This transcription is provided by artificial intelligence. We believe in technology but understand that even the smartest robots can sometimes get speech recognition wrong.

Newsday: Why Constraints Drive Innovation - AI Agents and the Jobs They Can’t Fill

I'm Bill Russell, creator of this week Health, where our mission is to transform healthcare one connection at a time. Welcome to Newsday, breaking Down the Health it headlines that matter most. Let's jump into the news.

Bill Russell : All right. It's Newsday and today we are joined by, I should just keep coming up with new adjectives for

Sarah Richardson: you. I know you say is goofy for me.

Bill Russell : I said goofy for direct support. I know. And I think incomparable for Sarah Richardson the last time, Actually, I'll put Chad, how

Sarah Richardson: about Dodger and Mariner fans?

Drex DeFord: To go by the time this airs, there might be a World Series going on between the two.

Bill Russell : No kidding. What's gonna be the bet.

Drex DeFord: I don't know. We'll have to come up with that.

Sarah Richardson: We'll tell something good. because football's totally in the can for me this year. I know we've got a decent record, but everybody is injured.

Bill Russell : game.

Yeah. I want it to be embarrassing. I was a Cardinals fan and a friend of mine was a Cubs fan, and we made a bet on a series and essentially I had to wear the Cubs jersey to the board meeting. Which was really embarrassing. I mean, you know, I had my suit on, but I had the Cubs jersey over top.

And then of course you have to answer the question. It's like, oh, why are you wearing a Cubs jersey? It's like,

Drex DeFord: oh

Bill Russell : c, I

Drex DeFord: gotta tell the story.

Bill Russell : Gotta tell the story. Well, hey, we're getting ready. I'm heading out to Nashville tomorrow. And then we head to our various summits. We have four summits going on this weekend.

I mean, by the time this airs, we will have been done the summits. What do you expect to hear at these summits? So we have two CIO summits and a CISO summit. We also have A-C-M-I-O summit going on that is actually chaired by the CMIOs, which is, kind of awesome. I love that. I mean, do you expect to hear anything a little different twist on what you normally hear at a CISO event?

So it'll be interesting to kind of hear and a lot of 'em are from academic based, you know, university. It's gonna be a really interesting conversation to see what they come up with. I think some of the chronic problems in healthcare are the chronic problems in healthcare, and we're probably gonna talk about those, but there will be some new people in the room.

And I know some of them and they can be pretty provocative, so I'm expecting them to unload it, that unload on the room.

Bill Russell : what I like about that group is when we sort of throw it out and say, Hey, talk about something that's working. You start to hear different things that people are doing to address the.

Drex DeFord: Yeah. Well, they talk about things like minimum viable hospital, and it's interesting because I think a lot of the conversations.

We, I use this analogy all the time about my car's in the ditch, right? Like a lot of the conversation sometimes is the woe is me, my car is in the ditch. Oh, my car's in the ditch too. And like, you can let 'em do that for a little while. There's a little bit of like misery loves company group therapy thing that goes on, but pretty quickly they start to talk about like.

Sometimes it's about, here's how I got my car out of the ditch. But sometimes it's like, these are the things I've tried and haven't worked and those are super valuable to everybody else in the room too, because it's like, I was thinking about trying that, but if that, why didn't that work for you? And we get into those conversations, which often leads to some new.

Alternative route to get the car out of the ditch.

But let's talk about, you know. What are we gonna do about this car on the ditch

Drex DeFord: solutions?

Bill Russell : Sarah, you have a little different twist on an event. You have next generation leaders in that room. What do I mean, what do you think next, not the CIOs themselves, what do you think the next generation's thinking about

Sarah Richardson: so, I love that we do the prep calls. because you get the one-on-one time with every single person who is coming and we're pretty specific. And having them tell us what are you thinking about? That's working. What are you thinking about that's not working? Because then we're also crowdsourcing. How much time do we wanna spend on some of these topics pervasively and people are proud of the usual things like, Hey, we gotta win, and getting certain scenes approved.

No, it's all integrated. There's all third party risk. So everyone's an expert now or has to be, and understanding the foundational aspects. How well you connected your CTO and your ciso, and then your applications can lay on top of that, some other aspects. But pervasively the themes coming into this week, which I keep thinking it's next week, it's not how AI powered workforce tools are being utilized in organizations, whether it's virtual nursing, virtual note taking, virtual activities for the doctors, the remote, like activities that are out there are winning more and more.

