89 – Venture Capital’s BIGGEST Bet on AI | Here’s Why It Could Fail!
Episode Notes
In this episode, Marcus and Vic begin with a reflection on the impact of a hurricane and move into a broad discussion of several topics. They explore the ongoing dock worker strikes, focusing on the role of automation and union demands. The conversation shifts to AI's effect on job displacement and societal change, highlighting the rapid development of AI tools. The hosts then cover inflation concerns linked to the strike, followed by a review of recent developments in healthcare hiring, corporate real estate issues, and venture capital activity in healthcare AI. They close with updates on virtual dementia care, physical therapy practice management software, and venture capital's cautious stance in growth-stage investing.
Stay Connected
Watch this Episode on YouTube
Episode Transcript
[00:00:00] Marcus: If you enjoy this content, please take a moment to rate and review it. Your feedback will greatly impact our ability to reach more people. Thank you. All right, um, I think Just quickly before we dig into everything, um, just an acknowledgment that there are so many, uh, families, um, individuals that are, uh, still, Yeah, from the hurricane.
[00:00:23] Marcus: Wondering where their loved ones are, you know, we're in Nashville, Tennessee, in the southeast, which is where the brunt of the storms, um, took place and, you know, Nashville was spared, but we, Yeah. Both lived through a flood in 2010, and so we know how devastating these things can be. And, you know, unfortunately, I think from a fatality perspective, uh, this is far worse than what we experienced in Nashville in 2010.
[00:00:51] Marcus: And so I just want to start by, uh, sort of, um, just saying that, that our, our thoughts with everyone who is, um, working [00:01:00] to continue to try to, uh, sort of restore power and access and, um, and find loved ones and friends, uh, in, in that region across North Carolina, South Carolina, Georgia, um, and parts of Florida.
[00:01:15] Vic: Yeah, that's right. It's sad and floods happen really quickly and they're just devastating.
[00:01:21] Marcus: Yeah. Yeah. So, uh, nothing more to say. Um, We have a lot to talk about. And so with that, we will dig in
[00:01:38] Marcus: Vic. I got to give it to you, man. Um, three ish months ago, you, you walked in with what you walked in with a set of stories and you had a story about the dock workers. And I was like, why are we covering the dock workers? And, uh, now I know why we were covering the dock workers. Holy cow. Um, it doesn't, [00:02:00] it's not something I'm, I want to brag about being early on this, but it's the thing I love about doing this with you every week is.
[00:02:08] Marcus: We are spending time studying the world and not just healthcare specifically. We're not
[00:02:13] Vic: shocked on Tuesday when they go, when they go on strike, like everyone else in the world. Yeah,
[00:02:17] Marcus: yeah, yeah. So anyway, I will tee it up to you. You've got a series of stories here. We're gonna kind of run through these, hopefully in rapid succession, but a lot has happened in this slowly building story over the last 90, 120 days.
[00:02:30] Vic: Yeah, a lot has happened and also, um, management and the union are not talking. And so there's lots of, um, media and we have a clip of Harold Daggett, who is the leader of the union and he is not going to back down in my opinion. And he did a 17 minute interview last week and he's very clear about his demands and they're going to be.
[00:02:59] Vic: [00:03:00] Difficult for the various companies to agree with no automation. Um, and a much higher raise, but I also, he is credible to me that he's not going to back down. Um, I will put a clip, we'll put, we'll put a link to this, but I think it's worth watching pieces of it just because to hear his language, he is.
[00:03:22] Vic: It's standing up for his, for his members and not willing to have, um, tracking cameras, tracking, you know, things on the trucks, tracking them, or using technology to replace humans. And while I understand that, I think it also is, you know, the, the economy is going to progress. We're going to, we're going to have, you know, tools.
[00:03:46] Vic: And so there's just this tension. It's really the, the tension between the, the workers and ownership or, or shareholders at the end of the day. And it's a hard issue because these companies are not [00:04:00] American companies. So it's, I find it hard, even as a capitalist, I find it hard to not. side with the emotion of the, of the workers.
[00:04:10] Marcus: Yeah. Look, this is, this is the beginning, right? To me, this is the true beginning, right? The first honest conversation coming from organized workers who actually can impact our daily lives about what we all know is coming. Right? Like, like, let's not, you know, I get, I get so sick. Yeah. I'm
[00:04:29] Vic: probably limited to the port.
[00:04:30] Marcus: No, no. Like, like, I get so sick of being in this tech bubble where people talk about, oh yeah, of course there's going to be some job losses, but all these new jobs will be created and there's all this flourishing and blah, blah, blah. And it's like, okay, but what's the reality of that process that you're just sort of, you know,
[00:04:47] Vic: Yeah, you're waving your hand and say it will, there'll be dislocation and we'll work it out.
[00:04:50] Vic: Right. Um, the, and the difference this time is that if you look at the. Industrial Revolution or you look even at the internet. This has [00:05:00] the AI tools have come up much faster. Oh, gosh. And it's going to cause a greater displacement of humans. It just is and they have to be retrained and not even clear what jobs will be available.
[00:05:13] Marcus: Yeah. So normally, uh, we sort of keep our AI to the end, but I think it's going to be throughout this. Yeah. I reviewed the storyline. I mean, there's a lot of stories where AI is, is a, is a focal point.
[00:05:28] Vic: So if you think the strikes will be over soon, watch Harold on YouTube. New York Times did a really good story sort of describing, yes, there's some pay components, um, that, you know, there's a thousand media stories about the strike this week.
[00:05:43] Vic: Right. Yeah. But we'll link to this one in the time. It's really, they're really talking about the automation piece. And my view is that's the sticking point that they, they broke off talks about. They just can't come together. The hourly wages are an issue, but I think management's used to [00:06:00] negotiating that.
[00:06:00] Vic: And they can, they can navigate that. The, the lack of any new automation is, is a really difficult point to negotiate. I
[00:06:07] Marcus: mean, look, just in the few minutes that I watched of, of, uh, Daggett's, um, you know, Candid conversation as it's framed on the ILA YouTube page, you know, the thing that that struck me was it's not just automation I mean he's talking about How technology has progressively been added to the docks.
