Jim Kean is a senior technology executive with an innovative career focused on making healthcare easier by pioneering disruptive consumer-oriented health, wellness, and fitness digital experiences. The consistent theme: creating user delight, loyalty, and retention while fundamentally re-engineering the cost of delivery.

Prior to joining Molecular You, Jim was initially General Manager of the Consumer Platform for Cambia Health; a Blue Cross Blue Shield insurer based out of Portland, OR. In this capacity he drove development of a next generation platform for omni-channel consumer engagement. Later he became General Manager of Value Management which focused on population health analytics.

Jim is Founder and former CEO of WellnessFx™ a venture backed San Francisco based company that made it easy and affordable for consumers to manage key areas of their healthcare lives by providing inexpensive, profile driven access to targeted health content, diagnostic and telehealth services, and HIPAA secure storage. January of 2013, the company was named 2012 Innovative Company of the Year by the Nutritional Business Journal. In December 2013 the company merged with Thorne Healthtech (THRN).

In 1995, Jim founded Sapient Health Network (SHN) which became the Consumer facing business of WebMD. As President of SHN, Jim pioneered the concept of online consumer health communities, building some of the largest and most loyal groups on the Web today. The company’s service was named one of the “Seven Best Websites of 1998” by Business Week. In 1999, SHN merged with WebMD and later Healtheon to form the current version of WebMD. SHN represented the consumer facing side of the merged companies. Today over 23 million consumers per month visit WebMD for health advice. After WebMD, Jim served as Entrepreneur in Residence with Alloy Ventures of Palo Alto.

Steve Martorano 
The Behavioral Corner is produced in partnership with Retreat Behavioral Health -- where healing happens.

The Behavioral Corner 
Hi, and welcome. I'm Steve Martorano, and this is the Behavioral Corner. You're invited to hang with us as we discuss how we live today, the choices we make, what we do, and how they affect our health and well-being. So you're on the corner, the Behavioral Corner. Please hang around for a while.

Steve Martorano 
Hi, everybody, welcome again, it's me, Steve Martorano hanging on the Behavioral Corner. I hope you know about us. But if you don't, I'll tell you briefly and modestly, this is a podcast about everything. Because everything is what affects our behavioral health. It is all made possible by our underwriting partners Retreat Behavioral Health, you can find more about them down the road. We're going to take a look at a fascinating topic that sounds brand new but in fact, it is a very, very old story. It's the intersection of technology and the practice of medicine. I dare say that those intersections are as old as Hippocrates. I mean, it includes everything from the stethoscope to MRIs. Whenever there's been a technical advance that could benefit or move medicine forward, medicine would adopt it and carry on from there. So there's nothing new about talking about the latest innovation technologically, that is going to impact not only medicine but everything. And that is artificial intelligence. But I do believe if we're having read some of what our guests are written, and what I read myself about the power of AI, we're maybe it's some kind of inflection point, because of the awesome potential for artificial intelligence to change everything. And we're going to take a look at what that means in terms of medicine. With our guest, Jim Kean, Jim is the CEO of a company called Molecular You. That's Y - O - U. We'll find out what that company does. He actually is a serial entrepreneur. I love that phrase. It sounds sinister, but it's not. And he's here because he has some strong feelings and some predictions about AI in medicine. Jim, thanks for joining us on the Corner.

Jim Kean 
You bet. Steve, glad to be here.

Steve Martorano  
So I think I got that, right that, you know, medicine has been looking for new and different things and different forms since the beginning of time. What makes AI different? Is it different?

Jim Kean 
I think the promise, and also potential danger of AI is that you're in a way, giving away decision-making authority. And it's how you really decide who you want to be responsible for different parts of say, in medicine service delivery, like, should the AI take the part of the doctor? Or should the AI just be the person who checks you in? Should the AI make the diagnosis? So there's all these things that you need to check the box on. Because medicine is actually a complex set of steps that happen. And you need to be super thoughtful about who you give authority to. Who has the license and makes the call? And then, you know, what parts do you think can free a doctor up? You know, because like the potential, there's a joke from a medical conference I sat at about 15 years ago. And they were talking about electronic health records and all that and the promise of them, and the doctor said, Yeah, you know, what's great about EHRs is I'm now doing tasks that take me a half hour that I never had to do before.

