3 Takeaways - Mark Cuban: The Hidden Companies Controlling Healthcare, the Future of AI, and Why Not Me? (part 2) (#307)

Episode Date: June 23, 2026

Most people think healthcare is broken by accident.Mark Cuban says it isn't.In part two of our conversation, he explains the hidden companies who really control healthcare, where AI is headed, and the biggest opportunities most people are missing.He also shares the advice he gives young people, the three “superpowers” he believes matter more than talent, network, or money, and the one question he thinks everyone should ask themselves.

Last time, Mark Cuban talked about what makes a great entrepreneur and why he started cost-less drugs. In part two, he gets into how a handful of massive companies dominate healthcare and why that's no accident. He also shares where he thinks AI is really headed and where the biggest opportunities are right now. So what does he tell kids when they ask if they have what it takes? And what are his three superpowers for success? Hi, everyone. I'm Lynn Toman, and this is three takeaways. On three takeaways, I talk with some of the world's best thinkers, business leaders, writers, writers, politicians, newsmakers, and scientists.

Each episode ends with three key takeaways to help us understand the world and maybe even ourselves a little better. Today, I'm excited to be with Marry. Mark Cuban, entrepreneur, investor, and former owner of the Dallas Mavericks. He went from getting fired at 25 to building a billion-dollar career and inspiring entrepreneurs through Shark Tank. And now he's working to disrupt health care with cost plus drugs. Mark, thank you for joining three takeaways today.

Thanks for having me, Lynn. I'm really excited to be on the show. What do most people not understand about how the health care industry works? It's so vertically integrated. The biggest companies do hundreds of billions of dollars in sales, as thousands of subsidiaries, do $160 billion in intercompany transfers, which allows them to gain the system. We hear stories about pharmacy benefit managers, about insurance companies,

but they own thousands of other companies that you don't even know they own. They may own the value-based care center, the emergency room. And by owning so many vertically integrated companies, they're able to define the economics of the entire healthcare industry. So when we talk about health care, we typically like our doctors and they do a good job for us. It's the economics of health care that we truly dislike. And that's defined and really regulated, let's call it regulatory capture by those big

companies. There's a bill that Elizabeth Warren and Josh Holly put out called the breakup big medicine bill, which I think is the only way we're going to get to an efficient market where the best doctors, best providers, best pricing is going to exist. And is this the three pharmacy benefit managers that own everything? Yeah, the big three BBM. Yeah. In two cases, they're owned by big insurance companies. In one case, they own a big insurance company. And they've gotten so big, A, they're two big to care. And B, it's basically regulatory capture, scale capture, where they're that big where they typically set the prices for everybody's health care, other than in certain hospital situations. And that's the reason that there's no transparency at all in health care?

Yeah, because transparency is their enemy. 60% of people get their health care through employers who self-insure, where the employer takes responsibility for all that cost of health care. And even with big companies, with thousands, tens, hundreds of thousands of employees, the PBMs don't give full information. In many cases, they won't even give all the claims information. So the company doesn't know, you know, if people are staying healthy or not, they don't know what anything costs. It's deliberate for them to really create this information asymmetry and opacity because that gives them an advantage when it comes to pricing. So what are you trying to accomplish now? What we're trying to accomplish is make it so that information is available to employers and employees

so they understand what everything costs. We created a website called cost plus wellness.com. It's a place where providers, doctors, hospitals can put their contracts or care so that mostly employers can use it to cut their costs and the care they provide to their employees. And the reason it exists is it's the only place where people can actually look at contracts before they make a commitment on what to pay for the care that their employees are getting. And so we're the first to do that as well. And it's really been amazing because employers don't trust the insurance company they work with. The employees don't trust it. We all go through it every day, right? Each year, there's open enrollment and we have to pick our plans and we make a determination based off the health of our families, what we can afford. And it's a tough

decision and there's always information that we didn't know about that we missed. And that should never be the case. But it's intentional. And it doesn't just apply to you, you know, with your employer or the ACA or Medicaid or Medicare, whatever it may be. It applies to your employer who's trying to make the best decision for everybody that works for them and their families. And these big companies do everything possible to prevent that. It's crazy that nobody knows what anything costs. I'll make a suggestion to everybody, when it comes to the insurance plan you pick, or if you're an employer, the insurance companies you work with or the PBMs you work with, take the different contracts or take the different plans that are offered to you and put it in your favorite AI model and just ask it,

is this good? Or what question should I ask? Or here's what's important to me. Am I missing anything? Because the one thing about those AI models, when you upload a PDF file or a text file, whatever it may be, they're good at reading the details. Much better than you, than we are, right? We all zone out halfway through. And so by having it, the first thing I always say is summarize it. And then the second thing is, what am I missing? You know, is there something I need to be concerned about that I might have otherwise missed? And they're not perfect, but they're really, really, really good. And it gives you an advantage as opposed to being at a disadvantage. Okay. Let me ask you more about AI. There are a handful of companies building powerful

proprietary models. How do you think this plays out? You know, nobody knows. We don't know, and the companies don't know if it's going to be like the search engine industry where Google dominates and everybody else has maybe 2%, or is it streaming where Netflix is the leader, but there's 20 different other streaming companies that have a chance. And so not only do we not know, but the AI companies themselves, you know, Google with Gemini, Open AI with ChatChief BT, my favorite Claude from Anthropic, et cetera. They don't know if it's going to be winner take all or not, so they have to go all in.