[00:06:27] Marcus: Yeah, and he talks about the surveillance You know cameras that are everywhere and and you know, basically says, you know, listen You can't even breathe on these things, you know, everywhere we go, we're being watched. Now, of course, it makes sense when you think about security and, and, you know, not, not just protecting against theft, but also like, you know, all sorts of things being placed into boxes and, you know, I mean,
[00:06:49] Vic: fentanyl sometimes is coming through the ports.
[00:06:51] Vic: There's lots
[00:06:52] Marcus: of reasons to have technology for sure. So I think it's less about should there be cameras and more about hearing his [00:07:00] position on it and realizing, no, there's a, there's a much more fundamental cultural disdain for technology, period. This is not isolated to automation. Automation just represents the, You know, the worst form of technology, but they've already been in their view enduring, you know, progressive additions of technology to their day in and day out working experience.
[00:07:25] Marcus: He goes
[00:07:25] Vic: back to the 1960s and like documents the every step along the way how technology was brought in that improved throughput, improved more freight coming in and reduced jobs and put pressure on his members and. He's right. I mean, it's a clearly a scripted. I mean, someone fed him questions, but it was very prepared.
[00:07:48] Vic: It was candid, but but pretty prepared. Yeah. Um, but it's, but there's a lot of truth to it.
[00:07:55] Marcus: Of course, there is a, a current sitting president. And, uh, [00:08:00] while that president is not running for reelection, uh, his vice president is running for election. And so therefore the incumbent, uh, is under pressure because it's happening on their watch
[00:08:10] Vic: when he could.
[00:08:11] Vic: Force the union to go back to work. The president can impose. A 90 day period where their union must go back to work in order to also have fruitful discussions, right? And kind of, I think two things are interesting from the political point of view. One is that the teamsters for the first time, I think, since the teamsters have been involved in politics, which is like 30, 40 years, did not endorse Harris.
[00:08:40] Vic: They didn't endorse either way, either way. They have always backed the Democrat and this year they And so for Biden to go against the unions and order them back to work, I think would worsen the percentage [00:09:00] of votes that Harris got from the, from the union agree area. So there hasn't to do that. But at the same time.
[00:09:08] Vic: We're already seeing shortages in stores of produce and some paper goods. It's going to, I mean, he describes it, the, the automotive industry will be hit sort of in 15 days. He kind of goes through, I mean, Daggett knows exactly how it's going to flow. And then the further thing that makes it difficult is the union has said, fine, we'll go back to work, but we are not going to work very hard.
[00:09:31] Vic: And so I think the Democrats are kind of stuck. You know, they could force them back and they might have to, but it's, it's going to be sort of either way is not good. It's a bit of a lose, lose
[00:09:42] Marcus: situation for that, for sure. Yeah. Um, and so you, you got some additional links in here that we'll share, uh, you know, one from a profile on Daggett in the wall street journal.
[00:09:52] Marcus: We don't need to really go into that one as much, but, uh, some really great links here. So in Florida, um, yeah, this afternoon,
[00:09:58] Vic: this was, so this was, uh, [00:10:00] You know, three hours ago, recording Thursday afternoon, around noon, East Coast time, Governor DeSantis in Florida himself has brought in the Florida National Guard to open up the ports.
[00:10:13] Vic: Now, whether they have the ability as a guardsman to, to turn on this stuff and actually unload the ships, I don't know, but he is attempting. I bet they can. I bet they can figure. I mean, I think they can figure out something that they won't be maybe as efficient, but I think they can figure out something.
[00:10:35] Vic: That's right.
[00:10:35] Marcus: And look, uh, just reading from his post on X here. It is unacceptable for the Biden Harris administration to allow supply chain interruptions to hurt people who are reeling from a Category 4 hurricane. I mean, It's hard to argue with that. That's a pretty good point. That's a pretty good point, right, from his perspective.
[00:10:57] Marcus: Yeah, I mean,
[00:10:58] Vic: he's governor of Florida. [00:11:00] Florida was not as bad as North Carolina, but they had plenty of damage. They have a lot of issues there. That's right.
[00:11:06] Marcus: They don't need this on top of it. Right. Right. So, um, Thanks. So, yeah, I mean, just really crazy scenes in the last, uh, the last few days around the stock worker situation.
[00:11:17] Vic: Yeah, and I think it's going to, um, my belief is that the inflation won't be affected for the next print. So in the September inflation, we, of course, we won't get it for another week or so.
[00:11:30] Marcus: Right, it's delayed.
[00:11:31] Vic: But I think it's going to be too new to really affect that. But I think we are going to see Inflation really not related to the Fed cutting rates, but due to the, due to the dock worker shortage.
[00:11:43] Vic: And so that's going to complicate the Fed's job around the whole thing.
[00:11:47] Marcus: All right, moving into healthcare, healthcare, hiring slows, raising concerns for the U S job market. So I was talking to Emily and she said her and Tom have been watching the BLS data and, and, and apparently they see some wage pressure starting [00:12:00] to sort of creep back in.
[00:12:01] Marcus: And so, um, perhaps they're, they're, uh, Data watches is also being reflected proactively on the slowdown of health care hiring.
[00:12:11] Vic: Yeah, I think there's some debate on, um, how much the hiring is slowing. Certainly this minor health care story. It was a very slow month of health care hiring. Um, but I think we still have a lot of volume and high acuity patients coming in.
[00:12:31] Vic: So I don't know if this is a trend, although it. It was a, you know, a concerning pretty low number of hiring.
[00:12:40] Marcus: Yeah, I mean, volume is a totally different metric, right? And, and there was a deficiency of hiring and then hiring sort of got stabilized. And they sort of got the labor thing back and back in order.
[00:12:53] Marcus: And now we're kind of back to the wage pressure issue, which is maybe more just kind of normal health care challenges, as [00:13:00] opposed to just not having the workers in place at all. I think that's right. And, uh, this article from Modern Healthcare is just sort of flagging that this could also reflect broader issues in the job market since healthcare is really the number one employer in the country.
[00:13:13] Marcus: Yeah. When you stitch it all together. All right. Go into the corporate, uh, real estate storyline. We cover a Chrysler building owner is poised to lose control of the iconic skyscraper. Wow. The Chrysler building.
[00:13:25] Vic: This is a, maybe the most faint, one of the most famous buildings in America. They cannot keep it because of the, you know, the whole real estate, it's, there's not a lot of tenants in it, and they can't keep up with the payments, so they're losing it.
