Karan Singhal

I lead the Health AI and AGI Benefits teams at OpenAI, working towards safely securing the benefits of AI.

Our work aims to:

– Improve human health with AI
make health intelligence accessible to everyone
improve performance and safety for 230M+ people who use AI for health weekly
enable new 0→1 use cases where greater intelligence will most improve human health
– Make models deeply beneficial: understand, measure, and improve the ways models can be made broadly and persistently beneficial to humanity
– Make tangible AGI benefits real: scale AI’s benefit to humanity across domains

Questions I’m interested in currently:

– Where will increases in intelligence most marginally improve human wellbeing?
– How can we differentially accelerate societal resilience to advanced AI?
– Where could today’s models be better, for patients, clinicians, healthcare, or scientists?

My previous work at Google includes Med-PaLM and Med-PaLM 2, launched in Google Cloud and featured in Nature, Scientific American, The Economist, BBC, and other outlets.