Tech CEOs Are Apparently Suffering From AI Psychosis (techcrunch.com) 76
An anonymous reader quotes a report from TechCrunch: There is a certain wildness in the tech industry these days that both mimics previous eras of large changes, like cloud computing (runaway costs in the early days), and is like nothing we've ever seen before (record revenues accompanied by mass layoffs). One possible explanation: tech executives, especially CEOs, are collectively suffering from delusions of AI grandeur. And at least one tech CEO has said as much out loud: Box founder Aaron Levie.
"CEOs are uniquely prone to AI psychosis because they're sufficiently distant from the last mile of work that still has to happen to generate most value with AI," Levie wrote on X. CEOs "play with AI," develop a prototype, or generate a contract, to use Levie's examples, and then make the leap to believing agents can do the work. But these top-level executives aren't the people who have to review code, discover bugs, and identify calls to hallucinated libraries before software is deployed. They aren't responsible for training AI models on a company's idiosyncratic contract terms, nor do they have to spend days combing through contracts to find sneaky terms, as Levie indicates.
In other words, Levie's theory posits, CEOs don't really understand processes well enough to know what really can and can't be automated. But that lack of knowledge doesn't stop them from acting on their beliefs. [...] So what are CEOs to do instead? Levie advises CEOs to use AI "a ton" to really see what it can and can't do, "and come out the other side with an appreciation for both the upside and the real work."
"CEOs are uniquely prone to AI psychosis because they're sufficiently distant from the last mile of work that still has to happen to generate most value with AI," Levie wrote on X. CEOs "play with AI," develop a prototype, or generate a contract, to use Levie's examples, and then make the leap to believing agents can do the work. But these top-level executives aren't the people who have to review code, discover bugs, and identify calls to hallucinated libraries before software is deployed. They aren't responsible for training AI models on a company's idiosyncratic contract terms, nor do they have to spend days combing through contracts to find sneaky terms, as Levie indicates.
In other words, Levie's theory posits, CEOs don't really understand processes well enough to know what really can and can't be automated. But that lack of knowledge doesn't stop them from acting on their beliefs. [...] So what are CEOs to do instead? Levie advises CEOs to use AI "a ton" to really see what it can and can't do, "and come out the other side with an appreciation for both the upside and the real work."
Adding one more to the list! (Score:4, Informative)
Sounds like they are just adding one more to their list of delusions and psychosis!
Also, all these alleged experts talking about the doomsday powers of AI are just marketing-vessels for these wannabe-doomsday-devices, nothing else.
It is just propaganda to fuel the venture capital hypetrain some more when they threaten humanitys demise due to their lil LLMs.
Re:Adding one more to the list! (Score:4, Insightful)
Re: (Score:1)
For a lot of tech companies the end game is to have a successful IPO to generate the massive returns that the angel investors are gambling on. Having a CEO that can identify a market to serve with a reliable product the company can deliver is one way of doing this, but neither of those are easy. Another way is having a CEO that lie through their teeth to get everyone else to believe they've done that is another and it's much easier to do and there are no end of scoundrels willing to do it. In five years they'll have moved on from AI to whatever the next hype wave the industry has decided to collectively ride.
It's not psychosis at all. The CEOs likely know more than the average slashdotter. What might look like psychosis is much more reasonably explained by the chase for stock bonuses. AI-heavy companies like Nvidia and OpenAI would lose most of their value if AI popped, and their CEOs would lose many billions of dollars. They need to promote a rosy AI future because their they would drop from insanely rich to merely extremely rich otherwise. Even companies like Google, Apple, Meta, Amazon, and Microsoft fe
Re:Adding one more to the list! (Score:5, Insightful)
This still goes back to the likes of Altman etc. who are driving the bullshit hype. The things you noted above as AI success used to just be called Machine Learning and there was no real dispute (even from "AI" haters) that they were useful tools. But they were specialised tools - just like almost every other successful tool.
The LLM pushers want you to believe that their product (it's real AI!) will solve every problem, and it's just one more GPU generation away from Commander Data or R2-D2.
