Descriptions:
The second half of a Cognitive Revolution live event brings together three guests to examine AI from the intersecting perspectives of biomedicine, research automation, and existential risk. Abhi Mahajan — founder of Noetic AI and known online as “owlposting” — opens the discussion on AI for cancer treatment selection, explaining how foundation models can help predict which patients will respond to which therapies. He describes the growing trend of acquiring pre-clinical drug assets from China’s biotech pipeline and expresses calibrated skepticism about many published AI-for-biology results while remaining optimistic about long-term trajectory.
Helen Toner of CSET presents findings from a closed-door workshop behind the organization’s “When AI Builds AI” report. Participants failed to reach any consensus expectation about how automated AI R&D will unfold, leading to the conclusion that it represents a major source of potential strategic surprise. Toner frames the uncertainty along three axes — the length of the lead-up period before rapid capability growth, the steepness of the S-curve, and the ultimate ceiling — and notes that participants tend to cluster into two polarized camps across all three dimensions simultaneously, leaving underexplored scenarios like a steep curve with a low ceiling or a gradual curve with a high ceiling.
Jeremie Harris closes the episode by arguing the safety field currently lacks both reliable technical control mechanisms for superhuman AI systems and the trust and coordination infrastructure needed for the US and China to address this collaboratively. The conversation also explores how each guest maintains situational awareness and how AI tools are reshaping personal productivity stacks.
📺 Source: Cognitive Revolution “How AI Changes Everything” · Published February 14, 2026
🏷️ Format: Podcast
