This is a YouTube video titled "The PhD Who Masters People Too | Agentic" containing a monologue or single-speaker commentary discussing the impact of Deepseek on the AI community, particularly its open-sourcing of reinforcement learning techniques previously held proprietary by OpenAI, and the implications for AI research and competition among leading organizations.
Deepseek moment is an amazing event that happened for the entire AI community. They open sourced the ideas that OpenAI was holding really proprietary for themselves. And that made OpenAI freak out because they were thinking that they are the only ones that are holding like this complex reinforcement learning steps that they were doing to design 0103 and the reasoning kind of generation of models. But Deepsea came in and did a basic RL actually showed that you can actually get to the same level of performance. It opened up massive amount of opportunities for research based on the capacity that OpenAI and anthropic teams already have. These guys are gonna stay on top simply because of the resources available to them, so they can scale the same ideas up and get to like better versions of intelligence.
The AI Agent Getting More Supplier Data in 5 Weeks Than a Team Did in 2 Years
