Today in health, it I'm going to talk a little bit about my experiment with AI. This might inform how you're thinking about AI. And I know it has changed my thinking. I know I said I was going to talk about leadership and I will over the next three days. Talk about leadership. But today I did a couple of projects and I just think it would cause me to think differently as a CIO.

And I want to see what you think. My name is bill Russell. I'm a former CIO for a 16 hospital system. And creative this week health set of channels and events dedicated to transform healthcare. One connection at a time. Today's episode is brought to you by Chrome O S imagine a healthcare system where technology works seamlessly in the background. Keeping your data secure your teams connected and your patients at the center of care.

Visit this week health.com/google-chrome O S to learn more. All right. Hey, we're excited about what we're doing with Alex's lemonade. Stand this year. As we talk about having a child with cancer, one of the most painful and difficult situations a family could face. We want to come alongside to them.

If you can be in Vegas, we would love for you to be a part of that. And that is the Sunday before hymns. We designed the golf tournament. To be fun, exciting whether you're a golfer or not. So don't let that stop you from. Being a part of it. It's going to be fantastic. If you're a health it professional, we invite you to join us just by showing up.

You will help these families and for partners. Please reach out to us for more information on sponsoring the term tournament. It's going to be an event you won't want to miss. All right. Hey, one last reminder, share this podcast with a friend or colleague use it as a foundation for daily or weekly discussions. On topics that are relevant to you in the industry, they can subscribe wherever you listen to podcasts. All right.

As I said Kate and Sarah will be back next week. I just wanted to open up this year with a few notes, gets you thinking and heading in the right direction. And over the break, I started going back and reading a lot of stuff by Jeff Sutherland. Now just Southern land. If you're not familiar is one of the co creators of scrum. And , if you haven't read his books there. They're fantastic.

How to do twice the work with. In half the time or something to that effect. But Jeff Sutherland, scrum will get you where you need to go. He has published some additional books since then.

And as I was looking through it, I was going through his posts on social media. And there's a rapid prototyping post out there. Heinrich. Kinniburgh I'm sure I'm mispronouncing that. But he made this video on rapid prototyping of On steroids with Claude. We go from napkin sketch to running app in 30 seconds and create four iterations within three minutes.

Another two minutes, we have a design doc and implementation plan for production for both web and iOS. And I'm looking at that going. All right. I read these things all the time. I want to get that experience and it turns out Jeff Sutherland ass actually commented on it. And he said, very cool. Claude has emerged as the lead programmer of my scrum AI team. But real development doesn't go as smoothly.

It took five week, weeks heads down with Claude and five other AIS. For them to produce a product that could independently create all my expense reports for four companies for an entire quarter in one second. When we were done, Claude said it never would have happened without the human daily pushing the AI's to follow all the good development practices, acceptance tests. Automatically run on every build. Test driven. Development pair, programming, code reviews, and so forth.

In real development, the AI is generate a lot more bugs. Then humans. The interesting thing is that when the design became too complicated and needed refactoring, We probably rewrote the entire code base from scratch every sprint for five, one week sprints. After the pain of the first release. The one human and six AI team learn to collaborate better. And the second release, surprise everyone because. It got done in one sprint, a 500% acceleration, similar to what. We saw on our first scrum team, we broke out the shame pain on that one. All right.

So you read all these things and you go, oh my gosh. That is pretty incredible. I wonder what it looks like to actually try this. Now I did this on a smaller scale. We have a bunch of WordPress sites and the team has been asking me for some additional features to gather information from the site and whatnot. And if you know how WordPress works, you go out there and you look for plugins and then you get something that's not quite the right fit, but it's close enough.

And you implement that because you could just go install, configure, boom. And away you go. And I thought, all right, software development is changing. And if this is true, then I should be able to act as the as the product owner. The process lead. If you will, the scrum master and get an AI team and a group of AI team members, I could have Claude as the lead developer.

And I have found that clutch development capabilities are really good. I can have testing QA done with another AI full-stack AI production with AI, you get the idea you could have you could essentially build out the AI team. So I did that. I again, through prompts and w a single session, I said, all right, we're going to develop this thing. Here are the requirements here, some of the stories around how it's going to be utilized. And I want to start developing this. And so what I had been doing in the past is I would say, Hey, develop this thing.

I develop it. I put it on the site. I tested. And I do a whole bunch of stuff and refactor it and then come back to it and say, Hey, I got this error, that kind of stuff.

This time, what I did is I said, all right, we're going to keep all the w we're gonna do things until we get to something that is functioning and working as much as possible on, on the AI itself. And so it had all those things. And I said, you're going to work as the lead developer. And then in another AI session, I said, you're going to be testing and QA of the code and explained. Testing and QA, what I was looking for in a coding practices and whatnot.

And so the lead developer kicked out some code. Code goes over to the second testing and QA testing, QA comes back and says, here is what we found. I dished that back to the original AI and the original AI said, oh, I see the error and whatnot. It went back and forth 33 times. I kid, you not went back and forth 33 times. And. When it was done. It worked.