[00:13:44] Marcus: Not much more to say other than keep your eyes on the corporate real estate shoe. Yeah, the real, yeah. I mean, we, you know, this is one of those things where, Uh, I think we need to keep tracking it because there is still the potential for a nasty [00:14:00] fallout in this space. Um, we're not there yet, clearly, but I mean, this is not a good story.
[00:14:05] Marcus: They missed 21 million dollars in ground rent payments. Yeah, right. That's something you're going to easily make up. Uh, okay, so moving into VC, the biggest opportunities for healthcare AI and the hurdles that stand in the way, Bessemer report. So we had Flair that did a report, um, like two weeks ago on, on medium.
[00:14:23] Marcus: And now Bessemer is coming out with a report on healthcare AI.
[00:14:27] Vic: Right. Yeah. And they, um. They did a good job with their report. It kind of goes back maybe 25 years and tracks the, um, the devices and also patents and FDA approvals. And, you know, as you would expect, there was sort of a low, low, but steady amount of AI or really machine learning or technology based that they're sort of with hindsight calling AI now.
[00:14:54] Vic: And then it's gone, not vertical, but exponential growth and since the last three [00:15:00] years. So they are covering in a slightly different way than Flare that I thought was interesting. And we know the best, you know, Bessemer Venture Partners is a good, good group.
[00:15:10] Marcus: So, uh, just more venture firms sort of narrowing the focus in on healthcare and AI, both Flare and Bessemer are good, good healthcare VCs, right?
[00:15:19] Marcus: So not coming from outside of the space. Yeah. Um, and then you, you've got a good look. Yeah, we have a
[00:15:24] Vic: link to their actual model.
[00:15:25] Marcus: Yep. All right. Uh, keep moving. All right. So now we're getting some VC deals. Ripple, R I P P L announces 23 million Series A to expand virtual dementia care.
[00:15:35] Vic: Yeah. So this is on the same, uh, trend that I've been seeing of, um, kind of focused disease state, um, virtual care.
[00:15:44] Vic: We had, we had a, we had a midi, which was a woman's health maybe last week, I think it was last week. So this is dementia care. Dementia care, I think deserves its own. services. It's very complex and hard to manage if you have a parent or loved one [00:16:00] with dementia. And it's a pretty big, pretty big space. So good to see them.
[00:16:04] Vic: Bring in a decent amount of money. They're expanding. They're already in place, but they're expanding geography.
[00:16:10] Marcus: Uh, the round was led by kin ventures and they're based in Seattle and other institutional investors from the round arch ventures, general catalyst GV, which is Google ventures F prime. So solid group of VCs around the table on this deal.
[00:16:23] Marcus: All right. Uh, next up we've got. Okay, this is a council capital deal, so it's based in Nashville, uh, not the company, but the VC firm and StrataPT secures 25 million for practice management software. So practice management always a bit of a tricky business to get into unless you've got a soft market that you can enter.
[00:16:42] Marcus: Uh, this is targeting StrataPT, the, the, you know, physical therapy, um, space and familiar name as the CEO, Paul Singh. So that's a blast from the past of our lives, uh, Vic.
[00:16:54] Vic: Yeah. I mean, he, he was a VC when we were starting. Yep. Um, and then he's [00:17:00] moved into being more of a, uh, operator.
[00:17:02] Marcus: Yeah. I mean, it's co founder of 500 So, I mean, you know, that's
[00:17:06] Vic: a real, he was in a real significant VC seat, and then he decided to be, to be a founder.
[00:17:12] Vic: This is maybe the, Second or third company he's done feels
[00:17:16] Marcus: like the third maybe the second in health care. Mm hmm. So he's been doing a lot of stuff in health care Bootstrapped it up until now and and apparently council capitals taking a minority stake. So
[00:17:27] Vic: yeah, and it's a focused Physician practice management really around physical therapy occupational therapy.
[00:17:33] Vic: They may broaden But that's where they're focused right now.
[00:17:37] Marcus: Yeah. Well, Paul's Paul's good operator and, you know, with the background as a VC knows how to structure a deal. So I'm sure he did. I'm sure he structured this deal quite well for himself. Uh, 25 million at a minority stake tells you something about the, you know, the valuation of the company.
[00:17:51] Marcus: Uh, all right. So CRV does something that. Is rare in venture capital. They gave money back and that is because, uh, [00:18:00] they do not believe that at the growth stages, you, there are sufficient opportunities to find good value at good prices and actually generate a return in terms of the surface area for exits.
[00:18:10] Marcus: So they're returning 275 million that they have not yet invested from their 500 million select fund, um, which is designed to back mature startups. And instead they're focusing on what. Right?
[00:18:24] Vic: Yeah. So a couple of things. Charles River Ventures is, you know, one of the earliest venture firms in Boston and on the Charles River.
[00:18:34] Vic: And the first generation of partners has retired, sailed off into the sunset. So this is CRV, which they can't call themselves Charles River Ventures anymore, I think in that transition, but it's the second group of general partners. Expanded into the growth stage, but it's always been an early stage specialist.
[00:18:56] Vic: Yep. And they're now getting out of that [00:19:00] business.
[00:19:00] Marcus: I mean, look, the idea that they are returning the capital is Um, but also they wouldn't return the capital if they thought they could generate a return on it. Right. Right. And so this is, you know, to me, people are talking about, people talk about venture capital as if it's one thing.
[00:19:18] Marcus: Right. And they say, Oh, you know, venture is getting to be a really hard business, blah, blah, blah, all this other kind of stuff. And it's like, it's not one thing, guys. Yeah. You know, yes. If you've been tracking this space, yes. Growth. VC super hard, really bad, right? Why I kind of laugh at why it's so hard and why it's so bad.
[00:19:38] Marcus: Right. To me, it's bad because it was easier for some of these firms to just raise greater amounts of capital, collect larger fees and wait upstream. For all the hard work to be done by the early stage VCs and getting the companies ready. And then they just, you know, take really strong positions in between early stage and private equity [00:20:00] or IPO.
[00:20:00] Marcus: Right. Um, but it's like, how much work were they really doing? Like not that much work. And so what happened, it became a very crowded space where you're. only real advantage was capital. Yeah. And now there's not as many opportunities because the IPO market is way less active than it was five years ago. Um, and the valuations were really inflated from, you know, from, from, uh, the, the, the zero interest rate era.
