How US Healthcare Actually Works, the WWII Tax Loophole that Broke it Forever, and How AI Can Fix it | Nikhil Krishnan, Out of Pocket

23:47However, the issue here is that a lot of those entities that represent the aggregate leverage point or whatever, also now go to these companies they're supposed to be negotiating against and say like, hey, if we choose you, we want to cut, basically, because they have all of the power in this situation. And this happens in a bunch of different places. So for example, if you as an individual, sign up for a health insurance plan, you might use a broker to do that, right? And be like, hey, the broker is going to guide you to a given plan, blah, blah, blah. They offer the service for free, but they're getting paid commissions, usually from the health insurance carrier, right?

24:26But now you might be, it's like a little bit of a weird situation, right? Where you're like, well, aren't you supposed to be helping me out? But you are getting your payment from the entity that theoretically either you're supposed to be negotiating against or are getting commissions paid on them. So maybe you only pick the entities that give you the best commission structure. Right. So, so the, the, and again, this is not unique to healthcare. This is any service where it is free for you and the seller is paying for it, right? Like any broker basically across the board in any industry will typically operate like this because people don't want to pay out of their own pocket for those kinds of services but you are paying in other ways you know i mean like you're you're paying in the form of either like for negotiation or not being offered all the options one of which that might be right for you yeah the the i mean the broker thing is

25:16Turner Novak:like a classic even just in finance more broadly like i like i think you've got like or like selling health insurance selling like a mutual fund there's always like i think my favorite and And just like finance, like your financial advisors, like advising you on this, like the different choices of funds. But then there's some where they pay out a percentage just right up front. Like you get paid 8 % of the deal value just right when it closes. And it's like, they're going to recommend you the investment that makes them the most money up front. Like no questions asked. Of course. Or, you know, like getting an apartment or buying a house or whatever, right?

25:50Like there are brokers involved in that and you think you're not paying, but like it's baked into the selling fee, right? Like, so you are paying, you just don't understand. It's like in a convoluted way.

26:01Turner Novak:Yeah. My favorite hack for buying real estate, I've done this twice, is like in residential real estate, I bought two houses and the realtor always gets 30%, sorry, 3%, one from the buyer, one from the seller. I don't have a realtor. I just find a house I want and I say, I want to buy this. You can be my realtor. Just tell me what I need to pay and I will, I want to just win this. and like you'll get double the commission. And it's like, it's like a hack to the system because of the way it's designed and you'll win. Like you'll, you'll have the highest bid because the realtor wants the money. Yeah.

26:38And also, yeah, I mean, there's lots of ways you can, you can game the broker stuff, right? But at the end of the day, I think it represents like a bad transaction mechanism. Oh yeah. For everybody. Yeah. Especially when I think of AI tools, for example, right? Like this is a great use case where it's like usually you're paying a broker to reduce the information asymmetry in the transaction between you and the seller, right? For whatever it is, right? You're buying a house, you're buying health insurance, whatever. You're like, this is complicated. I wish someone would guide me through it. But the reality is like now with AI tools, like it's actually much easier to do that with like a co-pilot actually that may not have those incentive issues that you deal with current brokers that make it like more straightforward to go to a co-pilot.

27:25Turner Novak:Yeah, maybe that's an interesting segue into just generally AI and healthcare. A pretty broad, like loaded topic. I mean, like everything could be AI. What are you kind of seeing right now in terms of what is actually getting adopted and actually is kind of working right now? So basically, most of the stuff that's actually getting adopted touches the payment rails in some capacity. So in healthcare, that's called revenue cycle management as a whole, which is the process by which a provider essentially takes the medical documentation, turns it into a bill that they send to the insurance company to get reimbursed, and the insurance company pays them or argues with them or whatever.

28:06right so that is an incredibly convoluted process it is a lot of transforming text from one place to transforming it into text in another place so as you can imagine like it's an excellent llm use case in a lot of ways but also simultaneously like it doesn't solve the core problem which is that payers and and doctors like don't want to pay each other right like that is just like it's like a Tom and Jerry fight right you gotta you can add all the tech you want to it but they're still fighting right so it's like you have the docs using AI bots to like call the payers and other payers are creating AI bots to like stop the doc bots right like we kind of ended up back in the same place a little bit but you know that's an area you can demonstrate ROI very quickly right it's like hey we'll come in and we you know your your accounts receivable are way lower than you think.

29:03And there's all this uncaptured revenue and codes you should be adding and we'll send it to the payer and you'll get more money and all this kind of stuff. And then you go to the payer and you're like, hey, you're overpaying for the stuff and blah, blah, blah. We'll use bots to prevent overbilling and all this. So that is the area with the clearest adoption currently. And there's a lot of steps in the process of revenue cycle management and so people are tackling different steps so even things like ai scribe which like you're recording a doctor's visit and they have the raw audio data and you can transform that then into the documentation you need or the bills you want all this kind of stuff but all of this is like the way you can get adoption quite quickly is by demonstrating rli on this payment side of things but you know again my personal opinion is like the that process is just very broken right Let's just like payer and provider fighting.

29:58So it's an area I'm like not as interested in personally, but I do think it's the area that it's getting the most in inroads currently. But I do think there's a much bigger opportunity and like actually totally rethinking how we do like care delivery from the ground up. So just like if you were to rethink a doctor's office from the ground up with like AI at its core, what would it look like? It would actually look probably quite different, right? your your the form you fill out before you go into the doc is probably gonna be more dynamic and like risks assesses risk assesses you and only pushes you to an in-person visit if it's serious enough or you know navigates you a telemedicine visit if it's not and all this kind of stuff so that stuff i think is particularly interesting there's a company called doktronic that is like in utah right now that's testing out fully autonomous prescription refills, basically.

30:53So if you are in a relatively low risk medication class, and you're getting a refill, an AI basically is just is just doing it. It's like rewriting the script for you. Like that is an example, I think of like, of like, where we are probably moving towards more clearly of like, hey, you know, can AI actually autonomously start doing some tasks and if the answer is yes there then you can really start rethinking like the workflow from the ground up rather than like every time you need a human in the loop basically you're just going back to like the most inefficient part of whatever the workflow

31:30Turner Novak:is right thinking about like that doctronic and how that interface is like you're talking about like giving some sort of care don't you need licenses and there's probably regulations around that stuff how does that intersect with all this yeah so right now like this is kind of one of the open questions is like where does liability lie right if the system messes up who's liable and all this kind of stuff like one way you can sort of work around this which like i don't have confirmation but i just assume this is how doktronic is doing it is that you can have basically a doc essentially put their license give the ai essentially access to their license essentially, where they absorb the liability.

32:12But then they also get the upside of if things work out, right? Where it's like, you know, if a doc wants to see 10x more patients and AI enables them to do that, they both absorb the liability of something going wrong as well as the upside financially if things go right, right? And that's like...

32:29Turner Novak:That's probably good, yeah. Yeah. And then it's on the doc to basically evaluate a lot of these AI systems to make sure it's doing the things that they wanted to. And that is a little bit of the gap. It's like, okay, can they do a good job doing that? Should it actually be the FDA who is checking to see if like these, are they more like medical devices or are they more like, you know, like nurse practitioners that like work underneath the doc? Actually, that's like not super clear which one it should be. And that is sort of the like open question is like, should this be regulated as if it's a medical device or as if it's a person?

33:01And it's like not, not that clear.

33:03Turner Novak:Interesting. what do you think is the right way to do it because when i think about this and i think about anytime i've gotten anything from the doctor it's literally there's like care staff whether it's like a nurse whatever that does all this like prep work they get it all ready for the doctor the doctor talks to you for like a couple minutes or whatever just makes a decision based off the information that's been collected and presented and then and then the nurse comes back in and does the rest so it's almost like you can speed that process up with more software technically right yeah i mean not only speed it up but also like you can have a more expansive visit right like if i if i can talk to the ai for an hour i can actually get more information down and also i can talk to it through the course of my life rather than only during the visit itself right so you can get very different information doing that right i don't think there's a right answer here like i think one of the benefits of having like 50 states is that a lot of different states are experimenting with different ways of regulating this, right?

