Speaker 0: Every year, Becker's annual meeting brings health care leaders together to unpack the most pressing issues facing the industry. And every year, those conversations shift in profound and unexpected ways. This April, more than 3,500 healthcare executives will return to Chicago for Becker's sixteenth annual meeting. Seven ninety five elite speakers will offer new lessons, new case studies, and predictions about what comes next. Join us April 13 through the sixteenth.

For the agenda and event details, visit beckershospitalreview.com and click on the events tab in the upper right.

Speaker 1: Welcome to the Becker's Healthcare podcast. I'm Chris Sosa, your host. I'm very happy today to be joined by doctor Giovanni Bialimonte. He's a professor of pediatrics, biochemistry, and molecular biology and a researcher at Tulane University. He also chaired his task force on the use of AI in research, which is our topic for our discussion today.

Doctor Bialimonte, thank you for joining us today.

Speaker 2: Thank you for inviting me, Chris. I really appreciate it.

Speaker 1: Wonderful. So doctor, you've been, researching a number of topics for better part of thirty years now. But for those in our audience who may not be, intimately familiar with yourself and your organization, could could you please introduce yourself and let us know

Speaker 2: a little bit about that? Yeah. I'm a physician scientist, basically, and I've been practicing respiratory medicine clinically for about three decades. And about two decades, I've been generously funded, for my research, in several areas, of biology and respiratory diseases by NIH. I I added also a significant component to administrative duties over time, serving as division chief first, the department chair for many years.

Then I spent seven years at Cleveland Clinic as the president of the Cleveland Clinic Children's Hospital. And most recently, I joined, Tulane. I worked for six years. I served as vice president for research. Recently, I stepped down for that and started my first sabbatical in my career, which I'm, thoroughly enjoying.

And, the sabbatical has given me, the opportunity to to to evolve new areas of research, but also, to, go deeper into my ongoing interest as now several years old, for artificial intelligence, quantum computing, and advanced technologies applied to health care that, as you know, I'm sure everybody will agree on this. I'm gonna completely change the the the the landscape, of how we think and how we, manage and how we deliver health care, for many, many years.

Speaker 1: You can say that again, doctor Bienle. There's so much to discuss. And as I mentioned, you chaired Tulane's task force on AI and research. So let's start there. So could you please take us through simply how the task force evolved, which I know could take a lot of time, but that's okay.

We got time. And just highlight the the most significant findings from the last couple years.

Speaker 2: Well, you know, it was a great experience because I had the opportunity to interact with a lot of, brilliant minds, and most of them actually with the little to do with health care, including, engineers, mathematicians, and the like. And so open, several different angles to read, you know, to read what the opportunities are and what the risks are for for for this new technology. And, and we wrote a blueprint that summarizes, some of these concept that emerged. And, to make a a very short summary, there is no doubt that, the specific area that we, were interested in, which was research, is gonna be impacted dramatically in any component by by the introduction of artificial intelligence, already is. I predict, and I'm a little bit radical in this, but I predict that, virtually every single component of what today is the administration management or grants, for example, is gonna become, virtually independent from human input.

And the reason for this is that most are repetitive tasks that the computers will do already do in some part, with with the in a fraction of the time, with minimal cost and with the incredible visibility. For example, the preparation of a grant, the the assembly of a grant, the submission of a grant, the verification of a grant, some other functions that already a lot of investigators are dealing with, but they're gonna become more and more, stringent. For example, recently in the last years, we have seen a significant increase in the cases of plagiarism or, publication of, of, data that were not, either accurate or completely fabricated. And the reason for that increase is is quite simple. A lot of the fabrication of data could escape at UNI, but will not escape at the visual intelligence and computerized.