Steve Martorano 
We have no problem accumulating this data. We just nobody pays any attention to it. Now, before I want to get into the weeds on that. But I've gotten ahead of us here a little bit. Tell us about Molecular You.

Jim Kean 
Yeah, so that kind of starts with a little bit of my backstory when I left business school in the early 90s, I decided I didn't want to go work for a large company. And I wanted to get involved in startups. And that wasn't very popular back then. So I joined up with a health tech startup, I went through that I discovered fraud at the startup and I ended up becoming the CEO at 29. Because I reported it to the board. And then I orchestrate the sale and I was kind of hooked on doing startups. But I finally got to do my own. I had this big checklist of things not to do. And that was this company called Sapien Health Network. And that became actually when we merged with them, the consumer-facing business to WebMD. And so the thing that's been great about digital health and just being in consumer digital health, is I've had the privilege of seeing every single step. So back in the 90s, the big step was let's democratize access to data by pasting a browser over it so that you don't have to go ask a doctor or nurse you know, what this means. I can go look it up myself. And that was fairly revolutionary back then. And so over time, we've had more and more of this digital health come into the fray. And I've gotten to see every step. Now, most recently, I decided to start a company called Wellness Effects in 2010. And that was to democratize access to diagnostics, blood diagnostics for consumers. And we ended up selling these packages, you didn't have to go ask a doctor, you could just book it online. And then you get bundled with a 20-minute telehealth console. So we're pretty groundbreaking at the time. And a lot of this work, we did lead to a lot of other companies being formed. And I ended up having to sell that company for a variety of reasons. And I went, I sold it, I was invited up as a guest speaker at the Precision Medicine conference in British Columbia. And I met a bunch of really amazing world-leading scientists, because after the genome was sequenced, you had to go sequence the transcriptome, which was RNA, and then that followed with the proteome, and then the metabolism. And so all those things are in the chain. And so the promise of molecular medicine really was highlighted by this amazing group of people. But back then it was like 10s of 1000s of dollars and multiple tubes of blood to run. But over time, I became an advisor to them for the past nine years. And I worked at Blue Cross Blue Shield after that, but I still kept in contact with them. And now it's checking in with them about a year ago. And they said, you know, we've had kind of tough going here in Canada. But we, you know, we've gotten the cost of this down sub $100. It's a small tube of blood, it's easy to draw. And so everything's gotten scalable. And I said, you know, the US has actually got a lot of places where this would be readily taken up. So that's how I became CEO, of the company. But I've known the scientist and the people there for over nine or ten years. And so the premise of doing this way to think is your genes are your architecture, right? Unless you're in a secret government program, your genes aren't changing. But they're a possibility. So then all the choices and that's about 10 to 15% of your total health outcomes are driven by your genes, the rest of it's actually the type of health care yet and the personal choices you make. And so that leads to the ongoing synthesis of your proteins or your stack of about 25,000 proteins and about 75,000 metabolites. And the way those readout on a second-by-second basis will indicate if you're going through a good time or a bad time, your life if you're making good dietary choices, if you're smoking, not smoking, drug abuse, all that stuff, it gets read out on your proteins in your metabolites. So the thing that's intrigued people forever, is what if you could cheaply and easily at the molecular level, read this stuff out, you would see things happening far in advance 5, 10, even 15 years before you actually had conventional symptomology come up from you know what we do right now. And so that's the thing that's always been the hook for me, and the fact that it's gotten cheap, and scalable. And now we're in this golden age, our current assay is kind of think of it like a chip has 250 biomarkers that have been validated, we're about to release the next version, it's going to have about 825 on it. So it's just this remarkable outpouring of scientific research, and this decade is going to become known in the future is probably the Golden Age when all these things started getting unlocked. And the promise here is it's going to personalize increasingly and make it easy to see what's going on. So I'll pause there, I just said a lot.

