AI Mania Is Eviscerating Corporate Decision-Making

July 31, 2026

AIBusinessSociety

Sources：Hermit Tech + Lobsters

Why You Should Read This Article

On July 18, 2026, an article titled AI Mania Is Eviscerating Global Decision-Making sent shockwaves through the tech community. Quietly passed around among Fortune 500 executives, it garnered over 120 upvotes on the professional developer forum Lobsters, with CEOs even calling it “the most widely circulated document in the boardroom.”

AI Mania — A Metaphor for Irrational FOMO Investment Figure: Global tech giants are projected to spend over $725 billion on AI infrastructure in 2026. Source: Unsplash

But this isn’t just an article about technology. It’s about you—every everyday worker anxious about whether to learn AI, every professional facing queries on “how your company uses AI” without knowing how to answer, and every employee watching their boss throw money at AI without producing meaningful results.

The author spent considerable time digesting the original article and over 300 professional comments to unpack a critical question: When everyone believes AI is changing everything, why does actual output fail to match the mania?

Rational Decision-Making Hijacked by FOMO — The Prisoner's Dilemma in Organizational Behavior Figure: FOMO-driven AI investment is turning into a global corporate performance. Source: Unsplash

I. The $725 Billion Gamble

First, consider a staggering figure: In 2026, four tech giants—Amazon, Microsoft, Google, and Meta—will spend $725 billion on AI infrastructure. That is higher than the annual U.S. defense budget during the Trump administration and several times the GDP of many medium and small nations.

Yet Goldman Sachs issued a stern warning: To justify this level of investment, the AI industry needs to generate at least $1 trillion in annual revenue. And what are the market’s most optimistic forecasts? Around $450 billion. More than half the shortfall remains unaccounted for.

Nikhil Suresh, author of the original piece and head of a data consulting firm, noted that over the past year and a half, his team witnessed every single AI project—literally every single one—either fail or hurtle toward failure. “We have seen a 0% success rate,” he wrote. “This wasn’t because we only saw bad projects. What we saw spanned across industries and clients of all sizes.”

Even more ironically, when he gave a simple data analysis demo using AI—essentially typing data descriptions into a prompt box—hesitant clients instantly clamored to buy. “It was like a dark force seized their limbs, reached into their chests, pulled out their beating hearts, and handed them over as credit cards.” He refused the deal, stating that taking money while knowing it wouldn’t deliver was fundamentally irresponsible.

On Lobsters, user aspensmonster commented: “Trillions of dollars being dumped into a technology that will yield a few tens of billions at most in actual returns—that’s the simple root of this entire insanity.” Another user, spc476, offered an apt analogy: “A trillion is a thousand billion. Converted into smaller terms, it’s like spending $1,000 a year to earn back $70. Long term, the economics just don’t add up.”

II. Who Is Forcing You to “Use AI”?

You might have heard interview questions like: “How do you use AI in your daily work?”

On Lobsters, developer lake shared his job-hunting experience. After months of interviews ending at the initial HR screening, he realized the primary culprit was his inability to express enthusiastic praise for “using AI to write code.” “Hesitating was wrong, and honestly saying ‘not much’ was 100% wrong. A voice inside told me I could lie, but what happens after I lie?”

Another user, jmmv, shared his situation: “I’m the guy at a big tech company trying to keep a ‘rational attitude towards AI’—and I was almost squeezed out. I had to force myself to use LLMs and alter my tone to sound less ‘objective.’ Even now, I’m walking on thin ice.”

This societal pressure trickles down from top executives to every single role. The original article disclosed a typical case: A Fortune 500 executive claimed internally that AI boosted productivity by 100x, but privately admitted it was purely to avoid alienating major clients who were making similar boasts. If he said “100x is impossible,” client executives would be called out, jeopardizing lucrative contracts.

As Nikhil put it: “Executives are pointing guns at each other across the world, and nobody dares put theirs down first.” It is a classic Prisoner’s Dilemma—nobody speaks the truth because the one who speaks out gets fired.

III. How FOMO Became an Organizational Performance

The real crisis isn’t whether AI has utility—it certainly holds value in specific domains, such as real-time transcription in healthcare. The problem is that “using AI” has been turned into a symbolic gesture: a test of corporate loyalty.

It sounds unbelievable, but the article revealed that many companies have instituted “token consumption leaderboards”—ranking engineers by how many AI tokens they consume, where higher is better. How did smart engineers respond? They wrote scripts to let AI talk to itself, burning through token quotas automatically while they watched Netflix. “Not a single one was ever caught,” an engineer remarked in the piece, “even when they admitted the AI-generated code wasn’t worth deploying.”

Another engineer was even more blunt: “I duplicated our company’s Go repository and had AI rewrite the entire codebase in another language just to burn our quota and save my job. I hate all of this.”

