GPT-4 Integration: Revolutionizing Healthcare in Hospitals

0:00Since its launch, one of the biggest areas that ChatGPT I believe can disrupt is in healthcare And today we're seeing a pretty big step in that direction. Epic, which is one of the biggest healthcare providers and I think hospitals in America, I think it is actually, you know, the biggest, it has the largest share of US acute care hospitals in that market. It has about 33 % according to a report. So Epic is a really big hospital healthcare company, and they are integrating GPT-4, and they have a pretty close partnership already with Microsoft that it would seem they're expanding. So today on the podcast, we're going to talk about what they're doing to actually integrate GPT-4 into healthcare, how that's going to change healthcare, what some of the implications are, and I think this is going to be an interesting one.

0:52So let's dive into the podcast. first off i will say that it's been five months now i think since microsoft first launched chat gpt so um what's interesting to know is usually around the six month mark after something like this comes out there is going to be a massive wave of companies that are finally in production right there's been a lot of stealth companies that have been developing in stealth and i know that chat gpt obviously can be integrated with an api a lot quicker into some saffs softwares than other things. But that being said, you know, really solid companies, sometimes it does take them a while to get out with some good products.

1:28And so I believe we're about to see a really massive wave of these products launching. And the reason why I think that that's relevant is because as we're also, well, that's happening, we're also seeing at the same time, some of these massive companies, right, one of the biggest hospitals in America is now integrating Chai GPT directly. And it's going to be interesting to see how that affects a lot of these startups and a lot of these SaaS services, if they're essentially, if they were, you know, essentially building a product that is now not going to be needed because hospitals are directly implementing, or if there's still going to be a place in the market for them, because I know there was a lot of healthcare SaaS companies coming out of ChatGPT.

2:07So that's going to be something interesting. That's not just with healthcare, that's with a lot of different areas. And I think it really comes down to if, you know, these companies are building features or fully fledged platforms and products, right so like if these if they're just doing like an app that does you know it like transcribes notes or summarizes something small for people it's probably going to get replaced or absorbed into a more you know a bigger platform or something that is a lot more robust of a software so that's going to be interesting but in any case epic is essentially launching this they're gonna it's gonna they're gonna be using a lot of the special stuff that microsoft has been building to the Azure platform.

2:47And apparently they've already begun piloting a lot of different open AI services at a few different healthcare locations, including UC San Diego Health, UW Health, and Stanford Healthcare. So a lot of the health systems have started using Microsoft's technologies to essentially automatically draft message responses. That's one thing that they're using it for. They actually said that we work with these organizations to identify a small subset of users who have been given training and context to leverage this feature in production the users are generally clinical information adjuncts and or clinical leads who see the promise ai has in shaping the future of healthcare and understand the considerations involved in being able to benefit from this technology effectively in these early phases right so they're just looking for early adopters people that see the see the potential and then they're using them to kind of test a lot of these features and once they find something that really works then they try to scale that out so as part of this kind of partnership that epic and microsoft are doing their microsoft is also going to bring its natural language queries and interactive data analyst tools to slicer dicer which is epic self-service reporting tool so this is kind of interesting right like epic already has its own self-reporting tool and now they're incorporating their own stuff so it just is one extra barrier to entry of any of these healthcare care companies or SaaS companies, you know, getting in with a big company like Epic when they already have their own SaaS tools.

4:13And now they're already building out, you know, chat GPT on top of it. And I mean, I don't think it's very farfetched to say that Epic could then go take slicer dicer and license that out to other hospitals and other healthcare providers. You know, essentially, it's a hospital, but now it's becoming a software company. It's kind of interesting spin. And I feel like we may see a lot of that in the future, especially when you look at companies that have data sets that they're using to help with those tools. So I don't obviously, you know, obviously, it's going to have a lot of really sensitive information and something like this.

4:47So I doubt that Epic is going to be using this, like they're using their using all of the people that are their hospitals data to train these tools and then licensing that out. But it's going to be interesting. I think there's a lot of cases, you know, we see with Bloomberg, how they made Bloomberg GPT based off of a lot of their internal data. And, you know, essentially they were a news company now they're turning into an ai more software company so it's going to be interesting to see what other companies make that sort of pivot and start incorporating software to help you know their bottom line increase so i think this suite of analytics and tools that was essentially designed to help physicians explore their hunches about medical questions that's what you know epic is currently working on as well as get answers about the operational efficacy of their hospitals.

5:38That's kind of what their software does. And so now integrating ChaiGPT is going to help this in a huge way. I think it's going to really just make it more, I think, conversational while a lot of these physicians are using these tools, which I think is one of the best user experiences you can add to software. If you can make it a lot more integrated and conversational, it's a lot easier to use than, you know, the software of the past, which was a lot of drop downs and buttons and things that are a lot less intuitive. So I think people are going to really appreciate that in healthcare. And at the same time, you know, I think this has a lot of different capabilities.

6:13Last month, Nuance, which is owned by Microsoft, also announced that it was integrating GPT-4 into its clinical note taking tool. So obviously, this is another area that healthcare is using this. They're integrating just different tools. obviously doctors when you go in they got to take a ton of notes about you and ask you a bunch of questions and I think that a lot of this is going to be um auto-filled you know you say something once and then it's going to fill it in a lot of different places or it's going to um help you come up with uh different things um apparently also with epic and what they're kind of integrating gpt4 in it's also to help physicians um ask their system they're building about the impact of chronic conditions on their local community using more of natural language queries.

6:58It's going to be really interesting to see, you know, how this builds out. People have said that GPT-4 is actually pretty good when it comes to medical information, and it's going to be interesting seeing physicians use this, especially, I mean, at least, you know, it's a physician using it hopefully just as a help and relying on their own background experience. You know, I think areas you'd want to watch out for is, you know, GPT hallucinating and giving a physician some sort of crazy medical piece of advice. But it would seem that that's less of the area that they're going into and more of more of just like helping them with a lot of the administrative tasks.

7:34But it does say that they're that they're creating features to explore data and making that more conversational for physicians. So that's going to be pretty interesting. And also, they're also using it for, you know, asking it about different conditions and medical things for communities. So I believe that would be an area that they would definitely want to keep a close eye on hallucinations in healthcare. Obviously, it's got a lot of really amazing potential, but you really want to watch out that this thing is being used well, especially when it comes to medical and people's health. I think this is going to be a big area that people are watching into the future.