufficiently distant from the last mile of work" required to turn AI outputs into reliable business tools.

A graduating student holds up a sign at a Boston University commencement ceremony on May 2023. Credit: The Guardian

When executives experiment with AI, Levie wrote, "they see the happy path results, often not considering the next 10 or 20 things that have to happen to get sustainable results from agents." An executive might vibe-code what looks like a disruptive product prototype, he argued, but never has to review the underlying code before it ships – or verify the legal terms in a contract an AI just generated.

The gap between a compelling demo and a production-grade, enterprise-reliable workflow is precisely where most AI projects go to die, and where the people closest to the actual work live every day.

– Aaron Levie (@levie) May 24, 2026

That disconnect is already carrying real consequences. An overwhelming majority of tech executives expect AI to trigger layoffs within their organizations, while tens of thousands of workers have already lost their jobs to fund new AI infrastructure projects.

Critics are accusing tech corporations of engaging in AI washing, blaming LLMs and chatbots for staff reductions that would have happened regardless. Meanwhile, AI agents have been caught wiping entire corporate databases – backups included, while some Big Tech employees are gaming the system by inflating and faking their AI tool usage to score better on internal productivity leaderboards.

Meanwhile, Microsoft is already eyeing AI agents as the next major licensing frontier. Levie's prescription for CEOs is straightforward: use AI extensively enough so that they can experience both the good and the ugly coming from the tech without betting their company's future on a single technology innovation right away.

It's worth noting that Levie is not an AI naysayer. The Box CEO is a regular AI evangelist on X, advocating for a future in which agents become the default mode of software development. He is also an active angel investor, with a portfolio concentrated in enterprise software, SaaS, cloud computing, AI, and cybersecurity. His concern, in other words, isn't with AI itself – it's with executives who stop short of understanding what it actually takes to make it work.

Permalink to story:

Tech executives are suffering from "AI psychosis," says Box CEO

 

[Squid Surprise]

"It's worth noting that Levie is not an AI naysayer. The Box CEO is a regular AI evangelist on X, advocating for a future in which agents become the default mode of software development. He is also an active angel investor, with a portfolio concentrated in enterprise software, SaaS, cloud computing, AI, and cybersecurity. His concern, in other words, isn't with AI itself – it's with executives who stop short of understanding what it actually takes to make it work."

You buried the lead

 

[Lowhangingfruit]

The chart says that "most companies haven't realized financial returns from AI" but then shows that over 40% of companies have either increased revenue or lowered costs or both. Yes that's technically not the majority, but it's a very significant proportion considering the technology has only been out for 3 years. And has really only gotten good for 1 year. Even cloud adoption wasn't this rapid.

This article is just title clickbait. I guess that's ok.

 

[wiyosaya]

Has any executive ever been in touch with those at lower levels?

 

[NumberNine]

Being out of touch with reality is the timeless curse of aristocracy.
They do not execute their own decisions.

 

[Hecate91]

Well, yes obviously the tech bro CEO's are having too much of the AI koolaid, making astronomical promises to their shareholders, and especially the tone deafness of it all, while AI hasn't done much of anything beneficial for the regular person, and hasn't made things cheaper or easier.
I'm sure the AI apologists will call this article "anti-AI clickbait" but in reality CEO's should either shut up, or perhaps if AI is so amazing CEO's should replace themselves with AI bots.

 

[Squid Surprise]

Well, yes obviously the tech bro CEO's are having too much of the AI koolaid, making astronomical promises to their shareholders, and especially the tone deafness of it all, while AI hasn't done much of anything beneficial for the regular person, and hasn't made things cheaper or easier.
I'm sure the AI apologists will call this article "anti-AI clickbait" but in reality CEO's should either shut up, or perhaps if AI is so amazing CEO's should replace themselves with AI bots.

Dunno about "AI-apologists"... what about the "anti-AI trolls" going on about how terrible AI is - when it is anything but?

