Today in health, it let's define the most used buzzword, I think in history, which is AI. It's about time since we were hearing it every day might want to have a definition for it. My name is bill Russell. I'm a former CIO for a 16 hospital system. And creator of this week health set of channels and events dedicated to transform healthcare.

One connection at a time. We want to thank our show sponsors who are investing in developing the next generation of health leaders. They are notable service now. Enterprise health parlance. Certify health and Panda health. Check them out at this week. health.com/today. Hey, check out our new service this week.

health.com/news. Rate's a great resource for you. Curated news. So you only get the articles that are relevant to you. All right. Hey, we raised another, I don't know, five, $6,000 for childhood cancer. If you want to be a part of that, hit our website. Now top of the website and the blue banner is a way just click on that and you can give today. We appreciate your generosity. Alright, one last thing, mentor someone share this podcast with a friend or colleague use it as a foundation for daily or weekly discussions on the topics that are relevant to you and the industry like this one.

What is the definition of AI? Anyway? Mentor them have conversations. They can subscribe wherever you listen to podcasts. All right. I was sitting with ZAFA Chadri during the one presentation. And we stopped counting. It was unbelievable. The number of times that AI was used by the four panelists. And it's amazing to me. That we are at this point where it's being used this much.

When I came. To do that, I sought out the experts and talked to them. James And David Linthicum and others who had been in cloud computing for awhile. In fact, I I was a part of buying David Linda comes company back in the day and he would speak, he wrote many books on architecture and cloud computing, and we grew a consulting practice out of that. In car.

In fact, that was my first foray into podcasting David Linthicum and I did a podcast, the cloud computing podcast back in the day. And that was a first time a microphone got placed in front of me and we put things out there and it was a, it was the start of this. I guess as we look at it when I talked to David. About about cloud computing.

He identified five. Six, six essential characteristics of cloud computing, as I think about it. The first was self-service provisioning. You didn't have to go to it. You could just provision it from anywhere. In fact, It was one of those things that it was a cloud model was internet based.

Therefore you can get, you could provision something from anywhere in the world. And so self-service provisioning was one of the first characteristics. Sec, second was rapid elasticity. You could scale up and scale down. We saw the string of the pandemic. You could scale it up, scale it down. People were saying, oh my gosh, we need more. Oh, virtual, we need more, virtual meetings teams and zoom and all that stuff.

Rapid elasticity allowed us to scale up and scale down very rapidly. The third thing was metered service pays you go. So you didn't have to sign that contract for a massive system that you didn't use. You paid for what you were going to use as you went. And so you can actually scale up and scale down on a monthly basis and a lot of cases with good cloud models, you can do that. The fourth thing was resource pooling.

That is that all these resources could be used for many things. And that was what most people think about as these massive data centers on the edge of town? That's where the resource pooling happens. The fifth is API based so we could automate it. And you could act if it was as a service,

so we could write. Into infrastructure. Like we do software, we actually made it into software. So we, it was API based. And then broad network access, which I touched on a little earlier, but essentially. You can get to cloud from anywhere. It's one of the things we like about cloud is that it gets us closer to our patients.

It gets us closer to our end users because cloud is ubiquitous it's everywhere. So it's broad network access, and there was a handful of service models. Back in the day, we used to talk about infrastructure as a service. That's where you just get your infrastructure. Platform as a service. This is where we were building things out there on top of it.

And Azure is a, it is a platform as a service. As is AWS, I'm not sure. Azure is Microsoft. AWS is, I don't know. I don't. It's just AWS is the platform. Anyway and the final was software as a service. Application as a service is called a bunch of different things, but it's software as a service.

That's. That's the Salesforce service now. Workday kind of model where you get the application and actually off Microsoft 365 as well. So we had three models. I'm sure there's other models today, but quite frankly, 90% of cloud can be those three models. And then we had deployment models.

You had private and public cloud. You had hybrid cloud. Where the deployment models. All right. So we had a definition. So as people were throwing this buzzword around, I'm like, okay, I could put into a category. I could say that software as a service and it should have these characteristics in order to be considered. Cloud based because a lot of things came out and it was a thing called cloud washing.

And so people would say, oh, this is cloud-based and then you could say, wait, it doesn't have self-service provisioning. It doesn't have rapidly less disease. It doesn't have any user services. You go. It doesn't have resource pooling, API APIs, or broad network access. That's not cloud that's cloud washing.

So you were able to see it. No it, and you were able to to make the determination. Is this cloud, is it not cloud? All right. So fast forward to today and hymns and Vive and every other conference you are going to go through this year. You're going to see the word AI. And I think that I don't have as clear a model for this yet, I'm going to work on it.

But I think the first thing to do is it define what makes this different from any other application. And most applications are based on algorithms. And AI refers to the, the ability for machines to act like it's artificial intelligence. So machines to act as humans. So let's go back to algorithms, right?

