
This episode discusses Anthropic's decision to hold back the release of their AI model Mythos, the implications of compute costs, and the marketing strategies involved.
Mark Andre explains that the high compute costs associated with Mythos, estimated to be 10 to 20 times that of Opus, may have influenced Anthropic's decision. He suggests that by delaying the release, they could focus on the upcoming Opus 4.7.
Andre also mentions that this strategy created a perception of scarcity and altruism, which served as a marketing advantage. Government interest in the model increased due to the belief that it was being held back for its impressive capabilities.
He acknowledges that there may be genuine altruistic reasons for the delay, as Mythos could reveal coding vulnerabilities that need to be addressed by companies with large code bases.
Ultimately, the discussion raises questions about the balance between commercial viability and the ethical considerations in AI development.
Anthropic's delay of Mythos raises questions about compute costs and marketing strategies.

They’re holding it back because it’s so amazing.Why did Anthropic hold back Mythos?
Mythos does reveal coding vulnerabilities that people didn’t know about before.Why did Anthropic hold back Mythos?