Richard Sutton, a leading figure in reinforcement learning, has reignited the AI development debate by reaffirming his 'bitter lesson' on May 19. Sutton advocates for focusing on scalable general methods powered by compute, rather than relying on human knowledge. This stance has drawn criticism from AI experts like Gary Marcus, who argues that current large models still depend heavily on human knowledge. Thomas G. Dietterich also contends that understanding human cognitive structures is crucial for identifying AI's limitations. The debate highlights differing views on the balance between computational power and human input in AI's evolution.