00:00at the end of 2022 our team at a16z
00:02asked dozens of Partners across the firm
00:04to Spotlight one big idea that startups
00:07in their fields could tackle in 2023
00:09emerging from this exercise came
00:11literally 40 plus builder-worthy
00:13Pursuits for the year and that ranged
00:15from entertainment franchise games to
00:17Precision delivery of medicines small
00:19modular reactors and of course loads of
00:23and in this two-part series we'll be
00:25covering 12 of those big ideas in
00:28today's part two We've Got Big Ideas
00:29spanning fintech American dynamism and
00:31bio and health featuring the voices and
00:33big ideas of Vijay pande Julie U Angela
00:36strange Anisha charya Ryan McIntosh and
00:39Michelle Vols as a reminder the content
00:43here is for informational purposes only
00:44should not be taken as legal business
00:46tax or investment advice or be used to
00:49evaluate any investment or security and
00:51is not directed at any investors or
00:53potential investors in any accz fund for
00:56more details please see a16z.com
01:08carrying over some of the themes around
01:10AI from part one let's start with a
01:12niche hi everyone I'm Anish from
01:14Andresen Horowitz I'm a GP on the
01:16fintech team and my big prediction is
01:19around GPT applied to credit counseling
01:21and that it's going to really unlock
01:22credit counseling at scale so here's the
01:25idea open AI makers of chat GPT has
01:29created this amazing new technology that
01:31you've all seen all over Twitter that's
01:32doing all these remarkable things and
01:34you know one of the things that's been
01:35under discussed is its ability to drive
01:39you product Cycles in fintech and
01:41financial services the one that I've
01:43been thinking about you know based on my
01:44own experience has been how it can
01:46improve credit coaching credit
01:48counseling and advice and the reason for
01:52this is that you know the technology
01:53sort of unlocks a labor Supply at 10x a
01:57lower cost than humans so if the only
02:00cost-effective way to deliver credit
02:02counseling in the past was to subsidize
02:04the human cost us with you know High
02:06fees high prices or providing lead gen
02:10to financial products which weren't that
02:12awesome for consumers we now actually
02:15have a technology that can make it
02:16possible for companies to offer the same
02:18quality of service without monetizing
02:22and really the the problem that's being
02:25solved here is that credit counseling
02:28and coaching is something that is a
02:31sort of expert system that no UI is able
02:35or has so far been able to perfectly
02:37capture it's really hard to build a UI
02:40that captures all the Nuance of you know
02:42having a 15-minute discussion with
02:44someone that really understands how the
02:45credit system works and can help guide
02:47you into how to improve your credit like
02:49that 15 minutes is invaluable and so far
02:51no user interface you know or technology
02:54has been able to perfectly capture it
02:55and with credit GPT you can really
02:58deliver that service and deliver it at
03:00scale and I think it's going to change a
03:02lot of people's credit and really change
03:04Financial Services more broadly as you
03:06know potentially the entire country and
03:09many people who engage with financial
03:11services become more credit worthy so
03:13Niche this is really interesting to
03:15really set the stage how big is this
03:17market and let's break that down not
03:18just in terms of the people who are
03:20currently using credit counseling but
03:22also the the folks who maybe could
03:23benefit from it but could not afford it
03:25in the past for sure so credit
03:27counseling there's a few ways to think
03:29about the market size there's no perfect
03:30answer you know there's the sort of what
03:32I call the dark heart of credit
03:34counseling which is you know you go to
03:36some mom and pop diner in some town in
03:40America and you see a stack of business
03:42cards for someone that promises to you
03:45know fix or improve your credit in just
03:47a week or a month and usually those
03:49people you know do it through some shady
03:51techniques and charge really high high
03:53prices surrounding that are you know the
03:56technology companies that really took
03:57the first swing at this and these
03:59companies were were successful I mean if
04:01you look at a company like Credit Karma
04:02they required for seven billion dollars
04:04and you know they've provided credit
04:06Improvement and credit coaching to over
04:09a hundred million people so that said if
04:11you take a look at the actual credit
04:13impact in terms of how many people
04:16scores got how much better of a service
04:18like that only a small subset of people
04:20that use Credit Karma actually saw
04:22dramatic like uplift in their scores if
04:24you think at the most broad level you
04:26know really every person in this country
04:31um which is the idea that you know that
04:33they're sort of seen as being more risky
04:35than they actually are because they
04:37don't know how to play the credit game
04:40um that's an enormous number of people
04:42certainly more than tens of millions of
04:44people who could potentially benefit
04:46from this and look lenders and creditors
04:48also stand to benefit as the people they
04:51lend to make better decisions and manage
04:53their money more effectively yeah I like
04:55that Framing and I I can definitely see
04:57how this could be a wonderful
04:58democratizing tool for those who can't
05:00access it today but I also have to think
05:04about how Finance as a space is as you
05:07could say less fault tolerant than maybe
05:09some other spaces that AI is currently
05:10being applied to so just as an example
05:13if I enter into chat gbt hey I'm looking
05:15for a recipe on how to make rosemary
05:17chicken well if that recipe tells me to
05:19add a little more salt then desired I'm