It worked flawlessly exactly how I wanted it to work. I was shocked. And it I produced a documentation documented the code correctly. So at each one of the AI, actually, I didn't have to use that many AIS, but each one of the AIS, if I was doing a if I was trying to refactor for different platforms and whatnot, I would have utilized some of the other things that he talked about in his post. The one human five AIS. Feature kind of thing in under an hour, but I don't have any deployment.

Really. It's pretty straightforward. But if you could imagine it went back and forth 33 times before I put the code on the site and tested it. And again it works exactly how we drew it up. And without any errors. Why do we have this conversation? The reason we have this the, so what of this conversation is the world is changing. And way too many people are underestimating.

What we can do with AI and way too many people are overestimating. What we can do safely with AI today. Is what I'm finding. And I don't want you to fall in either of those categories. And in the underestimation category, I find that too many people are not utilizing it to its full potential. And as I think about this first of all, My entire team. Again, I don't have a huge team, but it's a significant portion of our budget. It's not a significant it's. It's a mild portion of our budget.

When I think about it. My team all have access to to an AI tool. All of them have access to chatty fatigue periods when they come on board. I fired it up. They get added to the team. They all have access to that. Some of the team members have access to a second one and some of the team members have access to a third one. So second one being Claude depending on what roles they use. And then third one could be whatever they. Need or require it now I find I'm the one I've four of them. That I utilize, but they are each one of them is learning these tools. I find that's going to be critical. Over the next year.

I think that's going to be critical for each one to learn these tools, to get better at these tools. I know one of our team players, we gave them a task that they had never done before. And I said, use the AI to teach you how to do the task. And we also had some materials. There was a there's some books on what we're going to try to do.

And there was a specific book that we were going to follow their framework. And so I had her read that book, but then I had heard utilize AI to say, look, are you familiar with this book? You familiar with this framework? Are we are doing this and away. She went down this thing.

And within. I don't know, within 24 hours. She had generated what we needed for for four different hosts. For just it just months worth of work. And she had done it all within 24 hours. I believe that is indicative of what is going to be possible. With your team this year, if you allow them to use these tools. And it depends which tools you use.

Each tool has its strengths and its weaknesses. I've found myself using clod a lot more than I had in the past, but that could, because I'm leaning on the development side right now. I know the chats UBT is, was my go-to last year. I know Google Gemini is interesting to me and the things that they're doing in their notebook, LLM and other things. That I'm watching and keeping an eye on. The but it is making each individual much more effective. Again, so what's the, so what here?

I think it refactors what your team can do. And also I think a re factors what you're looking for in your next tire. Again, one of the biggest mistakes I think people make is that they hire for what they need today instead of what they're going to need. A year from now. Or six months from now. And you just naturally start with a legacy problem.

The minute you hire for something you were doing yesterday, instead of something you're going to be doing in the future. And so that person very quickly becomes somebody that you have to retool. And re-skill why not look for somebody who has the skills you're going to need a six months to a year from now.

I think every worker this year will recognize how much each one of these tools. Can impact their job. And how much more effective it will make them. I think the other thing is that software will fundamentally change. I think networking will fundamentally change. I think networks more and more will be driven by AI. I think routing tables, I think. Just the. Ability to respond to the attacks are going to require split-second.

And because of that, I think you're going to see more and more of these tools move to AI generated networks that are resilient, that can respond to these attacks very quickly. I think the same thing's true. Around development. I think software is going to fundamentally change this reasoning aspect.

I think. We've already seen it, starting to show up in all the, of these platforms, but I think more and more, we're going to figure out how to put the safeguards around these things how to put the guard rails around these things. And to utilize them effectively.

And now I realize we are in healthcare and we have to be a hundred percent accurate when it comes to clinical decision. Making and that kind of stuff. But that's not what I'm talking about today. I'm talking about your it team. I'm talking about all the players within your it team. And at 20 bucks a person to get access to chats, UBT per month, 240 bucks a year, that's nothing. And for an additional 20 bucks to get access to Claude per month, you're now talking 40 bucks a month. That's 480 bucks.

That's 500 bucks. I think that 500 bucks comes back to you in one week. Of them utilizing the tool more effectively. I think training people on using the tool. And by the way, that's just within the it team. I think there's an opportunity now. Anyway, I'm going to stop there. I wanted to get this into your thought process because it's changing the way I'm thinking about things.

One human five AIS features in under an hour kind of thing I think is pretty interesting. And to Divya. I'll be able to develop this software, deploy this software. And less than an hour is really incredible to me and game changing. You're talking about something that used to take me a couple of days to do, and I'm now doing it in under an hour. And giving these features to my team, which is going to save them. Ours in terms of how they. Collect information and whatnot from our website. And yeah. So it just fundamentally changes the game.

I think it can fundamentally change the game for you and your team. Just something to think about that's all for today. Don't forget. Share this podcast with a friend or colleague. We want to thank our partner, Google chromo, 📍 or us who is investing in our mission to develop the next generation of health leaders.

Check them out at this week.

health.com/chromo. S thanks for listening. That's all for now.