Steve Martorano 
No, no, it's good. It's good. It says we certainly know and understand why, why you're excited about what you see now. Because you've been looking at this for a very long period of time. Most of us have come in contact with AI overnight. And it's baffling and sometimes scary, and certainly not well understood. But I think I understand what you're saying this has been a process that started a long time ago. What's different now is that technology has changed. And as you said, made it vastly more robust, cheaper, and faster. And I think I understand now what you mean when you say that AI, potentially a goal, one of the goals is to move from being a very reactive practice to a predictive practice. Give us a little more explanation for that because, on some level, it sounds like fortune telling you what's funny about genetics. Most people approach genetics in the way I do. The family traits and genes I think I have that I think are valuable like maybe longevity or I still got my hair. I think that's cool. But other parts I go now that genes don't have anything to do with it. So if I understand you genes are important. but they're not utterly determinative about how will wind up.

Jim Kean 
A lot of your personal choices dictate how your body decides to express your overall genetic code. So imagine like, an architect delivered you a set of plans. And you got a contractor, let's say it was a contractor who used really substandard materials, built it on poor land, and then didn't take care of the house, you would end up with a really pretty crappy house, right? Let's say the architect, though, is a design genius, or not the architect, but the contractor really had a sense of artistry to use great materials and had a flair to it. And it became a signature house that, you know, everybody wanted to live in and buy. Those are the same house plans, right? But they got expressed differently. So that's almost the same process that I'm talking about biologically, as well. Now, the thing that other was the other breakthrough, just to say where the AI came in. Right now, there are about a million new pieces of creditable clinical and scientific research coming out that identify like, here's a marker, or here's something you didn't know about. And it's beyond human comprehension right now, to be able to constantly find that and make sure that some obscure scientist, you just made a discovery has that incorporated into our overall health, and so you couldn't do that without AI. And so the other part that the company really has revolutionized, I think, is in the last nine years, they put together this machine learning apparatus that sits on top of all the scientific journals and medical journals and if that keeps up all this data, so whenever a new marker is discovered, it doesn't just go into some dusty warehouse, digital warehouse, but it gets surfaced, and then we'll actually the way it makes it on our assays, we go out and see if we can replicate the results by running it through our testing system.

Steve Martorano 
So let me ask you that, when we got to get back to, you know, talking about the ghost in the machine, and what they're capable of what seems like magic, but where's the, in this specific case where does the doctor stand? I mean, the machines may know all this information and may in fact, even be able to predict it in a certain way. Where does the physician come in for that? Yeah.

Jim Kean 
So that was the other problem that company or challenge the company addressed in the last 10 years, because I took one of the first tests way back when, and I got my results. And there were unlike anything I've ever seen as far as depth and complexity. But the doctor and I, and I'm not a doctor, by any means. It just happened to have been in the field for a long time. I'm asking you to say it looks super interesting, I have no idea what it means. So right there, you can see the problem, right? If it's data, it's not information. And so part of the job has been now in the last 10 years, concurrent with all the discoveries of relevant markers, is how do you put it together, so it becomes practical, that just have in the everyday flow of a clinician, where they don't have to work hard. You know, because one thing, doctors and other medical professionals work really hard to get their license. And so generally, they're not going to do anything that makes them look silly in front of a patient. So if they aren't sure about the data they're getting, they're probably not going to bring it to you or use it. In fact, they'll just stick to what they really feel they know and wait for it to catch up. So there's an old saying, in medicine, medicine advances the death of one doctor at a time. And the other stat that always stuck in my mind is whatever treatment a doctor prescribes for, you know, strongly correlates with the year they got out of medical school. And it makes sense because you get out of medical school and you're like on the frontlines of Vietnam with bullets whizzing over, you're not going to go down and do a bunch of research or get current all the time. If you're super dedicated, you may pull down some scientific citations on one interesting case you saw that week. But mostly, you're just cranking through a bunch of 10-minute appointments.