05:24however if I'm asking for some of this
05:27credit counseling through an AI and it
05:29tells me to allocate my money in
05:30ineffective ways well that maybe again
05:33is something that is less desirable and
05:36something that people will have less of
05:38an appetite to withstand in terms of the
05:41error rate and so how do you think about
05:43that and as kind of tacking on to that
05:46how do you think regulation might play a
05:48role the truth is that credit coaching
05:51and most most Financial advice is
05:52actually less subjective than you might
05:54think so the amount of salt to put in a
05:56recipe is a little subjective you know a
05:58certain chef may like a little more a
06:00certain chef may like a little less
06:02whereas there's an answer albeit one
06:04that may be moderately complex to derive
06:06as to how to optimize any given person's
06:09credit and finances and I think that you
06:13know these large language model driven
06:15systems will perform I know they will
06:17perform very well against that problem
06:18set it's it's a it's a sort of closed
06:20domain answerable problem I do Wonder
06:22though do you think that there's going
06:23to be a couple errors that actually
06:28get attention that therefore prohibit
06:30the ability for people to really
06:31implement this at scale do you know what
06:33I mean like the people who are the first
06:36movers on this it feels like you need to
06:39get it really right at the beginning
06:40yeah I think that's right look I think
06:42that there are a subset of challenges
06:45within this space and a subset of
06:46customers you know who either
06:48desperately need the service and are
06:50getting it so wrong that you know any
06:53rational advice would be better than
06:54nothing or who have a set of problems
06:57that are actually you know very boxed
07:00and very easy to address so yeah look I
07:02do think that you need to have an
07:03incremental approach to this but I think
07:05the wrong reaction is that you know
07:07because this is actually coming from a
07:09piece of technology instead of from you
07:11know a subjective judgment driven human
07:14it's probably going to be wrong I think
07:16it's probably going to be much much more
07:18right you know something else that comes
07:19to mind is the incentives at play here
07:22so some people might say that these
07:25financial services companies they make
07:26money on these people with bad credit
07:28and perhaps more money than the people
07:31who have good credit and so maybe
07:32there's an incentive for them not to
07:34want to put this technology into place
07:36so what do you have to say to that I
07:38disagree I think that a lot of financial
07:40services companies charge the prices
07:42they need to charge in order to turn a
07:44profit to every customer based on their
07:47respective credit worthiness you know I
07:49think the painting of some companies you
07:52know companies as good or evil is just a
07:54fundamentally incorrect framing um you
07:56know they have to turn a profit they
07:58charge a price and the price they charge
08:00depends on the default rate of the
08:02consumer if that default rate goes down
08:04they can charge a lower price and uh and
08:07receive the same profit and more
08:09importantly they're subject to less
08:10volatility in markets like this when
08:12sort of my market cycle is changing and
08:15their business or book can get blown out
08:17by a sudden shift in consumer credit
08:19worthiness that's a good way to put it I
08:21think another framing I like from your
08:23big idea was this idea of just unlocking
08:26a 10x cheaper labor Supply and so where
08:29else in fintech do you think this maybe
08:30could apply and specifically looking to
08:322023 do you think there's any
08:34low-hanging fruit there's lots of
08:36low-hanging fruit I mean the most
08:38low-hanging fruit is anything that
08:39requires you know customer support
08:42I think one of the the most decadent
08:44features of getting to work with a
08:48um or being wealthy is the ability to
08:50email you know our phone to have your
08:52problem solved that way
08:54and that's going to get unlocked for
08:56everyone you know the having to use an
08:58app that doesn't quite doesn't have a
09:00field that you can or a drop down that
09:02can answer the problem that you have on
09:04that day with your bank is a very
09:05frustrating experience as is calling
09:07into a call center and waiting forever
09:09and that's going to be a thing of the
09:11past I think that everyone is going to
09:12have a private bank like set of
09:15interactions or at least interface to
09:17their interactions with their bank so
09:19that's one wealth management I think is
09:21another wealth management is tricky
09:22because it's one part counseling you
09:25know one part advice and it's just hard
09:27for people to for obvious reasons build
09:29empathy with the system so the therapist
09:31aspect is going to be difficult but the
09:33advice aspect is is going to be
09:35commoditized well I'm glad you brought
09:37up the therapist element because it's
09:39been fascinating to see how even
09:40companies like replica people are
09:43forging forming real relationships with
09:46technology sometimes it's at the
09:48Forefront where they know that it's an
09:50AI who's responding and interacting with
09:53them and sometimes it's invisible and so
09:55that's that's going to be fascinating to
09:57watch totally it's really cool I mean if
09:59you read if you read the newspaper
10:01everyone's very mad at technology if you
10:03look at the NPS of Google Apple Facebook
10:06all of these companies you know it's not
10:09just better than the MPS of you know
10:12their financial services companies and
10:15you know most media outlets and I mean
10:17if they're amongst the most beloved
10:19companies in the world so the idea that
10:21consumers would adopt this at scale
10:23despite the fact that their mental model