[00:20:26] Marcus: So,
[00:20:26] Vic: yeah, I think that's right. I mean, it was over capitalized. And so there were many firms that didn't lead deals and kind of followed around other firms and just would throw money around and then hope for an IPO or a big exit and the multiples have come down, the IPO markets are down. Maybe they're creaking open, but they're not like they used to be right.
[00:20:52] Vic: And so you have to actually add value and it is and take risk and take risk. Yeah, take real risk, [00:21:00] right? And um, many of those firms are not designed for that and charles river is Pretty good. I think they're good early stage investor. They're not giving money back on their early stage they basically are abandoning the growth stage because they Can't get the money to work in the risk return profile they want.
[00:21:21] Marcus: Yeah. And, and they're absolutely right. They're absolutely right about that. PE firm TPG buys majority stake in health information network. Sure. Scripts.
[00:21:28] Vic: I think this is healthy. Sure. Scripts is pretty dominant in the Um, electronic prescription, um, kind of processing space, they're, they're 50 percent owned by Express Scripts, 50 percent by Caremark, which is CVS, and 50 50 partnerships, you know, it's very difficult to get anything done.
[00:21:53] Vic: And so I think this is good. It's gonna, it's gonna make it a more normal company that [00:22:00] CVS TPG will drive, uh, which is going to be healthy.
[00:22:03] Marcus: Yep. Agree that this ownership structure is really funny. Uh, uh, half owned by the National Community of Pharmacy Association and the National Association of Chain Drug Stores and half owned by Express Scripts and CBS Care.
[00:22:17] Marcus: I mean, that's just designed to be dysfunctional. Yeah, that's just dysfunctional. Right. So, um, I guess maybe they, they thought that ownership model would create Switzerland,
[00:22:25] Vic: Well, I think they all contributed their Electronic processing thing to make this dominant position sure and then they had to let it sit for a couple of years before they are ready to sell it to someone else, something like that.
[00:22:40] Marcus: I, I mean, I, I suppose, but I think there's also, there's the reality that it's, there's the pharmacies and the drugstores on one side, and then the PBMs on another side, and they are largely aligned, but you know. Not always a line. Yeah. So, um, yeah, having an independent financial actor as a majority owner of the company is probably the [00:23:00] right thing for this, for this business.
[00:23:02] Marcus: Gosh, UPMC consistently in the news and not always favorable stuff. So, I mean, I just want to say, obviously, this is a large, important health system. You know, when you have sort of a string of stories like this, you sort of hope, okay, well, hopefully the, the bad stuff is getting worked through and, you know, better days are on the other side for this organization, maybe next year, you know, because, uh, the healthcare workers having a non compete lawsuit and the feds backing them on this.
[00:23:28] Marcus: I mean, again, just when we're talking about labor, um. It's just a labor friendly environment right now, um, and labor is kind of consistently winning. Now, this is not necessarily about wages. This is, you know, about non competes, but I think generally speaking, non competes are not that friendly, um, politically.
[00:23:47] Marcus: Um, people generally, generally don't, don't appreciate them. Business owners tend to think about them as, as being necessary, uh, to protect the value that you're putting into your employees. Um, but the general public is not business owners. And so this is [00:24:00] not a favorable thing politically. And, and, uh, You know, you just don't want to constantly be in the news for sort of negative things, whether it be missing budget.
[00:24:09] Marcus: I mean, they've had some big operating losses and now having this, this situation where the feds are, um, you know, back in their workers.
[00:24:16] Vic: Yeah. And they, um, have been accused and, and the, The feds are supporting it of kind of blacklisting doctors where like you're not allowed to be hired back and they really aggressive and then forcing nurses to repay training costs and, you know, almost not allowing people to move for a better career opportunity.
[00:24:40] Vic: Um, In a way, that's pretty, seems pretty egregious. I don't know the details of it. So,
[00:24:46] Marcus: yeah, so I mean, not much more to say there other than like, hopefully we're getting to the end of the bad stories with, with UPMC. Um, so there was a vice presidential debate, uh, this week and I gotta say, I, I watched a, about a half [00:25:00] hour of it.
[00:25:00] Marcus: Um, you know, whatever. I mean, it was fine. Uh, I think it was nice to see them be relatively cordial towards each other. So that was sort of, they actually
[00:25:13] Vic: had a pretty good discussion about various issues without name calling or mudslinging too much. That's right. That's right.
[00:25:21] Marcus: Um, I was on a group chat and There seem to be very different opinions on sort of, you know, who was winning or who this was good for.
[00:25:30] Marcus: I
[00:25:30] Vic: don't know that vice president debates really affect the election that much. I mean, incrementally maybe.
[00:25:38] Marcus: I'll give you my view on this one. I think you're generally right. Um, I think in this case, uh, Vance was really the, um, the lightning rod for the whole positioning, uh, that came from Tim Walz of these guys being weird, right?
[00:25:58] Marcus: Like, you can say Donald Trump is weird, but I mean, [00:26:00] that doesn't stick very well. But the J. D. Vance is weird thing, that was sticking pretty well, right? And I think My big takeaway from this, because I didn't think there was that much consequential in terms of what was being said. Uh, my big takeaway was that, uh, Vance is a polished, you know, Yale, Yale law school.
[00:26:20] Marcus: He was, he's polished in that venue and he did a good job of kind of offsetting that weird moniker. Right. I mean, I think, I think you can go back to other clips that you have of him online and different settings and things like that. And some of his positions on things, certainly, but. You know, in primetime on national broadcast television for more than an hour.
[00:26:42] Marcus: He stood up there and he didn't sound very weird, right?
[00:26:45] Vic: Yeah. He did a good job sticking to, uh, the talking points that will pull well, more
[00:26:52] Marcus: discipline than Trump.
[00:26:53] Vic: Yeah. Much more. So he came across as he's able to be an attack dog on a [00:27:00] podcast or a Trump rally and he can certainly play that role. I've seen him play that role, but he clearly.
[00:27:07] Vic: that he can play the middle, middle of the road, uh, sound like a normal human being too.
[00:27:14] Marcus: Yeah. And
[00:27:14] Vic: that's a
[00:27:15] Marcus: skill. Yeah. He can do the shapeshifter thing pretty good. I mean, I think. His weak point appears to be actually being in every man, like actually being in the general public. It seems to not do that well in those settings.
