From the course: AI Applications in Healthcare: A Conversation with Dr. Matthew Lungren

Why design proprietary AI healthcare systems?

- Building these proprietary systems within organizations to maintain data and to have greater control and a lot of it is also making it be more specific. Chat GPT is very general because it looks at all of the data for the entire internet, in the entire world. So to be able to really guide these systems in a specific way for the workflows and processes of a particular institution is so important and it increases efficiency even more, so. - Yeah, I mean, anyone who's using the tools, and hopefully some of the audience, they're already starting to get familiar with them, but the ability to chat with it, that's great, and you often unlock all kinds of insights, and it's really interesting, but the ability to put in a document, let's say a specific document, maybe it's your insurance contract, the terms, right? And then have a conversation with the document, that is just mind blowing, and that is so powerful. And so that's why having one of these systems in that firewall, so that if you are going to upload a document to have a conversation with, you don't have to worry, is there sensitive information. That document might actually, potentially having a breach or a problem with my compliance by sending it to, again, a commercially available solution versus one specifically designed for my organization or my health system. I think that takes that, that sort of ratchets down the stakes a little bit and allows you to experiment more freely with the model.

Contents