10:25have to may have to change around things
10:26like therapy in the context of financial
10:28advice isn't crazy next up we have
10:31Angela hi I'm Angela strange I'm a
10:33general partner on the fintech team here
10:35at Andreessen Horowitz and I believe
10:37that in 2023 doing compliance well Tech
10:41enabled is going to become a competitive
10:43Advantage versus an annoying thing that
10:45you have to do in the background
10:47so if software is eating the world it
10:49has not yet taken a big enough bite out
10:51of compliance post Dodd-Frank financial
10:53services companies face more than 50 000
10:56regulations across dozens of federal and
10:59state agencies and that's just in the
11:02the existing and very manual compliance
11:05policies and risk processes are failing
11:07at both large financial institutions and
11:10at fintech startups supported by sponsor
11:13furthermore while compliance is complex
11:15for businesses operating in just one
11:17geography it is even more difficult to
11:19manage across multiple countries and as
11:22more global companies embed fintech the
11:24need for Global compliance and risk
11:25infrastructure is increasing
11:28in 2023 companies of all sizes will turn
11:31to software to solve their challenges we
11:34expect to see more tools for sponsor
11:35Banks to manage third parties for
11:37fintech companies and companies
11:38embedding Financial Services to manage
11:40all aspects of risk and compliance and
11:43importantly more compliance
11:44infrastructure serving default global
11:46companies all right Angela you mentioned
11:4850 000 regulations that is a ton and I
11:52want to give our listeners a sense of
11:53the weight that this puts on companies
11:55Without Really commenting on whether
11:56this is a good or bad thing what kind of
11:58operational or economic again using the
12:01word weight do these regulations place
12:02on fintech firms like is it one percent
12:05of Opex 10 50 that they're having to
12:08allocate resources to in order to make
12:10sure that they're compliant compliance
12:12is both very expensive and the existing
12:16Solutions don't work very well as you
12:18can imagine the numbers vary widely but
12:20you know there's large banking surveys
12:22that go out every year estimates are
12:23between six to ten percent of banks
12:26revenue of their entire revenue is spent
12:30and they're spending all of this money
12:32and despite that you know we feel like
12:34we read a different headline every week
12:36about fines like for instance since the
12:39last financial crisis in 2008 there's
12:41been more than 250 billion dollars in
12:43fines for poor compliance procedures
12:46and if you look at like how does this
12:48actually impact operations right and so
12:51it's bad and it's getting increasingly
12:53bad like some of the larger Banks if you
12:55you read their annual reports 10 years
12:57ago four percent of their employees were
12:59in the compliance and risk functions and
13:01now it's up to 15 but also like you look
13:04at this and you look from our
13:05perspective like this is a huge
13:06opportunity but first it's helpful to
13:09understand why and primarily three
13:12reasons one that everybody points to is
13:15the Dodd-Frank Act that came in after
13:16the last Financial regulation right and
13:18so there used to be 30 000 regulations
13:21across a variety of uh of you know state
13:23and federal agencies and in a very short
13:26period of time that ballooned up to 50
13:28000 but this happened very quickly and
13:31in many cases didn't have enough time to
13:33become very clear which points to
13:36problem two right if you had compliance
13:40procedures that were working well in the
13:41old world and there was some software
13:43and many manual procedures and all of a
13:46sudden that complex actually doubled
13:48your software was just not set up to be
13:51modular accommodate new rules and so
13:53what did banks have to do they often
13:56just threw a lot of people at the
13:57program problem and so you talk to some
13:59of the large Banks and they have
14:00thousands of people often sitting
14:02offshore that are monitoring AML alerts
14:06ofac sanctions rulings so it's just a
14:09very very people-heavy business
14:12and then the third reason which is
14:14actually I think great for the
14:15industries the industry has evolved a
14:17ton in the last 10 years not that long
14:19ago all Financial Services was done by
14:21walking into a bank I'm fond of saying
14:24that every company is going to be a
14:25fintech company now we get our financial
14:27services from all sorts of different
14:28software companies we're way more Global
14:30we pay in different ways and there's
14:32lots different types of regulations that
14:34come along with that that all financial
14:36services are having to adopt to
14:38absolutely I mean you sent me an article
14:40where recently coinbase was fined 100
14:43million dollars for background check
14:45failures and coinbase is a large public
14:48company they're heavily capitalized and
14:50they're struggling with these compliance
14:52needs if they can't keep up how are
14:54smaller startups even going to navigate
14:56this listen and Coin and this is like
14:58coinbase takes compliance very seriously
15:01and this compliance had had nothing to
15:03do with crypto it was simply you know
15:07all Financial Services institutions
15:09have an obligation to kyc like
15:12understand who their customers are as
15:15they go through the onboarding process
15:17so imagine any company right like you're
15:19set up you've got your systems working
15:20well and in their case what happened is
15:24prices were rising in cryptocurrencies
15:26they're a like regulated uh
15:29cryptocurrency company in the US they do
15:31a very good job of what they do and all
15:33of a sudden they had 25 times more
15:36monthly transactions like how do systems
15:39that are not completely software based