34:04It's like some people are like, hey, this is a sandbox, go test it out and just like give it a shot. And then other states are like, no, you can't do that. And we're going to, you know, not, we think that there's too much downside. And I do think it's like, there's good, it's good that there's more experimentation here. And I think where areas where you have more like acute healthcare needs will be more willing to try this out. And so, yeah, I think that, That's probably where it'll end up landing. I do think there will be some shifts in how we think about malpractice insurance or liability insurance for software and stuff like that.

34:43That will probably change. But in terms of how it's regulated, I honestly don't know. And there's a version in which maybe you have to present a certain baseline level of competence through an FDA route. And then at the deployment level, it just lives underneath the practice or the doc or whoever, essentially.

35:05Turner Novak:And one thing I know you've kind of mentioned before is you feel like doctors are kind of like the most screwed in the health care system. So, I mean, you think doctors like there's like one of like the highest status roles in the world. Like that'd be amazing when it like so what's going on where doctors are kind of screwed? I mean, to be clear, Screwed is, you know, in the grand scheme of things, they still have, you know, a very secure job and then like they pays quite well. Yeah, I'm over. I'm clamoring. I'm clip farming here. I'm trying to get some clips. Of course. Yeah. You know, I think that the issue that the docs run into is that.

35:45it it they basically are like on a productivity hamster wheel a little bit where it's like you go work for a hospital the hospital basically controls your time as well as what you're allowed and not allowed to do and as a doc you come in and you basically are put on this hamster wheel like 15 30 minute visits or whatever it is and you just gotta bang this out and you have to hit this productivity metrics basically to get your pay and all this kind of stuff is it so there's

36:11Turner Novak:like quotas in the back end like you gotta visit these many people or like quotas is not the right way to put it it's like we have a system basically that's called rvus and they these are essentially they're essentially like scores put towards how you get reimbursed essentially so this is like and and it factors in like hey how hard is this is the procedure that you're gonna do what is the like liability risk of doing it how much equipment do you need to do it so there's like a scoring methodology essentially that says like hey for this given service this is like how many how how how much productivity it brings to the health care system and we'll reimburse you accordingly and so that is a metric that is actually calculated by the government for medicare payments essentially now it gets adapted for like different use cases and all this kind of stuff But as a doc, like you can get bonus basically based on a lot of these productivity metrics.

37:12And you know, you are sort of expected to do a mix of services that will hit a certain sort of level of productivity, essentially. So it's not like, hey, for every prescription, right, you get like whatever. It's not, it's not, you know, it's not like as maligned as that, you would think. But they are sort of incentivized to do like a mix of things that are sort of more productive for the hospital or whoever they're employed by, essentially. Which puts them in a tough spot, right? Because in a lot of those cases, your salary would maybe be a bonus, essentially. So you're not participating in like the upside if you owned your own practice and you were sort of like grinding this out, putting more hours, and it goes straight into your pocket, right?

38:02And someone else is deciding a lot of these things on your behalf. So, you know, I think it's just tough in that way. And then that's combined with, I think generally, unfortunately, the social status of being a doctor in society has just gone down quite a bit. Like people are just more skeptical generally of like the healthcare apparatus as a whole. and so as a result like patients are more bitter when they're coming in they're also much more complex when they end up in the hospital so the cases are like more complicated less patients have a relationship with their doctor and so as a result like it's not like you're building these like long-term relationships it's like more episodic in nature and like people don't feel the same social you know sort of social connection i think that they do a lot of the medical industry that they used to, frankly.

38:52Turner Novak:Yeah, it reminds me of, like, there's definitely been times where you go to the doctor and they walk in the room and they're like, hey, and, like, they look at the paper, they're like, Turner, and they're like, how's it going? Like, you just, like, they don't know you. Yeah, that's like my parents. They're doctors. They come in, they're like, hey, they kill, right? What are you? Are you our nephew, son? I can't remember. What's your relationship? Which one are you? It's like, I'm an elderly child. I'm the only one. No, I mean, yeah, but yes, like, they're, you know, they're trying to figure it out on the fly to you a little bit.

39:22And it's a bit that, you know,

39:24Turner Novak:the system doesn't encourage these long-term relationships nearly as much anymore. Yeah. So the thing you mentioned was RVU. That's, what does that stand for? Relative value units. Okay. Are there, so what does it ultimately incentivize? Like there are just, are there certain things that are more incentivized versus less incentivized based on, on that calculation? Yeah. Like procedure, procedures are definitely much more incentivized across the board. So this is like a like a surgery type of thing, like a medical intervention type of concept. Surgery is one example or like, you know, infusing a drug, like anything that is sort of a little bit more requires like some sort of like doing something to you rather than just like evaluating how you are.

40:06Basically, there's definitely like more bonus payment bonuses, productivity bonuses for doing that kind of stuff.

40:13Turner Novak:And it's essentially because those things cost more money and like the hospital or the health system is more profitable the more you do that kind of stuff. Yeah, those are just like the higher margin sort of things to do essentially as a whole. So like, you know, if you're a hospital, like primary care doesn't make you any money. And it's like a terrible, you know, it is bad economics. But a lot of times hospitals will buy primary care practices because they will feed into the higher margin things that a hospital wants to do, like procedures, for example. So, you know, again, like that, that is it.

40:46That is the reality of like living in our current healthcare system is there's definitely much more. There's definitely much more reimbursement incentives to towards procedures.

40:55Turner Novak:And is that ultimately where the profitability comes for most health systems is doing procedures of some kind or like, like administering drugs? Yeah, I mean, doing procedures. There's a lot of new ways that hospitals end up making money in strange convoluted schemes that like. Oh, really? yeah i mean this is probably too walkie to get into for like this podcast but like for example there are government programs where hospitals that serve mostly lower resource patients can also get drugs for much cheaper in a program called 340b and then get it reimbursed at like the normal insurance rates so if you acquire the drugs for much cheaper and get it reimbursed at normal rates you can make a huge spread right so there's a lot of like ways to do this there's a lot of different ways hospitals make money now it is not just like doing doing like hey you deliver a service we get paid but there's like a lot of you know basically i would say generally like subsidy schemes especially for hospitals that are serving lower resourced areas that like you can make money in other ways too for sure and by the way, I don't think it's like malicious.

42:08I think it's, it's, it's just one of those things where it's like, we have not quite figured out what are, how we think that payment should be structured for people who want to get care in lower resourced areas. And we're trying to figure out who should foot the bill for that. And there's like a lot of different discussions and arguments about what the right answer is there.

42:28Turner Novak:Well, it's probably also like, if you're just taking two contrast examples like a very a lot of net worth a lot of disposable income where they're just spending a ton of money on these like elective procedures versus like the most lowest income area in the country like if you're picking between those two areas you're going to build a hospital or the health care product for people who have disposable income and you won't build it for the people who can't pay for it so i guess you do sort of need to incentivize there to be care built there and like infrastructure there too so of course you need some program to be like hey don't cherry pick really healthy patients or really like easy patients right but but it's it's a it's a little bit of a tightrope you have to walk where it's like if you set up systems that are too profitable or get too much subsidies for taking care of patients that are sicker or lower resource or all this kind of stuff then people will gain it in very different ways too so you know it's a little of a lose-lose situation where you're like if you if you if you make it too easy if you make too many subsidies to do this it'll just get gamed in a different way is there anything about the the healthcare system that you think is like the most surprising for people who just like don't know how it works maybe it's one of the things we already talked about but is there anything like it's like a dinner party trick when you like drop it and like blows people's mind and be like wow that's that's just weird like that's i didn't i wouldn't have expected that i think for a lot of people, they don't understand how large the healthcare system is here and the sense of employment.

