Q&A: General Catalyst's Stephen Klasko on the future of healthcare

LOS ANGELES – At the Future of Health Annual Summit, former Jefferson Health CEO and General Catalyst executive-in-residence Stephen Klasko sat down with MobiHealthNews for an in-person interview to discuss the real potential of AI, why medical education remains stuck in the past, and how economics and malpractice fears slow true innovation. 

MobiHealthNews: What do you think the future of healthcare looks like?

Stephen Klasko: I think about this a lot because I had a chance to be an advisor to Apple back in the early 2000s.

What Steve Jobs was incredibly good at is saying, "Think about what is going to be obvious 10 years from now and do it today." I think we in healthcare get so caught up in what's the next incremental step, right? So if you go back to the Apple example, Steve said, "The old math is computers and operating systems. The new math is going to be this digital lifestyle." I mean, the people at Apple said, "Hey, dude, our entire revenue is computers and operating systems," back in the early 2000s.

When I got to Jefferson, I became the CEO of a two hospital system and the president of a one campus university. I did six acquisitions. I merged the number three design university in the country with a 200-year-old Health Science University who created the design of the human experience in healthcare, and I used the same old math and new math thing. I said, "Look, the old math is in-person care, outpatient care and in-person tuition. The new math is going to be radical collaborations with these whole digital, strategic partnerships." I got the same, "Hey, you're either crazy or on drugs." Steve Jobs was on drugs. He wasn't crazy. I am crazy and not on drugs. But the fact is, my thing was, let's create true partnerships with these digital health companies.

So, the way I got into the job I'm at now at General Catalyst is I got to meet Hemant Taneja, the CEO of General Catalyst, and he had written a book called Unscaled about their investments in Airbnb, Warby Parker and Stripe – global economy is unscaling. And I sort of came up, being what I am, and said, "Hey, what if a Silicon Valley entrepreneur and a CEO in academic medicine had a baby? What would that look like?" 

And what was cool is we wrote that book. We wrote a book called UnHealthcare: A Manifesto for Health Assurance, which has really become the blueprint for everything in health that GC has done. All of our funds are called Health Assurance Fund. The acquisition of Summa Health is, you know, the Health Assurance Transformation Company [HATco]. So, in some respects, I feel like the OG of that concept, but I think the key is we have to get out of what are things going to look like?

MHN: President Trump announced Project Stargate at the beginning of this year. They are building large data centers, and one of the promises during a press conference was that they are going to be able to cure diseases. Do you think AI is going to help cure diseases?

Klasko: I think it will help. I think we have to be realistic. I mean I want to be optimistic. I want to think fantasy-wise, but you have to be realistic. It is why I was never a big fan of the Cancer Moonshot because there are 8,900 moons. I mean the reason they said that is, you know, it was when Sputnik happened and we said, "We're going to get to the moon." Here's A and there's B.

We've done some good things. There's some things where you would have died before that you don't now, but we haven't cured cancer, and we're not going to cure cancer. AI is not going to cure cancer.

But I think that AI can do things, like one of the companies I'm working with is called Paradigm, and it's using AI to democratize clinical trials. So, those kinds of things, absolutely it will help. 

It will help pharma figure out which molecules they should go after and that will cure certain cancers. So, yes, we will speed up the cure of diseases. What I'm really interested in, partly because I'm, like, 50-years-older than anybody else in venture capital firms, is this whole piece of not necessarily longevity, but health span.

This conference is doing a lot of the work around, literally, they've taken data of people that live over 100 and here is what is fascinating. I forget what it was, but like 25% of the people that live over 100 smoke and drink. Mick Jagger and those guys are still alive. What is it? 

And AI is helping us find things like some FDA-approved drugs for different things that actually help longevity. Like, Metformin is something people use for diabetes, but if healthy people take it, there's some evidence that will help. GLP-1s. I mean, we might get to the point where there's GLP-1s in the water, right?

So, I think AI is going to really start to unlock some of those things of why did this person live to 100? Why did this person die at 70? 

I think where I'm really excited is around Alzheimer's, because we do not know what causes Alzheimer's, but we might not have to know what causes Alzheimer's if AI can figure out who has the genes, like who has gotten it versus who hasn't gotten it, and is there a way we can deactivate or delay that, not even knowing what causes it. 

So, things like Alzheimer's, and in my specialty of obstetrics, preeclampsia and pregnancy-induced hypertension, we don't really know what causes it, if we can have AI help us determine who might get it, predict who might get it, and what we can do, that's going to be a game changer. So, yes, a game changer just like antibiotics were a game changer.

One of the best ways that I look at AI is like fire–when cave people got fire. Before fire, it used to take you 60% of your day to get protein in your body. Think about this, if you go to a zoo, mammals are chewing all the time, gorillas, right? So what happened? Fire happens and it's not taking 60% of their day anymore. They can get their protein in 20 minutes. Yeah, they didn't go, "Hey, cave partner, you know, what are we going to do now?" It's just that things that they had to do while they were sitting around just chewing, they can now be more productive.