And today, there are plenty of softwares that allow people to actually verify whether, something has been, simply, built on somebody's imagination or whether it's real data. So there are millions of different aspects that artificial intelligence another one, for example, I do epidemiologic research and population research. And, the current study that I'm running is the first one in which we're gonna use artificial intelligence algorithms to analyze the data. And, already, we're seeing a lot of, driving, phenomena that, very unlikely it's very unlikely that we could have picked if we were just using our brains. So in an extreme world, basically, I will say that, a lot of experiments may become obsolete, because it can be can be reproduced in digital formats, and and, and therefore, problem solved in a fraction of the time.

And so all the all those things are great, and I expect a significant acceleration in the development, for example, of, new drugs, of, new therapeutics. I suspect that one of the areas of immediate progress, and and substantial progress is gonna be in robotic surgery, which already is, is evolving and so forth. It's gonna it's gonna be a a complete new word. I I a friend of mine, sometime ago said that to apply for, med school or law law school today is is stupid because by the time you graduate, the computers are gone and the and the robots are gonna do the work of doctors and, lawyers. It's a little bit extreme, but the the this history of artificial intelligence has shown us that, things that sound stupid, sound impossible, become very possible and very, almost simple, within years.

And so I think that definitely the new doctors and the new lawyers are gonna be collaborating primarily with the with the computers and robots, and the the the the scope and and and the and the applications of their work is gonna be completely, revolutionized. At the same time, it was an excellent opportunity to collaborating with my colleagues to assess what I think needs to be central in the in our analysis of, artificial intelligence and new technologies, and that are the risks that are associated with this, technology. And frankly, this is becoming a a main focus of my of my interest. And one of the things that we did was we tried to categorize, the different types of risks that are associated with the with the the introduction of, for example, Gen AI. And it specified that at the time we basically were discussing these matters, everybody was concentrating on Gen AI.

Today, most people concentrate on Gen AI with it is a totally different universe, and and, and, it's gonna make most or all of our conclusions, irrelevant because Genetec AI is not just an evolution of, Gen AI. It is a it's a a a change, a drastic, complete change in the way we are gonna interact with the with the with this technology. At at the time, talking about GenAI, we identified the the major types of risks as intellectual property risk, for example, and the confidentiality breach. These technologies are gonna make almost impossible to keep, data, confidential, safe, and so forth, not with the current technologies. That is gonna create a lot of problem for the, respect of, intellectual property.

It's gonna create a lot of problem with the attempts to block, individuals that are trying to get into critical databases. And that becomes, can become very problematic because, I give you an example. We know very well that there are very large databases already in existence, containing genomic data on on individuals. And, and this new technology is gonna make, are gonna make almost impossible to keep this safe, unless we develop a new, very complex strategies for, to secure those data. Another element is the biased and the balanced coverage.

It is true that, some people see in AI, a democratic force that is gonna put everybody in the same level. That that is not necessarily gonna happen. There is a very big risk that, the GI systems are trained, on a large amount of data, and the data is not, heavily distributed, does not give a fair, weight to all the components of society. So that can actually increase the risk of, inequalities. Another element is, the the cost.

Attention intelligence costs a lot of money, and it's gonna break a lot of banks, and it's gonna be out of the reach of a lot of people. And one of the most critical area of cost is gonna be energy. It's not, by, serendipity that, China's, run a very aggressive campaign for many years to increase their ability to have large sources of energy. Because, this is, an incredibly unprecedented, unprecedented requirement on the, power, need needs, that we had before AI. Another one is the, ecosystem volatility.

So in other terms, we were used to change and upgrade systems every certain number of years. Pretty soon, we are gonna go into the weeks. And, a lot of technologies are evolving at a speed that that is unprecedented. And, therefore, that is may create problems because, you know, you acquire a certain technology and becomes, deja vu within within months. So that creates, again, a significant cost problem.

And I left for last, the one that I think is most concerning in terms that the more we rely on artificial intelligence, the more we are at risk of the so called hallucinations. Artificial intelligence is not perfect. I think it's gonna become less and less imperfect over time, but, you understand that, you put, a lot of vital functions of the society and even human beings in the hands of artificial intelligence, you know, when you have an hallucination that can cost a lot, not only in terms of money, but also in terms of of human lives. So, I think the the best part of this experience was the understanding of the very delicate balance, very delicate balance existing between the advantages of this technology and the disadvantages of this technology. And and, again, all these considerations are gonna go on a different scale when, Agentic AI is gonna be, is gonna become more, prevalent.