 

[Steveb8189]

The chart says that "most companies haven't realized financial returns from AI" but then shows that over 40% of companies have either increased revenue or lowered costs or both. Yes that's technically not the majority, but it's a very significant proportion considering the technology has only been out for 3 years. And has really only gotten good for 1 year. Even cloud adoption wasn't this rapid.

This article is just title clickbait. I guess that's ok.

This, exactly. It's a lot more than the 23% who have increased costs or decreased revenue.

 

[Joeg1484]

The chart says that "most companies haven't realized financial returns from AI" but then shows that over 40% of companies have either increased revenue or lowered costs or both. Yes that's technically not the majority, but it's a very significant proportion considering the technology has only been out for 3 years. And has really only gotten good for 1 year. Even cloud adoption wasn't this rapid.

This article is just title clickbait. I guess that's ok.

considering the technology has only been out for 3 years

Just to educate a little... AI has been around since the late 1960s, so its not new. In fact, we went a round of AI enshitification in the 80s, 90s and now it looks like its making rounds again... This time, its taking allot of finance with it - and hardware. Seems to be more destructive these days than it ever was!

Someday, people will learn that the only intelligent life here on earth are humans!

AI is cool technology and all, but its just 1's and 0's - thats all... Nothing new here... move along!

 

[axiomatic13]

I truly think Alex Karp has lost his marbles. I find his manifesto to be completely unhinged.

 

[Tantor]

I wouldn't trust the rosy figures. There is enormous Groupthink pressure on CEOs to be pro-AI.

"Groupthink is a psychological phenomenon that occurs within a group of people in which the desire for harmony or conformity in the group results in an irrational or dysfunctional decision-making outcome. Cohesiveness, or the desire for cohesiveness, in a group may produce a tendency among its members to agree at all costs. .... The term was coined in 1952 by William H. Whyte Jr..... [who] derived the term from George Orwell's Nineteen Eighty-Four..." - https://en.wikipedia.org/wiki/Groupthink

Note that Groupthink is also behind some military disasters.

"Most of the initial research on groupthink was conducted by Irving Janis, a research psychologist from Yale University. ... Janis used the Bay of Pigs Invasion (the failed American invasion of Cuba in 1961) and the Japanese attack on Pearl Harbor in 1941 as his two prime case studies."

Groupthink is the key factor behind Trump's Iran war disaster.

 

[human7]

The chart says that "most companies haven't realized financial returns from AI" but then shows that over 40% of companies have either increased revenue or lowered costs or both. Yes that's technically not the majority, but it's a very significant proportion considering the technology has only been out for 3 years. And has really only gotten good for 1 year. Even cloud adoption wasn't this rapid.

This article is just title clickbait. I guess that's ok.

The chart shows most of the CEOs are oblivious to the cost, unless they are relying on exclusively free model providers (I doubt it, given the limitations). For costs to stay neutral, the AI has to eliminate costs, and for most that hasn't happened (how could it, without layoffs, and then it isn't really AI that's eliminating costs).

 

[human7]

considering the technology has only been out for 3 years

Just to educate a little... AI has been around since the late 1960s, so its not new. In fact, we went a round of AI enshitification in the 80s, 90s and now it looks like its making rounds again... This time, its taking allot of finance with it - and hardware. Seems to be more destructive these days than it ever was!

Someday, people will learn that the only intelligent life here on earth are humans!

AI is cool technology and all, but its just 1's and 0's - thats all... Nothing new here... move along!

But there's a bit of history you left out: every 5 or 10 years society (or some part of it: researchers, CEOs, marketing people, somebody anyways) decides to redefine what AI is, so the "AI" from prior decades doesn't count, and in a few years "generative AI" won't be AI either. Thus, AI is new! (I say this half jokingly).

 

[deathspeed]

Happening at my work right now...

 

[pnntmp]

So, today's Anti-AI article is based on the opinion of one guy, and some cherrypicked anecdotes.

Funny that the Cost-Revenue chart directly proves everything else in the article wrong.
It shows a positive result for 42% of the companies, and truly negative for just 1%. But .. it's based on year -old data. That's the year when AI made staggering progress, and also price per token fell ~10x. So the already fantastic 42% vs 1% old result is way more fantastic today.