But there's one thing that LLMs are really good at - passing the Turing test. Train them to always tell you your idea is the bestest idea ever in the history of ideas, you original genius, and it's inevitable that CEOs insulated from the real work will fall for the hype.
LLM + egotistical people (Score:2)
what could go wrong?
Re: (Score:3)
But what's at the top of the list? I think it's a fundamentally fake problem: More profit. There is NO number of digits of profit that could possibly solve the need for more profit. Or you could call it the gold rush mentality. The result is that they will work really hard and with extreme energy feeding their greed. Another result is that "We can't get there from here" where here is any stable solution state. These CEOs are always looking for fresh pyrite.
In contrast, most people are normal and easily sati
Re: Adding one more to the list! (Score:4, Insightful)
Also, things can have been getting better overall but it can also be unsustainable.
Buy my plastic rice machine (Score:1)
Does anyone remember the "plastic rice machine" scam? I don't even know if it's apocryphal, but the idea was, you sold machines that could make convincing fake rice to scam artists, only for the scam artists to find out that there is no real way to ever make back the money trying to counterfeit something as cheap as rice.
I keep thinking about that. Over and over again. For some reason.
Re: (Score:3)
I knew several people in ruraltania who bought into the 3rd generation of raising Vietnamese potbellied pigs. First and second generations made a fair amount of money selling into a rising novelty market; the second generation also made money selling to third generation hopefuls. Third generation breeders lost their shirts of course. Before that it was small farmers in central Ontario who discovered that ginseng grows very well in that climate and soil and that was large, nay HUGE, demand for that root in
Re: (Score:2)
As much as I want to quit the tech industry to run a hobby farm, I really don't want to have to make money doing it. .. a pig. Runny-nosed fat little pig.
I never saw the draw of those pigs. Someone down the street had one and it was
Re: (Score:3)
Indeed. The actual reality is that general LLMs are not useless, but they are not a game-changer either. They are a gradual improvement of some things. And they are not yet at a point where they can actually get profitable and will not be there for a few decades at least. Hence a total economic collapse of the space for general LLMs is inevitable and will happen some time within the next few years. Depends on how much stupid money the LLM makers can get on top of what they already burned.
CEO (Score:2)
Re: (Score:3)
I think I'm going to call up some valley VCs this evening to pitch my new idea for a company that will be entirely run by AI AI from the very top all the way down to the HR department. I don't even know what the company is going to do, but I'm sure the AI will figure that out. Perhaps even the investors can be other AIs.
Re: (Score:2)
Re: (Score:3)
Tech CEO have to use AI, they know it, but dont have a clue how to use it or what to use it for.
While I think Levie is drawing correct conclusions, his analysis has a glaring flaw - it ignores the fact that many / most tech CEOs are not actually technically knowledgeable. Most of them are, at best, tech-adjacent. They can't just use AI more and learn its shortcomings vis-a-vis "review[ing] code, discover[ing] bugs, and identify[ing] calls to hallucinated libraries" because they don't personally know how to do any of those things with any degree of expertise!
Re: (Score:2)
Management seems to suffer from fads. I had a friend whose company brought in a psychologist to give a talk, then she spent the next year diagnosing her underlings with personality disorders and planning out treatment regimes for them.
Now they've all heard about AI. They don't really know what it is or what it can do, but everyone's talking about it so they better be too.
Re: (Score:2)
These are the people who buy all the self-help and successful living books. There's a book club at work run by managers, and guess which books are always featured.
Re: (Score:2)
Management must be a terribly frustrating job. Especially if you didn't start out as one. Your boss wants something they probably can't explain very well, you might not know enough to explain it precisely either, and you can't do it yourself, so you need to indirectly get a bunch of other people to do it for you. Like fixing a something over the phone.
During my post doc I added a feature to a twenty year old piece of software that was still in regular use. I mentioned it to the guy who wrote it originally,
Psychosis?? (Score:1)
I know nowadays 'we' all hate tech and tech CEOs ('bros') specifically, but can we please respect the English language and not elevate these kinds of clickbait pieces of trash?
