By playing, you agree to uListen's Terms and Privacy Policy and YouTube's Terms of Service.
a16z2023 Big Ideas in Technology (Part 2)
Summaries, chapters, transcripts, and FAQs on this page are made by uListen, not YouTube.
EVERY SPOKEN WORD
70 min read · 13,591 words- 0:00 – 1:12
Intro
- SSSteph Smith
At the end of 2022, our team at a16z asked dozens of partners across the firm to spotlight one big idea that startups in their fields could tackle in 2023. Emerging from this exercise came literally 40-plus builder-worthy pursuits for the year, and that ranged from entertainment franchise games to precision delivery of medicine, small modular reactors, and of course, loads of AI applications. And in this 2-part series, we'll be covering 12 of those big ideas. In today's part two, we've got big ideas spanning Fintech, American dynamism, and bio & health, featuring the voices and big ideas of Vijay Pande, Julie Yoo, Angela Strange, Anish Acharya, Ryan McEntush, and Michelle Volz. As a reminder, the content here is for informational purposes only, should not be taken as legal business tax or investment advice or be used to evaluate any investment or security, and is not directed at any investors or potential investors in any a16z fund. For more details, please see a16z.com/disclosures. [upbeat music] Carrying over some of the themes around AI from part one, let's start
- 1:12 – 10:31
GPT Unlocks Credit Counseling - Anish Acharya
- SSSteph Smith
with Anish.
- AAAnish Acharya
Hi, everyone. I'm Anish from Andreessen Horowitz. I'm a GP on the Fintech team, and my big prediction is around GPT applied to credit counseling, and that it's gonna really unlock credit counseling at scale. So here's the idea. OpenAI, uh, makers of ChatGPT, um, has created this amazing new technology that you've all seen all over Twitter that's doing all of these remarkable things. And, you know, one of the things that's been under-discussed is its ability to, uh, drive new product cycles in Fintech and financial services. The one that I've been thinking about, you know, based on my own experience, has been how it can improve credit coaching, credit counseling, and advice. And the reason for this is that, you know, the technology sort of unlocks, uh, a labor supply at 10X, a lower cost than humans. So if the only cost-effective way to deliver cre-credit counseling in the past was to subsidize the human cost with, you know, high fees, high prices, um, or providing lead gen to financial products which weren't that awesome for consumers, we now actually have a technology that can make it possible for companies to offer the same quality of service without monetizing them in that way. Um, and really the, the problem that's being solved here is that credit counseling and coaching is something that is a moderately complex, um, sort of expert system that no UI is able or has so far been able to perfectly capture. It's really hard to build a UI that captures all the nuance of, you know, having a 15-minute discussion with someone that really understands how the credit system works and can help guide you into how to improve your credit. Like, that 15 minutes is invaluable and, and so far, no user interface, you know, or technology has been able to perfectly capture it. And with Credit GPT, um, you can really deliver that service and deliver it at scale, and I think it's gonna change a lot of people's credit, um, and really change financial services more broadly as, you know, potentially the entire country, um, and many people who engage with financial services become more credit worthy.
- SSSteph Smith
So Anish, this is really interesting. To really set the stage, how big is this market? And let's break that down not just in terms of the people who are currently using credit counseling, but also the folks who maybe could benefit from it but could not afford it in the past.
- AAAnish Acharya
For sure. So credit counseling, there's a few ways to think about the market size. There's no perfect answer. You know, there's the sort of what I call the dark heart of credit counseling, which is, you know, you go to some, uh, mom-and-pop diner in some town in America, and you see a stack of business cards for someone that promises to, you know, fix or improve your credit in just a week or a month. And usually those people, you know, do it through some shady techniques and charge really high, high prices. Surrounding that are, you know, the technology companies that really took the first swing at this, and these companies were, were successful. I mean, if you look at a company like Credit Karma, they were acquired for $7 billion and, you know, they've provided credit improvement and credit coaching to over 100 million people. You know, that said, if you take a look at the actual credit impact in terms of how many people's scores got how much better of a service like that, only a small subset of people that use Credit Karma actually saw dramatic uplift in their scores. If you think at the most broad level, you know, really every person in this country that is misscored, um, which is the idea that, you know, that they're sort of seen as being more risky than they actually are because they don't know how to play the credit game, um, that's an enormous number of people, certainly more than tens of millions of people, um, who could potentially benefit from this. And look, lenders and creditors also stand to benefit as the people they lend to make better decisions and manage their money more effectively.
- SSSteph Smith
Yeah, I like that framing, and I, I can definitely see how this could be a wonderful democratizing tool for those who can't access it today. But I also have to think about how finance as, as a space is, as you could say, less fault-tolerant than maybe some other spaces that AI is currently being applied to. So just as an example, if I enter into ChatGPT, "Hey, I'm looking for a recipe on how to make rosemary chicken," well, if that recipe tells me to add a little more salt than desired, I'm gonna be okay. Uh, however, if I'm asking for some of this credit counseling through an AI and it tells me to allocate my money in ineffective ways, well, that maybe, again, is something that is less desirable, um, and something that people will have less of an appetite to withstand in terms of the error rate. And so how do you think about that? And as kind of tacking onto that, how do you think regulation might play a role?
- AAAnish Acharya
I-- The truth is that credit coaching and mos-most financial advice is actually less subjective than you might think. So the amount of salt to put in a recipe is a little subjective. You know, a, a certain chef may like a little more, a certain chef may like a little less. Whereas there's an answer, albeit one that may be moderately complex to derive as to how to optimize any given person's, um, credit and finances. Um, and I think that, you know, these large language model-driven systems will perform-- I know they will perform very well against that problem set. It's, it's a, it's a sort of closed domain answerable problem.
- SSSteph Smith
I do wonder though, do you think that there's going to be a couple errors that actually get attention that therefore prohibit the ability for people to really implement this at scale? Do you know what I mean? Like, th-the people who are the first movers on this, it feels like you need to get it really right at the beginning.
- AAAnish Acharya
Yeah, I think that's right. Look, I think that there are a subset of challenges within this space and a subset of customers, you know, who either desperately need the service and are, are getting it so wrong that, you know, any rational advice would be better than nothing, or who have a set of problems that are actually, you know, v-very boxed and very easy to address. So yeah, look, I do think that you need to have an incremental approach to this, but I think the wrong reaction is that, you know, because this is actually coming from a piece of technology instead of from, you know, a, a subjective judgment-driven human, um, it's probably gonna be wrong. I think it's probably gonna be much, much more right.
- SSSteph Smith
You know, something else that comes to mind is the incentives at play here. So some people might say that these financial services companies, they make money on these people with bad credit, and perhaps more money than the people who have good credit. And so maybe there's an incentive for them not to want to put this technology into place. So what do you have to say to that?
- AAAnish Acharya
I disagree. I think that a lot of financial services companies charge the prices they need to charge in order to turn a profit to every customer based on their respective credit worthiness. You know, I think the painting of some companies, you know, companies as good or evil is just a fundamentally incorrect framing. Um, you know, they have to turn a profit, they charge a price, and the price they charge depends on the default rate of the consumer. If that default rate goes down, they can charge a lower price and, uh, and receive the same profit. And more importantly, they're subject to less volatility in markets like this when sort of m-market cycle is changing and their business or book can get blown out by a sudden shift in consumer credit worthiness.
- SSSteph Smith
That's a good way to put it. I think another framing I like from your big idea was this idea of just unlocking a ten x cheaper labor supply. And so where else in fintech do you think this maybe could apply? And specifically looking to 2023, do you think there's any low-hanging fruit?