43:59Like it is an implicit jobs program in the US. And I think that's the thing that people struggle with a little bit is like, hey, essentially manufacturing was replaced by healthcare as a jobs program. And it is the number one employer in this country. And it's number one employer in most states. And so if we want to see the productivity gains of tech, AI, whatever, we also have to figure out what is like the next jobs program that people have to go to. Because right now it can't, you know, it has to be something that basically people who don't have advanced degrees can make an income from. And currently that is healthcare, but it is, it's not clear what it should be next.

44:42So, I mean, there's lots of other things, right? Like, I mean, most, most people don't understand this whole employer involvement in U.S. healthcare. and they don't understand it's like a very unique thing to our country and not other countries. But, you know, I think another thing too is maybe this might surprise you. What percent of U.S. healthcare do you think is like pharma spend?

45:04Turner Novak:I mean, I feel like it's either really high or really low. I'm going to say like 62%. It's between like 9 % and 11 % somewhere there. Like it's really, like for example, it's just like much smaller than I think a lot of people think. and and so you know whenever i whenever i talk to people about this like i think that always surprises them because they're just like it it seems like it's much higher because you hear a lot more about like specific really expensive drugs and how high profile they are and a lot of people who can't afford them and it's like it obviously it's like bad and it sucks but actually relative to the rest of what we spend in health care it's actually like much smaller i think i saw a stat once i'm gonna butcher the exact statistic but it was something like 40 of all cable news ad revenue is like drugs basically like pharma it it's a really big number like that which is pretty crazy i feel like i have the sky rizzy the sky rizzy like you know yeah jingle branded in my brain unfortunately yeah that i probably like the funniest thing it's not even really related to health care or something but like you're watching like a basketball game and like these guys like peak physical specimens.

46:13Turner Novak:And then there's an ad for like, it's like Sky Rizzy or like Doritos or like some junk food. Or like PD pills or something like that. You know, it's kind of interesting. Like I thought it was having this debate with somebody recently where I was just like, you know, of all, of all the things that we want people to consume more of generally, right. actually like drug products are probably not as bad on the scale of things versus some other stuff oh really like how it's just like my i mean it's a debate right like i would say that like getting more people into like sports betting and like buying cbg products or whatever is probably more net negative to them relative to a lot of people who are for example underdiagnosed with heart disease or whatever, and probably might actually go see the doctor if it's rammed into their head enough that like, hey, you might have this condition.

47:12So I don't know, like, I'm sort of, I think, you know, direct to consumer advertising for pharma is like bad in a lot of ways. And like, you know, it is easy to figure, to make that case for why we shouldn't have that more. But I do think that there's like a lot of the things that people are upset about with pharma advertising or just upset about in consumption culture in general in the US and relative to other categories that get a lot of bad spend, actually, like, maybe pharma products are not the worst thing in the world. Yeah, and then when you think about, like, the effect of, like,

47:47Turner Novak:one of these drugs is, like, it cures your diabetes or whatever. Like, it makes you better. Just we'll pretend whether that's true or not. And then, like, you look at, like, an alcohol brand, and it's, like, partying, having fun, and you could say like it's actually really bad for you to be drinking alcohol but we get to we let them run these ads where it's like they actually think you're cool if you if you if you do this stuff or you drink this stuff and it actually is really bad for you it's kind of crazy to me that like pharma companies have to put all their side effects things in an ad and then like alcohol companies don't right yeah that would like isn't that sort of like a weird juxtaposition in a lot of ways yeah well it's like yeah the alcohol party seems way more fun because the former party is telling me that i'm gonna like shit myself to death at the end i literally might die like i literally might die from taking this drug exactly well you won't from the alcohol you'll be cool and you'll get to hang out with all the cool people and you know it'll make it make you awesome exactly but so talking about ai stuff that is actually kind of working in ai right now in healthcare are there certain areas that you feel like they're almost like the amount of attention versus actual effectiveness is like a little bit out of whack, like some things are maybe a little overhyped?

49:00Yeah, I mean, again, like, I think a lot of the stuff that's just like speeding up this revenue cycle process is very short term. Like, I just don't think that, yeah, I just don't think like speeding up, you know, hey, we're going to make how you submit a bill to insurance faster is actually going to change that much. Like, I do think that eventually the payers on the other side of this equation will build their own systems to slow this down. Right. Like the friction is sort of a point of the system in some ways. So I'm not like that excited about that kind of stuff. And I do think it's like there's a lot of money that goes to that, but it's not as durable, I think.

49:38And also quite. Quite interchangeable, too, like you can you could swap out the vendor under the hood and not a lot of people care. So I think that gets a lot more attention than what it's like actually probably that useful. So yeah, that area I think is like quite overhyped generally.

50:00Turner Novak:Are there any areas where they sound like they'd be really good startup ideas or like healthcare business opportunities that you just never end up working or usually kind of a trap? There's so many, dude. I want to hear them all. I need to know these because I want to make sure I'm not investing in them. yeah like medical tourism marketplaces like i get that all the time where it's like oh it's all cash pay and like people go abroad and what if we just did and it's like you know dude like all the good hospitals abroad have their own programs already to like shepherd people there they don't need to go to a marketplace they're pretty much full the rest of the places are quite bad health insurance here doesn't cover complications if you get a medical tourism and come back, right?

50:43So medical tourism marketplaces is one area that I'm just like, I just feel like I get it all the time. Another is like, hey, you know, we will analyze your bill. And if we discover issues, we'll take a percent of savings and you don't have to, you know, do anything. That sounds awesome. It sounds great, right? Like, and I actually think you can build a smaller business around this, especially today with like AI tools and stuff like that. But the problem in a lot of these cases is a lot of it is fighting the hospital and them basically just like not playing ball, essentially. So it has nothing to do with like tech and all that kind of stuff.

51:25It's like a lot of times just being like, hey, like prove it. And also you have to catch a patient at a very specific time in their journey, right? Like it's really hard to acquire a customer for something like this. and they'll usually only use you like once basically so just the making the economics of that work are really tough also because it's not recurring in a lot of cases very lumpy revenue you spend a lot of money to acquire a patient and then they also like don't stay with you in a lot of cases so that's like another one that i feel like i hear all the time um there's so many more dude i get like i get like you know pitch a week around like oh we make it easier to recruit patients to clinical trials and all this kind of stuff.

52:08And I'm just like, you know, a lot of these are a lot of these are also fine ideas, but they're just like not defensible at all. Or they're just like not venture businesses, which is totally fine, right? It's just like, hey, this is like a services business, and you should go do it. And actually, I think you can make a pretty good life out of this. But they all want to like raise$10 million. And I'm just like, it is, you're gonna try doing this and try to build a big business and like this is just the wrong capitalization

52:35Turner Novak:structure for this type of business and and it seems like coming back to a lot of this is like the like the profitable acquisition and retention of customers is like this longer term relationship so it's kind of like the importance of like the primary care practice for a health system where it literally doesn't make money aside from being the way you acquire people for these more profitable one-off procedures, essentially. Yeah. And I mean, again, like you could be profitable even in a one-time transaction. It's just people have to keep coming back to you. So again, like the other problem too, is let's say you offer a service that's covered by insurance for one thing, and then you leave your job and you switch to a different insurance and that doesn't cover the thing that you normally use.

53:22Suddenly you just lost them as a customer, right? It's like built-in churn, essentially. So again, it just depends on the service and what it is, but it is just really hard to deliver. Also, just remember that also for most people in the healthcare system, they are not using the healthcare system on a recurring basis. They're using it relatively infrequently. So to try and build a business that gets recurring payments is really hard because they're not using the healthcare system on a recurring cadence. So how do you build a business basically that has more predictable cash flow or like can build a predictable revenue business in that paradigm?

54:00Right. And that's really tough to do.

