Reactions from ranked influencers
2 posts
David Sacks on cyber guardrails. It's difficult to say how much of this is directed at Anthropic and OpenAI, and how much is directed at the administration.
Here’s another example: Hugging Face tried using American frontier models to analyze an AI-powered cyber attack. But the guardrails blocked requests containing real exploit payloads so they switched to GLM 5.2 running locally. The guardrails actually impaired defensive security. https://twitter.com/clementdelangue/status/2078987852495364398
2/
Here’s another example: Hugging Face tried using American frontier models to analyze an AI-powered cyber attack. But the guardrails blocked requests containing real exploit payloads so they switched to GLM 5.2 running locally. The guardrails actually impaired defensive security. https://twitter.com/clementdelangue/status/2078987852495364398