Veteran tech engineer **Steve Yegge — known for his years at Amazon and Google — is raising the alarm about what he calls “AI fatigue” among software developers as artificial intelligence tools become central to coding workflows. In a recent interview, Yegge described a “vampiric effect” where AI can initially boost productivity and excitement, but quickly drains energy and concentration, leaving engineers feeling exhausted rather than relieved. He says that despite AI’s promise, the mental cost of sustained high-intensity, AI-assisted work is becoming a serious real-world issue for developers.
Yegge argues that engineers should treat deep, high-output “vibe coding” sessions as a limited resource, suggesting that even the most productive developers may only be able to sustain about three hours of real coding work per day before burnout sets in. According to him, expecting longer periods of continuous, AI-driven productivity is unsustainable and can actually harm companies over time if leaders don’t adjust their expectations.
Part of the fatigue, Yegge says, comes from the constant acceleration and heightened output demanded by modern AI tools — which can make engineers feel like they must work harder and faster just to keep up. Some developers, including Yegge himself and his peers, have even joked about needing mid-day naps to cope with the intensity of their work under these conditions, underscoring how widespread this phenomenon has become in the tech community.
This concern over AI-related burnout echoes broader industry reporting about “AI fatigue” among engineers more generally, where productivity gains from AI tools can paradoxically lead to increased workloads, constant context switching, and mental strain rather than a lighter, more balanced work life. That broader trend is attracting attention as the tech sector — and beyond — adapts to the growing presence of AI in everyday professional tasks.