[00:27:30] Marcus: Um, but yeah, look on the debate stage. I, I, I thought he presented well, regardless of what the words that he was actually saying that that's kind of irrelevant. Um, it was how did he come off?
[00:27:43] Vic: Yeah, right. And then Waltz, I think, uh, He just is not as good as Kamala. So it was a really interesting thing.
[00:27:50] Vic: That's right. J. D. Vance was, I think, much better. Than Trump. Than Trump. I think, I mean, I think that he might have a [00:28:00] better chance of actually winning the presidency than Trump, but Trump has all this momentum and he's got a bunch of people that, you know, are always. The Trump brand is, I mean, I
[00:28:08] Marcus: disagree with that, but that's just because the Trump brand is a once in a generation brand.
[00:28:12] Marcus: It's ridiculous.
[00:28:13] Vic: But, but, uh, Waltz is clearly not as good as Kamala. Yeah. And so you could see the difference. It was not, um, it wasn't really a fair fight.
[00:28:21] Marcus: Yeah. So we've got a link here to a story of fierce health. We're talking about the key healthcare takeaways from
[00:28:26] Vic: healthcare point of view. I don't think it matters.
[00:28:27] Vic: I think
[00:28:28] Marcus: it was benign. Uh, I mean, I don't think anything I'm
[00:28:31] Vic: ready to get the election over with, but I think we will. We will navigate whether it is Republican or Democrat. I don't think there'll be a sweep where the, it's going to be divided government, I believe. See, it seems like it's lining up that way.
[00:28:46] Vic: I think the Senate's going to be Republican.
[00:28:49] Marcus: But it also looks like the Republicans are going to lose the House. Yes, that's what I think, yeah. Um, okay, so Gavin Newsom had a flood of AI bills coming into his office that he had to decide if he was going to [00:29:00] sign or if he was going to veto. Right. And he vetoed the big one, the big overarching one.
[00:29:04] Marcus: The
[00:29:05] Vic: big safety one was pretty nebulous and hard to define. Yeah. And he vetoed it, which he should have.
[00:29:11] Marcus: Yeah, over the top. But he did sign a couple of AI bills that were focused in healthcare.
[00:29:17] Vic: Yeah. Yeah, there were two bills, I think he signed 17 bills related to AI. Oh my gosh. Um, in the last, in September.
[00:29:26] Vic: And two of them directly are healthcare AI bills. One is if you use generative AI in emails or in chat messages or voice, you have to clearly disclose and I think, uh, there's even requirements for multiple times to disclose so that the, the patient. that it's not a human, which seems reasonable. And then there's a, um, Like, you can't use it in, um, [00:30:00] fighting claims.
[00:30:02] Vic: I think that was the next one. So there was, um, the automated tools for like prior off and
[00:30:11] Marcus: well, well, it's not that you can't use it. It's, there's all sorts of guardrails around avoiding how it's used. You need to avoid bias. You know, you cannot make determinations based on age, gender, race, disability, things like that.
[00:30:24] Marcus: So, you know, starting to put some guardrails around how deterministic the AI can actually be in terms of the, you know, what are the factors.
[00:30:32] Vic: Yeah. Some of those particularly age you might be important in health care.
[00:30:39] Marcus: Yeah, but there's I mean They're not saying you can't use that in your assessment of a claim.
[00:30:43] Marcus: They're saying you can't use AI. Yeah in an AI model Boy, Humana, what a rollercoaster they're going through first q4 of last year. Okay, everything's gonna be bad this year You know, we're gonna downrate we need we need to rewrite our projected budget, etc, [00:31:00] etc then the The FTC, uh, sorry, uh, in, in court, the whole CMS star thing gets, gets thrown out.
[00:31:07] Marcus: So they back up, you know, they get this allocation of capital and now their star ratings for just them, not, not across the board, just Humana have been dinged, uh, and their stock slid 10%, um, as they warned their 2026 payments can be impaired.
[00:31:23] Vic: Yeah, so they're gonna lose, um, significant amount of bonus payments.
[00:31:28] Vic: I think they lost, on average, an entire star. They went from like four and a half to three and a half. Um, so it, it is very negative to the financial model. I mean, you need those star ratings to be. So I think there may be other payers that also have similar reductions. Sure. My view is that him and, um, kind of comes out with this earlier and that there might be other ones as well, [00:32:00] but, but they.
[00:32:01] Vic: They took a big hit this week.
[00:32:03] Marcus: Boy, crazy scenes happening at CVS. So we've got Activist Investor pushing for big changes. We've got discussions via disclosure around the thoughts that CVS may break the business up. Um, And how long has CVS, Aetna been that merger? It's less than 10 years, right? Oh, yeah, yeah.
[00:32:26] Marcus: It's five, five, five years. Yeah. Yeah. So already discussions around potentially breaking it up. Um, and the business is there just to, to kind of do the bullet points. Um, CVS, which is the retail, Brand, um, Caremark, the, the PBM, uh, Aetna, the insurance business, um, and then do they have another big business here?
[00:32:51] Marcus: Well, they have,
[00:32:51] Vic: uh, they bought, uh,
[00:32:53] Marcus: One
[00:32:53] Vic: Medical? They bought Advanced Primary Care. Oh,
[00:32:56] Marcus: yeah, yeah, yeah, yeah. Which has
[00:32:58] Vic: not performed well. [00:33:00] Hold on, no, no, no. One Medical was Amazon, right? Okay, it's one, um, I can't remember which one it was. They bought, they bought one, maybe it was Oak Street. Oak Street, Oak Street Health, yeah.
[00:33:11] Vic: It was a lot of money. Over a billion dollars and it has not performed well. Yeah,
[00:33:16] Marcus: and any other large business units that they have over there? Yeah.
[00:33:21] Vic: No, I think those, those, I mean, they do manufacture some products, some healthspan products, stuff like that, but that's part of the retail. I put that in part of the retail.
[00:33:29] Vic: Okay. Okay. Yeah. Um, and when you think about those businesses, they're quite different, right? Yeah. No question. There's a lot of argument that those, the business, the strategies to grow and maximize the shareholder value of those businesses are not the same.
[00:33:47] Marcus: Well, I, I, I think it's, it's, um, you know, when they put it all together.
[00:33:54] Marcus: At a macro 50, 000 foot level, it sounds great, you know, but healthcare is a [00:34:00] really intricate, difficult business and you got to like get all the knobs and dials right in each business unit, right? You know, like why does HCA perform so well? Because it's a hyper focused business. Yeah. You know what I mean?
