We should be more tired than the model
In a reflective essay on Hacker News, Vicki Boykis critiques the AI industry's focus on scaling models without considering the human toll. She points out that while models like GPT-4 and Claude 3.5 Sonnet improve, the people who deploy, monitor, and maintain them face increasing burnout. Boykis notes that developers are expected to keep up with rapid releases, shifting APIs, and new infrastructure demands, often without adequate support. She argues that this unsustainable pace leads to decreased productivity and morale. The essay calls for a more balanced approach that values human well-being alongside technological advancement, suggesting that acknowledging fatigue could lead to better long-term outcomes for the industry.
Developers must recognize that AI progress comes with human costs, urging sustainable practices.