15:40actually scale to that
15:42and so you ask like how do small
15:45companies deal with this in some ways
15:47companies that are starting now have
15:50have a little bit of an advantage right
15:51because if you're starting something de
15:53novo versus you're trying to fix systems
15:56that you already have in place
15:58you know what you're going to be dealing
16:00and then two and this is the pieces
16:01exciting from an investment perspective
16:03the technology in the space has come a
16:05long way as have the teams that deeply
16:08understand the tech and really
16:09understand the compliance procedures
16:14um if you kyc somebody and you need to
16:16do what's called enhanced due diligence
16:18often you need to actually check you
16:20know a photo of their physical ID and
16:22there's lots of companies that have been
16:24doing this for a long time you might
16:26expect that they'd all be automated
16:27oftentimes they're sending pictures of
16:29your ID off to people who are manually
16:32checking is this ID real or not right
16:35like that doesn't scale now image
16:37recognition technologies have gotten
16:38much much better that technology can
16:40solve this type of problem and we're
16:43seeing this applied across many
16:44different areas of compliance well I'm
16:46glad you mentioned automation because
16:48something that really stood out to me in
16:50the article about coinbase was that they
16:52had a backlog of a hundred thousand
16:54transaction monitoring alerts and 14 000
16:56users requiring extra due diligence and
17:00so that is just an unwieldy amount of
17:02information for a company to pour
17:04through and it seems like again even a
17:06heavily capitalized company
17:09it's just not realistic and so I'm
17:11trying to understand whether these
17:12things can be automated with technology
17:14whether they can be actioned or
17:16regulated through code or if we're kind
17:19of resigned to this idea that we hire a
17:21bunch of offshore assistants to pour
17:24through the data manually like what what
17:26are we seeing in terms of the
17:27introduction of Technology into this
17:30space well if you ask chat gbt that
17:32um they will basically say surprise AI
17:35but let me tell you a little more
17:36specifically let's take any money
17:37laundering for instance right and
17:39there's there's again billions of
17:41dollars that are thrown at this problem
17:42depending on what estimation you believe
17:44it's only three percent of dollars that
17:47are actually captured there's two
17:48trillion dollars that are still
17:50laundered every year
17:51and so you look at the technology that's
17:54behind this and it's a very manual based
17:58alert systems so for instance anyone
18:00that does a transaction over ten
18:01thousand dollars you need to flag it
18:03somebody on the back end has to go check
18:05you know who exactly is this customer is
18:08this um you know is this laundering or
18:10not and so it's rule-based systems that
18:14as we've seen with other companies just
18:16don't scale now with the introduction
18:19with you know Ai and machine learning
18:22you can do things like uh two things one
18:25just better gather a little customer
18:27information that sits all across the
18:28bank thus vastly speeding the manual
18:31review and then two dramatically
18:33reducing the number of manual reviews
18:35that you actually have to do because
18:37these systems can learn based on what
18:39compliance officers are doing and so
18:41throw up far fewer false positives such
18:44that if it does need to go to manual
18:45review it's a much more reasonable
18:48number of things that a human could
18:50actually handle and I think that's just
18:51one simple example you can apply you
18:54know this and other Technologies across
18:56just large large areas across all
18:58financial services so when we're talking
19:00about these different regulations again
19:02we mentioned 50 000 of them and they
19:05span different types of Regulation as
19:07well and so when I think about companies
19:09in these spaces that are helping other
19:11small or large companies navigate this
19:14ecosystem deal comes to mind for
19:16cross-border compliance with respect to
19:18payroll sardine for kyc which we've
19:21already mentioned in AMR now are there
19:23other gaps that you see in this in this
19:25I'm going to call it this compliance web
19:27because it really does seem hard to
19:29navigate that you don't really see
19:31companies effectively servicing any
19:34industry that is hard to navigate has
19:36spawned a very large industry of
19:38necessary Consultants right and you talk
19:43um you know larger financial
19:44institutions earlier startups it's now
19:46very top of Mind people want to do this
19:48well and then they just see a web of
19:51vendors and what do you do and so I'd
19:54say there's there's really well there's
19:56many opportunities but three I would
19:58highlight one traditionally compliance
20:01has been very siled like you think you
20:04start with know your customer and then
20:06you get the customer on you've got to
20:07monitor their transactions and then you
20:09might have some separate system for
20:10fraud and none of these necessarily
20:13speak to each other you have all these
20:14different types of vendors you have new
20:16data sources that might come on board
20:17how do you integrate those and so now
20:19there's there's new companies you
20:21mentioned sardine that are really
20:23looking at this from a like fraud has a
20:26lot to do with who you on board has a
20:28lot to do with the transactions from a
20:30holistic perspective and just better
20:32feeding this data into a circle and
20:34making it much more modular to integrate
20:37new types of data sources so just
20:39looking at this problem from a holistic
20:41point of view opportunity two you know
20:43we we talked a little bit of coinbase
20:46um just in the U.S more than 10 percent