[00:34:15] Marcus: It's not a confused business. It knows exactly who it is and what it is and what it's doing, right?
[00:34:19] Vic: Yeah. Yeah. And, and Unite Health Group has. All of these, well, they don't have the retail, they don't have retail. They don't have the retail. But they, uh, are really good at integrating and data sharing between the different groups.
[00:34:32] Marcus: But importantly, they do not have the retail footprint. And that's a whole nother thing, right? You know, CVS is not just trying to deliver care. They're trying to sell bags of chips. Right, right. You know what I mean? Like, it's a, it's a pretty different business.
[00:34:48] Vic: Yeah. Yeah. Both CVS and Walgreens. I think the physical stores that were really valuable for a long time now are, are anchor.
[00:34:59] Vic: And [00:35:00] Albatross. Yeah. Yeah.
[00:35:01] Marcus: Yeah. For sure. Uh, so, you know, we'll, we'll just continue to watch this, but, um, lots of, lots of chatter going on around, around what the future of CBS, CBS is. Uh, okay, so we've been tracking what's happening in the GLP space, um, and one storyline was that because there was a perceived shortage, the FDA said it's allowable for compounders to enter the market to make sure that they fill the gap of those shortages.
[00:35:26] Vic: There is a lot of debate. I was talking to Emily about this. The shortage is no longer true. We don't have a shortage. Yep. And she thinks that there's going to be ways that a A well financed group like HIMSS will continue to say, well, there's shortages of some other component that they're using in their compounding.
[00:35:50] Vic: But there
[00:35:50] Marcus: is a progression here, right? Because they didn't do an enforcement action against the compounders, but they did do an alert talking about the safety issues. Because there [00:36:00] was a, there was a rise in people going to the ER. Yes. Because they were ODing on these compounds. Yes. Which they're not doing for ZepBound and Monjara, right?
[00:36:10] Marcus: Yeah, I haven't, I haven't seen that. Haven't seen that. Right. We haven't seen that.
[00:36:14] Vic: And
[00:36:14] Marcus: the FDA a, has not issued an alert about
[00:36:16] Vic: that, right? Yes, that's right. The warning was only for the compounding.
[00:36:18] Marcus: Yeah. So I guess what I'm just saying is they are clearly sort of moving in this direction, right? Where Yes, they're, they're, they're trying to build a case that the compounders are outside of the scope of the FDA.
[00:36:32] Marcus: Right? The, you know, they already say the, it's not FDA approved.
[00:36:35] Vic: Yes. The, the compounding is. Um, not FDA approved. I don't think it's ever FDA approved. Right. And it is allowed for, um, your patients. Yeah. if you need to compound something and you can't get manufactured drugs, I think for a public health reason.
[00:36:58] Vic: And so they are [00:37:00] saying there's been, there's an alert that some of these compounding services are not safe. And then now they're coming out saying there is no shortage, which I think you are right, is sort of laying the groundwork for future action.
[00:37:15] Marcus: I think they're building a case because I actually think there's quite a bit of risk to the FDA if they don't get out in front of this.
[00:37:20] Marcus: If you know, okay, you let a well funded outfit like him play. That just creates precedent for a bunch of not well funded outfits to say, why can they play in and I not play? And, you know, you could potentially have a lot of dangerous stuff happening out here in the streets. And the FDA, you know, rightly or wrongly is going to get tagged of what are you guys doing here?
[00:37:42] Marcus: You're supposed to be, you're supposed to ensure. These things are safe. So my sense is they're building towards actually doing some type of true, you know, action that resembles an injunction. So I, I, I think I, I think, yeah, I think they should, they're not jumping towards that, right. But they're, [00:38:00] they're building a case, and I think they're also trying to like signal like, Hey, listen, think twice about this because Right.
[00:38:06] Marcus: We're, we're letting, you know, we're not really in favor of this. Right. There's no other way to read, read. Yeah. And then I think there's
[00:38:15] Vic: our economy with, uh, you know, drug discovery and patent law and the FDA approval. It doesn't work if you just can go compound. That's right. So
[00:38:26] Marcus: that's right. Um, okay.
[00:38:28] Marcus: Here's a piece in the New York Times about how fentanyl drove a tsunami of death in America. We've talked about the street drug challenges in America. Um, we've been talking about him for a while. This is, uh, a great
[00:38:40] Vic: visual
[00:38:41] Marcus: article, right?
[00:38:42] Vic: So it really needs to be watched on YouTube. But for the people that aren't watching, uh, the, the scary thing about fentanyl is that it, you can die very quickly.
[00:38:51] Vic: Yes. And it's unlike other, um, Heroin or Oxycontin or something else where it's much more of a slow, [00:39:00] um, you certainly. can die of it, but it's much more over time. Right. Fentanyl, it's difficult to understand what you have in your hand, and so it leads to many more mistaken overdoses. Right. Um, what I thought was really interesting, and we go down to the next graphic, but they have a, uh, New York Times does a good job with the graphics.
[00:39:22] Vic: They, 2000, there were no deaths or very little deaths. from Fentanyl, of course, didn't exist yet. Um, and then you see it start to come in around 2010 ish, maybe a little later, 2012. Um, and surprisingly, it comes in through the northeast, through Maine, um, I think across the border of Maine. Uh, and then starts sort of progressing towards the west, and then later, it comes up through Mexico to Arizona and the southwest.
[00:39:52] Vic: And so, there's all of this noise around fentanyl coming in through the southern border, which I think certainly is true today. True [00:40:00] today. But it, It also can through Canada. Yeah. And wherever it comes from, we need to try to stop it.
[00:40:08] Marcus: Yep. So if you're interested in learning more about sort of the progression over the last two decades of fentanyl and, and sort of how it comes in, this is a great article to review.
[00:40:18] Marcus: Um, all right. Our AI rundown, uh, AI will drive broad deflation. Silicon Valley pioneer Vinod Khosla says Vinod agrees with you. We've, we've, you and I have had conversations about deflation and I, and I've, I've said that, um, You know, in the case where AI and automation or whatever is used to lower the cost of production in a, in a market where, um, availability is still limited.