54:02Turner Novak:yeah i mean it seems like the answer to that is like it's almost like this like stack ranking of like well you should sell to the employer because they will always just pay the individual and then you know maybe if you can't get the employer you sell to like the insurance company or something like that and then it's like i don't know i don't know how this how this ladder kind of goes but but if i'm thinking about this maybe that's where i would go selling to the doctor selling to the patient yeah like every single one of those is different trade-offs right so if you're selling to the employer, for example, most of the employers use benefits consultants or brokers to help figure out the vendors that they're choosing.

54:37Yeah, you're going to go to the brokers. But the brokers want their pound of flesh, and they also want a commission, and they also have their own vendors that they work with normally that they are comparing you against, right? And so now it's like you're selling to the broker because I'm selling it to the employer, and then you actually have to get the employees to use your thing. So now you've got to sell to the employees and so so you can see like it sounds great in theory and then it's like just a different set of trade-offs you're making by going down that distribution and then even like above the brokers

55:06Turner Novak:you might say well each there's like hundreds or thousands of brokers that work with each insurance company like should i just go straight to the insurers like you jump you jump over everything totally you go to the insurers and then they'll be like okay where's your data and then you'll you're they're gonna send you to probably one of their you know employer clients that they work with to test you out and then you're back at square one right so fair and but then there's kind of this whole movement is sort of like the consumerization of health care that's a pretty open-ended topic is that sort of like this evolution of people realizing that maybe there's like this other rung that's just like the individual person that's receiving the care like i just go right to them is that sort of what's been happening yeah i think in general that is where there's been more.

55:55There's just been more activity, right? Like, I think we swung quite far to like more paternalistic healthcare system. And now there's like a swing back to like a more consumer oriented healthcare system. Obviously, I think AI tools play a big role in this, right? You now have tools that give people much more expertise in their pocket to make decisions. It could do a lot of things that like, you know, typically clinical people could only do. So there's definitely much more there. And there's like more, you know, push, for example, to make it easier for patients to get their own data out of health records and put it into one place and give it context data and all this stuff.

56:35So like, the pieces are moving there for sure. and I think that's good I think like people should have more agency in the decisions that they make in the healthcare system I think in general also people should be allowed to take on more risk themselves in more risk in uh uh in in the risk reward trade-off of like services they're able to get right like I should be able to say like hey I want this AI to prescribe me this drug and it might get a wrong some of the time but i in exchange i get a cheaper service like i think that's sort of generally where things are trending towards but i will say that like there's also bad parts of this too right like i think that you could look at maybe like dental services as like an analog world here where you're like hey dental is extremely consumer oriented You know, like you are, it is very schedulable and shoppable.

57:32You can figure out the prices for things and they give you real time checks on your insurance and like how much it's going to cover and all this kind of stuff. But you see all the downsides of that system too, where it's like you're constantly being upsold on things. There's a lot of like, you know, bad actors that are telling you you have cavities that maybe you don't have. Right. So, you know, it's all the pros and cons of a consumerized system. and so the question is just like how far do we want health care to move in that direction and i you know i think the answer is like probably further along the consumer spectrum than we are today but how far along is like still an open question right i mean you probably saw this by the way like with with all these like glp ones for example right like there's all these like kind of shady compounded glp one things or it's like yeah like come get your drunk here like we'll help you lose weight and then there's like a whole spectrum within that of just like okay very consumer oriented maybe a little too consumer oriented yeah just don't ask how we make them or where they come from but don't trust us bro yeah like hey it's super cheap like don't ask why it's super cheap but like come here and get it right like that's probably maybe too far on the other end of the spectrum so it's like where where do we meet in the middle somewhere here basically

58:51Turner Novak:in the dentist thing it kind of reminds me like i don't know you probably like seen like the joke of like each time you go to a new dentist they just tell you you have a ton of cavities and like you got to get them all refilled each every dentist will tell you like you have a million things wrong with your teeth of course exactly and but now it's like hey maybe i can actually just take that same image put in an ai tool and feel like a more informed consumer now when i go to the dentist i feel more empowered and like having that discussion like that's actually probably a net good thing for everybody so with people able to just access this stuff in clod or chat gpt like are they is are they winning just like consumer healthcare ai like does everyone just say like you know you know here's my thing tell me what to do and like and they and they solve everything or what's going to happen well i mean again it's a few things right like one is again liability is sort of not an answered question here so it you need to be able to figure out like if it gives the wrong diagnosis and something messes up who's on the hook for it and i don't i don't think that's an answer to question and i think it'll tilt at least for now towards doctors or the medical establishment absorbing that liability and right now i don't think we have good frameworks for that for just a standalone lom today so i think they'll be limited in what they're able to do because of that right so it'll be hard for them to like prescribe a drug, right?

1:00:17Like, the thing is, like, you will hit a point where you're like, I need to do a next action, and I can't do it with just the LLM, right? That might be get labs, get a drug prescribed. That might be I need like someone to put hands on my body to like answer certain questions or like run certain tests or whatever. And I think, at least today, the LLMs can only get you so far. But if you need when you need the next action, you need sort of to interface with the healthcare system in some capacity. So I do think people that can offer those services have a better end damage there. Now that might change, obviously.

1:00:56And I still think the LLMs are great at providing interpretability for results you are already getting back, right? So a lot of people, for example, put their labs in to like get answers about like anomalous lab results. But in order to generate net new data or net new actions, there's still no like mechanism for the llms alone to do that there there's like a break in

1:01:19Turner Novak:terms of what they can do anytime there's like a real world action necessary whether it's like someone touching you someone sending you something you you receiving some physical thing in the real physical world like that will always probably have to be a separate product probably right and it should actually be under the supervision of a doctor in some capacity right like it whether or not the ai is doing one ordering or whatever is fine but like someone has to be the orchestrator of this in some capacity and also if it needs to be escalated that it goes to someone who is like the right touch point for that right so generally yeah i feel like one interesting area that i've a lot of people are talking about i still don't know what my opinion is honest, but the whole like peptides concept, like what is the, what's happening there for people who are not familiar?

1:02:13Turner Novak:And like, what do you think is going to happen over the next couple of years? Oh God. So, I mean, the high level is like a peptide is effectively an amino acid chain, right? But the way people talk about it today is effectively unregulated drugs. I'll call them right. Like unregulated pharmaceutical products. So, you know, typically when a drug comes to market and goes through a relatively rigorous clinical trial process to prove if it's working or not, that takes a very long time. It is very expensive. And so when it gets approved, it is a very expensive product at the end of this. Right. And also takes a while to get access to it.

1:02:51Turner Novak:But they're basically saying like this is safe to take. Like it's been rigorously proven. That's what the test is. Okay. Yes. so the other version of this which is sort of gaining some traction now is hey we should be able to get access to these compounds faster and take all the risks ourselves and like if it's not safe we absorb that risk if it's not whatever you know doesn't work we absorb that risk and all that kind of stuff but we want to be able to test it ourselves rather than wait for it to get approved by this like long sort of clinical trial process that, you know, we don't believe, we don't believe that we have to wait that long, essentially.

1:03:33And so there's a lot of products that are out there that people swear by and all this kind of stuff that do everything from repair your muscles to, you know, that are acting ozempic to whatever, like make you live 500 years and, you know.

1:03:49Turner Novak:There's a one that lets you live longer? Sleep upside down. No, no, I'm just kidding. I mean, I don't know. The longevity people claim anything. But anyway, there's a whole range, right? I'll be focused. I'm claiming to curate ADHD. It's impossible to know because you can kind of say whatever. That's the problem, I think, with this space a little bit. So that's like the general, like, whatever, the level setting of what peptides are. Are they pretty much all, like, injectables? Like you have to take a needle and inject it into your vein? No, there are some ingestibles, right? There are, you know, I think also introducing unproven products systemically through your body, which is the oral route also produces different risks too, versus maybe a localized injection version of it.