- AAAnish Acharya
There's lots of low-hanging fruit. I mean, the lo- most low-hanging fruit is anything that requires, you know, customer support. I think one of the, the most decadent features of getting to work with a private bank, um, or being wealthy is the ability to email, you know, or phone to have your problem solved that way, and that's gonna get unlocked for everyone. You know, the-- having to use an app that doesn't quite-- doesn't have a, a field that you can-- or a dropdown that can answer the problem that you have on that day with your bank is a very frustrating experience, as is calling into a call center and waiting forever, and that's gonna be a thing of the past. I think that everyone is gonna have a private bank-like set of interactions, or at least interface to their interactions with their bank. So that's one. Wealth management I think is another. Wealth management is tricky because it's one part counseling, you know, one part advice, and it's just hard for people to, for obvious reasons, build empathy with the system. So the therapist aspect is gonna be difficult, but the advice aspect is, is gonna be commoditized.
- SSSteph Smith
Well, I'm glad you brought up the therapist element because it's been fascinating to see how even companies like Replica, people are forging-- forming real relationships with technology. Sometimes it's at the forefront where they know that it's an AI who's responding and interacting with them, and sometimes it's invisible. And so that's, that's gonna be fascinating to watch.
- AAAnish Acharya
Totally. It's really cool. I mean, if you read, if you read the newspaper, everyone's very mad at technology. If you look at the NPS of Google, Apple, Facebook, all of these companies, you know, it's not just better than the NPS of, you know, their financial services companies and, you know, most media outlets and-- I mean, it, it-- they're amongst the most beloved companies in the world. So the idea that consumers would adopt this at scale, despite the fact that their mental model have to-- may have to change around things like therapy in the context of financial advice isn't crazy.
- SSSteph Smith
Next up we have Angela.
- 10:31 – 23:48
Compliance as a Competitive Advantage - Angela Strange
- ASAngela Strange
Hi, I'm Angela Strange. I'm a general partner on the fintech team here at Andreessen Horowitz. And I believe that in 2023, doing compliance well tech-enabled is gonna become a competitive advantage versus an annoying thing that you have to do in the background. So if software is eating the world, it has not yet taken a big enough bite out of compliance. Post Dodd-Frank, financial services companies face more than fifty thousand regulations across dozens of federal and state agencies, and that's just in the US. The existing and very manual compliance policies and risk processes are failing at both large financial institutions and at fintech startups supported by sponsor banks. Furthermore, while compliance is complex for businesses operating in just one geography, it is even more difficult to manage across multiple countries. And as more global companies embed fintech, the need for global compliance and risk infrastructure is increasing significantly. In 2023, companies of all sizes will turn to software to solve their challenges. We expect to see more tools for sponsor banks to manage third parties, for fintech companies and companies embedding financial services to manage all aspects of risk and compliance, and importantly, more compliance infrastructure serving default global companies.
- SSSteph Smith
All right, Angela, you mentioned fifty thousand regulations. That is a ton. And I wanna give our listeners a sense of the weight that this puts on companies. Without really commenting on whether this is a good or bad thing, what kind of operational or economic, again, using the word weight, do these regulations place on fintech firms? Like, is it one percent of OpEx, ten percent, fifty percent that they're having to allocate resources to in order to make sure that they're compliant?
- ASAngela Strange
Compliance is both very expensive And the existing solutions don't work very well. Um, as you can imagine, the numbers vary widely, but, uh, you know, there's, there's large banking surveys that go out every year. Estimates are between 6 to 10% of banks' revenue, of their entire revenue, is spent on compliance costs. And they're spending all of this money, and despite that, you know, we feel like we read a, a different headline every week about fines. Like, for instance, since the last financial crisis in 2008, there's been more than $250 billion in fines for poor compliance procedures. And if you look at, like, how does this actually impact operations, right? And the, the-- So it's bad, and it's getting increasingly bad. Like, some of the larger banks, if you, you read their annual reports, ten years ago 4% of their employees were in the compliance and risk functions, and now it's up to 15%. But also, like, you look at this and you look from our perspective, like, this is a huge opportunity. But first, it's helpful to understand why, and primarily three reasons. One that everybody points to is the Dodd-Frank Act that came in after the last financial regulation, right? And so there used to be 30,000 regulations across a variety of, uh, of, you know, state and federal agencies, and in a very short period of time, that ballooned up to 50,000. But this happened very quickly, and in many cases didn't have enough time to become very clear. Which points to problem two, right? If you had compliance procedures that were working well in the old world, and there was some software and many manual procedures, and all of a sudden that complexity doubled, your software was just not set up to be modular, accommodate new rules. And so what did banks have to do? They often just threw a lot of people at the program problem. And so you talk to some of the large banks, and they have thousands of people often sitting offshore that are monitoring, uh, AML alerts, OFAC, uh, sanctions rulings. Um, so it's just a very, very people-heavy business. And then the, the third reason, which is actually, I think, great for the industry, is the industry has evolved a ton in the last ten years. Not that long ago, all financial services was done by walking into a bank. I'm fond of saying that every company is gonna be a fintech company. Now we get our financial services from all sorts of different software companies. We're way more global. We pay in different ways, and there's lots different types of regulations that come along with that that all financial services are having to adopt to.
- SSSteph Smith
Absolutely. I mean, you sent me an article where recently Coinbase was fined $100 million for background check failures, and Coinbase is a large public company. They're heavily capitalized, and they're struggling with these compliance needs. If they can't keep up, how are smaller startups even going to navigate this?
- ASAngela Strange
Listen, and Coin- and this is, like, Coinbase takes compliance very seriously, and this compliance had, had nothing to do with crypto. It was simply, you know, all financial services institutions [lips smack] have an obligation to KYC, like, understand who their customers are as they go through the onboarding process. So imagine any company, right? Like, you're set up, you've got your systems working well, and in their case, what happened is prices were rising in cryptocurrencies. They're a, like, regulated, uh, cryptocurrency company in the US. They do a very good job of what they do. And all of a sudden they had 25 times more monthly transactions. Like, how do systems that are not completely software based actually scale to that? And so you ask, like, how do small companies deal with this? In some ways, companies that are starting now have, have a little bit of an advantage, right? 'Cause if you're starting something de novo versus you're trying to fix systems that you already have in place, you know what you're going to be dealing with. And then two, and this is the piece that's exciting from an investment perspective, the technology in this space has come a long way, as have the teams that deeply understand the tech and really understand the compliance procedures. Like, for instance, um, if you KYC somebody and you need to do what's called enhanced due diligence, often you need to actually check, you know, a photo of their physical ID. And there's lots of companies that have been doing this for a long time. You might expect that they'd all be automated. Oftentimes, they're sending pictures of your ID off to people who are manually checking is this ID real or not, right? Like, that doesn't scale. Now, image recognition technologies have gotten much, much better that technology can solve this type of problem, and we're seeing this applied across many different areas of compliance.
- SSSteph Smith
Well, I'm glad you mentioned automation, because something that really stood out to me in the article about Coinbase was that they had a backlog of 100,000 transaction monitoring alerts and 14,000 users requiring extra due diligence. And so that is just an unwieldy amount of information for a company to pore through. And it seems like, again, even a heavily capitalized company, that's just not realistic. And so I'm trying to understand whether these things can be automated with technology, whether they can be actioned or regulated through code, or if we're kind of resigned to this idea that we hire a bunch of offshore assistants to pore through the data manually. Like, what, what are we seeing in terms of the introduction of technology into this space?