Steve Martorano 
I mean, I see the diplomas on the wall, but I don't know where they finished in our class. You know, I mean, they may have just made it I don't know. So the technology is only going to be as useful as the people who can interpret it, and then make it practical. I certainly get that. Right now, let's talk about some areas that maybe you know, an ordinary person getting a medical checkup might benefit from, What's going on right now that AI has changed?

Jim Kean 
So you get some strong opinions. And there are two kinds of general schools of thought around AI. One was what I call the Babylon Health from five or six years ago where Babylon said, alright, this spot is going to replace the doctor ultimately, now they changed that a bit but back in the day when they first came up there that was the whole idea is that you don't need a doctor anymore, eventually AI is gonna get so good that they'll be able to diagnose you and you know, in a lot of cases, and then the AI can sort through diagnosis is pretty well. The second school is called human lib. And it means that the Babylon bot should have actually been the assistant to a human doctor and that the human doctor has the call, ultimately, but uses the information and it helps him shortcut and get through things really quickly. So from a practical standpoint, I think the apparatus can be constructed around a doctor to free them up from administrivia, and other sorting through things and they can just focus on being people with their patients, and almost have that coach-athlete relationship that generally gets the best results. And so if we can free doctors from being administrators or entering things in the computer, or this and that, to practice medicine, I think that's going to be a win-win for everybody. If it becomes that we just get rid of humans and in your doctor bot will see you now. I think that's a little bit more dystopian, in my opinion.

Steve Martorano 
So it seems to me, I'm wondering whether you agree that the practice of medicine is part science and part art. Do you think that AI has the potential to narrow or change the balance of that what you described, is that in effect, it seems to me the doctor, is the human player in this the machine is the tool -- will that process be enhanced? Or will that art-science divide be eliminated?

Jim Kean 
I think if you set it up to where the AI has guardrails, and it knows its role, and that the human knows the role of the AI, it will be a net benefit, because you'll have expanded medical care and access to professionals. And instead of saying the human has to do 100% of the task, you take away the 90% that can be automated or done by a bot. And that 10% that is left is the really intense, smart, creative art part of the that you just described, of the practice. And the other thing is, when you really get down to it, and part of our testing results, the most fascinating part of that has been that we are now seeing almost an explicit link because when we're running this test, it's holistic across all your body systems. And you can generally see that a lot of conditions actually have a cognitive and behavioral link. And so for example, we have our autism test has a 99% accuracy profile, based on the biomarkers. And what we see in a lot of those patients is they have comorbidities, you know, and it turns out that there's a lot of the metabolites that come out of the dietary processes, to show some toxic compounds are fed and all that and so it and it basically affected the brain formation and, and whatnot. So you're having this gut mind link. And then it's surfacing early stages, and we still have to validate a lot of this, but there is definitely a link to what's going on with the rest of your body and cognition, behavioral health.

Steve Martorano 
Our guest is Jim Kean. He's the CEO of Molecular You, he's here to tell us to give us the benefit of his expertise the in the area of technology and medicine and the intersection, and what artificial intelligence may, you know, present in the future here. You know, when what were the doctors, when Bones would come in and someone was lying on this? On the gurney? He would do his diagnostics...

Jim Kean 
With his wand!

Steve Martorano 
With the wand. And the wand would tell him everything that was wrong. Is that science fiction? Or is that a potential future for us?

Jim Kean 
I mean, you kind of do that with MRIs and other imaging right now, it's just a really intense process, and you don't have a hand MRI and whatnot. So you can see logically if you get really good with the magnetics and the other aspects of the hardware that's where it could go. And the imaging, you know, for example, I saw a really interesting company about five years ago, and this entrepreneur had collected, I think about five or 6 million scans, MRIs of somebody's brains, and they have unified medical records that are so they're able to match it to qualitative assessments of mental capabilities as far as a progression from like early stage dementia all the way to full-blown Alzheimer's, and then he matched it up to brain shape. And so that was pretty interesting, but the only way he was able to do that was by using AI machine learning and AI technologies to rapidly classify brain shape like the shape of a brain It seems to correspond with early-stage dementia. So I think that's the perfect area where you can do that. Now getting back to Dr. McCoy. You know, you've seen the march of how we've reduced how many tubes of blood are needed and whatnot down to almost like a Theranos. Like, a mythical small tube...