You may keep trying to 'prove' that AI isn't working, but it's getting harder and harder to convince even the prejudiced.

 

[TooQik]

Has any executive ever been in touch with those at lower levels?

Yes...but HR usually has to get involved to cut off the harassment case.

 

[TechSavy101]

They underestimate the power of consumers & workers.
Ai is the biggest scam of 2020s decade. When the bubble bursts most of these companies will go under or get absorbed by bigger ones. Day of Ai doom will come.

 

[Dirtronix]

“Technology is dominated by two types of people: those who understand what they do not manage and those who manage what they do not understand.” -- Putt's Law

 

[Squid Surprise]

You may keep trying to 'prove' that AI isn't working, but it's getting harder and harder to convince even the prejudiced.

Alas, there are still plenty of people who believe articles like this...

This was from 1889 about how terrible electricity would be... human nature doesn't change - just our technology

 

[Hecate91]

And yet, the AI apologists keep posting the same article for some reason.
It has me wondering, is some AI company paying for this narrative, or are you looking to smack people over the head with that agenda until people believe you?
Electricity benefits humanity as a whole, with a higher quality standard of living and convenience. AI isn't anything new, yet the AI tech bros are trying to sell it as the best thing ever that will make everything cheaper,easier, and better.
However it's doing quite the opposite for the normal person, causing electricity rates to go up, water to become polluted, and tech products which are essential for everyday modern life to skyrocket in cost.
But I'm sure I'll be called an "anti-AI troll" because not jumping headfirst into the AI BS and instead being Actually Intelligent must mean anyone is an AI hater, because that seems to be the online discourse these days, if you aren't fanboying for something you must hate it.

 

[Gravitytr1]

C suite should be called parasite class. Work the least and benefit the most and have the least risk.

These people statistically have higher rates of being disassociated from reality. They live in their own bubble, both literally and metaphorically.

The most work they do is when they get together to conspire and monopolize against the people and the planet.

 

[Gravitytr1]

But there's a bit of history you left out: every 5 or 10 years society (or some part of it: researchers, CEOs, marketing people, somebody anyways) decides to redefine what AI is, so the "AI" from prior decades doesn't count, and in a few years "generative AI" won't be AI either. Thus, AI is new! (I say this half jokingly).

Yes. Changing definitions of common words is much easier now with global central authorities who have strong influence over their respective fields. Whether that medical (we saw a lot of this during COVID), sociological (we still see this today), technological etc.

I remember a dos program as a kid 20 or 30 years ago that communicated with you like an LLM in 2026 would, without the petabytes of pirated data of course.

 

[TechSavy101]

Alas, there are still plenty of people who believe articles like this...

This was from 1889 about how terrible electricity would be... human nature doesn't change - just our technology

Lol this guy. We dont have real Ai as they are forcing everyone to believe it. Its just machine learning algorithm that cross references answers. No independent thinking or logic. This has become an excuse for these rich companies to turn everything into subscriptions. Gaming, routers, GPUs, TV networks, Door camera subscriptions etc. They want to take away ownership, it started with Home market and now its expanding into everything. If left unchecked soon you wont even own your clothes, you will need to pay a monthly subscription or go naked.

 

[jbright]

There's a whole class of people right now who have convinced themselves they understand AI capabilities because they've used ChatGPT to draft emails and generate a slide deck. That's like claiming you understand surgery because you've used WebMD.

 

[p51d007]

THAT is the difference between "the west" and China.
Most companies want to generate revenue from the money spent on "AI".
China isn't too worried about the money they are spending on AI.
Why? It's about control. They hope with the amount of money they are
funneling into the anti AI data center push in the west that companies in
the west will give up on the AI data centers since there isn't any return on
the investment. Meanwhile, the CCP will continue to build out more and
more data centers, continue to reverse engineer chips, steal chips, bribe
and payoff more and more people to help them "conqueror" the AI war.
If they gain control over AI, it's one more piece of the global puzzle they
will own, giving them more control.