- ASAngela Strange
Well, if you ask ChatGPT that answer, um, they will basically say, "Surprise, AI." But let me tell you a little more specifically. Let's take anti-money laundering, for instance, right? And there's, there's, again, billions of dollars that are thrown at this problem. Depending on what estimation you believe, it's only 3% of dollars that are actually captured. There's $2 trillion that are still laundered every year. And so you look at the technology that's behind this, and it's, uh, very manual-based alert systems. So for instance, anyone that does a transaction over $10,000, you need to flag it. Somebody on the back end has to go check, you know, who exactly is this customer? Is this, um, you know, is this laundering or not? And so it's rule-based systems that, as we've seen with other companies, just don't scale. Now, with the introduction with, you know, AI and machine learning- You can do things like, uh, two things. One, just better gather all the customer information that sits all across the bank, thus vastly speeding the manual review. And then two, dramatically reducing the number of manual reviews that you actually have to do because these systems can learn based on what compliance officers are doing, and so throw up far fewer false positives, such that if it does need to go to manual review, it's a much more reasonable number of things that a human could actually handle. And I think that's just one simple example. You can apply, you know, this and other technologies across just large, large areas, um, across all financial services.
- SSSteph Smith
So when we're talking about these different regulations, again, we, we mentioned 50,000 of them, and they span different types of regulation as well. And so when I think about companies in these spaces that are helping other small or large companies navigate this ecosystem, Deel comes to mind for cross-border compliance, um, with respect to payroll, Sardine for KYC, which we've already mentioned, and AML. Are there other gaps that you see in this, in this, I'm gonna call it this compliance web, because it really does seem hard to navigate, that you don't really see companies effectively servicing?
- ASAngela Strange
Any industry that is hard to navigate has spawned a very large industry of necessary consultants, right? And you talk to well-intentioned, um, you know, larger financial institutions, earlier startups. It's now very top of mind. People want to do this well, and then they just see a web of vendors. And, and what do you do? And so I'd say there's, there's really... Well, there's many opportunities, but three I would highlight. One, traditionally compliance has been very siloed. Like you think you start with know your customer, and then you get the customer on, you've got to monitor their transactions, and then you might have some separate system for fraud, and none of these necessarily speak to each other. You have all these different types of vendors. You have new data sources that might come on board. How do you integrate those? And so now there's, there's new companies, you mentioned Sardine, that are, are really looking at this from a, like fraud has a lot to do with who you onboard, has a lot to do with the transactions from a holistic perspective, and just better feeding this data into a circle and making it much more modular to integrate new types of data sources. So just looking at this problem from a, a holistic point of view. Opportunity two, you know, we, we, we talked a little bit Coinbase. Um, just in the US, more than 10% of people have a crypto wallet. And so does it make sense that you would have entirely separate on-chain compliance systems from the, you know, off-chain world if oftentimes the, the people are the same? And so that's an opportunity to bring more data into the set and to improve compliance. And then the third part I'm, I'm pretty interested in, uh, which has a, a coordination challenge, but everyone that deals with financial services is sitting on often very siloed pieces of data. So for instance, if I suspect as a bank that one of my customers is a money launderer, and I wanna get information from another bank to be able to check that out, there's a long compliance procedure, um, enabled to do that. So just how do we better enable data sharing, which can help elevate every financial institution in doing a better job in compliance?
- SSSteph Smith
Yeah, I'm going through this right now as a customer of several banks and financial services firms where I'm changing my residency and I'm having to go through the exact same very lengthy, very arduous process with every single one. And of course, there's privacy concerns about what information is shared between these different entities, but again, as a consumer, I'm experiencing this problem right now. One thing I wanna ask you about as we close this out is how compliance relates to competition, what role it plays there, and as we're seeing these increased number of regulations, whether companies are choosing to go to other jurisdictions where maybe this isn't the case.
- ASAngela Strange
You know, we've invested in, in many global company is, like every country has pretty significant compliance regimes. Um, and they're, they're very different. And I... So I think from a just how does this look going forward, I think there's opportunities for new compliance companies to just help keep up with what are all the different changing regulations in all sorts of different jurisdictions, right? So if I'm a company in the US and I want to expand to Brazil, I have to learn an entirely new different compliance regime. Wouldn't it be great if I had an infrastructure provider that would help make sure that I was compliant and help stay continuously compliant? So I think that the opportunities here are really twofold. One, obviously new companies that help rethink this, you know, highly manual procedure in different jurisdictions in different ways. And then two, I think new companies that, you know, aren't compliance companies but need to be compliant are gonna think about what are their compliance systems and what are the, what is their approach to compliance very much from day one in the same way that they think about many other elements of their infrastructure and bake it into their strategy to, one, often provide a better customer experience, but then two, keep these costs really down so that they can spend their profits on other things.
- SSSteph Smith
Yeah. I mean, coming back to what you mentioned before, 10% of revenue, that's, that's a lot. That's not 10% of profit. Um, so I think this will be top of mind-
- ASAngela Strange
Think of all the new products you could be building with 10% of your revenue.
- SSSteph Smith
Next up we have Michelle.
- 23:48 – 34:38
Small Modular Reactors Advance the Nuclear Renaissance - Michelle Volz
- MVMichelle Volz
Hi, I'm Michelle Volz. I'm a partner on the American Dynamism team, and this is my big idea for 2023. Small modular reactors advance the nuclear renaissance. Though nuclear energy accounts for 20% of the US's electricity, it's commonly misconceived as a dangerous and non-viable option when it comes to adding reliable sources of carbon f- free energy. But recently, nuclear energy has been having a bit of a renaissance with the Inflation Reduction Act earmarking $30 billion for tax credits towards existing nuclear reactors, a first for the US. The timing is right now to usher in new innovation in this space. There are opportunities across the nuclear supply chain, from fuel sources to mining to manufacturing vendors and beyond. Um, and one area I am particularly excited about is small modular reactors or SMRs. By leveraging advanced manufacturing techniques and modular design, SMRs can be quickly and efficiently mass-produced, bringing down costs significantly. This can make nuclear energy more accessible for a variety of applications, including providing clean and reliable power to remote communities or even one day in space. While there's still a way to go in reforming the regulatory frameworks for these types of reactors, SMRs and the broader nuclear industry are likely poised for growth in the year ahead.
- SSSteph Smith
So I feel like energy has been top of mind with some of the geopolitical shifts that have happened, especially in the last year. But I feel like within that conversation, nuclear has become a more prominent topic, but at least in my world, not SMRs. So can you go a little more deeply into what SMRs are and maybe how they differentiate from a traditional nuclear plant, whether it's the difference in size, cost, time it takes to actually build.
- MVMichelle Volz
SMRs or small modular reactors are basically exactly what they're called. They're small and modular reactors. Um, what that does is that means they can de- be deployed in places that traditional nuclear plants can't. Um, they won't ever provide the scale that like a traditional nuclear power plant would, but they require significantly less capital expenditures, um, and can be essentially shipped to wherever they're needed. Um, so it can be like remote communities or, um, on ships or in space. Um, and, and because of their size and because of the timelines, it just like unlocks a lot of new opportunities within a timeframe that's actually reasonable. So we could get this within years instead of within decades.
- SSSteph Smith
Can we get specific on those years? Like, how long does it typically take to set up a traditional nuclear power plant versus an SMR?
- MVMichelle Volz
Uh, the bulk of the time in a traditional nuclear power plant is actually just like getting the site location, going through all the specs, getting, um, all the designs approved, and that like, that just takes a ton of time. Um, there's just a lot of hold ups, and we haven't seen a new nuclear plant get built in decades. Um, with SMRs, we think it can be done in years. Like, there's still regulatory things to get through because this hasn't been done yet. There hasn't been a, like, design framework on the regulatory side to, to say like, "This is how these should be built. This is like the timeline." Um, but with-- there are a number of startups like actually building today and getting designs approved and, um, starting, starting on the path to get these built. So, um, we don't know exactly what the timelines are, but I think we are excited in what could happen in the next, um, couple years or, or before 2030.