20:48of people have a crypto wallet and so
20:52does it make sense that you would have
20:54entirely separate on-chain compliance
20:57systems from the you know off-chain
21:00world if oftentimes the people are the
21:03same and so that's an opportunity to
21:05bring more data into the sentence to
21:07improve compliance and then the third
21:09part I'm pretty interested in uh which
21:12has a a coordination challenge but
21:15everyone that deals with financial
21:16services is sitting on often very siled
21:20pieces of data so for instance if I
21:21suspect as a bank that one of my
21:23customers is a money launderer and I
21:26want to get information from another
21:27bank to be able to check that out
21:29there's a long compliance procedure
21:33um and able to do that so just how do we
21:36better enable data sharing which can
21:38help Elevate every financial institution
21:41in doing a better job in compliance yeah
21:43I'm going through this right now as a
21:45customer of several Banks and financial
21:47services firms where I'm changing my
21:49residency and I'm having to go through
21:51the exact same very lengthy very arduous
21:53process with every single one and of
21:55course there's privacy concerns about
21:56what information is shared between these
21:58different entities but again as a
22:00consumer I'm experiencing this problem
22:02right now one thing I want to ask you
22:04about as we close this out is how
22:06compliance relates to competition what
22:09role it plays there and as we're seeing
22:11these increased number of regulations
22:13whether companies are choosing to go to
22:16other jurisdictions where maybe this
22:17isn't the case you know we've invested
22:19in in many global companies as like
22:21every country has pretty significant
22:26um and they're they're very different
22:28and I so I think from a just how does
22:31this look going forwards I think there's
22:33opportunities for new compliance
22:35companies to just help keep up with what
22:38are all the different changing
22:39regulations in all sorts of different
22:40jurisdictions right so if I'm a company
22:42in the US and I want to expand to Brazil
22:44I have to learn an entirely new
22:46different compliance regime wouldn't it
22:48be great if I had an infrastructure
22:49provider that would help make sure that
22:52I was compliant and help stay
22:53continuously compliant so I think that
22:55the opportunities here are really
22:57twofold one obviously new companies that
23:00help rethink this you know highly manual
23:03procedure in different jurisdictions in
23:05different ways and then two
23:08I think new companies that you know
23:09aren't compliance companies but need to
23:11be compliant are going to think about
23:14what are their compliance systems and
23:16what are the what is their approach to
23:17compliance very much from day one in the
23:20same way that they think about many
23:21other elements of their infrastructure
23:23and they get into their strategy to one
23:27often provide a better customer
23:28experience but then two keep these costs
23:31really down to the thing to spend their
23:32profits on other things
23:34yeah I mean coming back to what you
23:36mentioned before 10 of Revenue that's
23:39that's a lot that's not 10 of profit
23:41um so I think this will be top of think
23:43of all the new products you could be
23:44building with 10 of your Revenue next up
23:47we have Michelle hi I'm Michelle Vols
23:49I'm a partner on the American dynamism
23:51team and this is my big idea for 2023.
23:54small modular reactors Advance the
23:58though nuclear energy accounts for 20 of
24:01the US's electricity it's commonly
24:03misconceived as a dangerous and
24:05non-viable option when it comes to
24:07adding reliable sources of carbon free
24:10but recently nuclear energy has been
24:13having a bit of a Renaissance with the
24:16inflation reduction act earmarking 30
24:18billion dollars for tax credits towards
24:20existing nuclear reactors a first for
24:23the U.S the timing is right now to usher
24:27in new innovation in this space
24:30there are opportunities across the
24:32nuclear supply chain from fuel sources
24:34to mining to manufacturing vendors and
24:37Beyond and one area I am particularly
24:41excited about is small modular reactors
24:43or smrs by leveraging Advanced
24:47manufacturing techniques and modular
24:48design smrs can be quickly and
24:51efficiently mass-produced bringing down
24:53costs significantly this can make
24:55nuclear energy more accessible for a
24:57variety of applications including
24:59providing clean and reliable power to
25:01remote communities or even one day in
25:04while there's still a way to go in
25:06reforming the regulatory Frameworks for
25:08these types of reactors smrs in the
25:10broader nuclear industry are likely
25:12poised for growth in the year ahead so I
25:14feel like energy has been top of mind
25:16with some of the geopolitical shifts
25:18that have happened especially in the
25:19last year but I feel like within that
25:22conversation nuclear has become a more
25:25prominent topic but at least in my world
25:27not smrs so can you go a little more
25:30deeply into what smrs are and maybe how
25:33they differentiate from a traditional
25:34nuclear plant whether it's the
25:36difference in size cost time it takes to
25:38actually build smrs or small modular
25:41reactors are basically exactly what
25:43they're called they're small and modular
25:46reactors and what that does is that
25:48means they can be deployed in places
25:50that traditional nuclear plants can they
25:54won't ever provide the scale that like a
25:57traditional nuclear power plant would
25:58but they require significantly less
26:01Capital expenditures and can be
26:04essentially shipped to wherever they're
26:05needed so it could be like remote
26:07communities or on ships or in space and
26:13and because of their size and because of
26:15the timeline it just like unlocks a lot
