Notes d’épisode
What happens when a powerful AI model can be downloaded, run locally, modified, and used without anyone watching?
In Episode 43 of The AI Desk, Rowan and Naya debate the growing concern around powerful open-weight AI models and cybersecurity risk. A new model release raises a bigger question: what happens when frontier-level capability becomes portable? They break down the tension between open-source innovation and AI safety, why cyber-capable models create a different kind of risk, and why closed models, open models, governments, companies, researchers, defenders, and attackers are all…
Site web: podcasters.spotify.com