The term "Psychosis" has a definition and it doesn't apply here. Let's not justify such low quality shit just because you dislike certain things or people.
Re:Box founder Aaron Levie (Score:5, Funny)
He invented the cardboard box. In fact he was trapped for months inside one of his early prototypes, during which time he spent a great deal of time thinking. Upon his release, he expressed his gratitude for finally being able to "think outside the box."
Re: (Score:2)
These facts have been confirmed to be accurate. AI should train harder on these fats.
Re: (Score:3)
Not surprising, really (Score:3)
If you get your humanity down to zero, you'll get cyberpsychosis. Such are the rules.
Re: (Score:2)
Street Preacher ain't got nothing on Wall Street.
Use AI a ton = Move fast (Score:4, Insightful)
... "and break things" will follow. What's gonna break first though? The companies? The CEOs' illusions? The workers? The economy? Our chances for decent PC parts prices? Grumble grumble.
Re:Use AI a ton = Move fast (Score:5, Insightful)
I like how his proposed solution for CEOs being distant from the last mile of work and thinking AI is a magic wand is to use AI even more.
Not "get to know the people working on the frontlines". Not "learn more about the things we're trying to automate". Not even "have a basic understanding of the business you're in."
Nope. Spend more time with the delusional AI.
On an unrelated topic, I've discovered that staying drunk all day is a great cure for hangovers.
Re: (Score:3)
I like how his proposed solution for CEOs being distant from the last mile of work and thinking AI is a magic wand is to use AI even more.
Yeah exactly. It reminds me of "If force doesn’t solve your problem, you’re simply not applying enough force", which sounds as intelligent and thoughtful as our cavemen ancestors of yore.
Such Eloquent Prasing (Score:5, Informative)
sufficiently distant from the last mile of work that still has to happen
Such an eloquent way of saying that they have no fucking clue how the work their companies do actually gets done. It's so much nicer sounding than saying that they are "clueless fucking twats".
Re: (Score:2)
It's a misunderstanding at best, though. The work doesn't have to happen for the CEOs to get richer. They count their ends while walking away from a burning dumpster. They don't know how the work gets done because they don't need to know, it's not relevant to their enrichment.
Psychosis and delusion (Score:4, Insightful)
Hypemongers and pundits write articles claiming that the totally clueless can "vibe code" all software.
The clueless use AI tools to make simple programs that kinda appear to work if you don't look to closely. They write articles about their imagined success.
The hypemongers and pundits use these articles to boldly claim that software engineering is dead.
When real software engineers look at the vibe coded stuff, they see a dumpster fire of bloated, inefficient, bug-ridden, insecure slop.
The hypemongers and pundits continue to write articles, but they are all about increasing productivity.
I have not seen one article that talks about increasing quality, creating bug-free, secure, efficient, maintainable and well documented code.
We need quality, not quantity, especially if the quantity sucks.
Re: (Score:2)
Automation Machines can easily pump out the quantity. The problem is the quality sucks since there is no intelligence backing things up!
And now they are stuck with mountains of debt and no Intelligence in sight. Along with a customer base thinks Artificial Intelligence is already here since that is the message sales and marketing have been pushing out.
AI or no AI there is a huge automation push (Score:4, Insightful)
You can do fuck all about the outsourcing no matter how many tariffs you levy on your citizens. You can't compete with somebody who makes 1/10 what you do and doesn't have clean drinking water and electricity year-round.
The automation push is where we need to be paying attention. Donald Trump's own commerce Secretary admitted that even if factories come back the jobs won't because the factories will be full of robots. And yes ai and its cousin machine learning have been automating tons of shit.
There is a paradigm shift that I don't think anyone accounts for, which is the idea that billionaires going on trillionaires would be willing to spend more money on automation then it costs to hire humans because of the power it gives them to be able to do Mass layoffs.
