A software engineer has sounded the alarm about a growing issue in the tech world: AI fatigue. Siddhant Khare, a software engineer at ONA, shared his experience in a candid essay titled "AI fatigue is real and nobody talks about it," describing how AI tools have boosted his productivity while simultaneously leaving him more drained than ever. Khare noted that while he shipped more code in a single quarter than ever before, he also felt more exhausted than at any point in his career. He likened his current role to that of a perpetual reviewer, constantly checking and refining code generated by AI, like a judge at an endless assembly line. "Every time it feels like you are a judge at an assembly line and that assembly line is never-ending, you just keep stamping those PRs," he wrote. The paradox he identifies is that AI reduces the cost of writing code but dramatically increases the cost of coordination, review, and decision-making. Instead of focusing deeply on one problem, Khare now juggles up to six different tasks at once, each supposedly solvable in an hour with AI. But the mental toll of constant context switching is immense. "The AI doesn't get tired between problems. I do," he said. His story resonated widely across developer communities on platforms like X, Hacker News, and Lobsters. Many engineers echoed his sentiments, describing how AI has turned coding into a cycle of prompts, waiting, and editing—often leading to distraction and mental fatigue. One user joked about following a "Claude Code workout plan," incorporating movement during coding sessions to stay alert. Others described their days as chaotic, filled with the "C'mon do something" meme energy, waiting for AI agents to respond before moving on. A recent Harvard Business Review report based on an eight-week study at a U.S. tech firm found that AI tools didn’t reduce workloads—they intensified them. After the initial excitement fades, employees often find their responsibilities have quietly expanded, leading to cognitive fatigue, burnout, and poorer decision-making. Khare also highlighted the pressure to keep up with rapid AI advancements from companies like OpenAI and Anthropic. He admitted to spending weekends evaluating new tools, reading changelogs, and watching demos—driven by fear of falling behind. This constant FOMO, he said, is exhausting. Perhaps most concerning, he reflected on how AI is changing his core abilities. He struggled when asked to solve a concurrency problem on a whiteboard without a computer or AI. "It's like GPS and navigation. Before GPS, you built mental maps. After years of GPS, you can't navigate without it," he wrote. The skill atrophied from disuse. Andrej Karpathy, former AI lead at Tesla and the coiner of "vibe coding," has also expressed concern about the erosion of manual coding skills. He admitted he’s starting to lose his ability to write code from memory. To combat this, Khare has established personal rules—like limiting AI use to 30-minute intervals and taking a full 14-day break from AI discussions. He also believes AI companies have a responsibility to build in safeguards, comparing the addictive nature of AI prompts to game design tactics meant to keep users engaged. "You need to keep some sort of guardrails for the humans, so they don't self-distruct themselves," he said.