[00:40:49] Marcus: I think that availability will trump the lowering cost. It's just going to increase margins. So we were talking about houses in this instance, and I was like, okay, if they lower the cost of the The production of a [00:41:00] house, I don't think that's necessary. You know, if the, if the output is a very high quality house in a great neighborhood, I don't think that's going to decrease the price of the house.
[00:41:08] Marcus: That was, that was sort of my, my, my position there. But what he's talking about is the price of expertise.
[00:41:16] Vic: And the services economy and everything related to that. And I can already see it. I mean, like the, uh, the quality that I got out of AI, I need fewer. Interns and if you're a consultant, and I think if you are a consultant, you can do four times as much work.
[00:41:38] Marcus: I mean, right now. We're still in the era of let's let's speak. I think realistically. Yeah, we'll still we're still in the era of JNI being okay, but not that great. Yeah, right. It's it's it's like, it's not nothing. It's certainly a usable work tool. It's not. It's not like over the top great. Like you can't just [00:42:00] put in a prompt and get like an unbelievable report that you can write to the public.
[00:42:03] Marcus: That's not, that's not real today. Right. Um, also we're in the pre agent world. I know there's a lot of people building agents, talking about agents, but like no one, no one's deploying agents at scale that we can sort of point to yet. That's all building
[00:42:16] Vic: the first, um, GPTO reasoning thing that could lead to agents is barely out in beta.
[00:42:28] Vic: Right. I don't even have access to it. And you have to be on the inside with open AI or pay them a bunch of money to get it. Um, but it's coming, I think in the next year, the next two years where you can have, like in our business, you could have an IP agent that helps us understand intellectual property strategies.
[00:42:49] Vic: And is this company have a good chance to get your patient patented? Are there other licenses that we should be aware of? That's just one example. So in my head of. An agent that I could have in my computer, [00:43:00] then instead of hiring an IP law firm to go and look at this stuff, it would just look at every deal and give me a report.
[00:43:08] Vic: That's right. Now that's not here. You're right. That's not here yet, but that's what Vinod's talking about. That the amount of these expertise, consulting services, tools that you could use for the marginal cost of compute time is pretty significant. Gonna be significant.
[00:43:28] Marcus: I think I think that's right. Yeah, I think that's right.
[00:43:30] Marcus: And you know, of course They say a new age of abundance can begin but it's like he's positioning it that way. Yeah, it's going to be The
[00:43:39] Vic: disruptive devastating before this abundance devastating. Yeah, it's gonna be devastating.
[00:43:44] Marcus: Open AI nearly doubles valuation to 157 billion in funding round. Apple ended up bailing, I think, based on the valuation.
[00:43:52] Marcus: Um, and, uh, they raised 6. 6 billion.
[00:43:55] Vic: Yeah. And I was not asked to invest, but [00:44:00] I do not think this valuation is a good point to invest.
[00:44:05] Marcus: Uh, I'm going to make sure we include it. I listened to a fantastic podcast that gave me a lot of good, uh, scaffolding points, I would say for evaluating AI over the next 45 to 60 days, because that's about as long as you can sort of hold on to any scaffolding in this space.
[00:44:23] Marcus: Um, but 20 VC, Harry Strebling, uh, I, I like some of his episodes. Some of them like drive me nuts. Uh, but I haven't heard it yet, but, uh, would you invest
[00:44:33] Vic: in this, in this round?
[00:44:34] Marcus: Uh, I probably, that's a really good question, um, 157 billion valuation and 6 billion in the round. You know, so let me, let me tell you some numbers that, that actually made me think maybe I would invest in this.
[00:44:54] Marcus: Um, so OpenAI right now is burning 5 billion [00:45:00] a year, but they are going to generate 3. 5 billion this year. And as we've discussed, they are in healthcare. In fact, we, Oh yeah. In fact, we know that they are really more in health care today. They're getting more in health care than people have any idea. Um, and you know, like burning, if you were to swap out the B with an M, right?
[00:45:24] Marcus: Okay. I'm generating 3. 5 million. But I'm burning 5 million, but what I'm introducing into the market is not like incremental, it's, it's fully transformational and disruptive. I'd probably look at that and say, that's pretty good proof point right now. You know, um, now evaluation of 175. Uh, 57 billion. So I mean, what is that?
[00:45:51] Marcus: That's basically 50 x on their current revenue.
[00:45:55] Vic: Mm-Hmm. .
[00:45:56] Marcus: I think that's probably a decent valuation. I think [00:46:00] I, I, I think so. I, I think a 50 x on their current revenue's probably a good, that's probably a good, that's, yeah. Yeah. I, I take that bet.
[00:46:08] Vic: Yeah. So we, we disagree. I, I haven't listened to the podcast.
[00:46:12] Vic: Which I will, but my view is that OpenAI has the best product in the market, hands down right now. Yep. But I don't think they'll be able to protect it in a three year view. There's too many open source Models. Now they may, I think Sam Altman's point would be, we have more users, we have more training data, we have more ability to grow and we'll keep our lead.
[00:46:41] Vic: I just don't agree with that.
[00:46:42] Marcus: I, I, I guess what I think is that. Making the step change advancements in AI, um, is going to, is going to bifurcate from what open source models will provide, [00:47:00] like they're going to provide two totally different things. Like, I think that what OpenAI is creating is going to be the equivalent of like differentiated hardware.
[00:47:09] Marcus: You know, that, that is, that's truly unique in, in what its capabilities are. And it's something that you embed in the applications that need it at the highest level. There'll be a whole bunch of applications that don't need it. Yeah. And those applications will utilize whatever they'll utilize, you know, llama, because llama is certainly good enough.
[00:47:27] Marcus: Um, but I think we're open AI is going, you know, do I think basically here's the bet here, here's the, here's the bet. Do I think open AI can become a 1. 5 trillion company? Yeah. That gives
[00:47:40] Vic: you a 10x from here.
[00:47:41] Marcus: That's a 10x. Right. Right. I absolutely think OpenAI can become a 1. 5 trillion company. Um, I didn't get to finish what I was saying about the podcast, which is Brett Taylor, who is on the board of OpenAI, who was the co CEO of Salesforce, who was the CEO and founder of Quip, which [00:48:00] Salesforce purchased for almost a billion dollars.
