
Sign up to save your podcasts
Or
In this episode we cover a concerning pre-print which shows that GPT-4 can exploit 87% of one-day vulnerabilities, a stealthy release of "gpt2-chatbot" that we know nothing about, and Danti's new AI that allows analysts to search satellite imagery with natural language. All in just three minutes.
In this episode we cover a concerning pre-print which shows that GPT-4 can exploit 87% of one-day vulnerabilities, a stealthy release of "gpt2-chatbot" that we know nothing about, and Danti's new AI that allows analysts to search satellite imagery with natural language. All in just three minutes.