26:17of new opportunities within a time frame
26:20that's actually reasonable so we could
26:21get this within years instead of within
26:23decades can we get specific on those
26:25years like how long does it typically
26:27take to set up a traditional nuclear
26:30power plant versus an SMR the bulk of
26:33the time in a traditional nuclear power
26:34plant is actually just like getting the
26:36site location going through all the
26:38specs getting all the designs approved
26:41and that like that just takes a ton of
26:46um there's just a lot of hold ups and we
26:48haven't seen a new nuclear plant get
26:55we think it can be done in years like
26:57there's still regulatory things to get
26:59through because this hasn't been done
27:01yet there hasn't been a like design
27:03framework on the regulatory side to to
27:06say like this is how they should be
27:07built this is like the timeline
27:09um but with there are a number of
27:10startups like actually building today
27:13and getting designs approved and um
27:15starting starting on the path to get
27:20we don't know exactly what the timelines
27:22are but I think we are excited in what
27:26could happen in the next couple years or
27:29before 2030. that's exciting and I like
27:31the framing of decades versus years
27:34something that you just spoke to there
27:36which has come up in my very limited
27:38Research into this space is regulation
27:40it seems like the time it takes to get
27:44all of the necessary approvals to build
27:47a plant is astronomical or at least more
27:50than I think most people would think and
27:52it sounds like maybe there's some
27:54regulatory reform happening you also
27:56mentioned the IRA which is allocating
27:58funds towards nuclear what should people
28:00be looking for here in terms of Reform
28:02like what kind of Reform do we need
28:04whether it's with respect to the
28:06traditional nuclear power plant or these
28:07smrs which sound like they're a little
28:09newer and maybe their form is less
28:11Concrete in the U.S the nuclear
28:14Regulatory Commission or the NRC is the
28:17the body that would approve any new
28:20um or or nuclear reactor and it's often
28:23cited as one of the top bottlenecks in
28:25progress towards building more reactors
28:28there are super high fees like million
28:31multi-million dollars just to get
28:32designs looked at and approved there's
28:35very long timelines to go through the
28:37end-to-end approval process at all
28:39there's disputed Frameworks for how it
28:43regulates like radiation which slows
28:45things down even further
28:48um and as I mentioned before like
28:49critically there's a lack of the
28:51regulatory framework for smrs in
28:54particular like they haven't really been
28:56built before so there are no guidelines
28:58yet lowering the fees to approve the
29:01designs and accelerating the timelines
29:03overall are both the I think the
29:05essential steps if we want to see more
29:08smrs come into production within the
29:11yeah I was really surprised to hear that
29:14in the case of some of these regulations
29:17like there isn't actually even a
29:18threshold of the appropriate amount of
29:20radiation which is allowed it's just
29:23kind of this vague sentiment of do
29:25whatever you can to limit radiation
29:27which is not something that many people
29:30can action off of are there examples of
29:32other countries that have maybe set up
29:34more legible maybe more actionable
29:38legal Frameworks or regulatory
29:40Frameworks with nuclear and what have we
29:42seen in terms of those other countries
29:44in terms of maybe what we can learn so
29:46other countries are are kind of at
29:48various stages of development and
29:50support for for smrs in particular for
29:52nuclear overall we've seen some progress
29:56in Europe towards approving like
29:58building sites more I know China's
30:00putting a lot of money in into nuclear
30:02Innovation but but in but at the end of
30:05the day the countries that are lowering
30:07the barriers towards actually like
30:08building and testing these reactors and
30:11getting plans to start development are
30:15the ones that we should be looking
30:16towards and and following along with and
30:19overall like International coordination
30:22is going to be really critical here in
30:25um once something gets approved in one
30:27country it should help lessen the burden
30:29in other countries and so um there are
30:32some organizations kind of working
30:34towards coordinating across countries
30:36and internationally but I think think
30:38that will be something we should look
30:41towards and and really participate in in
30:44as people listen to this there may be
30:46potential Founders some of them won't
30:48want to be the entity that goes and sets
30:51up a whole nother nuclear plant or an
30:53SMR but something you spoke to in your
30:56big idea was this idea of the nuclear
30:58supply chain and all of these other
31:00businesses that kind of ladder up to the
31:02end product whether that's mining or
31:04manufacturing can you tease this a
31:06little bit more in terms of
31:08the different types of businesses are
31:10things or gaps that people might explore
31:12if they're actually looking to be a part
31:13of this industry in 2023 as we hopefully
31:17start building more reactors there's
31:19going to be material needs and things
31:21like uranium or the cooling materials
31:23depending on how the smrs need to be
31:26cooled and built there'll be needs
31:28around the manufacturing and transport
31:30of the reactors they're small modular so
31:32you can move them around but there's
31:34still some things to Think Through of
31:36how to move this sensitive material and
31:40then there'll be needs around like
31:41disposing of nuclear waste and thinking
31:44through the recycling process maybe on
31:46the more exciting side there's
31:48potentially new areas unlocked like
31:50decommissioning or repurposing other