1:04:43So you're making different trade-offs there. Again, like not very well studied. So it's very hard to figure out what those risks are. Thank you. But, you know, again, like, I think this goes back a little bit to what I was talking about before, which is like people feel like they want to be able to take on that risk themselves and make those choices. People want agency in their healthcare decision making. This is one part of it, right? I think that some of these drugs are going to cause really bad side effects. I think that there's now like, you know, there's sort of like trends to try and standardize a little bit of like how these drugs are at least procured and like manufactured and that kind of stuff.

1:05:27But even then sourcing has been really shady a little bit. But yeah, I don't know, man, like I personally wouldn't wouldn't do it. Right. And that's just me. I'm not like casting judgment on people that choose to do it or not. Like if you're in a lot of pain, for example, and there hasn't been alternatives for you to do and then there's someone tells you like, hey, this peptide can cure you of that knee pain that has been relatively incurable for a long time. Like, I understand why people would explore doing that. I do think that people should think a little bit more about the risk reward trade off that they are making for some of these things, especially if they are younger and relatively healthy and all this kind of stuff.

1:06:07But I understand why this trend is happening at a high level.

1:06:12Turner Novak:Yeah, that's a good that's a good way to encapsulate it all. I feel like I'm, I feel like I'm in a similar bow where it's, I think it's good that there's just new things, new options for people that do things like you, and you can make a choice if you want to use this. And I'm kind of someone where I feel like with a lot of this new stuff, there's, there's always side effects to these things that we don't know about, like with cigarettes, like took us like 30 years for the cigarette companies to basically admit that they literally kill you so i'm not saying that this stuff is specifically like cigarettes and they will literally kill you but there may be some analogies that are related to that i feel like oh totally i mean by the way like it's funny to think about the fda as like a mechanism because actually the way we review drugs and food are pretty much like this which is like for food you can bring anything to market as long as it's considered generally regarded as safe and for drugs there's a much higher burden to to bring a drug to market even though it actually affects a much smaller swap of the population and so peptides arguably are actually regulated more like food where it's like hey like this thing is like sort of safe ish like let's just introduce it to the population and see what happens and so you know i don't know like maybe maybe maybe that's like just a different way of thinking about it like you know for what it's worth there is like Like, you know, there's like a philosophical question around like how much evidence do we need to generate about drugs before we can consider putting them into the market, right?

1:07:49Like we test safety, we test efficacy, and then we test those things of larger population sizes as like the three levels of doing this. but yeah maybe there is an argument or like we should just test safety and let other people choose you know when they want to take the drug based on how much data they want to wait on to prove efficacy right so maybe actually after safety we just bring it straight to market and higher risk taking people can take it if they really want to and then we'll see how it goes and then people who are less more risk averse take it later or whatever right now again like we have randomized control trials to prove these things for a reason so that you don't have a self-selection thing like that but some people argue like hey actually like bringing taking so

1:08:32Turner Novak:long to bring the drug to market is doing more harm than that than good and we should just go to this other system i don't know like again oh everything is trade-offs you know yeah well it's kind of like with mcdonald's it's like it won't kill you necessarily like you can eat it as much as you want it's probably not the best for you but like you can eat it every day three times a day and just see how it goes for you. It's like one of those options. It's like we know that that's probably not a good idea, but you can choose to do that if you want. Yeah, exactly. It's up to you, man. You want to eat that Doritos flavor that is a mystery flavor and consume all the things you wanted that?

1:09:13You want a Baja Blast? Salute, dude. That's all you.

1:09:17Turner Novak:That was one of my favorite things in college, like the Taco Bell run at 2 a.m., Baja Blast. It's a drive-thru, so good. Yeah, it's like if you're putting that in your body, you don't need to worry about the peptides. No. So another healthcare related question. I don't really know how related to any of this it is, but I wanted to ask you, do you know why all medical software is just so bad? Like, why does it all kind of suck? Because I feel like that's like a general tech person thing is you go to a hospital and you look at the screen and it's like the ugliest thing you've ever seen in your life.

1:09:50Turner Novak:Like, why is all medical software so bad? You have the MRI machine that's advanced physics and it's running on MS-DOS. And you're just like, how can these two things be in the same room? There's a few things. One is, remember, all enterprise software sucks. So it's not just healthcare. Have you ever used SAP Concur? I've used SAP. It's brutal. It's crazy how it works. Right. Any enterprise software sucks. and healthcare is very concentrated, right? So you have a lot more large enterprises. So I would say at a baseline, like it's just enterprise software and it all looks quite bad, right? The second thing I would say is there's a, you know, healthcare is very rife with like this principal agent problem generally, where it's like the users of tools and software and all this kind of stuff are very rarely the people paying for it, right?

1:10:49So therefore, the people who are users have a certain set of things that they would love from software, but the buyers who are hospital admin or whatever are choosing on very different metrics, right, that they care about. And usability is probably much lower on that priority list. And so as a result, you have generally like, I would say, you know, worse looking and less usable software because it's just not a priority. But again, I also think most of this is actually just like healthcare is a lot of really complicated workflows and all enterprise software kind of sucks. Tim, you have to like account for a billion different workflows.

1:11:31And, you know, my hope is actually that like now with a lot of these new AI tools, like the GUIs and the user interface is going to change a lot for people who can basically use different tools that they can interface with. And then under the hood, you know, you have whatever database or enterprise software or whatever. And, you know, your user interface will just look very different. Right. So that's my hope. But yeah, DVD. We'll see.

1:11:57Turner Novak:Yeah. What have you seen works the best in terms of when you're navigating that sales cycle with a healthcare system? What's the correct way to approach that for maybe somebody listening to this? Whether they do it already or they're thinking about doing it or they're trying to learn more. How does it generally work? I mean, one is don't do it. But two is if you're going to do it, I mean, there's a couple of things. One is the ROI timelines you now have to prove have to be much shorter. Like basically within a year, you have to prove ROI. And there has to be a very clear case on how you're going to do that, right?

1:12:31One of the mistakes I think a lot of people make is they assume hospitals have budget for software, which is in general, like not the case, right? Like the budget for software is very small and it is generally not ROI producing. And so you typically will have to target some service line in cell capacity where there's headcount budget or where there's like financial transactions that are happening, et cetera. so to prove ROI it's much easier to do it through those vectors so I could see the the revenue

1:13:00Turner Novak:management where there's headcount and there's also payment kind of flows where you can jump into that and intersect that in some way yeah exactly like that's like a very easy place to go right like a lot of mistakes people will make is like oh we'll actually increase the productivity per worker for example but like that is very like you know roundabout like most hospitals want to a more direct line of sight into what the ROI is going to look like. And it's much easier to do that if you're like, hey, you don't have to increase your headcount by this much per year if we actually implement this correctly.

1:13:33And then you have to find the department that is your champion. And so every health system is structured very differently in terms of how departments are structured, who owns budget, who makes decisions, how is implementation chosen, and all this kind of stuff. And you have to find someone who's basically going to, like, run you through the many committees it's going to take to get approval to do this. And it should try to be, like, one person who's going to be your champion across all of those things. So those are, like, I mean, high-level tips. And, I mean, there's other ways you can go about it.

1:14:09Like, one thing that I think is quite interesting today is that there are more companies trying to do this, like, bottoms-up adoption motion that is been quite rare in healthcare where they're trying to get employees to just use the tool themselves. So like open evidence is a good example of this where it's like, hey, chat you can keep a doctor. Sounds great. Like, let's just get it in the hands of people and blah, blah, blah. And it's like so used by people that you just have much more leverage when it comes to talking to your hospitals and all this kind of stuff after that. So that I think is a slightly new motion that I've been seeing a little bit more of.