Steve Martorano 
That was my next question, I would most people find out a lot about their condition through bloodwork joining a blood couple of vials, we know the Elizabeth Holmes debacle, the fraud, and the claims that she made concerning her technology that was going to panel all kinds of things with the drop of blood turned out to not be true. Will AI make her vision real if it's not already?

Jim Kean 
So if you think about the dominant paradigm on blood testing for many, many decades has been you'll take the sample blood, and you'll run a bunch of chemicals against it to get the diagnostic outreach, like cholesterol or glucose or insulin, things like that. And it's been a very single biomarker, so you'll look at your glucose and say, Oh, well, it kind of looks like you're diabetic. But it's not seeing, okay, you're diabetic. But you also have all these underlying comorbidities that are arising. And there's very little comparison about the interrelationships between the biomarkers. We take a totally different approach because there are so many markers and that influence, we see them as clusters, or families of markers between 10 and 20, for example, and Alzheimer's, and how all those relate to each other. You know, the mathematical precision of where each marker is, and then where it is relative to everything else, it's more like a bio print. And there's just no way you can hand calculate that you need to use your eyes to go through these sophisticated prints that come out to give you insight into conditions like Alzheimer's, depression and other ones rheumatoid arthritis, various cancers across the board, and increasingly, they all in or relate with each other. And you couldn't do that without AI either.

Steve Martorano 
Is it safe to say that the things that we have been working on looking for causes and treatments would have taken decades and decades longer to be successful on were it not for the emergence of very powerful artificial intelligence machines?

Jim Kean 
Yeah, and I think you're, you know, everybody's kind of apprehensive, or not quite sure where it's gonna go. But you got to realize it's just really at the early stage. And they're just really beginning to harness this, but also, it's gotten really inexpensive to run, like proteome metabolome tests. And that shows you the function and what's going on right now. And I can give you a great example of one of our biggest wins in the last year is we have one of our longtime subscribers. And she was very, practice a lot of just classic health and wellness approaches. And when she gets suggestions after she got her test results from us, she'd work on her things to move those markers. And then her last test came in, and the markers lined up, and it spits out a high-risk profile for pancreatic cancer. And that one, I don't know if you know of it, but it is a masked cancer. And it generally doesn't surface until you're about three, three months from dying. So it's known as a death sentence for cancer. And this one, we found, she had three small lesions, it was surgical intervention was likely or you didn't have to chemoradiation. And it was caught early. It was stage one. In fact, the clinician said I've never seen a stage one pancreatic cancer. And so just running that regular screening surface that and she had the intervention and we just reran that test with her to see what her markers were doing post-surgery, and they've all gone back to normal.

Steve Martorano 
It's incredibly exciting.

Jim Kean 
Yeah, but think about it. If you're looking healthy on the surface, you don't get asked to get tested. And if there's no symptoms, you're not gonna go probe further, because medicine is kind of in a hurry. That's how they get paid. So if you think about it, we need to kind of completely shift to What if you got a test like this once or twice a year, and you had this long-running since the set of tests and they cost about 100 bucks, and it would go over 26 to 50 biological systems in your body, give you the readout tell you how they're interacting with each other. And anything starts forming even five or 10 years, you know, early-stage pre-symptomatic, you would basically save the system a lot of money and you would also avoid a lot of individual human suffering.

Steve Martorano 
Fascinating. Jim Keane, CEO of Molecular You, Jim, I could go on forever, but just this last Seeing there's a recent article in The New Yorker about the impact or the potential for artificial intelligence to treat mental health, mental disorders. And from the article they talked about, there are millions of people now who have apps on their phones Apify and Replicate -- cognitive behavior programs that help people with either depression or anxiety. Do you see a role for artificial intelligence in the treatment of mental health?