And that is gonna be Agentic AI is gonna be the introduction and and the need needed bridge into, embodied AI, which means the production robots. Those two are the much more disruptive potential on our society and our technologies and our, operations than Gen AI. Gen AI depends on us. Gen AI depends critically on the input of data, on the on the programming. The human component is still dominant in Gen AI.

In Genetec AI, that flips. And that is where, I think, that that is the reason I say that we have not seen not even the beginning of this revolution. When we are gonna see the revolution, it's gonna be when a GenTIC AI is gonna become more sophisticated, more prevalent. And, I have reasons to believe that 2026 is gonna be an historic year because the 2025 was kind of eventful. It was, seemed like, you know, a a breaking point, but I think it was just an introduction to what is gonna happen in 2026 and 2027 when the technology that has been maturing over time is gonna come to full fruition.

And at that point, really, the world is gonna change in ways that are in great part unpredictable.

Speaker 1: Thank you for laying all that out, doctor. There's so much that's exciting and scary at the same time and really unavoidable in terms of how AI is gonna affect all of us, including, health systems and health care in general. I do wanna ask I'm curious. So given how much you've been embedded in AI and its research, what would you say your comfort level is with AI in terms of how how well do you trust it? You mentioned hallucinations.

And and what do you think has to happen for humans to trust it more than they do now? Does that make sense?

Speaker 2: Yes. It does. It's a great question. And, I may surprise you, but hallucinations are not the part of AI that scares me the most. The part of AI that scares me the most is the risk of skill reprogramming and skill down regulation in humans.

And I give you an example. I'm old enough that when I was in school, they you know, I became very good at mathematics. I didn't like it at all, but, you know, I knew how to do, basic calculations at least, by end with no support. And then we start getting the calculators, and then the calculators get better and better. I don't know if today I I will be able to run a square root analysis or even basic mathematical skills.

Why? Because I don't need to. Now if you expand that to all the incredible potential of artificial intelligence and to the knowledge that, I don't think there are many students today that don't rely on artificial intelligence and Gen AI to to do their, their their work. I think that we risk to, generations of people who are gonna forget how to read, they're gonna forget how to think, they're gonna forget how to calculate, they're gonna forget how to strategize, because all those things can be done very quickly and very easily by, by this tool. And so, the the this this is a risk.

In fact, I I'm I'm in a very tough positions because I'm really trying to understand all aspects, relevant to my area of interest, concerning AI, Gen AI, and and the related technologies. But I try to use it the the least I can, if it makes any sense to you.

Speaker 1: Oh, certainly. Certainly.

Speaker 2: Because I want my brain to go keep working. And and and and, and, I don't want to be in a situation like the same way I lost my calculations still. I think, for example, I don't write, my penmanship is not the same as it was. I don't need it. I don't I I am afraid even that my memory is gonna, be, is gonna, be reduced because that's the nature of, Darwinian systems and evolution.

So in a world where all those things, all the things that we do, all the expression of our, of our intelligence can be achieved in a much easier way, What is gonna be left? That that is that is an enormous concern that I have. Another concern is, what this is gonna do to our workforce. And there is a little bit of pride in me because, at the con at the Baker's conference, a couple of years back, when everybody was cheering, and, we're partying and we're so excited about all the potential of artificial intelligence, I said, be careful because it's gonna kill a lot of jobs. And I I got criticized, and several of my colleagues said, you know, you're you're being over pessimistic.