1:14:47but yeah that's that's another another access is open evidence going to hit a wall in terms of just like the system you know squashing them i i know i don't really know a ton about them i just know

1:15:02Turner Novak:one when i had a checkup a couple months ago my doctor was using it and i was like oh i've seen the fundraising announcements about this company like it's cool he's using it but like in terms of bottom-ups adoption like aren't the powers that be going to be like wall like not allowed it's a good question i i really am very curious how this whole thing plays out like there's this triangle that's sort of forming a little bit where it's like there's like the scribe companies right that are recording the visit and turning in documentation and this is interesting because this is like that we're collecting data like we're we're building this like proprietary data that maybe doesn't exist anywhere else and you can then use this to like do work with like we can help you get stuff done yeah like like that is like the data generation side is like hey the visit is being recorded and we're going to generate data about this visit like that is happening but then you need to figure out what next steps are and so you might do a literature review so you might like look up like articles to be like what what do you think is the best thing to do or you know also sometimes co-locally called clinical decision support.

1:16:09Like what is Silva? This would be like an open evidence, for example, like, Hey, you know, based on what this patient has, what do you think is like the right course of action? Right. But then you have the system of record, which is the electronic medical record itself that has all the history and context data about the patient. And so all three of these are in like a dance a little bit, right. Where it's like, They all want to do what the other one does and also need each other to actually come to like a optimal conclusion. And so the question then just becomes like, who has the most leverage in that scenario?

1:16:44And that's kind of the fight that's going to play out, I think, in the next few years. It's just like the EMR companies are going to try and roll out their own scribes and literature review things. And like, they're probably going to be bad consumer products because that's not the muscle that they have built. but they have much more leverage because they are the system of record but then the scribe companies might build different tools and then the literature review companies might build into the scribes or like integrate into the ehrs or you know doctors love using them they want to demand more from you know so there's like a lot of i think it's going to be you know it's a little bit like that office meme where like everyone's like gums point to each other a little bit you know that that i think it's like going to be the sort of interesting fight to play out

1:17:25Turner Novak:What do you think is going to be the most important position? Does the system of record probably win because it's so widely ingrained already and they have political leverage? It's a good question. And again, I think other industries might actually have better. This fight might happen faster there. You even see with Salesforce and a lot of these other companies right where it's like are they gonna are they gonna launch their own tools are they just gonna be like mcp servers that like charge attacks every time you hit them like what is the role gonna look like even outside healthcare right it's not super clear right and there's a lot of big lawsuits happening right now on the in the healthcare land of like people using agents basically to like log in as humans and like extract the data and all this kind of stuff and you know they're like hey it's breaking terms of service and the people are like, your users want this and you're blocking them from doing it.

1:18:25And technically we're using it as if the user. So, and, you know, it's not super clear where this is going to land up. I, you know, I think again, like the bigger issue in my head is more around the consolidation in healthcare where it's like, it's like, of course, like a large hospital is going to give a lot more, is going to not switch their system of record company overnight, right? It's just like not going to happen. So the system of record has a lot more leverage in that scenario. So the bigger question is like, why aren't there more like, quote unquote, startups in healthcare? And I say that more as like private practices or small pharmacies or smaller insurance and whatever it is, right?

1:19:05Like, why isn't there more competition on the small, medium business side of the scale where maybe you might see new operating systems or you might see new systems of record or whatever? And that I think is like a more fundamental question. And I think maybe my hope is actually like building AI native operating systems might make it more attractive for people to start new things, basically, because you're like, hey, I can go work at this. I'm making this up, right? But it's like, I'm a doctor. I don't like working at this hospital. But it's really hard to go independent. But here's this company that is offering like, hey, we'll help you build a private practice and you'll be able to see 10x more patients and get the financial upside.

1:19:47And we build all the agents out for you and blah, blah, blah. Like that looks like a totally different system of record, operating system, all this kind of stuff. And maybe we'll make it more attractive for people to leave and start their own thing. Now, this doesn't solve a lot of the other problems of like how do they get patients or how do you get new customers? Or like if you're a new insurance company, you need risk capital. Like there's stuff like that that is not super solved. But I think if you can make it more attractive for people to start small, medium sized businesses, you will solve the like enterprise software system of record being bad kind of problem.

1:20:24Turner Novak:It almost sounds like Shopify for doctors or something like the most simple way of describing it where you just like enable them to create their own business. Yeah, essentially. And there have been a million people who have tried to start that company with like business in a box, start your own practice thing. And the problem has always just been that it doesn't solve the much harder parts of this, which is getting patients and all this kind of stuff. But again, if patients demands change themselves, where you're like, hey, I'm making this up. And it's like, I really want a doctor who offers is me, an AI agent that can see me really quickly and prescribe me, you know, my medication refills much faster, or I want a doctor who will actually interpret my wearable data and make it a part of the care or whatever, then actually maybe you might solve some of the acquisition problem also.

1:21:19So again, like there's a lot of moving pieces to this. I think physician independence is like one of the most important topics we have to solve in US healthcare today. and there's a lot of different problems and blockers for how that happens but that you know this might be one version of solving it it seems like you'd you'd need to build like a pretty full stack like not just like the system of record but also like the customer

1:21:46Turner Novak:acquisition the i guess like with the shopify analogy the way that got solved for was like you can run Facebook ads and push to the Shopify store. And then you maybe have these, like, a lot of these brands got, you know, do a lot of content online to like build narrative. So you can always need these, like, maybe not exactly the same, but like a social media first doctor where like, they're making content that helps them acquire local patients or something. I feel like you maybe see that in therapy, actually. I feel like in therapy, that's kind of happened where you day like making videos about it that you like get them into the back office and like it's a lot of like telehealth it's a virtual so maybe that's like that just was possible sooner maybe covid pushed it it's very cash pay oriented right that's the other component to this it's like in shopify e-commerce stores you're paying out of your own pocket pocket the other big issue that doctors wanting to start new practices face is that they could bad insurance reimbursement rates out the gate, right?

1:22:50And so it's really hard to make this economics work because you have bad negotiated rates out of the gate. But in more cash pay oriented areas, that's not necessarily the case. So therapy is a good example. All these longevity docs are a great example of this, which is like they're very strong social media presences and have strong opinions on like how they think care should be delivered in their side of things. But it's also very cash pay business, right? And so the question is like, how do you solve this for the insurance piece? because obviously not everyone's gonna pay cash or how do you bring the cash pay rates down so much that people are gonna pay out of pocket instead of using insurance or alongside their insurance.

1:23:27So, you know, there's like a whole movement of stuff called direct primary care now, for example, which is like cash pay primary care, essentially. So you pay a monthly subscription to a primary care physician and you get a certain set of services underneath that basically, whether it's like 24 seven texting or, hey, we get your meds at wholesale prices and we basically just charge you at cost or whatever it is, right? There's different sort of services you can offer underneath that. But the idea is also like you want to bring costs down for people, right? And make it more attractive for them and just be a part of that.

1:24:06Turner Novak:Interesting. Yeah, I feel like it's a good way to think what the physician independence, I think was the way you described it. Just like enabling more people to start their own healthcare-related company providing care to people? Yeah, man. I mean, I really feel for the doctors who are going to practice now. They feel like they don't have a lot of agency in their work. In a lot of cases, they're not practicing the way that they would love to practice, for example. And they come out with a lot of debt, so it feels like they have to be in a system that they didn't really necessarily want to be a part of.

1:24:47It's actually very surprising to me. Again, this is a little bit anecdotal, but it is surprising to me how many people I meet now who are residents. So they've gone through the really gamut of stuff. They're residents, and they don't want to practice full-time. They want to do two days a week or even less. And they're asking me about health tech product jobs. And I'm like, something has gone horribly wrong if people who have gone through all this training and are very, you know, they're highly skilled and they're in their own craft are now coming to, like, health tech because they're so jaded about their, you know, what they're doing.

1:25:28And, like, I think that's, like, a sad state of things.