Some call this “AI laundering”—completing all the work manually but claiming it was done by Claude because upper management wants to see AI adoption. When requesting headcount, managers must first prove they “attempted to replace the role with AI.” If you dare say you tried and AI fell short, you’re labeled “anti-AI” and placed at risk of layoffs.

This isn’t an isolated dysfunction. It is a global ritual of compliance. As one reader observed: “It’s like returning to the Inquisition. You must publicly proclaim your faith that AI changes everything. Say ‘no,’ and you’ll be burned at the stake.”

IV. Where Is the Real “Villain”?

Reading this, you might wonder: Is AI worth investing in at all?

The answer is nuanced. AI has brought genuine breakthroughs to specific fields. In software engineering, AI-assisted code generation improves efficiency under certain conditions. Medical transcription, content summarization, and customer support workflows are benefiting from AI integration.

The core issue lies in how AI is deployed.

On Lobsters, user srcrip pointed out: “AI will permanently change a few fields, such as copywriting, concept art, and maybe programming. But it won’t change most of the economy. The problem is every company is pretending it’s omnipotent.”

Veteran developer emk offered a brilliant classification of two types of AI believers: The first type believes AI can perform specific practical tasks—a belief backed by empirical evidence. The second type believes that “recreating human intelligence on silicon is inherently good.” While agreeing with the first, he strongly opposes the second: “I don’t think we should spend trillions of dollars desperately trying to render the human brain economically redundant. The steam engine didn’t create more jobs for horses, and human-level AI won’t for people either.”

This highlights the true villain: FOMO-driven decision-making mechanisms. When a CEO sees a competitor declare “AI boosted productivity 10x,” he cannot calmly say “Let’s review the ROI.” He feels compelled to announce a breakthrough of his own. Board members panic, investors panic, and the whole ecosystem enters a vicious cycle where nobody verifies actual results, yet everyone is forced to perform.

V. When Rationality Is Consumed by Mania

Lobsters user Yogthos left a highly upvoted comment: “This is because LLMs speak the language of corporate executives. In the world of management, there is no real reality check mechanism. These people essentially just tell stories to each other. LLMs are like drugs to them—because they say what executives want to hear, sounding plausible when executives lack the technical ability to verify whether it actually makes sense.”

This hits the nail on the head. Organizational behavior demonstrates that distorted incentives lead individuals to make choices directly opposing their long-term interests. When employees are rewarded merely for “showing enthusiasm for AI,” rationality takes a backseat. Companies start hiring candidates who tell the best AI anecdotes rather than those who solve real problems, pursuing things that “look like innovation” instead of sustainable business models.

This distortion isn’t unique to AI. During the blockchain rush, adding “blockchain” to a company name caused stock prices to surge; during the dot-com bubble, a “.com” suffix secured funding. But the scale of AI dwarfing prior manias—with $720+ billion committed—means that when the bubble bursts, the fallout will be far more severe.

VI. A Survival Guide for the Rest of Us

If you feel suffocated in this corporate environment, the original article concludes with candid advice:

Do not challenge the narrative publicly. Avoid questioning statements like “AI changes everything” unless you are prepared to face career consequences. Solve problems rather than taking on the system.
Go freelance if it becomes unbearable. Freelancing at least comes with concrete deliverables and deadlines, sparing you from court politics.
Cut back on AI news. The author noted he stopped reading AI news altogether “just to stay sane.” It’s fine if colleagues vent about AI, but remind them it’s just off-the-clock venting.
Start job hunting if asked to review piles of AI-generated slop. You will be burned out and eventually let go—it’s an inevitable outcome.

Developer arp242 shared his experience: “My last two jobs were found via HN’s ‘Who is Hiring’ thread. Nobody asked if I used AI. During the interview, they asked, ‘Do you need a Claude subscription?’ I said ‘No.’ They said ‘Okay,’ and that was it.”

Companies that refrain from AI theater do exist; they are often too small to feature prominently on major job platforms, but they are out there.

Epilogue

Returning to the article that sparked global debate: Nikhil’s team now turns down all AI implementation requests. The reason is simple—every AI project they’ve seen across every company has failed. He wrote: “All of our existing contracts would be completely unaffected even if OpenAI went under tomorrow.”

That might sound extreme. But when someone connected to over 300 technical networks and close to Fortune 500 boardrooms makes such a statement, it warrants pause: Does your company really need that AI chatbot?

Or is your boss afraid that without preaching “AI-native transformation” to the board, someone else will be sitting in his chair tomorrow?

If it’s the latter, the problem isn’t AI. The problem is that the organization has lost the capacity for rational decision-making. And that is what is truly alarming.

References: