AI promised less work. I've never worked more

Your daily engineering output has likely increased due to AI agents, even if your core responsibilities haven't changed.

Orestis Ioannou, a VP of Technology, has observed a significant increase in his workload since adopting AI agents for engineering tasks. While AI was pitched as a tool to reduce work, Ioannou's GitHub contribution data from 2019 to 2026 shows a sharp rise in activity in 2026, nearing his busiest period as a full-stack engineer in 2020. He is now handling multiple AI agents concurrently, each requiring intermittent attention for approval, review, or analysis. This constant context switching between numerous tasks, unlike the previous model of handling two to three focused tasks, has become mentally draining. Despite a career shift towards architecture and strategy, the low cost of AI implementation has enabled him to pursue more ideas and fixes, resulting in bursts of intense activity followed by periods of burnout, with even the "low seasons" showing higher output than before.

The proliferation of AI agents is fundamentally altering engineers' workflows, shifting the bottleneck from task execution to cognitive load management. Previously, engineers juggled a few complex projects; now, they manage numerous asynchronous AI-driven tasks, leading to context-switching fatigue. This suggests that while AI dramatically lowers the cost of implementation, the human capacity to manage and direct this parallelized work is becoming the limiting factor.

The observed pattern of "bursts and burnout" indicates that current AI adoption strategies are not sustainable for maintaining consistent, high-level output without risking engineer well-being. Companies need to develop frameworks for managing AI-augmented workloads, focusing on reducing context switching and prioritizing tasks to prevent burnout, rather than simply enabling more work to be done.