Why I Believe Entire Companies Right Now Are Under AI Psychosis

The pressure to adopt artificial intelligence has reached fever pitch. Board meetings now open with AI mandates. Quarterly earnings calls feature promises of AI-driven transformation. Marketing departments rebrand existing products with “AI-powered” labels. Yet beneath this frantic activity, a quieter concern is spreading among technology leaders: I believe entire companies right now are operating under what amounts to collective psychosis around AI—a state where organizational rationality has been suspended in favor of hype-driven decision-making.

This isn’t hyperbole from luddites resisting change. This is an observation emerging from experienced technologists watching their peers make decisions that contradict decades of software engineering wisdom. The concern isn’t that AI adoption is happening. The concern is that it’s happening divorced from any clear business rationale, technical feasibility assessment, or honest cost-benefit analysis.

What we’re witnessing is organizational behavior driven by fear of irrelevance rather than strategy. When an entire company pivots toward AI not because customers demand it or because it solves a real problem, but because competitors mentioned it in earnings calls, that’s not innovation. That’s panic wearing a technology costume.

The Current State: How Entire Companies Became Convinced They Need AI Everything

The timeline matters here. Through 2023 and into 2024, generative AI moved from research curiosity to mainstream consciousness. ChatGPT hit 100 million users faster than any application in history. Every major cloud provider released their own AI models. Venture capital flooded the sector with unprecedented funding. The narrative solidified: AI is the future, and companies that don’t adopt it will die.

This narrative contains truth. AI will reshape industries. Some companies will gain competitive advantage through intelligent deployment of machine learning. But somewhere between “AI matters” and “AI must be in everything,” organizational decision-making broke down.

I believe entire companies right now are experiencing this breakdown in real time. Consider what’s actually happening in many enterprises: IT departments are tasked with identifying “AI opportunities” without clear criteria for what constitutes an actual opportunity. Product teams are retrofitting AI into roadmaps where it wasn’t previously planned. Executives are committing budget to AI initiatives that lack measurable success metrics. The question changed from “does this problem need AI?” to “how do we add AI to this?”

The pressure compounds from multiple directions simultaneously. Investors ask about AI strategy. Competitors announce AI initiatives. Industry analysts publish reports ranking companies by AI maturity. Consulting firms pitch expensive AI transformation programs. Employees worry their skills will become obsolete. In this environment, the rational response—pause, assess, plan deliberately—feels like the reckless response. Moving fast and breaking things, once a startup ethos, has become a corporate survival instinct.

What’s particularly revealing is how this plays out in practice. Companies hire AI teams without clear reporting lines or integration with existing product development. They purchase expensive AI platforms without understanding what problems those platforms solve. They launch AI features that users never requested because the feature exists, not because it addresses a genuine need. The outputs of these decisions often look identical to what the company was already doing, just with “AI-enhanced” in the marketing copy.

Why This Matters: The Hidden Costs of Organizational AI Psychosis

The consequences of this collective delusion extend beyond wasted budget and embarrassing product launches. I believe entire companies right now are experiencing genuine organizational damage from their AI psychosis, and the damage is often invisible until it’s severe.

First, there’s the opportunity cost. Every engineer assigned to retrofit AI into a product that doesn’t need it is not working on features users actually want. Every dollar spent on AI infrastructure that sits mostly idle is not available for core business improvements. Every meeting spent discussing AI strategy instead of customer problems is time not spent solving those problems. Companies are burning resources on the technology equivalent of cargo cults—mimicking the forms of AI adoption without understanding the function.

Second, there’s the talent distortion. The best engineers want to work on meaningful problems. When a company’s stated priority becomes “add AI everywhere,” talented people either leave for companies with clearer missions or stay and become cynical. The engineers who thrive in this environment are often the ones most comfortable with ambiguity and politics, not necessarily the ones best equipped to build solid systems. Over time, this selects for organizational dysfunction.

Third, there’s the customer experience degradation. When a company forces AI into products where it doesn’t belong, users notice. They experience worse performance, unexpected behavior, or features that solve problems they don’t have. The company’s reputation suffers. Trust erodes. And the company blames “AI adoption challenges” rather than their own decision-making.

Fourth, there’s the technical debt accumulation. AI systems, particularly large language models, are expensive to run and maintain. They require continuous retraining. They have failure modes that are hard to predict. When a company deploys AI systems without genuine use cases, they’re not just wasting money in the moment—they’re creating ongoing operational liabilities. The technical debt from this period will haunt these organizations for years.

Perhaps most importantly, there’s the strategic clarity loss. Companies that are genuinely thinking about where AI creates value are making different decisions than companies operating under psychosis. The former are being selective, building AI capabilities in specific domains where they have competitive advantage. The latter are trying to be AI companies, which is a different and usually worse strategy. By the time the psychosis breaks, years have passed and the company’s strategic position has deteriorated.

How This Plays Out: The Mechanics of Corporate AI Delusion

Understanding how companies arrive at this state requires looking at the actual decision-making processes. The mechanism isn’t malice or stupidity. It’s organizational psychology meeting market pressure meeting information asymmetry.