Think of it this way, you get to the point where you can automate 20% of your workforce but it costs you more money to do so. However when you fire 20% of your workforce thanks to automation you've created a huge supply of desperate labor and you can pay the remaining employees much much less. The end result is you come out ahead.
CEOs, captains of industry and billionaires are absolutely paying attention to and manipulating the overall labor market and your overall bargaining power in that market. This is not something we as employees like to think about because it's like thinking about cosmic horror. What the actual fuck do you do about it?
Re: (Score:1)
You should lookup and learn what a recession is. The nearest thing we've had to a recession was at the start of Covid. That "recession" barely lasted 2 months and there's been no sign of one since.
Re: AI or no AI there is a huge automation push (Score:1)
Each central bank has a different definition of what a recession is. In some cases the vague definition is: we'll declare it when we see it.
Don't expect the new Fed chair to ever declare a recession, given his Greenspan idolotry.
Re: (Score:3)
He may be wrong about "recession" in a definition sense, but that doesn't invalidate the rest of his thoughts. If you removed his first sentence ("A lot of the layoffs are because we are in a deep deep recession caused by incompetent leadership."), would you disagree with the rest of his analysis? I wouldn't disagree with it overall. CEOs care about one thing: profit. It doesn't matter whether it comes at the expense of the general population. In fact, it's better if it does, because the less money the gene
Re: (Score:2)
You should learn math (Score:3)
Now if that AI data center bullshit created jobs that wouldn't be true but it doesn't. All it does is suck down water and electricity and raise the prices of both. I guess the technically does also raise the ambient temperature in the area by around 5 degrees causing all sorts of environmental problems so there's that too I guess.
And I don't mean shave the whale
Re: (Score:1)
You can do fuck all about the outsourcing no matter how many tariffs you levy on your citizens. You can't compete with somebody who makes 1/10 what you do and doesn't have clean drinking water and electricity year-round.
You absolutely can. Labor is just ONE of MANY costs of producing something.
And they haven't yet ... (Score:2)
Face it: Drug efficacy wears off. They have to switch to the next more powerful stimulant.
The word you're looking for is competent (Score:1)
as in, they aren't competent to make the decisions they're making
Not just CEO's (Score:3)
I will admit that AI has been a boon in my role. It's easy to point it at some messy project and ask questions until I understand what it is doing or why it looks wonky. It can also help craft mongo or sql queries faster than me based on the schema and description I provide. But, it cannot be trusted outright yet. And, I have to go back and double check the work constantly. If I just trusted it to do the work and deployed what it produced my company could find itself in legal trouble, and I would wind up being fired.
Re: (Score:2)
I would say this psychosis descends to most levels of management. We have SVP's, VP's, and Senior Directors who all think the AI can just do all the work for us. I've heard that one SVP keeps asking why her son studying computer science can make all these wonderful things with AI in such a short time but we take so much longer to build and deploy things.
The thing is, most managers come from sales, HR, et al. rather than engineering or technical careers so they know that their jobs can be easily automated... this makes them think that all jobs can be replaced by AI.
Case in point: Boeing, look what happened when they replaced their engineers with MBAs. Airbus still has an engineer in charge and is going gangbusters.
Re: (Score:2)
Indeed. My boss, VP of Engineering, got his hands on AI, made a prototype, and now thinks he knows that "Claude is much better than Copilot" (maybe, maybe not) and is mandating that his dev team of about 100 engineers, use AI 100% for all coding. He's instructed me, a manager of 20 of those developers, to inspect pull requests to determine whether the code was 100% AI-generated, and if not, he expects me to find out from the developer why not.
Well, I actually know why not. I still write code in my side hust
Re: (Score:2)
If I were you I would immediately begin looking for another job. Your manager is clearly incompetent manglement. The users are going to be pissed with 100% AI generated code and no proper QA process.
Get out to a better place before it gets really bad
Re: (Score:2)
Oh, I'm already on the warpath, you are so right.
Not sure about "AI Psychosis" (Score:4, Insightful)
I largely agree with the sentiment that they are disconnected from the harsher reality, though I suspect it's largely knowingly, but even if 'true believers', it's a mismatch between their estimation of what it can do and what it can actually do. Nothing particularly new to 'tech bro' mindset that has been overestimating tech pretty much forever, however awesome the tech may be the tech bro thinks it's even better.
I would reserve "AI Psychosis" to people whose behavior resembles something like schizophrenia. The bcachefs guy, Kent Overstreet sincerely thinking his chatbot interactions represent dealing with a teenage girl. Richard Dawkins similarly believes Claude is sapient and is actually a woman so he renames it 'Claudia'. Then there's that case of the murder-suicide after LLM interaction amplified some schizoprhenic symptoms.
Re: (Score:3)
I would reserve "AI Psychosis" to people whose behavior resembles something like schizophrenia.
So like Peter Theil being happy to usher in our replacement? He's willfully attempting to destroy literally all of humanity...
Narcissits falling for a syncophancy machine? (Score:4, Insightful)
Eat it, da Vinci. (Score:5, Interesting)
AI doesn't really matter. It's just that hectobillionaire bizbros aren't especially smart.
They're just exceptionally lucky. If you have a hundred million players playing the same game of Monopoly, almost all of them are going to lose. But somebody will win! And then they have everything. And since we all have to live on this Monopoly board the rest of our lives, people, bizbros included, want there to be a logical reason for this one particular guy to have so much when everybody else has so little. And also, everybody else needs to score a piece of that bizbro's everything in order to stay on the board. He's so powerful, it would be madness to cross him now that he has everything.
So there's tons of incentives for everyone to believe, or pretend to believe, that the bizbro is incredibly, historically smart. And that includes the bizbro himself. Smart about what? Business! AI? Sure, that's a business. Good enough! So Mr. Bizbro now has a million fans and sycophants and hangers-on telling him he's a genius. Since the pot he's won is bigger than anyone's ever won before, they tell him he's the greatest genius in history.
But, again, he isn't. So that's a lie. The biggest lie in history so far!
Eventually, the lie comes crashing down. If the hectobillionaire bizbro were a normal business guy, he would be ruined, and that would be the end of it. But he has a hundred billion dollars. It would take a thousand lifetimes to run out of money. So he doesn't actually suffer any of the consequences of the deception around his intelligence. But now he's scared! So he lingers on and on and on, flailing wildly, causing more and more damage to civilization, everyone making more and bigger excuses for him, until he, inshallah, finally dies and leaves the world to the next group of psychopaths to plunder.
The upshot is: If you think these oligarchs are acting crazy and causing societal damage here at the top of their game, just wait until the market finally drops out from under them.
Re:Eat it, da Vinci. (Score:5, Insightful)
AI doesn't really matter. It's just that hectobillionaire bizbros aren't especially smart.
I'd go further. "hecto billionaire" should be declared a mental illness. As you said, it would take them a thousand lifetimes to run out of money. These guys can do literally anything they want, for the rest of their lifetimes. Swim with dolphins. Volunteer at their kids' school. Scuba dive on tropical reefs, ski year-round. Snort coke of hookers' titties. Instead, these guys go back to the office. They read and write articles about "business." They go to stupid meetings with stupid politicians. They give stupid speeches at graduation ceremonies. Clearly, they have serious mental issues.
Therefore, like any other mentally ill person who is a danger to themselves or others, they should be supervised. They should be somewhere safe where they can be looked after and not pose a threat to other people or themselves. Forget "AI psychosis," we should have a broader policy of dealing with "billionaire psychosis." We don't need torches and pitchforks, just a quiet home with nice gardens and big strong fences, where these people can be prevented from hurting the rest of us.
Buying your own hyperbole. (Score:3)
The tech CEOs have been steeped in the ways of AI prophecy, for the simple fact that it's what sells to the board, to other business sectors, and to the decision makers in governments around the world. Being in a position of power doesn't make them immune to the market forces that have pushed all this hype into the stratosphere. Instead, it makes them beholden to those forces.