1:25:30Turner Novak:Is it that the process of becoming a doctor is so hard and long and people just, they're like, I don't want to do this anymore? Or, like, what's causing that? I mean, that might be one part of it for some people, but I really think it's just like they they look at. You know, they look at working at hospitals and all this kind of stuff, and it just doesn't bring them the like level of fulfillment maybe they thought they would have or they want to do more things. Right. So they're like, you know, there's all this cool stuff happening in AI and tech and all this kind of stuff. And they want to like be a part of that.

1:26:06And that's cool. And it's awesome. but maybe there should be like cool awesome things also happening in like traditional medical establishment or working your traditional medical career that makes you excited about that too right um so i don't know like i think it's hard to pinpoint exactly what it is obviously keep in mind that like if people are coming to me it's a very self-selecting group of docs that like are probably particularly jaded or particularly interested in health tech as a whole but i think Part of it is just like, you know, they they they don't want medicine to be the only thing that they do in their life.

1:26:40And it maybe doesn't bring them the level of fulfillment they expected it to.

1:26:43Turner Novak:That's fair. I think it's an interesting transition. I know we want to talk about just like starting a business like content creation. So so for out of pocket, what was like the 30 second kind of run up into how this thing got started? Because you've kind of you spent a ton of time in like health care industry analysis kind of. What was like the original story behind how it all got started? You know, it's fun. So, I mean, I was at a company called CB Insights and we were doing a bunch of content about, you know, different tech things. And I was running the research team around healthcare there.

1:27:15So that's kind of where I started reading a lot of this stuff and started writing. And then I joined a clinical trial startup after that for a little bit. Oh, I forgot about that. And then I left. Yeah, I know. I did. For a hot second, I was an operator. And I, like, wasn't very good at that.

1:27:28Turner Novak:Failed operator. Failed operator. but then I left in Feb 2020 and you know what a time I know seriously I was like the world's my oyster and you know I started writing again because I just wanted to like build that muscle back up the original idea was actually make the daily show for healthcare so I was actually really interested in you know I think Jon Stewart was like an amazing influence in my life earlier on of just like making topics that I consider very boring, very interesting. And I was like, can I do this for healthcare? Turns out video is really, really hard. And like, you know, we filmed some sketches and I was just like, oh my God, I cannot do that.

1:28:12Turner Novak:Didn't you like post some of these at some point? I feel like I've seen stuff over the years. Yeah, yeah. And I ended up posting a few of them because I was just like, we already did, they'd had the footage and I was just like, all right, let me just do it. But like to make a business out of that and to like do it consistently, like I have such deep respect for like, you know video people now because i'm just like that is so hard you know um and i was better at writing anyway so i was like let me just keep writing um and obviously like the sub stack economy was sort of like getting formed around then you probably got an insane boost of like starting in the beginning of covid when everyone was like you know give me that give me this like subscribe to that subscribe to that yeah i mean people were also like locked in their house with nothing to do but read you know and healthcare was like topic du jour so it sort of worked in my favor in a lot of ways um but you know it's been six years since then and so it's evolved quite a bit we you know i would say are now really all in on in-person stuff right we you know we i've been doing the newsletter thing i still do it and we do a bunch of virtual courses and all this kind of stuff but it's very clear that like we are we are you know content on the internet i think is like changing very rapidly in ways in a lot of ways that i don't like and so spamming out in that sea of like terrible ai slop feels like an uphill battle in some ways and so we're all in on in-person stuff now and so we're doing more events we're doing hackathons we're doing these like micro conferences that are really focused on builders who are, you know, sharing across different companies.

1:29:56We run a bunch of healthcare co working spaces now. So like, you know, this is it's evolved quite a bit over the years. But but yeah, no, I do content as like, the bat signal to put out into the universe. I was just like, here's a topic area I think it's interesting let's all coalesce around it and meet in a place yeah and you almost think of it is like it's the top of funnel basically yeah it's the top of funnel it's also like a good self-selection mechanism like i want people who can read to come to our things yeah that's a good

1:30:29Turner Novak:like you know the if you only if you're like a short form new media type you know the brain short circuits when you try to read something you know what's so funny dude it's like there are a lot of people that I think make content and they would never want to make they would never want to meet the people that they'd make the content for you know what I mean where it's like you make content for people that you're you know you can monetize well but you would hate spending time with them and I'm just like I could never do that you know like for me the fun of this is like let me put content out there to find my people even if that group is much smaller you know and so I think of that as like a hey like let's hang out like let's do fun stuff together content is not the like be all end all for me yeah that's a good that's a good way to think about it because i've i've like almost accidentally figured that out like i really have not done that intentionally but like a lot of like most of my content is either like really dumb and just kind of making fun of like this whole thing that we all live in or it's i mean it's really kind of dialed in on this like super serious, I mean, I don't know how serious this conversation was, but these like pretty long, just conversations where it's like, I just try to learn stuff.

1:31:39Turner Novak:And it's like, you know, there's all these different angles of like how you package this stuff up. But I, most of the people that I meet that like listen to the podcast, like, oh, that person's like cooler than I thought they were. Or like that the people who follow you on Twitter, on the internet, from like the dumb jokes you make, they're like actually pretty smart and actually kind of interesting to get to know them. So yeah, that's a good, that's a good point. I had never thought about it like that, but that's a good way to think about it. Yeah. I mean, again, like I think if you plan on meeting people as the course of your actual job, you end up gravitating towards that anyway.

1:32:13So, but there's a lot of people who make content and they like, that is the end form for that. They're like, I make content and I monetize the content. and I don't want to actually meet people who we do this for. And that's a totally separate business. And that's fine. I respect that and all that kind of stuff. But that's not who I am.

1:32:33Turner Novak:And so what is sort of the business model for out-of-pocket, like the different revenue streams? Like you make the newsletter. I think you do some sponsor stuff in the newsletter for the events. Like do you make money off of the events? I think you do some investing too. Like do you have a fund or is it all just kind of angel investing? What does sort of like the whole like system look like? Yeah, yeah. So we do have some content, a lot of it is sponsored in some capacity. Like we have stuff that I write and they have some banner ads and we do some full page sponsorship. We co-develop courses with companies where they're experts in a thing to try and explain stuff to people.

1:33:10And that is like, those are sponsored courses. Events, we have sponsors for it. And we also charge tickets depending on what the event is. so we do we do make money off those and you know i do some investing but it's typically angel investing now investing is like my least favorite of the jobs and hats that i that i that i you know then put on i know i know it's your whole your whole shtick but for me i'm just like man i i don't have the like delayed gratification mechanisms that you must have but you know i i I see so many companies at this point that like, I like vouching for people in a lot of ways.

1:33:47And one way to like share your credibility or like give someone a boost is by putting a little bit of money behind it. And so I view that as more of like a, you know, thing to do to give people some credibility in the space where maybe they don't have it. But yeah, so there's a lot of different things. I mean, we have these co-working spaces and all this kind of stuff. So like a little bit, little bit, it's like a whole, you know, the hodgepodge of side quests underneath a bigger umbrella. But, you know, I run a Slack community too. And, you know, there's some monetization there, but it's like, you know, it's purposely diversified, I would say.

1:34:22So that like, you know, we're not too concentrated on any one stream. But we're all in on event stuff specifically.

1:34:30Turner Novak:So I think one interesting thing, do you do the courses for free? Like, are they mostly free to sign up for? I think I was looking at them or are some of them paid? How does that work? Yeah, so some of them are paid. So for example, I teach a healthcare one-on-one course, and that is people pay directly. Most of the times employers are signing up their employees to get up to speed on different things. And then most of the rest of the courses are free for people to attend. And the sponsor typically is paying us to basically co-develop the course and co-market it for them. and then we do a lot of on-site workshops now with companies to teach them more specific things in different areas.

1:35:09So for example, we do these courses on claims data analysis. So this is a very common data type that people have to work with that they're in healthcare on a data team, et cetera. We will typically go on site and teach people the basics of it. So like 60%, 70 % is like pretty similar where it's like, this is what a claim is. Here's what's in a data analysis, all that. and then we'll customize the last amount for whoever we're working with. You know, we found that just like for a lot of these things, it's just really hard to keep people's attention for long periods of time over Zoom. And so it's easier just to go on site and like treat it as a training basically.