- SSSteph Smith
That's exciting, and I like the framing of decades versus years. Something that you just spoke to there, which has come up in my very limited research into this space, is regulation. It seems like the time it takes to get all of the necessary approvals to build a plant is astronomical, or at least more than I think most people would think. And it sounds like maybe there's some regulatory reform happening. You also mentioned the IRA, which is allocating funds towards nuclear. What should people be looking for here in terms of reform? Like what kind of reform do we need, whether it's with respect to the traditional nuclear power plant or these SMRs, which sound like they're a little newer and maybe the reform is less concrete.
- MVMichelle Volz
In the US, the Nuclear Regulatory Commission, or the NRC, is the, the body that would approve any new nuclear site, um, or, or nuclear reactor, and it's often cited as one of the top bottlenecks in progress towards building more reactors. Um, there are super high fees, like million, multi-million dollars just to get designs looked at and approved. There's very long timelines to go through the end-to-end approval process, uh, at all. There's disputed frameworks for how it regulates like radiation, which slows things down even further. Um, and as I mentioned before, like critically, there's a lack of the regulatory framework for SMRs in particular. Like, they haven't really been built before, so there are no guidelines yet. Lowering the fees to approve the designs and accelerating the timelines overall are both the, I think, the essential steps if we wanna see more SMRs come into production within the next few years.
- SSSteph Smith
Yeah, I was really surprised to hear that in the case of some of these regulations, like there isn't actually even a threshold of the appropriate amount of radiation which is allowed. It's, it's just kind of this vague sentiment of do whatever you can to limit radiation, which is not something that many people can action off of. Um, are there examples of other countries that have maybe set up more legible, maybe more actionable legal frameworks or regulatory frameworks with nuclear? And what have we seen in terms of those other countries in terms of maybe what we can learn?
- MVMichelle Volz
So other countries are, are kind of at various stages of development and support for, for SMRs in particular, for nuclear overall. Um, we've seen some progress in Europe towards approving like building sites more. I know China's putting a lot of money in- into nuclear innovation. Um, but the, but again, but at the end of the day, the countries that are lowering the barriers towards actually like building and testing these reactors and getting, um, plans to start development are the ones that we should be looking towards and, and following along with. And overall, like international coordination is gonna be really critical here in this domain. Um, once something gets approved in one country, it should help lessen the burden in other countries. And so, um, there are some organizations kind of working towards coordinating across countries and internationally, but I think that will be, um, something we should look towards and, and really participate in in the US.
- SSSteph Smith
As people listen to this, there may be potential founders. Some of them won't want to be the entity that goes and sets up a whole nother nuclear plant or an SMR. But something you spoke to in your big idea was this idea of the nuclear supply chain and all of these other businesses that kind of ladder up to the end product, whether that's mining or manufacturing. Can you tease this a little bit more in terms of the different types of businesses or things or gaps that people might explore if they're actually looking to be a part of this industry in 2023?
- MVMichelle Volz
As we hopefully start building more reactors, there's gonna be material needs in things like uranium or the cooling materials, depending on how the SMRs need to be cooled and built. Um, there'll be needs around the manufacturing and transport of the reactors. They're small and modular, so you can move them around, but there's still some, um, things to think through of how to move this, uh, sensitive material. And then there'll be needs around like disposing of nuclear waste and thinking through the recycling process. Maybe on the more exciting side, there's potentially new areas unlocked like decommissioning or repurposing other types of energy plants, like coal plants, into nuclear plants. Um, and so I think there's gonna be tons of opportunities across the whole chain and even like software to coordinate all of this or to track and monitor emissions and things like that.
- SSSteph Smith
On that note, nuclear, we're currently talking about it in the stance of fission, which people are more familiar with, but there have been other advancements in fusion, which I think is a little further out. I don't wanna get too ahead of our skis in terms of talking about its economic viability today. Um, but 2022 was kind of this exciting year where at least one group was able to demonstrate net energy gain, and I think 2021 was a banner year in terms of investment into nuclear fusion companies. And so although these are two very different things in terms of where they're at, in terms of their ability to be implemented today, do you have any thoughts around, you know, where that's going, what fusion's role is in this wider matrix, or how we might think about that?
- MVMichelle Volz
So I think the overall energy needs over the next few decades and beyond are big enough that there can be many solutions, and I would love to see more progress on fusion. I think it could be something that's absolutely game changing if we can get there. Unfortunately, I think it is further away than we'd like and that people are hoping, especially if you think about any- anything close to, um, a scalable solution or something that could be powering an entire power plant. One of the main reasons I'm so excited about SMRs is that we actually aren't waiting on any new scientific breakthroughs, and these things are getting slated to be built in the near term. Um, again, on the order of years, we could see, um, these in production and actually producing the energy we need. So, um, while fusion is very exciting and I definitely hope scientists keep working in that area, I think, um, in the near term, it's this fission and it's the, the smaller modularity that's gonna be the breakthrough as we see.
- SSSteph Smith
I feel like I have to ask this question after you brought up this idea that the technology already exists. If we look out, let's say, five, 10 years from now, and we're still at the place where nuclear i- has not really made any advancements, it's still taking decades to create our plants, we still don't have the right regulatory infrastructure in place, like what would cause that? What is, is causing us to be at kind of this like friction-filled spot in the arc of history as it relates to nuclear?
- MVMichelle Volz
I think a lot of it is on the regulatory side. Like, we need to decide, um, as a country that we want to invest here and we want to push this forward. It's not gonna be about were we able to build it, it's did we decide that we wanted
- 34:38 – 46:36
Overhauling the Space Supply Chain - Ryan McEntush
- MVMichelle Volz
to invest in this? Did we decide that we wanted nuclear? The biggest barrier is just humans, not, not the innovation, just us getting in our own way.
- SSSteph Smith
Next up, we have Ryan.
- RMRyan McEntush
Hi. Uh, I'm Ryan McEntush. I am a partner on the American Dynamism team. Overhauling the space supply chain. Civilization's ancient past is segmented by advancements in mining and metallurgy, the Stone, Bronze, and Iron Ages. In each period, success over your rivals was determined by your ability to collect resources and produce increasingly useful technology. Today, we are in the Space Age, and the same rules apply. Space supremacy will be the measuring stick of industrial and military power for the foreseeable future. It's already supporting our digital economies and guiding our autonomous systems, but the true space industrial base is much broader. We not only need to construct advanced rockets and satellites but gather the materials and industrial capacity to do it reliably at scale. Space does not begin at liftoff. A complex global supply chain from mining to launchpad must be overhauled and secured within our global alliance networks. Beyond Earth, infrastructure must be built to service existing assets in orbit and power more ambitious missions to deep space. In 2023, the space industrial base will continue to grow in size, birthing critical companies serving our nation's interests. At the heart of this resurgence, inspiring founders are pursuing difficult problems in materials, manufacturing, and space infrastructure.
- SSSteph Smith
All right, Ryan, I feel like that's a fascinating prediction or big idea, but I feel like we have to start with the why question on this one. Why must we build advanced rockets and satellites, using your words, reliably and at scale?
- RMRyan McEntush
Well, I think most importantly, the United States should be the leader in space. This actually goes back to a, a JFK quote, uh, when he gave his famous moon speech at Rice University, where he said, uh, "No nation which expects to be the leader of other nations can expect to stay behind in the race for space." And I think that's Incredibly important, not only back then, but today as we have new, uh, near peer rivals, uh, entering the domain. Why is this important? Satellites already drive much of our economy. Uh, you know, this telecommunications infrastructure, this is GPS. Um, these are things that are relevant not only for the commercial side, but for the military side as well. Speaking of military side, in wartime, as example of, uh, Ukraine, we have sort of near real time Earth observation and signal intelligence coming from that. And this is essentially, you know, the ultimate height advantage in war. Um, but even something like climate change, having that Earth observation, being able to see where the minute emissions are coming from, being able to detect weather and variations in climate, it's all critically important to solving those problems. And perhaps lastly, a- and, and perhaps most long-term, there's essentially infinite resources in space. So a lot of these things we think about, uh, mining or even just accessing energy, uh, there's all of that in space. So in the long run, if we want to actually continue to grow as a species, we're eventually gonna have to figure out how to harvest and utilize space resources.