31:52types of energy plants like coal plants
31:57um and so I think there's going to be
31:58tons of opportunities across the whole
32:00chain and even like software to
32:02coordinate all of this or to track and
32:04monitor emissions and things like that
32:05on that note nuclear
32:08we're currently talking about it in the
32:10stance of fission which people are more
32:12familiar with but there have been other
32:14advancements in Fusion which I think is
32:16a little further out I don't want to get
32:17too ahead of our skis in terms of
32:19talking about its economic viability
32:22um but 2022 was kind of this exciting
32:25year where at least one group was able
32:28to demonstrate net energy gain and I
32:30think 2021 was a banner year in terms of
32:32investment into nuclear fusion companies
32:35and so although these are two very
32:37different things in terms of where
32:39they're at in terms of their ability to
32:41be implemented today do you have any
32:43thoughts around you know where that's
32:45going what Fusion's role is in this
32:47wider Matrix or how we might think about
32:48that so I think the overall energy needs
32:51over the next few decades and Beyond are
32:54big enough that there can be many
32:55solutions and I would love to see more
32:58progress on Fusion I think it could be
33:00something that's absolutely game
33:01changing if we can get there
33:03unfortunately I think it is further away
33:05than we'd like in that people people are
33:08hoping especially if you think about any
33:11anything close to a scalable solution or
33:15something that could be powering an
33:16entire power plant one of the main
33:19reasons I'm so excited about smrs is
33:21that we actually aren't waiting on any
33:23new scientific breakthroughs and these
33:25things are getting slated to be built in
33:28the near term again on the order of
33:30years we could see these in production
33:33and actually producing the energy we
33:37um while Fusion is very exciting and I
33:40definitely hope scientists keep working
33:42in that area I think in the near term
33:46it's this vision and it's the the
33:49smaller modularity that's going to be
33:51the breakthroughs we see I feel like I
33:54have to ask this question after you
33:55brought up this idea that the technology
33:56already exists if we look out let's say
34:00five ten years from now and we're still
34:02at the place where nuclear it has not
34:04really made any advancements it's still
34:06taking decades to create our plants we
34:08still don't have the right regulatory
34:09infrastructure in place like what would
34:12cause that what is is causing us to be
34:14at kind of this like friction-filled
34:16spot in the Arc of history as it relates
34:20to nuclear I think a lot of it is on the
34:23regulatory side like we need to decide
34:27um as a country that we want to invest
34:30here and we want to push this forward
34:32it's not going to be about were we able
34:35to build it it's did we decide that we
34:37wanted to invest in this did we say that
34:40we wanted nuclear the biggest barrier is
34:42just humans not not the Innovation just
34:45us getting in our own way next up we
34:48have Ryan hi I'm Ryan McIntosh I'm a
34:51partner on the American dynamism team
34:53overhauling the space supply chain
34:55civilizations ancient past is segmented
34:58by advancements in mining and Metallurgy
35:00the stone bronze and iron ages in each
35:03period success over your Rivals was
35:05determined by your ability to collect
35:07resources and produce increasingly
35:10today we are in the space age and the
35:14space Supremacy will be the measuring
35:16stick of industrial and military power
35:17for the foreseeable future
35:19it's already supporting our digital
35:21economies and guiding our autonomous
35:22systems but the true space industrial
35:25base is much broader we then only need
35:27to construct Advanced rockets and
35:28satellites but gather the materials and
35:31Industrial capacity to do it reliably at
35:34space does not begin at liftoff a
35:37complex Global supply chain for mining
35:39to launch pad must be overhauled and
35:41secured within our Global Alliance
35:44beyond Earth infrastructure must be
35:46built to service existing Assets in
35:47orbit Empower more ambitious missions to
35:51in 2023 the space industrial base will
35:54continue to grow in size birthing
35:56critical companies serving our nation's
35:58at the heart of this Resurgence
36:00inspiring Founders are pursuing
36:02difficult problems in materials
36:03manufacturing and space infrastructure
36:05all right Ryan I feel like that's a
36:08fascinating prediction or big idea but I
36:11feel like we have to start with the why
36:12question on this one why must we build
36:15Advanced rockets and satellites using
36:17your words reliably and at scale I think
36:19most importantly the United States
36:21should be the leader in space this
36:24actually goes back to a JFK quote when
36:26he gave his famous Moon speech at Rice
36:28University where he said no nation which
36:30expects to be the leader of other
36:31nations can expect to stay behind in the
36:34race for space and I think that's
36:36incredibly important not only back then
36:38but today as we have new near peer
36:40Rivals entering the domain why is this
36:43important satellites already Drive much
36:45of our economy you know this
36:46telecommunications infrastructure this
36:48is GPS these are things that are
36:50relevant not only for the commercial
36:51side but for the military side as well
36:53speaking of military side and wartime as
36:56an example of Ukraine we have sort of
36:59near real-time Earth observation and
37:01Signal Intelligence coming from that
37:03this is essentially you know the
37:04ultimate height advantage in war but
37:07even something like climate change
37:08having that Earth observation be able to
37:10see where the minute emissions are
37:12coming from being able to detect weather