And so they're raising hundreds of billions of dollars, investing in aggregate trillions of dollars because they have no choice. If you're second or third in this winner take all, any amount you've invested is wasted. On the good side, that's great for the country in terms of technology available. It's great in terms of competing

with other countries like China and elsewhere, but it's taking a lot of money and a lot of capital into these five, six companies. And we don't really know how it's going to turn out. It used to be in the PC industry. You always wanted the fastest PC. And whenever a new PC came out, this has the Intel 80 megahertz, 90 megahertz, 120, 200 megahertz. It's faster.

And you noticed the difference when you used it, right? Do you remember the days when it would be faster, faster, and you wanted that faster PC, or a new phone came out and it had more features? And then you got to a point of diminishing returns. And we don't know with AI, with those big models, if or where those diminishing returns are. And will we get to a point where we don't need the other 10, 20, 30% that they're trying to add, and somebody smaller were pop in. We don't know if there's going to be different or better ways.

I'm old enough to remember the days of WordStar and Word perfect, long before Microsoft Word was big. And they were dominant. It everybody just used them and that's what you learned in school until they weren't. And there's different technologies for AI called a worldview because with today's AI, you can give it pictures and you can give a text, but you can't, it doesn't learn from video. And it definitely doesn't learn from real-time video. And it definitely doesn't understand the consequences of its actions or recommendations. Will it get that smart?

It's going to depend on how we define smart. But there's a lot of things that have to happen. I'm a huge fan of AI. I try to use it for as much as I possibly can. It's a great democratizer of knowledge, but you have to know that it doesn't know everything. You know, you can't use it so you don't have to learn anything. You have to use it so you can learn anything. And I think once you understand that, it's an amazing tool. Yeah, I just hope that kids don't lean on it too much and that they learn critical thinking. I remember cheating on a test when I was in high school. We're always looking for shortcuts at various levels. I think it's going to help students once teachers adapt. Because right now, like I've gone to talk to teachers' groups. And I always ask them, how hard is it to get the light in every kid's eyes to shine bright about a topic in your class?

And they're like, it's impossible. Because we always teach the same subject matter to each child and try to get them to have the right answers. As opposed to adapting to each one in a way that really captures their curiosity and attention. That's what AI allows you to do. Once for a history class in sixth grade, we change things so that you're teaching about the Silver War and you can work with the AI to adapt to each student and have the AI review the responses and give different questions to support what they need as opposed to everybody getting the same questions and expecting the same answers.

As you start to do that, I think you're going to be able to get kids more excited about learning. If you can get that kid to say, hey, I want to learn more. about George Washington. That AI is going to teach some. And if that kid says, hey, I'm 10 years old, would you put together a study guide so I can do better in school, make it so that I can learn all the subject matter I need for school? The AI is going to do it without help from mom, dad, or the teacher. That's just such a huge step forward once we adapt and understand. Doesn't mean kids aren't going to cheat. Doesn't mean kids aren't going to take the path of least resistance or parents won't either or employees won't. That's just human nature. But once we get that attitude and

kind of reframe how we teach, I think kids will get smarter, not dumber. I hope so. I love the idea of AI generating that spark in kids. That is the whole goal. Five years out, what do you think AI changes? What will be different? You know, if you would have asked me this three months ago, I would have said it was dramatically different. But I think now, because it changes so rapidly and you learn stuff every day. I think it's going to be things that won't be quite as different as some people think. I'm not a dumer. I don't think everybody's going to lose their job. I went to school at Indiana University and Tara out in Indiana. I remember famously made all the records. RCA records was there. Then they made all the CDs. And there was a company that made the tuners. Remember when you had to

stand up to change the channel on TV made all the tuners? And all of them are gone now. And yet all those people lost their jobs and hopefully they found others and we survive. There are going to be people whose jobs are just yes or no answers. Do I do this yes or no? Is this right? Yes or no. They're probably going to lose their jobs unless they learn the skill sets that allow them to apply and figure out other jobs. In a big company, 20 people could ask the company's AI the same general question. Like, how should I do this and get 20 different answers? And unless you have, the ability to understand and use judgment and communicate, you can't cover every set of circumstances. Like we talked about agility in business, right? Big businesses have to be agile as well. Within

departments, they have to be agile. Well, how do you be agile after you've already trained your AI? Now, it can learn some things, not like a human can learn, but it can figure some things out based off of what is trained. But somebody's got to have the judgment when things zig and zag to know, no, this is the right answer. You've got to be able to spar with it to say, no, that's not the right answer. Maybe you need to talk to your boss, which means you are going to have a boss to make that determination because there'll be so many people using AI. The level competition will increase. It won't be sixth grade basketball. It'll be NBA level basketball, but you'll have the tools to be able to do that. If AI makes knowledge and expertise abundant, do you think

the personal relationships and businesses that offer personal services will be in demand? Yeah, of course, I've even made some investments along those lines. People are going to crave dealing with other people. But I don't think that companies are going to cut as many jobs as people think. The challenge will be the innovator's AI dilemma. When you have competition that's AI native and tries to compete with you with much fewer employees and more automation, then how do you respond to that? And some companies will try to just cut people. I don't think that's necessarily the right approach.

