Ammar Sabilarrohman
The tech industry is currently gripped by a peculiar duality. On one hand, generative AI is being hailed by boardroom executives as the most transformative economic force since the advent of the internet. On the other, a growing chorus of users, employees, and even industry veterans is pushing back, signaling a deep-seated exhaustion with the relentless, often unpolished, integration of AI into daily digital life.
At the center of this friction is a provocative critique from Box founder Aaron Levie, who recently suggested that tech CEOs are suffering from "AI psychosis"—a state of detachment from the actual, granular utility of the tools they are mandating across their organizations. This sentiment has become a focal point of debate on TechCrunch’s Equity podcast, where hosts Kirsten Korosec, Sean O’Kane, and Anthony Ha recently dissected the growing chasm between executive vision and user reality.
The Main Facts: The "Psychosis" Diagnosis
The term "AI psychosis," as coined by Levie, does not suggest a clinical mental health crisis, but rather a strategic one. It describes a phenomenon where leadership, insulated from the "last mile" of labor—the actual, tedious work required to make AI functional and reliable—becomes enamored with high-level projections of efficiency.
Levie’s argument is not that AI is useless, but that executives are failing to engage with the tools they preach. When leadership views AI through the lens of investor slide decks rather than the lens of a daily user, they lose sight of the friction points, the hallucinations, and the genuine user experience degradation that often accompanies "AI-first" product updates.
Chronology of the Backlash
The current state of "AI fatigue" did not emerge in a vacuum. It is the culmination of several distinct phases of industry behavior over the past 18 months:
- Phase 1: The Gold Rush (Late 2023–Early 2024): Following the massive success of LLMs, almost every major tech firm pivoted to an "AI-first" strategy, often prioritizing feature launches over product maturity.
- Phase 2: The "Bad Vibes" Era (Mid-2024): As widespread layoffs coincided with record-breaking investments in AI infrastructure, employees began to view AI not as a productivity multiplier, but as a direct threat to job security.
- Phase 3: The User Revolt (2026): Recent public reaction has shifted from curiosity to frustration. Events such as college commencement speakers being booed for mentioning AI, and a notable surge in users fleeing major platforms for privacy-focused or "non-AI" alternatives, suggest the tide is turning.
- Phase 4: The Pivot to "Anti-AI" Marketing: Search engines like DuckDuckGo have recently seen a 30% surge in installs, a trend attributed to users rejecting the "force-fed" AI search results now dominating Google’s interface.
Supporting Data: The Cost of Over-Innovation
The friction between corporate ambition and user preference is visible in current market data. Google, the titan of information retrieval, serves as the most prominent case study. By integrating generative AI into the core search experience, the company aimed to stay ahead of the curve. However, the result has been a dilution of the very utility users rely on.
When a user asks a simple question—such as "How many Ps are in the word Google?"—and receives a hallucinated answer, the brand equity of the search engine suffers. This is not just a technical error; it is a fundamental miscalculation of what users value.
Data points from the current market climate include:
- Search Migration: DuckDuckGo’s 30% increase in installs serves as a proxy for a significant, albeit minority, segment of the population that is actively seeking a "pre-AI" search experience.
- The "10 Blue Links" Demand: Even as Google pushes its "AI Overviews," it has been forced to maintain traditional search formats, acknowledging that a large portion of its user base still requires the accuracy of direct information retrieval over the synthesis of an LLM.
Official Responses and Industry Perspectives
On the Equity podcast, the hosts highlighted the contrast between Google’s reactive approach and the more focused strategy favored by companies like Anthropic.
Sean O’Kane noted, "There is a collapsing towards Anthropic’s approach—really trying to understand what it is you want to offer people and sticking to that." In contrast, Google appears to be suffering from "feature creep," attempting to solve commercial problems—like shopping and booking—within a space that users primarily view as an information repository.
Anthony Ha added that the polarization is real: "You can feel a little crazy because everybody’s using it and everybody loves it, but also no one’s using it and everybody hates it at the same time." This creates a paradox for developers: build for the AI-enthusiast, and you alienate the skeptical majority; ignore AI, and you risk appearing obsolete to investors.
Implications for Startups and the Future of Work
The "anti-AI" sentiment is not necessarily a death knell for the technology, but it does signal a massive market opportunity.
1. The Opportunity for "Anti-AI" Startups
As major incumbents prioritize AI at all costs, there is a clear "lane" for startups to offer "Human-First" or "Utility-First" tools. Companies that market themselves by explicitly stating, "We are not integrating AI into your workflow unless it is strictly necessary," are finding traction with users who are exhausted by the current tech landscape.
2. The Shift in Workforce Dynamics
The "AI-driven" layoffs that have defined much of the recent labor market are often top-down mandates. Unlike previous technological shifts—where tools like Slack or Excel were adopted bottom-up because they made workers’ lives easier—AI is often imposed from the executive level based on theoretical productivity gains.
As Kirsten Korosec pointed out, the disconnect lies in the fact that executives often don’t touch the "last mile" of work. If management views their staff as line items to be reduced through "AI efficiency," it creates a toxic internal culture. The true value of AI will likely not be found in downsizing, but in empowering workers—a nuance that is currently lost in the rush to cut costs.
3. The Future of Search and Information
The long-term implication for companies like Google is a potential loss of trust. If a search engine is no longer a reliable source of information, it becomes a "content generator," which is an entirely different (and potentially less valuable) service. The "AI psychosis" identified by Levie is essentially a loss of product identity.
Conclusion: The Path Forward
The tech industry is currently in a "trough of disillusionment" regarding the immediate application of generative AI. The way out is not more AI, but more intentionality.
Tech CEOs would do well to heed Aaron Levie’s advice: stop looking at the slide decks and start using the tools. They must bridge the gap between the board room and the workstation. For the industry to recover its equilibrium, it must move away from the "AI-at-all-costs" mentality and return to the foundational principles of user experience, accuracy, and utility.
The current "anti-AI" moment is not just a passing trend; it is a correction. It is a signal from the market that while users are open to innovation, they are not willing to sacrifice the reliability and clarity of their digital tools for the sake of an executive’s buzzword-filled roadmap. The companies that thrive in the next decade will be the ones that understand the difference between using AI to improve a product and using AI to replace the value that users actually came for.
Azzam Bilal Chamdy
Sagoh