MHN: Do you think AI is going to replace people in the workforce?

Klasko: I don't think AI is going to take over all these human jobs, but I do believe what we've always wanted, the Holy Grail. I'm a child of the 60s, so beyond the free love and dope, the one like Holy Grail thing was four-day work weeks. I think that's where we might get to. We will have a four-day work week, or less of a work week, because AI is doing all those things, and I can do some of the things that are more creative.

I think almost everybody has something they want to do, where they go if only I wasn't working 60 hours a week. Well, what if you are getting the same productivity at MobiHealthNews or wherever, at 30 hours a week, but you're not giving up any productivity, but your robot next to you is doing that while you're doing what you like to do.

That is the future I see, and I think we over worry that AI is going to steal everybody's job. If you think back, the number one job of semi-skilled people was being an operator. Most jobs in the country were, "Call 411. I need the phone number." There's now one operator, or maybe two: Siri and Alexa. But what happened is, the technology created AT&T and T-Mobile and those operators in semi-skilled jobs, got more skilled and now are working at AT&T and T-Mobile and Verizon. So, I think we'll start to see that now.

The reason it was cool having a university president here today, we have to train people differently. I mean, this is what drives me nuts. I've been the dean of two different medical schools and the president of a university–we still accept medical students based on GPA, MCATS and organic chemistry grades, and somehow we are just amazed doctors are not more empathetic, communicative and creative. Like, duh, right? It is asinine that it matters if I'm not going to be a physician scientist at Yale, whether I can memorize the Krebs cycle.

And especially when you think about the disparities, there are so many doctors that would go back to their community, people of color or whatever, that because they got a C in organic chemistry, never became a doctor. Who gives a s***?

MHN: Also, why are we turning down people who want to be doctors? We already have a shortage of doctors.

Klasko: The real tale of two cities are people that spend $150,000 for little Johnny or little Mary to get three Princeton reviews and a tutor, and there is another kid with a Barron's book. So, it is just not really fair.

MHN: So you don't think that we'll ever be in a tube where it scans us and it just heals us?

Klasko: It's a great question. It's really a great question. There's a problem with that. One of the things we haven't talked about is the economics. So, as this stuff happens, you know, we have great drugs now (CRISPR drugs and that kind of thing) but they're $250,000 a year. Do we give it to everybody? Do we just give it to wealthy people? It's the same kind of thing with full body scans. The problem with full body scans is, yes, I think we'll have them but there's a few different things. We have this weird economic model in the United States where it is hard to get somebody to do something when their salary depends upon them not doing it.

So, unless the economic model totally changes, and the malpractice environment! So, I'm an OB/GYN, and you get a full body scan and you have a small cyst, and that's probably not a big deal, but now I've seen it, and if I don't do anything, God forbid anything happens, there are going to be 12 malpractice suits. Well, you saw that.

In Europe, for every two breast biopsies, there's one cancer. In the United States, it is like for every 20 breast biopsies. Why? Because, as OB/GYNs, we do a yearly exam, we check your breasts, and if there is even, like, a dense breast or whatever, I have to get an MRI. And the radiologist is instructed to sort of not give a definitive answer because they don't want to get sued. If they don't give a definitive answer, then the surgeon feels they have to do a biopsy because they don't want to get sued.

Like, once we started to do 3D and 4D ultrasounds on babies, we found stuff that we didn't really know what to do with that just added to the concern of the mom and maybe some procedures that they didn't need.

I think there's going to have to be a balance between having people have too many things, and especially if you don't know what to do with it. Like, even now there's these tests of whether or not I have five cancer cells roaming through my body. Do I really want to know that unless I know: A) Where it is, and B) Can you do anything about it?

MHN: Exactly. Do I really want to know this?

Klasko: I'm 71, and there probably are five cancer cells.

MHN: I don't want to know if I'm going to die tomorrow.

Klasko: That's another really, really, really interesting one, because there are all these centers for human longevity.

These people pay like, $50,000 to look at their nutriome, biome, and every genome and everything, and let's assume that they could actually say, within this range, this is where you're gonna pass, between the age of 72 and 75 or whatever it is.

I always ask people, under 50 about 90% of people say, "I don't want to know." I think what happens is people over 65 say, "Yeah, I wouldn't mind knowing," because you figure, like, it's a different psyche. I'm 71, right? I mean, I run marathons, but if I knew in all probability that I was going to make it to 85 it's different than in all probability I was going to make it to 75, you know. how much time I would spend doing certain things. So, I think some of that will be good, as long as you have your choice. 

Now, what you have to worry about is that it's not being used by your life insurance company. Because now you get these notes, we've decided not to insure you, and we're not going to tell you why. You can imagine if there was a blood test. It's because you're going to die in ten years.