OpenAI Podcast Episode 16 - Building AI for Life Sciences
85 snips
Apr 16, 2026 
Yunyun Wang, product lead on OpenAI’s life sciences models focusing on bio-risk mitigation. Joy Jiao, research lead linking AI to wet labs and systems biology. They discuss building biology-focused models, safeguards and differentiated access, AI-driven lab automation and autonomous labs, and near-term impacts like drug repurposing and assay-driven evaluation.
AI Snips
Chapters
Transcript
Episode notes
Life Sciences Model Series For Early Discovery
- OpenAI built a life sciences model series focused on mechanistic genomics and protein understanding to accelerate early discovery.
- The team adds model orchestration and templatized research plugins (50+ skills) to make repeatable workflows one-click deployable for translational bio users.
GPT-5 Designed Experiments That Made Protein
- Joy recounts a GPT-5 and Ginkgo Bioworks collaboration where the model designed experiments that produced a non-zero amount of protein.
- That first wet‑lab success shifted the team's belief from uncertainty to clear evidence that models can accelerate real biology.
Dual Use Makes Bio Safeguards Hard And Necessary
- Bio capabilities raise dual-use risks because benign workflow steps mirror malicious ones, making information-hazard detection hard.
- OpenAI uses a risk‑averse stance, layered mitigations, and is moving toward differentiated access for trusted professional users.