The comfortable assumption that technology will always create more jobs than it destroys is facing a fundamental challenge. On The Diary Of A CEO, former OpenAI safety researcher Daniel Kokotajlo detailed why the dawn of general superintelligence breaks historical economic rules. Meanwhile, NVIDIA has expanded open-weight enterprise capabilities with a new family of high-performance embedding models designed to solve context errors in AI retrieval and memory systems.

We’ll talk about:

• Why traditional safety-net logic fails when AI reaches superintelligence and becomes cheaper, faster, and more adaptable than humans at every emerging task.
• A new open-weight embedding model taking top honors on the RTEB retrieval leaderboard, built specifically to eliminate search loops and lower to ...