I think coming up with ways to use that judgment people have to counter the judgment that AI doesn't have, I don't know for certain, but may be the way to go. It's essentially rethinking a business from the ground up because there are new competitors coming in. Yeah, and it's always been the case. I remember selling PCs to companies, and the big deal was being able to do,

what if on Lotus 1, 2, 3, which was just a spreadsheet. And buying a spreadsheet in the 1980s cost $500. And I literally had to go get trained in order to be allowed to sell a spreadsheet, which sounds hysterical now, just be on the pale. But everybody who got that PC and a spreadsheet had a huge advantage until everybody had a spreadsheet and learned how to use those spreadsheets. And I think AI will be the same thing. You can have it calculate what if all day long.

But if you don't know which what if is the right if, you haven't accomplished anything. Yeah, it's the what if for the traditional companies, like the car companies. They didn't see that an electric car is essentially an iPhone on wheels. Yeah, exactly right. Great way to put it. I'm stealing that. And you're right because they were car companies and they wanted to act like car companies. And the only similarity between an EV and a gas car are the wheels.

in the general shape. Mark, you talk to many kids, including young kids, elementary school kids. What do you tell them? I'm glad you asked me that. I always say the exact same thing. Whenever I'm talking about I talk about, but I always end with the same thing. I say, look around you. And typically there's a big screen TV or a big, you know, drop down big screen, and there's

lights. And I'm saying, you see that screen there? One day it didn't exist. Then one day someone came up with the idea made it happen. chairs you're sitting on, they didn't always look like that. Someone decided they had a better idea for chairs and they made it work and now you're sitting on somebody's idea. You see what you're wearing completely different.

So I want you to ask yourself a question. Why can't you come up with the next great idea? Why can't you come up with something that changes the world? So when you're second guessing yourself and wondering, can you do it? Ask yourself the ultimate question. Why not me? why can't I be the one to change the world? Because inside of everybody, there's something. The hard part is just finding it. And of all the things that you touch, that you buy, that you wear, that you see,

those people aren't famous. They weren't people who, you know, oh, that person's going to change the world. No. It's just somebody who had an idea, asked themselves, why not me? And then answered, why not me? And went through with it. That is lovely. What advice. do you have for young people starting their careers today? Learn AI. Learn everything you can about AI. It's like carrying the library that you went to school with in your back pocket or in your front pocket. And because you're just graduating, you're pretty much AI native.

You're the first college. My daughter is getting ready to graduate from college this week. And that's what I've been telling her and she's been picking up on it. And, you know, I'm proud of the fact that it's working for her because she initially doubted me and now it's, you know, but learn AI, learn agentic AI in particular, because there's 33 million small sole proprietorship, small type businesses who most of them have no clue about AI, can't afford to hire somebody, but they all have a to do list. And there's always things on that to do list that they can't get to and can't afford to hire somebody to do. And using a gentic AI, you may not be able to do all of them, but you can walk in, charge $25, $50, $100 an hour, whatever it is. And, do those things for them and then charge them some amount per month to come in and check on them to make sure it's all still working. And you do that and become that expert. You can work for yourself or you can go get a job working for a consulting company or you can go work for even a bigger company,

but don't, you know, historically the goal was to get a job with as bigger companies you can, get paid the most amount of money. But those big companies are the ones going through the biggest adjustments. And those jobs there that you typically expected in the past probably aren't there as they try to figure out. And probably because they have older employees that they're retraining or already have, you know, IT experience that they're getting up to speed because they don't want to just fire them. And so go to those small companies that need that help, understand agentsic AI and everything you can, and also realize you have more time to yourself now than you ever will for the rest of your life. And AI is changing so rapidly, you'll be in a unique position to be able to

keep up with it where so many other people have to worry about all the other things adults worry about. You'll have a big advantage. And finally, your three takeaways for someone listening to this conversation and who only remembers three things, which should those three things be? Curiosity is a superpower. Being nice is a superpower. And not being afraid and being excited to use AI is a superpower. Because once you can break down that barrier and realize it's a friend, not foe, then you are going to be able to do things far more efficiently

and fill your curiosities far more often than you ever dream possible. That is lovely. This has been a pleasure, Mark. Thank you. Yeah, I really enjoyed it. Thank you. Great questions. I really enjoyed this interview.

I really enjoyed it too. And if you missed part one of our conversation with Mark, I hope you'll go back and listen. If you're enjoying the podcast, and I really hope you are, please review us on Apple Podcasts or Spotify or wherever you get your podcasts. It really helps get the word out. If you're interested, you can also sign up for the Three Takeaways newsletter at Three Takeaways.com, where you can also listen to previous episodes. You can also follow us on LinkedIn, X, Instagram, and Facebook. I'm Lynn Toman, and this is Three Takeaways. Thanks for listening.