- SSSteph Smith
I like that you've illustrated the why, and I think an important aspect as we explore the idea of pursuing space more intensely is, as you mentioned, the supply chain. A lot of people think about the last mile there, which is actually launching a rocket into space, but there's a lot that leads up to that. Um, and as we explore that, that full space supply chain, as you've said, from mining to launchpad, I wanna understand what part of this chain you think is most at risk, or we could actually reframe that to say, what part of that supply chain has the most entrepreneurial opportunity? What part of that has the largest gaps for new critical companies to be created?
- RMRyan McEntush
Sure. Um, so the way I think about space and space supply chain holistically is that space is sort of the... Well, it is, you know, the final frontier. It's where our industrial capacity and the stuff that we can build, um, can actually extend into. And so it's not just, you know, the launchpad, it's not just the rockets going to space and the satellites, but it's everything else that allows us to do that. And so these are things like mining, uh, so actually getting the resources that we need to build the stuff. And it's not just the stuff that goes up into space, which is predominantly aluminum. A lot of our stuff is built out of aluminum, titanium. Uh, but it's everything else, like actually building the factories, the computer chips, um, even things like energy generation. So how are we actually gonna get the energy to do this stuff a- and build, you know, thousands of satellites or thousands of rockets? These are all things we need to think about. Um, so mining is the first one, and specifically within mining, I look at midstream refining. So not just mining the materials, but turning them into workable metals. And there are specific metals that are highly relevant, you know, some of the battery metals and stuff for energy storage and, and generation. Uh, but even things like rare earths, um, that are used in magnets as well. The second one is actually manufacturing space components. Um, so this is one of portfolio companies, Apex. They're actually building satellite buses. So a lot of companies are looking at potentially scaling out the space infrastructure that we have today. Whereas today it's, it's sort of bespoke building rockets, um, building satellites, uh, building whatever you need. If we're gonna do thousands of those and scale it up, we need to be able to develop a mass manufacturing process, and that might require separating out these vertically integrated businesses like SpaceX that does everything, and doing each individual component and building a separate company that handles just that. Last piece is what's called OSAM, so like onser- uh, on orbit, uh, servicing and manufacturing. Basically activity, uh, that services this market of satellites and space stations. If we're gonna have 30,000 satellites in orbit, we need to actually have the, the sort of companies that are able to repair them, fuel them, service them, de-orbit them. Um, and so I think those three areas are the most, um, immediate opportunities.
- SSSteph Smith
Whenever exploring these space opportunities, they are inherently moonshot opportunities. I mean, uh, you take a company like SpaceX, and SpaceX I think has provided a good example of, again, a company that has moonshot ambition to, you know, make us a multi-planetary species, yet also has kind of brought that back down to Earth, you could say, by having a business model that funds their progress, in SpaceX's case, by providing internet to people on Earth today. So again, not just betting on tomorrow, but actually showing that they're providing value through the revenue that they're driving, um, towards that eventual moonshot. And so I'm curious to know if there are examples, um, that you could provide of other companies, uh, perhaps that have been able to build that bridge. Um, because I think perhaps one of the reasons that maybe more founders aren't pursuing space companies is due to this moonshot nature, it's kind of like build or bust. Um, and I, I'd love to share different examples where maybe that's not the case.
- RMRyan McEntush
Well, I think the most poignant example is the company I mentioned a second ago, Apex, where today they're building satellite buses. So they're sort of riding the curve of everyone's building satellites, everything's getting launched up into space. Um, some companies are gonna find it a lot cheaper to outsource that and have another company build satellite buses. Um, so they're able to do that today, but in the long run they're gonna expand into building a lot of other stuff that's gonna be used in space. So the actual product is this mass manufacturing of space hardware. Um, but today they're focusing on what is immediately addressable, which is satellite buses. More broadly, there's sort of two areas that I think are interesting, uh, in terms of, um, how you're approaching the market. Uh, the first is what I call like Jamestown businesses. Jamestown was something I used in my essay, but if you remember historically, Jamestown was the first successful settlement, um, on the East- on the Eastern seaboard of the United States. Um, the reason they were successful was they had a competitive advantage of growing tobacco there and selling it back in England and, and Europe. And that's sort of what we've seen with space, particularly in the, the mature Earth observation market, where they were able to put satellites up and they had a competitive advantage because they were the only... That was the only way to get satellite imagery and, and to handle these sort of telecom networks. They had the immediate market for that, and they were able to build that. In the long run, those companies likely expand into other domains, but how you build that company in the beginning is service that, that market. The second one is, uh, tracking military contracts. The, the first sort of, uh, customer for a lot of space, uh, pushing the frontier if you will, is often government. Uh, SpaceX is still largely, it derives a lot of the revenue from government. Uh, but looking at areas like space domain awareness, which essentially is knowing what else is around you. Wha- spec- specifically if it's like a, a very important satellite, being able to see what's around you, whether that be, uh, you know, orbital debris or an enemy satellite. That's very important, uh, for, for Space Force and, and for other government, uh, programs. Um, also space tugs, so being able to take something from a lower energy orbit and push it up higher. Now, there's companies like Northrop which have proven this, um, what's called life extension, which is essentially taking a very expensive satellite and pushing it up a little bit more so it, you know, it can, it can last longer in space before it falls to Earth. Um, and then for Artemis and some of these programs going to the moon, you're gonna need something to actually take it once it's in space, that long mile, um, you know, to, to the moon. Uh, and then lastly, there's a l- bunch of other areas like tactical response that have sort of more military applications. Um, but looking for the military and government contracts as sort of the, the pathway for revenue, and then when the commercial opportunities catch up to that, you're gonna be there and have the technology ready to go.
- SSSteph Smith
You'll be ready to ride that wave. I, I feel like, um, I would be remiss if I did not cover this question, um, as we close this out, which is just, you know, you in your big idea kind of framed this idea of space providing the, you could say, the, this measuring stick of power. Um, and I think some people would say, "No, Ryan, that's still on Earth," right? Like, access to land, water, critical materials like cobalt. That's really what is going to be the measuring stick of the different countries or entities that succeed. What would you say to that?
- RMRyan McEntush
I think people don't realize how important and how critical space already is to even land battles. And, and Ukraine is a good example of this, where the United States is able to provide sort of real-time intelligence of troop movements, uh, signal intelligence, being able to detect everything that's said, every message that's, uh, that's sent, um, and targeting capabilities. So being able to actually know where people are, then also be able to send missiles to where they are. All of that assets to be able to provide that infrastructure takes place in space. So even, even though they're, they're land-based battles, the infrastructure to facilitate that and to actually win wars in the future are require, you know, th- these assets in space. So these are not only important for winning, but also they're one of the first targets if there was a real conflict between the United States and China or Russia, they would go out for the satellites first. Um, so, so that is, that is absolutely critical. And also, in terms of minerals, mineral discovery companies like one of our, uh, portfolio companies, Cobalt, they utilize these data sources from satellite imagery as well. So even in terms of minerals, they're, they're heavily relevant. Um, a- a- and more broadly, looking at resources, there's, there's, um, just an insane amount of material available in asteroids and being able to actually mine some of these other celestial objects. I think some of the numbers I've seen were like a small M type asteroid might have twice as much nickel and iron that is produced on Earth in a single year. So a country who's able to actually do this at scale, uh, solves a lot of these material component problems. And, and I think in the long run, we likely see a lot of these, once we start getting a lot of resources from space, we might move other industrial, very high energy intensive activities into space. You don't have to worry about nuclear radiation if you have a nuclear reactor in space and you're doing other sort of CO2 emitting, uh, processes in space. So you might be able to do a lot of industrial stuff and then send it back down to Earth. And that'd be the long-term plan.