Jim Kean  
I kind of go both ways here. So whenever I think about AIs, what you're talking about is forming a brain. And the way a good AI gets formed is through a training set. And so the richer and deeper the training set is so giving you a good example, Walmart archives, trailing 20 years of phone calls to Ask a Pharmacist. And if you convert it that, because they get millions of calls a year, if you converted all that you would have a pretty almost touring, like _____ to answer pharmacy questions, because that's a great training set, right? Yep. But if you have a limited training set, then you're really quickly can go off the rails as far as bias, you know, maybe you go into a practice where all they do is treat one type of patient. And there are only about 200 of them. Right? So it really is the training set part.

Steve Martorano 
And it's still garbage in garbage out?

Jim Kean 
Yeah. And then the other thing, if you think about it, like chat GPT everybody's on fire about that. And then in June, you had a couple of studies posted because you were starting to see the advent of poisoned AIs. And what the term meant was when they analyzed that, they realized that the AIs in their training set had started gobbling up, you know, bot-generated content. And...

Steve Martorano 
Bots were talking to other bots.

Jim Kean  
Yeah. So because now websites that are getting crawled for content on a regular basis to find the trainee set for the general AI is increasingly having a shift of bot-generated content like copywriting is done on product sites...

Steve Martorano 
Right. We can't have the bots talking to the box as being the only source of information, right?

Jim Kean 
Yeah, and a certain percentage that poisons the AI really, results start going off the rails and they start hallucinating more, and like when a computer science friend of mine was fairly pungent turned a voice. He said, Yeah, you had the bots running around eating other bots' shit, you know? Right? So it's a poor diet. And What they concluded is, that this is where it should give you hope that a high-quality AI can only it's like having a tree growing over a pure spring of water. Right? That spring water is human content generated by humans, we still produce the most outlandish, creative, subtle, and deep things. And the bots need that to be healthy bots and to grow and not be poisoned by the AIs. So it really comes down to back to your question is, where are they getting their training sets? Is it wide enough to get rid of bias and statistical, you know, kind of, I guess, fascinate on a particular topic and whatnot? And I don't know where they would get the CBT or anything like that, to train the bot and then have responsive questions. So I kind of, I guess I'm a show me state type of guy on that one.

Steve Martorano 
Do is going Jim Kean People want more information about Molecular You what you did do?

Jim Kean 
Well, we're a Vancouver-based company. I live in Portland, Oregon right now. And I'm in the process of setting up our US operations to come to market here. So we're not commercially available in the United States. But if you want to learn what we're doing or websites great, it's www.molecularyou. That's molecular y-o-u.com. And we will be making our market entry in the next couple of years. And we're available in Canada right now.

Steve Martorano 
Jim, thanks so much. There's so much to learn. And it's often hard to get your head around it, but it is something absolutely worth keeping an eye on artificial intelligence, given how you know, actual intelligence seems to be failing us. So thanks so much for joining us on the Corner. I'd love to have you back real soon.

Jim Kean 
Steve, anytime. This is delightful and I really enjoyed getting in with you as well.

Steve Martorano 
Appreciate it. Thanks, Jim. Hey, the rest of you. Thank you as well. Don't forget, follow us like us. Subscribe to the whole thing. We appreciate it on the Behavioral Corner. See you next time. Bye bye.

Retreat Behavioral Health 
Retreat Behavioral Health has proudly been serving the community for over ten years. Here at Retreat, we believe in the power of connection and quality care. We offer comprehensive, holistic, and compassionate treatment from industry-leading experts. Call 855-802-6600 or visit us at www.retreatbehavioralhealth.com to begin your journey today.

The Behavioral Corner 
That's it for now. Make it a habit to hang out at the Behavioral Corner, and when we're not hanging out, follow us on Facebook, Instagram, and Twitter at the Behavioral Corner.