Wes Roth discusses Anthropic’s Fable 5 and Mythos 5 release, comparing reported coding and reasoning benchmarks and describing game demonstrations. He treats the benchmark claims as results to test independentlyEvaluation is the systematic assessment of a system's behaviour or results against explicit criteria..
He also reviews reported biology and cybersecurity risks, classifier-based routing and observations from the system card, including interactions between parallel agents. His comments distinguish the public Fable release from restricted Mythos accessGovernance sets responsibilities, rules and oversight for decisions about how AI systems are used..
Watch on YouTube
