Chris Willis, Domo CDO, argues the tech industry's fear-driven AI rush creates more problems than solutions. Speaking from San Francisco's AI hub, he expresses bafflement at the public's lack of resentment toward companies pushing anxiety-inducing technology. He contends fear is not a sustainable innovation strategy, and the current marketing approach leaves organizations confused and leaders feeling threatened.

Willis identifies the core issue as AI models being marketed as products without a clear specification. Unlike traditional products designed for specific users and tasks, large language models promise to do anything for anyone, anywhere. This ambiguity leads to 'tokenmaxxing'—buying access and pushing usage without strategy—which he labels as theater rather than genuine innovation, failing to improve the bottom line despite increased individual productivity.

Companies err by treating AI as the solution itself instead of a tool to enable solutions, resulting in non-durable proof-of-concepts. Willis cites Klarna's failed attempt to replace customer service staff with AI, only to reinstate humans, as a cautionary tale. He stresses starting with simple automations, like invoice anomaly detection, and understanding where human judgment is irreplaceable.

His advice is to abandon moonshot goals and proceed methodically. The coming budget scrutiny from CFOs, questioning massive AI spend without returns, will force a reckoning. Organizations must do the hard work of aligning AI with actual business processes, or risk driving their business 'with the lights off, at night' using a powerful but poorly understood engine.