Well, it was very interesting, yesterday. I follow closely the that was, meetings, the the to hear that CEO, the largest bank in the in the world, I believe, or in The United States for sure, that is, Jamie, Dimon of j j JP JP Morgan Chase. And Jamie basically said exactly the same things. Jamie, in in a nutshell, said that if we do not I'm not very careful with the implementation, the launch, the use of artificial intelligence, particularly in the in the in the new generation of artificial intelligence algorithms, we're gonna lose an enormous number of jobs, and we have to be proactive in thinking what these people are gonna do. For example, in the work in the world of health care, as I said before, the opportunities are endless.

But when you look at any, large health care system, there is a certain small number or a fairly small number of doctors. All the rest are people that actually do highly repetitive tasks. You know, the doctors do highly repetitive tasks. But Sometimes. Yeah.

When you look when you look at, for example, verification of insurance, checking of the patients, analysis and submission of the claims, all the work we already know can be done much, much, faster, easier by system that don't get sick, and and they're very cheap. And another thing, that, Jamie said, yesterday is But, unfortunately, this is not a a process that can be modulated that much. Why? Because let's assume that you are the CEO of a of a larger care system, and you decide for your principles that you don't want to get involved with in AI, and not in the implementation AI in your health care system, you're gonna be killed on a financial basis because all your competitors are gonna leap in front of you and achieve the same performance or much higher performance with much lower cost. So, this is not gonna stop.

There is no no no possibility of stopping this. So I think that one of the most challenging aspect of, particularly, Genentech AI, is going to be before we launch on very large scale these technologies, we have to start thinking about what to do with people. All the people that are going to lose their jobs. You know, one of the things I noticed, there is a lot of attempts. I understand to reduce the, how can I say, the panic?

Mhmm. But but, I also think that we need to be realistic. There are very few things that, robots, artificial intelligence are not gonna be able to do. One of the most shocking experiences as a physician was I'm a a a respirologist, and, you know, for many years, I thought that you always needed a a physician to do a bronchoscopy, for example, any type of endoscopy. Today, there are several systems, that do bronchoscopies, in a totally robotics way, requiring minimal, additional effort from humans.

And, you know, it's just a matter of time that they're gonna be completely independent. And, there are already, hospitals in Japan where the entire and also in China, where the entire staff, nursing staff, and physician staff is made of robots. It's just the beginning. When you project those trends and you realize that the technology is there, we better be at least be ready be ready, for example, to a world in which humans probably are not gonna work five days, but only, three days a week. If that.

And one says, well, great. I mean, the more time for us. Yes. But then we have to find the way to pay these people. And already in Europe, there are a lot of, there's a lot of discussion about, providing basic salaries, you know, from the state.

In France, for example, they have a project like that because they already know that a lot of people are not gonna have a job. So and you cannot have a the majority of the population will have a job No. Because they get very angry. And that is the reason of the revolutions around the that occurred, around the the the first industrial revolution. But the first industrial revolution was nothing in the present to what we are walking to.

You know, there were now machines that were able to do the work of hundreds of people. Now we have computers that are gonna be able to do the work of tens of thousands of hundreds of thousands of people. And you already see several companies that are laying off, big segments of their workforce. They are doing it gradually. But, again, you just projected this trend forward, realize the speed at which, artificial intelligence is growing.

When you see the difference between the first version of chat GPT and and the current version of chat GPT are two completely different words. Then you realize that we have a significant problem. And I am I feel a little bit vindicated in a way that that that CEO of the largest bank in The United States, yes, they said exactly the same things. And what they did what they did was be careful. We need to time this process.

Otherwise, we're gonna have a big, big significant social problem in our hands, which is gonna be very dangerous from on a political level and a social level.

Speaker 1: Doctor, that's all just fascinating. I mean, I especially, key in on a part where so much of it comes down to how much we're willing to give up in terms of thinking and strategy. Right? And certainly the human element is is very important now, and we to put it out there, I don't expect any of us to have an answer to this question in this moment, but we have to decide how much of it is is worth it. Right?

So it's it's just fascinating and scary, and it's it's, again, still exciting in a lot of ways.

Speaker 2: But don't don't forget the element of greed, though.

Speaker 1: I I wouldn't. It's impossible.