So an algorithm is a set of computer instructions and sometimes rules and. It's code to perform specific task and solve a particular problem. Algorithms are sequences of well-defined steps. Operations that take some input and produce an output period. That's what it is that you put stuff into the algorithm, something comes out simplest algorithm is a math equation, right?

So you put something in, you put two numbers in five and five, five times five equals 25. That's an algorithm it's not going to change. It is the same. Every time you do it. So that's a simple algorithm and then you have more complex algorithms that do. A really fun things. That's an example of an algorithm.

It doesn't change over time. You could put a hundred different equations in there and algorithms the same. If it's a multiplication equation, if you give it two numbers, it's going to multiply those numbers. It's going to come out with something. That's an algorithm. How is that different from AI? AI refers to simulating human intelligence in machines, in computers that are programmed to think like humans and mimic their actions.

The term can also apply to, to machines that exhibit traits associated with the human mind, such as learning and problem solving. All right. So learning and problem solving, I think are two of the most important aspects when people are talking about AI. Is the system getting smarter as it does more computation. All right.

The more data that runs across it's transom, is it getting smarter? Is it changing? Is it adapting? Okay. So AI encompasses a lot of things. It's machine learning, it's natural language processing. It's a do. No, there's just tons of systems that now. Our learning systems, they get better.

The more machine learning, like the more you pass over it, it's transom. The more it's able to say, that's a cup, that's a ball. That's a person smiling. The more data that more clean data that goes across its system, the smarter it gets. And there's lots of examples of this. Of things that get smarter as it goes over computer vision falls into that category.

The more images that something sees if it's truly not an algorithm and it's truly machine learning. Or if it's true, truly those kinds of models, every time it sees a new image, it's going to get smarter and say, oh, that's that. That's what we're seeing. So let me talk about the difference of these things. A learning system versus a static system algorithm is static sequences of instructions.

AI is a complex system that can learn and adapt. So AI uses algorithms. But let's be clear. AI uses algorithms as building blocks, but combines them in a way that allows the machine to learn from the data. And so one of the key questions I'm always asking these vendors is. Does your system get better over time?

Does it adapt? Does it change? So algorithms are designed for specific tasks and they do not deviate from those instructions. It will be the same algorithm a hundred years from now. If it's not touched, it will be the same hour of them a hundred years from now, as it is today. And it will do the same exact task a hundred years from now as it does today. AI aims to replicate. The human mind, human intelligence. And can adjust its actions based on new information.

So if all of a sudden it recognizes that a cup is a cup and it's not a pool. It will then adjust. It'll go back. And every time it looks new. Pictures of cups and pools. It'll say, no, that's not a pool. That's a cup. Or that it is not a cup. It's a pool. Because when you think about it from an image standpoint, sometimes those two things can look very similar. But it can adapt.

It can change. This is the key. Concept, which is an hour of them is algorithms algorithm. It is not going to change. However, machine learning. NLP and these kinds of technologies are written in such a way that it will adapt as it gets more information. So adaptability. Traditional algorithms do not change unless the programmer modifies them. And AI systems can improve and adapt their behavior over time. Through learning. Essentially intelligence through. Adapting to the information that's coming across.

It's transit. That's the, that's one of the big differences. And so as we hear these different things, as we hear AI after AI, What's happening now is anytime somebody uses an AI R uses. Generative AI it's referred to as AI. And one of the biggest problems we have right now is when people say AI, we don't have that framework that I gave you for cloud.

Characteristics surface models, deployment models and I should probably do something. I will likely do something around this. Just broaden it a little bit more because AI is one of those buzz words that we have to get in front of. It's going to be used a jillion times this year. And it's going to be, it's going to continue to be used. And in order to speak intelligently about its promise to healthcare, we have to understand. The central characteristics of AI. We have to understand the deployment models of AI. And and the service models and generative AI is just one. Of the service models that is out there.

There are many other service models. And, some are going to be applicable to your health system and some are not going to be applicable to your health system. And anyway, that's a start. Let's call that. I start to defining the word AI as we move forward. And if you have some things that you've put together around this, please forward it to me.

I'd love to see him. And love to a. Love to start to incorporate that, incorporate them into the broader thinking of the industry so that we as an industry can speak more intelligently on what we are no more intelligently. What we're saying when we say the word AI. All right. That's all for today. Tomorrow, I'm going to talk about him.

So if you want to get the HIMS update, I will give it to you on Friday. Don't forget to share this podcast with a friend or colleague. Use it as foundation to keep the conversation going. And as a foundation for mentoring. We want to thank our channel sponsors who are investing in our mission to develop the next generation of health leaders. Notable service now, enterprise south by like notable great example of of artificial intelligence service now, enterprise health parlance, certified health and 📍 Panda health.

Check them out at this week. Health. Dot com slash today. Thanks for listening. That's all for now.