- SSSteph Smith
Next up, we have Vijay.
- 46:36 – 53:29
The Biggest Company in the World - Vijay Pande
- VPVijay Pande
Hi, I'm Vijay Pande, general partner of Bio & Health at a16z, and this is my biggest idea. The biggest company in the world will be a consumer health tech company. This may sound crazy to some, but why shouldn't this be true? Four of the top five biggest companies in the world are consumer companies, and healthcare is one of the nation's biggest industries. Much larger, in fact, than the size of the global advertising industry in which consumer giants like Google and Meta operate. From that standpoint, the number one slot should be a consumer health company. So we see two paths to a consumer health company becoming the biggest company in the world. The first is a vertically integrated path of being a so-called payvidor, a combined payer and provider that eventually owns most care. Imagine UnitedHealth Group, but with the user interface of Apple. Who wouldn't choose this insurance plan and provider? The second is a horizontal path of building a consumer marketplace or infrastructure layer that enables all other care delivery companies. Imagine Amazon or Visa, but of healthcare. We'd go as far to say that there is almost, almost infinite room to improve the consumer experience in healthcare and build massive companies as a result, and we expect consumer healthcare to be front and center in 2023.
- SSSteph Smith
All right, Vijay, this is a big prediction. So before we jump into the nitty-gritty, I wanna hear from you just how big is this healthcare industry?
- VPVijay Pande
Yeah. Actually, people forget just how huge it is. It's approaching 25% of US GDP. It's about, like, 22 right now. It was 20% in 2020, and it's just, uh, been gradually increasing. It's what's the healthcare crisis that we, we worry about. And so in that sense, it's 10 times bigger the marketing budget of consumer, uh, giants like Google and Meta.
- SSSteph Smith
Well, I'm glad you brought them up because some of these large companies, like Google, Apple, Facebook, and Amazon, are pursuing healthcare to some degree, and they have so much capital to pour into this s- nascent space for them. But to date, they have not succeeded. So why do you think these super heavily capitalized companies have not come out on top yet?
- VPVijay Pande
Yeah, I think there's a, a couple different ways to think about it. From purely a tech lens, you could imagine the days of Google fighting Microsoft. Like, how could something like a little startup known as Google, uh, uh, sort of compete against a giant like Microsoft? And it did because, uh, they were, uh, web native. Uh, they were sort of built upon a, a different set of rails and, uh, with founders that think of things differently. And so even just from a tech company perspective, uh, it's something that's so different is hard for giants to go after just because of the way they're built, and this is basically classic disruption theory. Healthcare's also different than just tech. You can't come in with, I think, a naive understanding of how healthcare works, either on the medical side, the life science side, or even just how the system works, and come in and, and make big changes. And so a combination of those two things, I think, will make it very difficult for the existing incumbents, at least on the tech side, to dominate.
- SSSteph Smith
Yeah, so we've covered the incumbents on the tech side, but there's also large healthcare companies out there today. One that you've mentioned is UnitedHealth Group, which is the largest healthcare company in the world. I think it's also the eighth largest company overall. What do you think is stopping them from dominating in the way that you're describing?
- VPVijay Pande
Well, I, I think just like we talked about for tech incumbents, I think there's the same issue now for healthcare incumbents in that, uh, the winner is gonna combine, uh, a deep knowledge of tech and a deep knowledge of healthcare. And it's p- just as it's very difficult to sort of build that into a tech company, it's very difficult to build tech into a healthcare company the same way. Uh, it's just not the way these companies are built, and so it, it's, it's the analogous problem on both sides.
- SSSteph Smith
And so something that you seem to be alluding to is the fact that we need new founders pursuing this problem. So if you were a founder today, what angle would you tackle this from?
- VPVijay Pande
I, I think the key thing is that the founding team ideally would bring together people that have, uh, this deep knowledge in tech and in healthcare. I think now sitting in 2023, we're in a different world where people have grown up with tech their whole lives. So if you're graduating from medical school now, um, or if you've been working the healthcare system, tech is not something that is unfamiliar. It's part of your, uh, your daily life and has been part of your daily life now for decades.
- SSSteph Smith
It's important for us to pay attention to these inflection points, and so it sounds like technology has changed in a way where there's new opportunities in this space. But I wanna really drill into this idea that you said that there's infinite room to improve the consumer experience in healthcare, and I feel like this has been true for quite some time. And so perhaps in addition to some of the technical changes or the talent changes, why haven't we seen many of these developments yet?
- VPVijay Pande
It, it's actually, uh, analogous to some areas in tech where people from a consumer background brought, came into enterprise, and enterprise software started having the look and feel of a really smooth UI from... you'd expect from Meta or from Google or something like that. A- and similarly, it, it just takes a different type of founder to really understand how to create that experience on the tech side, which is user interface and so on. But it's also, um, NPS is more than just about the software. It's about what you can actually do and what you can accomplish. And, uh, I, I bet there, there really isn't the same focus on user experience often in healthcare because, frankly, often the user, the patient, isn't the payer. That's really, I think, one of the key differences for a consumer-facing healthcare company, where finally now we'll be involved, and maybe to some degree things will be out-of-pocket and that will change things. But even still, I think someone with consumer mindset where the patient, um, is really paramount and that the patient's user experience is maybe even a part of how we think about their healthcare, uh, that's a fundamental shift.
- SSSteph Smith
Vijay, I think this is a really interesting concept, but pushback that I expect from this idea of the biggest company in the world being in healthcare is that there shouldn't be these massive companies profiting off of patient health. So what would you say to that?
- VPVijay Pande
Yeah. Well, first off, I think if you think about what a human endeavor should be, taking care of, of, of the sick is one of the most important things that we could be doing. It's important to realize that a com- company that's very successful at doing that is creating so much value that for it to become a large company is a very positive thing just in general. But also, I would add that there's different ways that companies become big, and one is through rent-seeking, where they have some advantage that they just take it, uh, they, they, they leverage, and they don't really contribute or innovate, or... And this is the spirit that we see in a lot of tech companies, where, uh, they actually bring huge degrees of innovation, and that innovation is, uh, really dramatically improving the patient experience surrounding, uh, dramatically improving outcomes. And as a side effect, this is building the biggest company in the world. That's the vision that we would love to build towards, where, uh, the improving, uh, the human condition, improving he- human healthcare is considered to be one of the most important things we could be doing and would be, in some sense, represented by the fact that such a huge company could be built.
- SSSteph Smith
Next
- 53:29 – 1:12:03
The Value-Based Care Stack - Julie Yoo
- SSSteph Smith
up, we have Julie.
- JYJulie Yoo
Hi, everyone. I'm Julie Yoo, general partner on the a16z Bio & Health team, and this is my big idea for 2023. The value-based care stack. Um, as we sit somewhere in the murky middle of the adoption curve and hype cycle of value-based care, or VBC, we are unabashedly VBC optimists. But we are also eyes wide open that many value-based models haven't delivered value yet. And one of the major reasons for this is that the clinical and operational models from the legacy fee-for-service world have simply been transplanted into value-based paradigms, resulting in the gaming of the system versus a fundamental reorientation of care models to be focused on value from the ground up. And so VBC done right demands purpose-built approaches that will be built on a fundamentally different stack. This emerging stack will support new entrant value-based digital health players and incumbents alike around unique requirements to deliver higher value care, things like data aggregation and activation, actuarial modeling, contracting, adjudication, panel management, continuous care, workflow support, and provider ecosystem integration around referrals, co-management, and network design, just to name a few. So if the first generation of the digital health tech stack, which we wrote about in 2020, was about enabling administrative and operational efficiency for virtual first providers, the next generation will be about helping providers bear risk, and, and enabling payers to collaborate in a more integrated fashion with their risk-bearing provider networks. We see this stack emerging in many different product phenotypes from SaaS platforms to solution marketplaces to MSOs, to service an entire range of buyer segments and levels of technical sophistication.
