AI progress continues, with Anthropic's Michael Gerstenhaber stating that advancements in language models are driven by self-correction and reasoning capabilities. Despite concerns from AI skeptics, the company reports steady improvements, unlocking new use cases with each model revision.

Source: https://www.zdnet.com/article/ai-isnt-hitting-a-wall-its-just-getting-too-smart-for-benchmarks-says-anthropic/