Andy Halliday
๐ค SpeakerAppearances Over Time
Podcast Appearances
He was working on tools for the people inside Meta
to help them build for Instagram and Facebook groups.
And then he moved over to Anthropic, and then he, as something to help himself, he used Anthropic's tools to build Claude Code, and then...
It's it propagated within Anthropic and then became a whole product line like went out to the world.
So now people will remember Boris Cherny's name in the annals of coding.
And he just, you know, he followed a trajectory up to including building something that's really quite remarkable.
Yeah, a couple other points on the healthcare front.
One is that just after OpenAI launched their GPT healthcare, Anthropic just launched Claude for healthcare.
So it's also a HIPAA compliant offering for providers and for insurers and for patients.
Uh, and it connects to medical systems.
Uh, but you know, I haven't used either chat GPT health or Claude for healthcare.
Um, and it'll be interesting to do that.
One of the things that is starting to be enabled by this collection of information from multiple healthcare providers into your chat GPK health or other sort of AI-assisted interpretation of all of those factors is the predictive abilities of analysis across a very large assemblage of data across many, many, ultimately millions of people.
And so I want to hearken to another thing that I read about over the weekend, which is that Stanford did something with AI with all of the sleep study information that they've done from Stanford's, you know, sleep study system.
So you can go to a hospital and you go for an overnight sleep study and they hook you up to all kinds of monitors and they wash what the patterns are of your sleep.
Let's note that, you know, as Beth mentioned when she talked about sleep studies a few days ago.
You don't sleep that well when you're hooked up to all these monitors.
Nonetheless, that information is now correlatable with all of the health information that Stanford has.
that you can now have access to, obviously, for your own account.