President of Microsoft Science saw ChatGPT coming (and now he predicts how it will change healthcare) | Peter Lee

In this episode of Lifers, Christina Farr and Peter Lee, President of Microsoft Science explore the evolution of AI from its early "neural net" origins to the groundbreaking scale of GPT-4. Peter shares insider insights on Microsoft’s strategic partnership with OpenAI and how AI is transforming specialized fields into "full stack" generalist roles. They dive deep into the future of healthcare, highlighting how AI could eliminate administrative "coordination" layers while significantly increasing clinical patient flow. Dive into the Granola notes from this episode: https://notes.granola.ai/t/58d5e88a-5b17-46fa-85be-c3fde5d3e8f3-009c2hma
—
SPONSOR: Granola
Granola AI, The AI notepad for people in back-to-back meetings: https://granola.ai/lifers with code: LIFERS

Interested in sponsoring the show? lifers@a16zstudios.com 

—
LINKS:
Lake Nona Impact Forum: https://lakenonaimpactforum.org/ 
Microsoft Research Podcast with Chrissy and Dave deBronkart: https://www.microsoft.com/en-us/research/podcast/the-ai-revolution-in-medicine-revisited-empowering-patients-and-healthcare-consumers-in-the-age-of-generative-ai/ 
Attention is All You Need (Google Research): https://arxiv.org/pdf/1706.03762 
The Pause Letter: https://futureoflife.org/open-letter/pause-giant-ai-experiments/ 
The 2028 Global Intelligence Crisis: https://www.citriniresearch.com/p/2028gic?source=queue 
Lifers episode with Dr. Bob Wachter: https://www.youtube.com/watch?v=p6U1fX1JgVk 

Chrissy Farr’s Website: https://www.chrissyfarr.com/ 
Subscribe to the Second Opinion Newsletter: https://secondopinion.media/ 
Chrissy’s Book: The Storyteller's Advantage: https://www.chrissyfarr.com/books 

Lifers with Christina Farr on YouTube: https://www.youtube.com/@LiferswithChristinaFarr 

—
FOLLOW:
Peter:
https://www.linkedin.com/in/peterlee4/ 
https://x.com/peteratmsr 

Chrissy:
https://www.linkedin.com/in/christinafarr/ 
https://x.com/chrissyfarr 

—
TIMESTAMPS:
(00:00) Preview
(01:23) Intro
(03:12) Evolution from deep neural nets to modern AI
(04:47) Demystifying model weights and the shift to training
(07:04) Breakthroughs in speech recognition and computer vision
(09:25) The arrival of transformers and the attention mechanism
(11:28) Why Microsoft invested when the research community ridiculed
(13:54) Addressing the Pause Letter and the need for caution
(16:44) AI as the moral equivalent of the transistor
(19:27) Sponsor: Granola
(20:06) How AI native teams restructure software development
(22:29) Empowering everyone from junior devs to the CEO
(28:01) Integrating AI into frontline clinical encounters and avatars
(32:41) The hidden costs and upcoding risks of ambient scribes
(35:48) Eliminating the coordination drag to improve patient throughput
(47:45) Displacing administrative and payer-side adjudication roles
(49:53) Predicting the decline of medical subspecialties
(53:05) Distinguishing between mathematical proof and human creativity
(56:50) Wrap