A principle in artificial intelligence asserting that general methods leveraging computation and data, rather than human-designed domain knowledge, ultimately yield the most effective and scalable results.
Overview
- Rich Sutton’s 2019 essay ‘The Bitter Lesson’ argues that general methods leveraging computation are ultimately the most effective in AI, citing computer chess as a primary example where brute-force search defeated human-designed strategies. (Source: Rich Sutton, via AI Daily Brief, 2026-08-24)
- Bloomberg GPT, a 50-billion-parameter large language model purpose-built for finance, was outperformed by general-purpose models, illustrating the ‘bitter lesson’ that general methods leveraging computation are most effective. (Source: Host (AI Daily Brief), via AI Daily Brief, 2026-08-24)