[00:48:01] Marcus: Yeah. Who was the CTO at Facebook and is currently starting a new business in the U. S. That's all about agents has about an hour long podcast with Harry from 20 VC and I think Brett Taylor is phenomenal. I haven't
[00:48:16] Vic: heard this podcast, but he's easily one of the smartest
[00:48:18] Marcus: people in Silicon Valley and his very pragmatic operator.
[00:48:25] Marcus: Mindset, um, points on this from a person who's on the board at OpenAI, I think are really important to listen to. So this is like an episode I'm trying, because a lot of people just don't know Brett Taylor. So, you know, they may not know the guys on the board of OpenAI. Like, this is the guy who Benioff pulled up a chair next to him and said, help me run Salesforce.
[00:48:45] Marcus: I mean, serious, serious player and was an engineer too, like before he became an operator. So anyway, um. Yeah.
[00:48:54] Vic: Yeah. So we'll, we'll link to that podcast and this, uh, next story is [00:49:00] why I have my, that's why I wouldn't invest. So NVIDIA just dropped their open source large language model and it is. Competitive with, with open AI model.
[00:49:13] Vic: It has the, um, vision and voice, all the multimodal, uh, it doesn't have reasoning. So that's where open AI is ahead, but I believe that reasoning is an architecture thing that open source. Projects will co opt now. Maybe they won't do it as well But I just think they're competing against the open community in a way That's gonna be hard to stay ahead of I I mean look that could be that might be wrong
[00:49:42] Marcus: I mean look here.
[00:49:43] Marcus: Here's here's an example. I am constantly shocked at how resilient and high performing Oracle has been yeah, like Oracle has no business being resilient and [00:50:00] high performing as a software company with all the open source software that's out there, right? And yet they are, right? I think that's what I'm talking about.
[00:50:10] Marcus: Like, the enterprise is going to want to buy open AI, full stop. They're not going to want llama. They're just not. But I do think it's cool. Well, I think it's interesting that NVIDIA is also rolling out their own AI models. I mean, NVIDIA
[00:50:30] Vic: views themselves, I think, rightly, as competing against OpenAI and Google while they're also selling them.
[00:50:38] Vic: That's right. I mean, they're going full stack.
[00:50:40] Marcus: They're going full stack, and I think it's going to be interesting to see whether or not NVIDIA really, um, Invest in the cloud in a true AI cloud offering, right? Um, because you got to think that that's the long term opportunity for them as opposed to just like constantly selling chips, selling chips to [00:51:00] other cloud providers, right?
[00:51:01] Marcus: Because the other cloud providers are trying to create their own chips like, you know, at some point you're gonna have to get, you're gonna have to have a full service layer. Yeah. That's not just transactional. Right. Um, and to me, this, this demonstrates their ambition in that regard, and they've got all the money in the world to do it.
[00:51:17] Marcus: Right. Yes. You know. And final story, MIT spinoff Liquid debuts non transformer AI models and they're already state of the art. So this is another topic that comes up briefly in the Brett Taylor podcast. So again, we'll put a link, you guys should listen to it, but, um, Vic, thoughts on this.
[00:51:35] Vic: This is over my head, technically.
[00:51:38] Vic: I wanted to highlight that really smart people are not stopping with the transformer architecture and they're pushing the boundary of the next thing. The next thing, the next thing, which will be additive, right? I think these things will be combined together. You use their liquid architecture in some use [00:52:00] cases, use a transformer and other use cases.
[00:52:02] Vic: You'll have this agent model sort of additional layer. And it's. Maybe a little bit complex, but it also is really excited to see the, the advancements continue.
[00:52:14] Marcus: Yep. Yep. I, I've, I think this was an important story for you to pull out. So I wanted to kind of end on it because I think people don't understand that transformer is a particular model of technology that all of our current gen AI stuff is built upon actually created at Google.
[00:52:31] Vic: Yeah. There's one insight at created at Google that they just gave to the world for free. Um, Which has created, I mean it's what the T is in GPT. That's right. And for five years, that was all anyone was focused on. 'cause it made such big advances. And so to me it was really interesting to see that there's new innovations as well at MIT and I'm sure other places.
[00:52:59] Marcus: Yeah. Well, you know, [00:53:00] there there's gonna be different ways that we're gonna be able to scale ai. Yeah. You know, one is energy. One is just load up a bunch of chips, right? You know, so there's these different, but one is like, what's the core underlying technology, right? This is
[00:53:14] Vic: cheaper to run. I think that
[00:53:16] Marcus: that would be the point, right?
[00:53:18] Marcus: It's like, can you create a, a fundamentally cheaper, more compressed version Of the technology that just changes the model changes the business implications of it completely and that that could be incredibly disruptive to NVIDIA, right? Yeah, you know, now, of course, you could run it on this thing and maybe like you get all this additional.
[00:53:37] Marcus: Power, but what if you know you start to commoditize the hardware that's actually needed to run these models at all, right? That's that's a that's a crazy thing Everyone's got a weakness and yeah, that could be a potential weakness. Yeah here
[00:53:53] Vic: Yeah, and I think there might be I mean we might get to AGI and have new [00:54:00] models that are inventing new ways to train themselves and an ongoing.
[00:54:06] Vic: We also might get to the point where we don't have any more data to consume and the models kind of get to diminishing returns where you, you can put More powerful tips on there, and you can run it to more billions of parameters. But we have, we have sucked in every data point that man has ever created.
[00:54:26] Vic: I haven't seen that the, um, create the AI's creating data sets on their own has really led to a lot of advancement. So the data access, just, there's not enough. I think most models have already consumed the entire internet, um, except for places where they're barred and they can't get access. So I don't know, it'd just be interesting to see how it all evolves.
[00:54:48] Vic: All
[00:54:48] Marcus: right. Uh, so just a heads up next week, we are not going to be coming with a Roundup episode. Uh, Vic and I are headed down to the Country Music Hall of [00:55:00] Fame for Nashville healthcare sessions, and we're going to be. Live there recording sessions with folks. So a
[00:55:06] Vic: bunch of really good people come in town and or they're here already.
[00:55:10] Marcus: Yeah. So we're really excited about that. And, um, instead of the normal roundup next week, uh, you're gonna kind of get a rollout of those interviews and then we'll come back the following week, two weeks from now. Um, and we'll, we'll try to compress two weeks of, of news into one week. But, um, just wanna give you that heads up.
[00:55:28] Marcus: And as always, thanks for listening and, uh, we'll see you on the next episode. Yeah.