Speaker 2: Again, artificial intelligence is gonna have major financial implications. So the risk, which is what, Jamie, Timon said, yes. The risk is that basically people go looking, go forward. They're looking exclusively at the bottom line, and that is gonna be devastating. And and then I basically ended with the with a a word of of of hope, saying that humans are still gonna be relevant and important.

There is no doubt about that, but our many humans are gonna be relevant. And he he talked also about, reskilling. Yeah. And, Chris, I'm I'm afraid that they're taking the issue with scaling a little bit too lightly because Yeah. It's not easy to get a 50 year old, that has done a certain job all his life or all her life and basically convert it into a Google engineer.

And there is there is

Speaker 1: a certain For sure.

Speaker 2: There is a certain limit. I mean, for the new generations, maybe. But what are you gonna do with all all those people that I don't think I remember one thing that, you know, I I found very interesting. I I was, at the Cleveland Clinic. I participated to the launch of, an upgrade that, of Epic.

I remember distinctly that a lot of the older physicians just quit. They said they want to to deal with the you know? They were very comfortable with paper charts. Obviously, we know that, you know, electronic medical record systems are much more efficient, much I mean, enormous advantages. But I'm old enough to remember when we didn't have that, and, you know, a lot of people spend most of their career.

It was a absolutely curl a cultural shock when when, electronic medical systems were introduced. And, again, I think there were a lot of problems when the first, EMS were implemented. So we have to continue, I think, at least make every attempt to put the human being at the center because artificial intelligence, Gen AI, and all the derivatives are going to be helpful and useful only for as much are gonna make people happier. And so to put the humans at the service of the machines is gonna be a a a major strategic, ever.

Speaker 1: That's well said, doctor Villalemonte. And and certainly as it relates to health care, you speak to any health care leader, and they always say that the patients and their staff is front and center. And we have every reason to to believe that's true. Right? I I believe that when when they tell me that.

They they wouldn't be in this this business, this area, this sector, otherwise. Right? So so that is encouraging, certainly. Lastly, doctor, I simply wanna ask you. You are on a sabbatical now as you mentioned.

Well deserved, clearly. But what do you think is is next for you? I mean, you have plenty that you to draw on. So so what do you look forward to doing in the next year, five years, whatever, time span, you know, you you see appropriate?

Speaker 2: Well, if I don't retire, which is supposedly, I definitely want to continue my research. I want definitely want to continue to do what I've done all my life, which is taking care of patients, which I still enjoy for as long as the robots are gonna allow me to. But in terms of this particular field of interest, I am particularly excited about how these technologies are going to, talk to each other, how they're gonna interact. And particularly at the crossroads between artificial intelligence, particularly agentic AI, quantum computing that is gonna change dramatically the speed of calculation. Therefore, it's gonna increase enormously the ability of artificial intelligence to function.

And, I believe clouds that basically are gonna allow these technologies to talk to each other. And as you know, a lot of hybrid clouds are are, proprietary. So Yeah. That's where the competition is, in my opinion, is gonna be. Not that much on the quantum side, not that much on the, Ajentic AI, but to have the clouds, where most of the data are gonna be stored or all of the data, and they're gonna be allowing this, this crosstalking.

That is where that is where I think the competition is gonna be, and that's where we are gonna see the most extraordinary manifestation of this technological, progress. Again, a lot of excitement, but but also the I think we have to be realistic, when this, this perfect storm is gonna is gonna is gonna hit. I hope that we are prepared because it's not gonna be the same world. It's not gonna be the same society. It's not gonna be the same organization.

It's not gonna be the same culture. It's not gonna be the same way of doing things.

Speaker 1: I hope the very same, doctor. Thank you so much for bringing all your your decades worth of challenge and expertise and insight on on AI to our Becker's audience today. As I've mentioned before, we love seeing you at conferences. I'm sure we will again. Yeah.

Enjoy your sabbatical, and, thank you again, and until next time.

Speaker 2: Thank you very much, Chris.