- SSSteph Smith
Can you just break down for the listeners what the difference is between some of the value-based care that you're speaking of and the legacy fee for service systems in healthcare?
- JYJulie Yoo
[lip smack] Of course. The first thing I'll call out is that you would hope that, you know, all the healthcare that you receive is value-based. So it's sort of telling that we've had to concoct this term to qualify what is value-based versus not. Um, but, you know, the premise of value-based care is that, you know, so much of the bloat of our healthcare system is the result of the sort of massively misaligned incentives due to the third-party payment model that is the dominant way for how healthcare is paid for across this $4 trillion industry in our country every year. And, um, it's probably actually not even accurate to call the fee-for-service system legacy because it's still the dominant model today. But, um, the fee-for-service system is one in which, as the name indicates, providers essentially get paid for what they do. Like, so for every visit that they do, every test that they run, every procedure they perform, they are submitting this atomic claim for each of those tasks and getting paid a fee for each of those tasks. And so as you can imagine, you know, that kind of system incentivizes doing more stuff and, um, and also incentivizes kind of a focus on treating people who are sick as opposed to this value-based system in which providers are getting paid for kind of managing the health of the population of their patients and preventing sickness, um, you know, versus treating it. The purest form of value-based care is where providers get effectively a set budget on an annual basis to manage their patient population. And, you know, any overages that you have on that budget you have to eat, but also any underages on that budget, you get to keep a share of that savings. And so again, it incentivizes a different care model associated with that, which is, uh, you wanna be much more pr- uh, sort of proactive about getting ahead of any escalation of health issues, um, to avoid all those catastrophic costs that might come from people really getting sick and ending up in the ED and things like that.
- SSSteph Smith
So what you're describing sounds great. It sounds like we're driving efficiency, proactivity, we're dealing with sickness before it happens. And so what have been the key roadblocks to actually achieve this reality that you're painting?
- JYJulie Yoo
[lip smack] Yeah. So I mean, fee-for-service has been the dominant payment model for many decades. Um, I think technically it was like the 1960s when the first, like, billing codes were implemented that led to this, uh, this type of proliferation. And so it's fundamentally just inherently hard to, like, change the business model of an entire industry. So I think there's just some inherent, uh, inertia that, you know, the entire way that our system operates has been optimized for fee-for-service. Everyone knows how to game it. Everyone knows how to benefit from it. Um, there's an entire, like, $150 billion industry dedicated to just submitting and processing claims so that you optimize your revenue flow. And so, you know, just because so much money's on the line, there's a, a ton of friction, um, against, you know, changing the status quo. I, I liken it to, like, the, the le- the legal industry where, you know, everyone's familiar with kind of the minute-by-minute bill-billing system that law firms use, and how many times have we heard about some upstart, you know, law firm who's trying to move away from time-based billing to fixed fee billing? But as quickly as it comes up, it goes away because there's just, you know, so much inertia in the system. Um, and so therefore, you know, I think it would really just have to take like a top-down government mandate to force an, an industry to make that kind of switch. And while it's, it's not like what I would call a full-out mandate, we effectively have had that, where we've had a number of significant government initiatives that have been put into place that create financial incentive to get, um, into the value-based care game for providers and payers and take full risk, as it's called, uh, in the form of these, you know, sort of budget-based payment models. Um, we've had a number of examples of that. It's been, you know, Medicare, um, Advantage, I think is a, a primary example that was started in 2003. We've had accountable care organizations or ACOs, which were initiated as part of Obamacare in 2012 and, uh, some subseq-subsequent flavors of that. So, you know, we're still in super early days, I would say, as far as the seeing the impact of those top-down mandates. That's really what it took to kind of shake the system up and say there's a, a different way of doing things.
- SSSteph Smith
Yeah. So it sounds like some of those initiatives were over a decade old, and so what is it about today in 2023 that you see changing, whether it's through legislation, through the founders who are stepping up to change things, or through maybe just the step-by-step changes that are happening to kind of reorient the inertia that we've seen in the past? What are you seeing that makes you hopeful about maybe some of the shifts today?
- JYJulie Yoo
Definitely at a baseline, just the general sort of like it take, it takes time for these things to take hold. I mean, it's the same way that we talk about technology shifts, right? Like even, you know, the internet having been invented, quote unquote, "on a mainstream basis" in the '90s, it's, it's really only now that we're seeing some of the fruits of that innovation. So I think there's just some law of physics type things that are fundamental like that that apply here. But, um, you know, I think it has to go, um, without saying that certainly the pandemic and everything that's happened in the last couple of years has had a huge impact on frankly like the financial beating down of providers that has created this huge breeding platform for providers to diversify their revenue and just create much more resilient ways to survive. And value-based payment is one way to do that, where you actually have a recurring revenue stream versus these atomic payments that rely on patients coming into your office. And so that, I think, has been a huge, um, tailwind for, for the adoption of value-based care models. You know, you also see a lot of decentralization of care away from traditional hospital and clinic-based settings. Um, you know, virtual care, home-based care, community-based care, all of these things are now very much getting adopted because they are fundamentally lower cost, you know, more convenient and much more reliable than, you know, having to force patients to go to these centralized settings. So, um, you know, I think that those are, are some of the key, uh, obvious, um, drivers of, of seeing kind of the why now that's occurring right now. Um, you know, that said, I think there's still-- we're still looking at, depending on what reports you read, you, you... We're, we're only at roughly about 20% of- All healthcare payments being in some kind of risk-based model. Um, so, you know, to use the crossing of the chasm framework, we're probably still in kind of like the early adopter, maybe early majority phase, but again, it comes back to that, that concept of we're just still fair- fairly early in, um, in the kind of the fallout from some of these top-down government mandates and, and what impact they can have on the industry.
- SSSteph Smith
Yeah, completely. And I think to your point, the change in inertia doesn't happen right away. I like how you've broken down the healthcare tech stack into three parts in the past. So that includes care delivery services, the back office admin, which is required to run a practice, and then the front office, which interfaces with consumers or customers. Is there a part of that tech stack that you think is really, um, particularly ripe for disruption today?
- JYJulie Yoo
Yeah. Yeah. And just for context on what you mentioned, you know, we had written a couple of years ago about this new tech stack for virtual-first care, and, you know, the premise of that thesis was that, um, all of these new virtual care companies are very ill-served by existing IT solutions, um, because those IT solutions were very much predicated on fee for service, you know, payment rails, um, s- sporadic facility-based patient encounters, so their entire data model was predicated on kind of these, you know, again, these sporadic visits, and, um, the workflows were basically all provider centric, not patient centric. And so we predicted that there would emerge this new tech stack of horizontal platform companies that are designed to deliver value-based care, longitudinal continuous care models, and treat the patient as a primary end user. And so that did come to bear. You know, the first wave of companies that came out started to build, um, what I would call kind of the low-hanging fruit capabilities in that stack, so the operational infrastructure for billing, things like financial services, CRM type capabilities, um, more kind of administrative workflows. And so, you know, to your question, I think, um, the ne- next wave of innovation is gonna be more on the kind of the higher levels o- of, uh, of sophistication of that stack on the clinical side as well as on kind of the risk management side. Um, and so what I, what I mean by that is on the clinical side, you know, I think we're all waiting for the promise of things like AI and ML to help supercharge the way that clinicians make decisions about patients and enable them to access in real time, you know, all of the intelligence that is available to us as society about what works and what doesn't work for a given patient population versus being beholden to, you know, the very heterogeneous distribution of knowledge and experience that inherently exists in any provider population. So I think the clinical aspect of that tech stack is, uh, one area that we think is very ripe for innovation. Um, and then I think the risk piece, uh, as we were, um, sort of discussing earlier, is, you know, providers are basically gonna have to learn new skills. If they're gonna be taking risks on these kind of budget-based payment models, they have to both learn how to deal with payers in a different way, the insurance companies, and, um, be able to design and negotiate these contracts in a way that they will be successful, um, and also, again, understand their populations in a fundamentally different way to avoid these, you know, clinical catastrophes that could lead to insolvency under these models. And so I think that's another area where things like actuarial, you know, capabilities, um, ways to underwrite risk and proactively, uh, intervene in populations at, at a, in a scalable fashion, um, is gonna be another area where we, we need, um, you know, to see innovation within this tech stack.