The technology leaders that can pull on Uncle Drex, that can tell a story effectively. They're getting the funding and the resources that others aren't because they know how to weave all of those different components into really thoughtful conversation with the people that they need to have on board to be able to make it work.

Bill Russell : I think I'm gonna get some sort of swag made with Uncle Drex on it. I think that's a thing. You know how on it's not Jeopardy wheel of Fortune. On Wheel of Fortune, when they go to the final thing, they always give 'em like a bunch of, it's like, we're gonna give you a e.

I mean budget, well, budget or finance or you know, the governance conversation is usually not like how do you stand up governance? It's usually how do you create the, the mechanism to control the amount and velocity of projects coming at it.

Drex DeFord: That's definitely one of the cars in the ditch conversations in all of our, you know, like there's unlimited demand and there's only so much of me and like, you know, how do I deal with that?

Sarah Richardson: That's a 30 year conversation. I've literally been having that conversation for 30 years, which means it will always be something we need to be thinking about how to morph and adapt and be more effective at understanding, especially with. Proliferation of tools we have today. What if you can create all the capacity and demand that you need because of ai as an example?

So then what,

One of the things we have done is we've held off on hiring certain roles. because we're like, look. I think we're going to stand up these agents that are gonna take this old job and this old job, and this old job, which used to be like a full-time person. And I don't wanna let somebody go. So it's like, let's just, let's run at the edge for now because we're gonna be rolling these things out in the next three months.

And these agents it's an interesting conversation that I'm hearing around these agents from this perspective. Some people are talking about treating them as if you're hiring a another person. So you give them a higher date, you give them reviews, just like you give an employee review. You know, are they leaning towards bias?

Are they, you know, are they still fulfilling what they should, are they making, what's their error rate? And those kind of things. Same way you would, you know, bring on a new employee and onboarding's a thing as well. You have to give it context, right? So you have to,

Drex DeFord: it's training records. You gotta put 'em through training.

Yeah.

But I think in healthcare, what does it look like over the next two years as you begin to bring up those agents? Because look, Microsoft's gonna give 'em to you. CrowdStrike's gonna give 'em to you. You know, Workday's gonna give 'em to you. ServiceNow's gonna give 'em to you. We're gonna be standing up agents, I believe all over the place.

Let's start there. Do you believe that we will be standing up agents? Across healthcare in the different disciplines that we oversee.

Drex DeFord: Yeah, no, for sure. And it's gonna be a lot like today, where you open up your. Whatever as a service application, and there's a new button and it's an AI button and you can put things in there and it'll generate stuff for you.

Outsourcing part of your do job to this AI assistant who can then do a bunch of, like grindy hate it, boring work, but the assistant is brilliant at it. It's gonna be really interesting to kind of see how we retrofit our teams to be able to handle that kind of new employee interaction.

Bill Russell : Well, the other thing I've been thinking about is, so it's gonna be really important for us to document the workflow, to understand our data and to understand our processes because I think what's gonna happen is the technology's gonna keep changing.

That is watching how you work and it's gonna say, , do you want me to do that for you? And Apple's been a little behind on this, but I think we're starting to feel it a little bit on our phones, but I think we will start to feel it on our PCs. Look, computer Vision can watch all the things we're doing today and mimic what we're doing.

But the thing is, the way we do them is sort of stuck in the past. Like, we go to a browser, we log in, we did, and with APIs and whatnot, it's like. You know, the AI agent would say why are you doing that? I'm just gonna go over here, update this record, pull this across, and, you know, update everything.

And then next time you look at it on your phone, it's gonna be updated.

Drex DeFord: you're gonna wind up with I mean, agents that are going to work well with other agents too. Right, right. Across applications. The Workday agent's gonna work really well with I don't know, the Epic agent or something. They will.

Where is your data? What are you doing with that data? Normally, what does good look like compared to all the other customers that agents like me, service in the healthcare industry, what does good look like? because I can actually crowdsource that great behavior and try to help you get to that point faster.

It's really interesting.

Bill Russell : Sarah, you talk a lot about transparency, and I think it's the right topic here because you know, when we hire a new employee, we're able to see their work and we're able to sort of monitor their work and we're able to sit down with them and go, Hey, why don't you evaluate yourself?

How do you, I mean, transparency's gonna be critical.

Sarah Richardson: The thing that worries me the most, and I love technology, I'm a big fan of, you know, anything that can make your life a little bit easier. But I also was at a recent event, a pretty big investor meeting, and one of the. People was like, Hey, when can we go back to paper?

I'm thinking like I remember phone numbers when I was 10 years old, but I can't, I don't know either one of your phone numbers at all. Like I don't even know what the first, what your area code is anymore. And my biggest concern with the transparency of usage of agent AI and beyond is. How do we teach the humans that are part of it?

The critical thinking skills to be able to effectively manage an agent, because very quickly an agent can be quote unquote, smarter than you. The other thing is when you displace that many humans from things that make them feel like they matter, and I'll steal a quote from David, Jim Bruno, that humans are meaning making machines.

If you have thousands, hundreds of thousands of people that no longer have purpose or meaning, because there's not necessarily a job that's fulfilling their full capability. What happens to that civil aspect, to that civic aspect of what we are responsible for as the purveyors of all of this technology?

And I was like, okay. But I can't tell you the last time I like pulled out the formula for statistical probability. I mean, I just, I don't do it anymore. And, you know, I, and same thing with the calculator. It's like, you know, it's like the smarter the calculator, the more they were like, oh no, that has that knows the formula.

Like you have to memorize, we're cheating. Yeah, you're cheating. I'm not worried about the fact that you don't have to know phone numbers. I mean, there's any number of a million ways, and it's one of those things that I think is in terms of practicing at the top of your license, I think it's one of the lower things in terms of practicing at the top of your license.

I am worried about the second one, which is humanity and meaning for people. But I think this is part of Maslow's hierarchy. I think It's like, okay, we're gonna take care of this. You can now move up to this and start thinking about. Higher level things. When you think about it,

It'd be nice if we put all of our energy on fixing the failed system instead of, oh, you know, if we could do this, we could tweak this and get more money from the insurance carrier and get paid a little bit more over here. And you're like, yeah. Let's fix, let's just fix that thing altogether would be a great source of our thinking and energy, I think.

Sarah Richardson: But you don't fix that as the CIO of a healthcare system That fixed comes at the policy and insurance level. To an degree you are having shelf insured plans,

Bill Russell : but I'm saying I think everybody moves up a level. I think everybody, you know, the person who was typing on the keyboard is now overseeing agents.

because the reality is you know, congress and the Senate and the whole, that whole system has a proven not good at solving the problem that ails us, which is health. Yeah,

Sarah Richardson: but you're gonna keep seeing a divide. So there's the whole group you just talked about that's gonna be up here in this quadrant.

I firmly believe there are more people in that lower quadrant, and that's gonna create a bigger economic and social divide than it ever has before. It's been a slow burn over the last few decades, but it's accelerating faster than ever. There's a lot of people who I don't believe can get there.

And so we're then responsible for that wellbeing or caring for those individuals in society and otherwise, and I'm talking about underdeveloped nations as much as what we see here in our own country. And so y all we're doing is create a bigger divide. So what are we doing about the middle? This is not a political conversation.

Bill Russell : . you're gonna drag me into this. I don't care about the divide.

I just care that the lower is high enough that they have a really good quality of life and the life they wanna live. I don't care that there's a massive gap between the top and the bottom. I just care that the bottom is high enough that they have a good quality. There's some kind of floor Yeah. That everybody's above the floor.

Yeah.

Nobody's in the basement.

Right, exactly. And the reason people are flooding into the United States and Europe and other things is because the basement in some of the places they're coming from is so low that they want to get to our basement. Yeah.

Drex DeFord: Yeah.

Sarah Richardson: And technology can raise that bar to your point, then we're going to be achieving more than one aspect of advanced technology.

Drex DeFord: Yeah.

Bill Russell : We see these organizations doing deals with power companies and, you know, standing up Three Mile Island again and doing that kinda stuff. How concerned should healthcare organizations be about their AI usage and how do they balance that with not falling behind in, in the world that they live in?

Drex DeFord: Man, it's, I mean, it's such a hard we were talking about you know, the new open AI Sora that came out, and so, you know, you think about every time you make a Sora video, how much power do you push through an NVIDIA processor to make that. Goofy dancing elephant thing, whatever it is that you're making,

Bill Russell : you know, I think I made fun of you for that.

Drex DeFord: Talking babies, like, you know,

We're gonna build a big power facility here. And the community saying, no, you are not, we're not having that here. You're not going to use up all of our water for that. You know, I don't know what the answer to it is. Ultimately, if we're going to go down this path, or we're gonna do a bunch of AI and there's a ton of chips running in giant data centers.

Ultimately the power has to come from somewhere and the cooling has to come from somewhere. Yeah, I don't know that we have a great answer for that right now. There are a lot of communities who are involved in these brawls that aren't happy with how things are going. But maybe there's a happy medium there somewhere in the discussion between the companies who are trying to build this stuff and the communities where they have to live.

Bill Russell : That's interesting. Almost like a utility. And the utilities are, you know, state by state, region by region

Do they need to grow? Do they need to shrink? What kind of specialties do they need? Can they. Handle supporting that kind of a you know, a new workforce. And can it move fast enough to be able to support that kind of a workforce and that kind of explosive growth that sometimes happens when these kind of facilities are built.

So there's a lot of things I think for us to think about for health systems. A lot of it is watching your own local community, watching your own state and local community, seeing where these things are happening and thinking about that. How's that going to affect our growth? Plan do we have facilities there?

Should we be putting a hos? Maybe that's an opportunity to buy a rural hospital if you're a big health system, because there's gonna be a great payer mix there two years from now. So you gotta think about all of that as a

Bill Russell : system. Sarah, you keep bringing this topic up and thank you for doing that, by the way.

Sarah Richardson: Would you remember people used to protest Walmart going in because it would decimate a town. It's not a dissimilar conversation. And so what was that happy medium? If there even was one that came out of some of that capitalistic aspects of why we say yes to certain decisions.

Bill Russell : And it did decimate towns like you could, it did. You can drive through some of these towns in the Midwest and see where the old shopping centers used to be. Now some of them have revived, but some of them are still completely decimated.

Sarah Richardson: I live in a town where there's probably more things that close than open.

On a regular basis, and I'm talking about services. Why? Because there's e, there's Costco on either side within a 30 mile radius and probably Walmart's bar refuse to go to Walmart. And yes, I'll publicly state that every chance that I get. Just because lots of reasons. So all of that being said, if I won't go to Walmart, am I using ai?

Those are the pieces that get left out. We were all fortunate enough to be raised in conversations with households that have those conversations. If you do this, then that sort of like the and or like fact to programming language, if this, then that, and here's some of your options. What option do you choose when you're informed?

And then that goes back to how we responsibly make sure that ai it's more than just accuracy. So if it's about energy efficiency, it's also about bias mitigation. It's also about how it's benefiting what we already talked about earlier, that underserved population. But how are you sourcing the information to make decisions?

That itself is not just an algorithm driven to the things that you already believe. You have to be exceptionally curious and really wanna know, how is my usage of X, Y, and Z going to impact the environment? It's like you're using plastic. Are you using, you know, CFCs? Like is the hole in the ozone that came up in the eighties?

We all need to know. How what we are utilizing for the advancement is also gonna play out long term. I don't even have kids, and I think about this stuff all the time. Y'all have grandkids, like what is the world they are going to inherit and how are you making it better or worse for them? Because you know what, in 40 years I'm out if I'm lucky to go 40 more years.

So why do I care so much? Well because still my responsibility to the human being to make it better than I found it.

Bill Russell : I do care about the environment. I care what we do to the environment, and there's a certain there's a lot of principles around that, you know, leave it better than you found it.

Yeah. And the companies that generate power are gonna have to figure out how to get AC closer to those data centers so you don't have 70 to 80% power loss across an AC line before it actually gets to where it's going to be used. Now it just think about that we have 70 to 80% loss because we decide to go AC instead of dc.

In terms of power. And if we had gone DC. We would've had to put in a million more power plants, but the power plant would be right next to or in the neighborhood where you lived, and we would not have had as much power loss. Well, I don't know. We should revisit some of those things. I always come back to this and I, you guys aren't as much fans as you used to be, but I come back to my Tesla car and it's not that, it's a ev.

I'm like, it's 15 years old but that they get into and they go, man, this is a really nice car. One of the things Elon Musk did is he simplified the crap outta that car to the point where 15 years later you still get into it and go, this is nice.

I could use some new seats, like my seats are getting, not ratty, they're old, but outside of that, it held up and it runs and people are like, well, what about the battery isn't, you know, eventually you're gonna put that battery in a landfill. I'm like, well, you know, it started at 280 miles.

It's now down to 2 25 miles. But I still find myself, I don't drive 200 miles.

But I think the real definition of innovation. Might be the thing that ultimately gets us out of this. It's gonna take too much power to run AI and all of that. When you don't have resources, that is the time that you are the most innovative because you have to figure out how to do something and you don't have anything to do it with.

So you can be super ingenious. Figuring out ways to make that work. And as you sort of alluded to, Nvidia re-engineering chips to be more power efficient to consume less power. They're going, I mean, they don't have any choice. There's no way there's enough power. Plants are gonna be built to meet the demand.

Drex DeFord: And then they rewrote the model so that the model could work better with the chips that they had.

Constraints force innovation to happen. It's not. The plethora of resources that make innovation happen. Honestly, I think it's the lack of resources that makes really good innovation happen.

Sarah Richardson: We talk about this all the time in our. Summits with our CIOs and CXOs, if you can optimize your workload to reduce the computational waste. Like fix the process and then put the technology on top of it.

If that is synonymous with also how we're thinking about more responsible consumption of energy or utilization, you've gotta win on both parts. That goes back to that transparency factor. I'm not just gonna automate something that sucks. I'm gonna automate something that I've optimized and I have to throw in a little plug for Toyota.

Drex DeFord: it applies to everything, right? Simple is better, simpler is easier to operate.

It takes less energy to run. It's easier to understand when something's going wrong, what actually has gone wrong. And it's easier to innovate on it because you understand how it works. So you can figure out that, oh, if I do this one more thing, I'm gonna be able to get one more mile to the gallon.

Bill Russell : All of that. Well, direction. It's a secretive turnarounds right. You go in and one of the first things you do is we're gonna simplify this environment and like, simplify. A year later, people are like, oh my gosh, you guys are so much, you know, less downtime and you're so much more efficient.

And what I said two years ago isn't valid anymore.

Drex DeFord: Yeah. You have to re-look all the time and say, my belief SY system is tethered to anchors that I put into the ground. Five years ago or six months ago, are those really the places I wanna be anchored? And you have to be willing to like let go of that.

And that change is really, Sarah talks about this all the time, but a change is super hard for most people. And if you think about our healthcare system in general, to kind of go back to the original conversation, how did we get here? What a mess. How could we have let this happen? It's because it's just been one thing piled on the next thing piled on the next Le Regi, you know, legislation piled on new regulations piled on state by state decisions about how to run things, and suddenly we have this.

Bill Russell : And the crazy thing is insurance didn't even exist. Like, what is it? 60 years ago, 16. Yeah, but it, I mean, it's like a new concept and everyone thinks like, oh, everyone has to have it. Well, it started off as just a benefit for certain employees and now it's like, oh no, this is like the entire system.

And that's exactly what happened over the years. It's like, oh, this is good. We should expand it to this. Oh, we should do this. We should expand it to this. Oh no, it should cover everybody. No should, and oh, the government should do this. You know, and you just look at it over the years, you're like. Oh my gosh.

This is Frankenstein's monster.

Kinda works

Drex DeFord: everywhere. Yeah. I used to talk to my team. All the, this emper fi always faithful is like the Marine Corps mantra. And for my teams, I had a tiny little man, he was green, I called his Emper. Gumby. That was our motto. Always flexible. And it was always about trying to figure out the best way to do the thing that we need to do today.

Bill Russell : Uncle Drex is back and we will, there he is on that.

Sarah Richardson: And there we are.

Bill Russell : Love it. Thanks for listening, everybody.

That's all for now.

That's Newsday. Stay informed between episodes with our Daily Insights email. And remember, every healthcare leader needs a community they can lean on and learn from. Subscribe at this week, health.com/subscribe. Thanks for listening. That's all for now.