Alex Champandard Claims Anthropic Leads via Book Data
Researcher says OpenAI relies on compute to match knowledge benchmarks.
TLDR
Alex J. Champandard stated that Anthropic models outperform on benchmarks requiring knowledge because they scanned millions of books. He contrasted this with OpenAI, which he said can only attempt to catch up through extra compute. The remark appears in a post from the AI researcher, developer, and entrepreneur who co-founded CreativeAI and focuses on game AI, ML tools, and creative applications.
Combined views
1K
1 Source, first seen 27d ago