1:35:49And so we've been doing more of those. We've been doing a few AI ones too where showing people like the basics of Claude or ChefGPT or whatever their internal tool is, but then also like interesting ways to use it that maybe they haven't tried. So, for example, connecting it to, you know, an MCP for XYZ vendor and just show them like, hey, you could do more interesting things. A lot of people just like need some handholding to do that. And so we've been going on site to do that.

1:36:14Turner Novak:And it's like a bit of a test that we've been doing. Yeah, because like a lot of people don't understand how big sort of the like L &D budget is. Like paying a vendor to help get your employees smarter. Like that's a good investment in a lot of cases. So it's a little bit of a mix. We found it was like a little bit of a mixed bag. So in the sense of there is L &D money, L &D money at larger companies tends to be earmarked for certain things. So whether it is they already have a platform that houses all the learning and development modules that they need for something, and then you have to be a part of that platform, which means you have to pay take rates and all that kind of stuff for that platform in exchange for distribution.

1:36:59or you have to be with some name brand thing that basically they can point to as like good, right? You have some certification or some whatever that proves like, hey, you're good. You know, it's always a little frustrating for me where it's like we offer courses or like training stuff for like a fraction of the cost of like a lot of these like quote unquote like highly certified places. And like I've seen their content and I think ours is like 10X better, but we're not certified for whatever thing that this enterprise needs to consider us good, right? So it's always, you know, it's a little bit of a frustrating process and I get it.

1:37:33They probably have used a lot of non-certified people and their content sucks. And like now we have to pay the price of doing it, but that's where we are.

1:37:41Turner Novak:It's interesting because when you talk about swinging more to in-person, like just with how the internet has changed, content on the internet has been changing, especially with AI, To what extent have you been enjoying it, using it? How do you think about how AI is kind of changing things lately? Oh, yeah. I mean, at a personal level, I use it all the time. It's extremely useful, for example, of just running business. Even just plugging into all of our stuff so I can get a better sense of what's going on in the business is incredibly useful. right so when we do uh whatever xyz events like we have a whole database of people in the out-of-pocket ecosystem and it's now really easy for me to be like hey who's the best person for this talk that we're doing that i can look through like hey this person worked at this company that's very relevant to what you're talking about and last time you emailed them it was very well so you know you can get all these context things super quickly right and also as a person who has to get up to speed very quickly on different topics within healthcare.

1:38:49I'm lucky that I have like built the like smell test ability to know when it's kind of bullshitting me or not. But it's very useful for getting up to speed on certain things quickly. Like one example is, you know, I read like primary source docs a lot more nowadays. So for example, lawsuits are great ways to like learn about the mechanics of like a specific slice of healthcare, right? But also at the same time, the lawsuits are extremely long and super like, you know, excessive. They have a lot of stuff that I don't really care that much about. So I have a clawed skill that basically will extract just the parts and nose that I like from a lawsuit.

1:39:28So removing all the like legal proceeding stuff, but just the mechanics of like, what is interesting about the suit itself. And so that has made me get up to speed much more quickly on different topics by just having a tool that makes it easier for me to like learn essentially. So on a personal level, I think it's awesome. And like, I use it all the time, but you know, I, I'm more worried about people that use this to essentially like pollute the commons of the internet or replace their own critical thinking with, you know, every time I see someone like at Grok, like, tell me why that should care about this.

1:40:02I'm just like, Oh my God. Like, you know, what are we doing here?

1:40:06Turner Novak:Yeah. Do you feel like that's happening a lot? Like, do you feel like people using AI to create content on the internet is happening more than most people think? Or do you feel like it's really obvious? Like most people realize that it's going on and they have an opinion whether it's good or bad. I think for the people that are on the internet is happening more. But I think there are not that many people who are posting on the internet. We have to remember that. It's like a very small slice of people who actually post. So most people are not. But the other thing too is most people, I actually don't think care if something's written by AI or not.

1:40:41In fact, I think a lot of people prefer it because it's like much more digestible and in the cadence that they like, etc. And actually don't care that it's slump. Actually, they are totally fine with that. I think one of the things that is probably frustrating to people like you and me is like, if you care about the craft and blah, blah, blah, you're like, oh, that's AI slump. Everyone will be able to tell. And it's like, not only can they not tell, a lot of people actually like that. And so your competition is basically, you know, the people's preferences. Which, by the way, it's not that different than like when we used to probably spark note Shakespeare back in the day.

1:41:17Where it's like, listen, I just like, I can't read this. I just like, don't understand. The prose is not there for me. I'm not getting it. I just got to read a simpler version of this. And I'm sure Shakespeare would be rolling around in his grave knowing that, you know, we were doing that. But it like worked for us, right? So is it really that different? I don't know.

1:41:36Turner Novak:I mean, like imagine Shakespeare reading Gen Z slang. Like, you know, like there's one, like a kid with like a broccoli haircut talking about like, you know, clavicular frame, mogging some foids and like, you know, bussing, whatever. Yeah, yeah, yeah, yeah. You know, Macbeth was, you know, whatever, like that court boxing kind of thing. So then how do you think someone should be thinking about just like posting on the internet like making content on the internet like do you do you think like our followers super important like should you be focused more on like creating like a brand or reputation or something like how would you advise somebody to say like if i'm like hey i've never i have no following i want to start like what how should i think about this i i just think like people so underrate reputation like i you know i think that that is the most important thing that matters like smart people really care about finding other smart people.

1:42:31And it's not about rage baiting. It's not about getting eyeballs and not like that. It's like, you know, for me, for example, like people send me really thoughtful emails and no one else can see them. Right. Only I can see them. And they just do it because they're genuinely intellectually interested in the stuff. And I reply back to every single one of them, you know, because it is just like so much more. I care so much more about people that are like they are intellectually curious and blah, blah, blah. And I think that if you're posting it doesn't even need to be posting publicly online it's just like you should be thinking about this and how do I find the people I really want to meet that I respect a lot and what is the best avenue to do that and like maybe it's not you know posting online maybe it's some other thing that you do but like your reputation I think really matters and I would just like host as if your reputation matters and like there's different ways you could do that like you know I think that matters a lot

1:43:25Turner Novak:well it's been a lot of fun thanks for taking the time to do this Yeah, man, I appreciate you having me. You're pretty active on Twitter. Unfortunately. Outofpocket.health. Yeah, outofpocket.health. We're doing some fun stuff in healthcare. So if you're interested in healthcare, regardless of how much previous knowledge you have about healthcare, like we, you know, come through. We try to teach people how all this stuff works. And I think it's fun. I think it's entertaining. And yeah, we're doing some cool stuff. And thank you for listening. Make sure to check out Nikhil and everything he's doing in Outofpocket.

1:43:57Turner Novak:Thanks again to this episode's sponsors, Flex. Upgrade to Flex Elite and get$1 ,000 in the description. Numeral, put your sales tax on autopilot at numeral.com. And Amplitude, for AI analytics, just ask Amplitude. If you enjoyed this conversation, please like, comment, subscribe, and share this episode with a friend who wants to know more about how the U.S. healthcare system actually works. Make sure to check out the back catalog of over 100 episodes with founders of companies like Robinhood, Sweetgreen, and Mercury. Tune in over the next few weeks for guests like Alex Israel and Metropolis, the parking company who's quietly pulled off one of the most successful versions of the AI-enabled roll-up growth buyout strategy that everyone was talking about last year, and Jim Belosik at SendCutSend, who's one of the pioneers helping reshore manufacturing back to the US.

1:44:45Turner Novak:If you want to miss any of these, subscribe to my newsletter, the split link in the description to get each episode plus a transcript emailed directly to your inbox every week. Thanks again for listening. See you next time.