Tech CEOs are apparently suffering from AI psychosis

The tech industry is currently experiencing a tumultuous phase reminiscent of past revolutionary shifts, such as the early days of cloud computing. However, this period is marked by unprecedented phenomena, including soaring revenues alongside significant layoffs. A prevailing theory suggests that many tech executives, particularly CEOs, may be falling victim to what some are calling 'AI psychosis'. Box founder Aaron Levie has been vocal about this issue, arguing that CEOs are particularly susceptible to inflated perceptions of AI's capabilities. He points out that these leaders often remain distanced from the intricate realities of AI implementation. For instance, they may experiment with AI tools, develop prototypes, or draft contracts, leading them to mistakenly believe that AI can autonomously handle all necessary tasks. In reality, these executives do not engage in the nitty-gritty work of code reviews, bug detection, or the meticulous training of AI models tailored to their company's specific needs, as Levie suggests. Despite his critical viewpoint, Levie is not against AI. He actively promotes AI advancements to his 2.7 million followers on social media and invests in AI startups, demonstrating his belief in the technology's potential. He recommends that CEOs immerse themselves in AI usage to genuinely grasp its capabilities and limitations, urging them to engage deeply with the technology to understand both its advantages and the substantial groundwork required. The urgency of this advice is underscored by the staggering wave of layoffs across the tech sector. In just the first five months of 2026, over 115,430 employees have lost their jobs across 152 tech firms, nearing the total layoffs recorded for all of 2025, which was 124,636 across 275 companies. Many firms cite AI as a justification for these redundancies, sparking debates about whether major companies are merely using AI as a scapegoat to mask deeper issues behind workforce reductions. A striking example comes from Zeb Evans, CEO of ClickUp, who announced the dismissal of 22% of his staff after deploying around 3,000 AI agents for internal tasks. He insists that the layoffs were not cost-driven but part of a vision for an organization focused on individuals who manage AI agents. However, current research offers a more cautious perspective on AI's impact on productivity. A meta-analysis published in UC Berkeley's California Management Review found little correlation between AI adoption and significant productivity gains. Moreover, findings by the National Bureau of Economic Research indicate that while AI may enhance productivity, there's a discrepancy between perceived and actual gains. Research from MIT suggests that AI agents still struggle to match human performance in many areas, with a forecast that they will only achieve basic competence by 2029. As AI becomes more prevalent, the burden of managing the output created by AI will likely shift to executives, potentially leading to chaos if they are unprepared to handle it. In summary, while the promise of AI is vast, the ongoing disconnect between tech CEOs and the operational realities of AI poses a serious risk to organizational stability.