Start with the C-suite. Executives are told by consultants, analysts, and the business press that AI is essential. They lack the technical depth to evaluate these claims independently. So they accept them as true. They then ask their technical teams, “What’s our AI strategy?” The technical teams, facing the same pressure, propose initiatives. Because the executives don’t have criteria for evaluating these proposals beyond “does it sound like AI?”, most proposals get approved.

Add board pressure. Public company boards now include questions about AI in every meeting. Investors ask about AI moats and AI-driven competitive advantage. Private equity firms pitch AI-driven operational improvements. The board pressure flows downward as executive compensation increasingly tied to AI metrics—number of AI initiatives launched, percentage of products with AI features, AI team size. These metrics measure activity, not value, but they’re quantifiable and therefore easy to manage toward.

Layer in competitive anxiety. When a competitor announces an AI initiative, the response is immediate panic followed by imitation. If Competitor A launches an “AI-powered customer service bot,” Competitor B must launch one too, regardless of whether customers actually want it. This creates a prisoner’s dilemma where all companies are compelled to make suboptimal decisions because not making them feels riskier than making them.

Add vendor incentives. Cloud providers, software vendors, and consulting firms all benefit from corporate AI spending. They have strong incentives to convince companies that they need AI solutions. These vendors employ some of the industry’s best salespeople, and they’re armed with research reports, case studies, and analyst relations programs designed to create demand. The information asymmetry is severe—vendors know far more about what actually works than the companies they’re selling to.

Finally, add individual career incentives. The engineer who proposes a risky AI initiative and it fails can blame “AI adoption challenges.” The executive who commits budget to AI can claim they’re “forward-thinking.” The consultant who recommends expensive AI transformation gets paid regardless of outcomes. The individual incentives all point toward more AI spending, not toward honest assessment of whether AI spending creates value.

In this environment, I believe entire companies right now are trapped in a system where rational actors are making individually rational decisions that collectively produce irrational outcomes. This is the definition of psychosis at the organizational level.

What Industry Observers Are Saying: Reality Checks From Experienced Voices

The concern about corporate AI psychosis isn’t fringe thinking. Experienced technologists are increasingly vocal about what they’re observing. Mitchell Hashimoto, founder of HashiCorp, articulated the concern directly: the observation that entire companies are operating under AI psychosis reflects a real pattern that multiple observers have noticed independently.

The concern comes from people with credibility. These are founders who’ve built companies, engineers who’ve shipped products at scale, investors who’ve funded dozens of startups. They’re not anti-AI. They’re pro-reality. What they’re seeing is companies making decisions that contradict what they know about building successful products.

Enterprise software leaders report similar observations. Companies are implementing AI solutions that don’t integrate with their existing workflows. They’re training employees on AI tools that won’t actually be used. They’re reorganizing teams around AI initiatives that lack clear objectives. The pattern is consistent across industries and company sizes.

What’s notable is the lack of disagreement from technical communities. When experienced engineers observe that something is broken, they typically defend it or explain why it’s necessary. In this case, the response from the technical community has been largely silent acknowledgment. People recognize the pattern because they’re living it.

What Comes Next: When Psychosis Breaks and Reality Reasserts

Collective organizational psychosis doesn’t last forever. Reality eventually reasserts itself. The question is what happens when it does.

The most likely scenario is a gradual deflation rather than a sudden crash. Companies will continue spending on AI, but the growth rate will slow. Some AI initiatives will be quietly cancelled. Others will be rebranded as “optimization” or “automation” to avoid admitting they were AI projects that didn’t work. Executives will shift from “we’re an AI company” to “we use AI where it makes sense.” The language will change before the behavior does.

The companies best positioned for this transition are those that maintained some skepticism throughout the hype cycle. They asked hard questions about ROI. They didn’t staff up massive AI teams. They treated AI as a tool, not a religion. When the psychosis breaks, they’ll barely notice because they never fully bought in.

The companies that will suffer most are those that made AI central to their identity. They hired aggressively for AI roles. They reorganized around AI. They committed multi-year budgets to AI initiatives. When the psychosis breaks, they’ll face difficult conversations about why they have AI teams with nothing to do and products that don’t need the AI they’ve built.

For the broader industry, the aftermath of this cycle will likely produce valuable lessons about how organizations make technology decisions under uncertainty. It will probably spawn a new wave of writing about organizational decision-making, risk management, and the dangers of herd behavior in technology. The cycle will repeat with the next technology—quantum computing, perhaps, or whatever comes after—and companies will make similar mistakes because the underlying organizational dynamics haven’t changed.

The more optimistic scenario is that some companies learn from this period and develop better decision-making frameworks. They might implement requirements that new technology initiatives must solve specific, measurable problems. They might separate the hype cycle from actual product development. They might hire skeptics and give them real power. These companies would be better positioned not just for the next technology cycle, but for all future decisions.

FAQ

Conclusion

I believe entire companies right now are operating under AI psychosis not because AI lacks value, but because organizational decision-making has become untethered from reality. The pressure to adopt AI has overridden the careful analysis that built successful technology companies in the first place. This period will eventually end, and when it does, the companies that maintained some skepticism will be better positioned than those that bought in completely. The broader lesson is about organizational decision-making under uncertainty—a problem that will outlast any particular technology hype cycle.

The question for companies now isn’t whether to adopt AI, but whether to do so thoughtfully or thoughtlessly. The distinction will determine which companies thrive in the years ahead.

– Auburn AI editorial