- SSSteph Smith
You know, something you touched on there is the focus in the past on provider-based care versus patient-based care. Do you see that changing, as in do you see healthcare becoming more of a direct to consumer model, or do you think that, again, using the term inertia that we've had in the existing system where care is focused on a middle layer of the provider, do you see that changing?
- JYJulie Yoo
Yeah. I mean, we actually have written a couple of pieces recently about this notion of our definition of direct to consumer care. I think the traditional, um, you know, sort of, uh, common, uh, interpretation of that phrase in healthcare has been, you know, out-of-pocket payments where you as a patient are paying for a service. Um, we argue that the definition is much broader than that. It's really any service in which the patient has a direct, uh, loyalty, you know, to an entity that is either delivering a service and/or facilitating payments, but not necessarily one that they're paying for out of pocket. So in that sense, uh, absolutely yes. Um, and I think it's a necessary component of value-based care that the patient be engaged in their healthcare because that is really the predominant way by which we're able to catch risk early. And so, you know, you hear about things like remote patient monitoring or continuous measurement where you're putting devices into people's homes and, you know, they are participating in the creation of data sets that are novel and also create a whole new dimension of visibility into how patients are doing. And I think that's one, you know, very, um, kind of prevailing example of how patients are taking more accountability and just more participation, um, in these, in these care models. So, so I would, yeah, I would definitely expand the definition of direct to consumer beyond what we t- we sort of typically think about in the kind of the e-commerce sense, um, to one in which patients are aware of, you know, the services that they're receiving and contributing, you know, novel insights and data sets to, uh, ensure that their, their providers can effectively be successful in these, in these kind of value-based orientations.
- SSSteph Smith
I've also heard you speak about the increasing affinity that consumers sometimes have with some of these companies that are building more of a relationship, even if the consumer is not the direct payer for that service. Could you speak to maybe just one or two examples of companies that have been able to facilitate this successfully?
- JYJulie Yoo
The first company that comes to mind is Firefly Health, which is a value-based care company, so they, um, you know, a lot of the examples that I mentioned earlier were in the government-sponsored insurance space, Medicare, Medicaid, et cetera. Um, one of the areas that we think is, is extremely ripe for value-based care innovation is the commercial, um, insurance space, so, uh, employer-sponsored insurance in particular in this case. And so Firefly effectively sells, um, a solution to employers that helps employers manage their healthcare expenses, uh, in a much more effective way. Um, and as all of us know, uh, insurance premiums continue to just skyrocket year over year and, um, especially given the macro environment right now. Um, you know, there's, this is just a, a burning platform issue for CFOs and CEOs across the board. So Firefly is kind of ins- inserting itself into that equation. But the way that they, um, manage their, their patient populations is through a very concierge type model where you can literally text, you know, their staff on a, a real-time basis twenty-four/seven and be able to, um, have clinically meaningful encounters with that staff about everything from appointment, you know, scheduling to pharmacy needs to just questions about whether I need to take my kid into the, uh, clinic or not, uh, and everything in between. And, um, the phenomenal stat that I always high- uh, highlight for Firefly... Actually, I'll ask you this question, Steph. How many, how, so you, you know, all of us kind of typically see your primary care doctor maybe once a year at most. Um, how many annual encounters do you think, clinically meaningful encounters do you think Firefly has on average with its patient population?
- SSSteph Smith
Well, now I feel like it's high, [laughs] 'cause I feel like it's more than one. But, um, this is per patient, you said?
- JYJulie Yoo
Yeah.
- SSSteph Smith
Um, I'm gonna guess-
- JYJulie Yoo
Take a wild guess
- SSSteph Smith
... three? [chuckles]
- JYJulie Yoo
So 45.
- SSSteph Smith
What?
- JYJulie Yoo
Um, 45 clinc- cl- clinically meaningful interactions that they have on an annual basis with their patient population. These are, you know, commercially insured, right? So these are people who have full-time jobs where they're receiving healthcare benefits. So typically, you're speaking about a much more healthy population than you would see in, like, the Medicare or Medicaid populations. And so because of that context, it's even more phenomenal that they're getting that level of engagement. But it's, um... You know, they, they use th- this text messaging modality. They do proactive outreach to their patients. They also provide remote monitoring devices. They will send providers into the home when there needs to be kind of hands-on on the body to do some kind of assessment or treatment. So, um, because they're taking this multimodal, uh, approach to delivering primary care, they're able to earn the right, you know, to, um, have these, these very frequent touchpoints with their member population, which not only helps with kind of just consumer experience, frankly, and, uh, uh, the development of loyalty towards them, but also obviously from a clinical lens that allows their clinicians to really be proactive about, um, highlighting, you know, opportunities to intervene early in their care journey. So that, that to me is a kind of a primary example of the way that you can implement using technology in really systematic ways, and not just relying on throwing bodies at the problem as traditional providers have done, and, you know, which obviously is inherently unscalable, um, to be able to, um, you know, provide a, a delightful experience to the patient while also delivering clinical outcomes and lower cost.
- SSSteph Smith
That's insane. 45 versus, as I think you said, less than one encounter for most people. And I assume that the price point of that is, is much lower as well through the facilitation of technology.
- JYJulie Yoo
And that's exactly right. The insight here is that only because of the technology that now exists to be able to facilitate multimodal care, meaning virtual care, home-based care, care in your community, as well as, you know, when needed, um, you know, steering you towards in-person care, um, because they're able to diversify the sites of care that, a- across which they're able to deliver their services, the, the cost structure of their clinic versus a tr- a traditional clinic is, is a fraction, right? So that ultimately I would call out is also another key driver and key unlock for making value-based care work, is the ability to fundamentally reduce the cost structure of care delivery, and that I would... You know, I would go as far as to say it's not even possible to do that without the technology that we now have today, whether it be mobile, whether it be these connected devices, um, et cetera. And even just frankly, technology adoption amongst providers, which did not exist until very recently, for better or for worse, and their ability to communicate directly with each other as they're managing collectively, um, these patients in these, in these budget-based, uh, risk models. So, um, you know, that, that is absolutely a fundamental part of the unlock that, you know, Firefly and many other companies have been able to take advantage of.
- SSSteph Smith
I hope you enjoyed this 2-part series. And if you're looking for the full list of 40 plus ideas, you can head over to a16z.com and you'll find it on the homepage. See you next time. [upbeat music] Thanks for listening to the a16z Podcast. If you like this episode, don't forget to subscribe here on YouTube to get our exclusive video content. We'll see you next time. [upbeat music]
Episode duration: 1:12:03
In the uListen app, the transcript lights up word by word as you listen. Get uListen
Transcript of episode m55YESd1HKU