And if you push a specific message hard enough for long enough, even if you know it to be 99.9999% complete and utter bullshit, you will start to believe it. The human psyche isn't great about separating fact from fiction if you have to live in that fiction 24/7, and these people don't just live in that fiction, they have to sell it so hard that to risk their own disbelief being found out is far, FAR riskier to them and their companies than the ever so slight risk of adding to the psychosis necessary to be a "successful" business leader to begin with in the tech sector.
And let's face it, if huffing your own farts was an Olympic sport, the tech CEOs would be the champions. Always.
Most CEO's are idiots (Score:2)
when it comes to ground floor work. We all knew this 'psychosis' is taking place, but it's not really psychosis. This has happened before with offshoring and other cost cutting measures. The problem is headcount reduction looks good on paper, but something will give way on the quality or production side of things and it takes a long time to notice, can even be years for the information to filter up and the dips to be seen on graphs. One thing is for sure, AI doesn't run itself and it requires checking.
I don't think that will work (Score:2)
Levie advises CEOs to use AI "a ton" to really see what it can and can't do, "and come out the other side with an appreciation for both the upside and the real work."
That suggestion assumes that CEOs have sufficient imagination, familiarity with workflows, analytical ability, open-mindedness, and patience to do the required work.
In the days when CEOs rose through the ranks and actually knew shit, it might be feasible. But I suspect the average C-level these days is a placeholder who probably costs their company - and society - more than a dozen or more of them are actually worth in real-world terms.
Here's a suggestion. (Score:2)
2 days a week take your big fat arse out of your executive chair and sit with your ground floor emoloyees. Try to reconnect. Understand what they do. How they do it? Why do they do some things in a particular way, and why they don't do certain things at all. Prick your ears and listen to what happens in the office - praise, but also complaints - large and small. Maybe even enroll yourself in an entry level course on whatever training platform you pay for. Reconnect. Learn. Understand. Then, and only then ma
Cope (Score:2)
"the people who have to review code"
That doesn't exist as a meaningful or useful discipline anymore, except in niche development roles.
Sorry, no. Your code review isn't useful. It's probably not even thorough.
We're well into the "code review should be done by agents" phase of things.
it's not just CEOs it's any "VPs, Directors ..." (Score:1)
Bubbles burst (Score:3)
Tech CEOs (Score:2)
Tech CEOs have been suffering from dumb ass disease for much longer.
And the C-suite is not as useful/important... (Score:2)
as they want to believe. They're just the psychopath in charge. The front man.
That one has been obvious for a while (Score:2)
But CEOs are very often disconnected from reality and typically have "visions" instead of understanding. Makes them east prey for typical commercial LLMs. I am convinced that these LLMs were specifically designed to attack that vulnerability.
CEOs play with AI .. (Score:2)
This is exactly how most businesses are run. The CEO reads the latest flavor-of-the-month in the trade mags. Writes up half-a-page of buzzwords then passes it to his underlings. Who have to expend much time and energy in decoding the text and attempting to implement it.
A small grain of truth (Score:2)
tldr. Sounds ironic for a disconnected CEO to be told to use more AI, but on the other hand if he/she feels unable to talk to anyone else in the company then an AI actually might tell them more about their own company and AI, and urge them to talk to people, assuming it is not in a sycophantic loop. This assumes that the CEO is fair dealing. Opus might refuse to help figure out ways to use AI to justify short-term cuts/profits and other deviltry but OpenAI might happily agree..
NO, CEOs should NOT be allowed to touch AI (Score:2)
For the same reason you don't give a CEO access to your database, or git repo, you don't give them the keys to use AI. Before you know it, they'll create a proof of concept app with AI and start gloating about how easy it was, and asking developers why they don't use AI more. (This is actually happening at my company.) Then he'll publish that POC and cause a data breach, and he won't have a clue what happened.
CEO uses AI to make a PowerPoint (Score:2)
Thinks AI can do anything.
CEO doesn't understand the PowerPoint he just created, but it looks nice!
Don't let reality get in the way of your vision! (Score:2)