37:14and variations in climate it's all
37:16critically important to solving those
37:17problems and perhaps lastly and and
37:20perhaps most long-term there's
37:22essentially infinite resources in space
37:24so a lot of these things we think about
37:26mining or even just accessing energy
37:29there's all of that in space so in the
37:31long run if we want to actually continue
37:33to grow as a species we're eventually
37:34going to have to figure out how to
37:36harvest and utilize space resources I
37:38like that you've Illustrated the why and
37:40I think an important aspect as we
37:42explore the idea of pursuing space more
37:45intensely is as you mentioned the supply
37:47chain a lot of people think about the
37:49last mile there which is actually
37:50launching a rocket into space but
37:52there's a lot that leads up to that and
37:55as we explore that that full space
37:57supply chain as you've said from mining
37:58to Launchpad I want to understand what
38:00part of this chain you think is most at
38:02risk or we could actually reframe that
38:04to say what part of that supply chain
38:06has the most entrepreneurial opportunity
38:08what part of that has the largest gaps
38:11for new critical companies to be created
38:13sure so the way I think about space and
38:17space supply chain holistically is that
38:19space is sort of the well it is you know
38:22the final frontier it's where our
38:24industrial capacity and the stuff that
38:27we can build can actually extend into
38:29and so it's not just you know the
38:32Launchpad it's not just the Rockets go
38:33into space and the satellites but it's
38:35everything else that allows us to do
38:36that and so these are things like mining
38:38so actually getting the resources that
38:40we need to build the stuff and it's not
38:42just the stuff that goes up into space
38:43which is predominantly aluminum so a lot
38:45of our stuff is built out of aluminum
38:47titanium but it's everything else like
38:49actually building the factories the
38:52um even things like energy generation so
38:55how are we actually going to get the
38:56energy to do this stuff and build you
38:58know thousands of satellites or
39:00thousands of rockets these are all
39:01things we need to think about so mining
39:03is the first one and specifically within
39:05mining I look at Midstream refining so
39:08not just mine the materials but turning
39:10them into workable metals and there are
39:12specific metals that are highly relevant
39:14you know some of the battery metals and
39:15stuff for energy storage and generation
39:17but even things like rares that are used
39:20in magnets as well the second one is
39:23actually manufacturing space components
39:26um so this is one of portfolio companies
39:27Apex they're actually building satellite
39:30buses so a lot of companies are looking
39:33at potentially scaling out the space
39:36infrastructure that we have today
39:37whereas today it's it's sort of bespoke
39:42building satellites building whatever
39:45you need if we're going to do thousands
39:47of those and scale it up we need to be
39:48able to develop a mass manufacturing
39:50process and that might require
39:52separating out these vertically
39:53integrated businesses like SpaceX that
39:56does everything and doing each
39:57individual component and building a
39:59separate company that handles just that
40:01last piece is what's called osam so like
40:03onser on orbit servicing and
40:06Manufacturing basically activity that
40:09services this Market of satellites and
40:11space stations if we're going to have 30
40:14000 satellites in orbit we need to
40:16actually have the the sort of companies
40:18that are able to repair them fuel them
40:20service them deorbit them and so so I
40:22think those three areas are the most
40:24immediate opportunities whenever
40:26exploring these space opportunities they
40:29are inherently moonshot opportunities I
40:31mean you take a company like SpaceX and
40:35SpaceX I think has provided a good
40:37example of again a company that has
40:40moonshot ambition to you know make us a
40:43multi-planetary species yet also has
40:45kind of brought that back down to earth
40:47you could say by having a business model
40:49that funds their progress in spacex's
40:52case by providing internet to people on
40:54Earth today so again not just betting on
40:56tomorrow but actually showing that
40:58they're providing value through the
40:59revenue that they're driving towards
41:02that eventual moonshot and so I'm
41:04curious to know if there are examples
41:05that you could provide of other
41:07companies perhaps that have been able to
41:09build that bridge because I think
41:11perhaps one of the reasons that maybe
41:13more Founders aren't pursuing space
41:15companies is due to this moonshot nature
41:18it's kind of like build or bust and I
41:21I'd love to share different examples
41:23where maybe that's not the case well I
41:25think the most poignant example is the
41:26company I mentioned a second ago Apex
41:27where today they're building satellite
41:30buses so they're sort of riding the
41:31curve of everyone's building satellites
41:33everything is getting washed up into
41:35space some companies are going to find
41:37it a lot cheaper to Outsource that have
41:39another company build satellite buses so
41:42they're able to do that today but in the
41:43long run they're going to expand into
41:46building a lot of other stuff that's
41:47going to be used in space so the actual
41:49product is this Matt's manufacturing of
41:52but today they're focusing on what is
41:54immediately addressable which is
41:55satellite buses more broadly there's
41:58sort of two areas that I think are
41:59interesting in terms of how you're
42:03approaching the market the first is what
42:05I call like Jamestown businesses
42:07Jamestown was something I used in my
42:08essay but if you remember historically
42:10Jamestown was the first successful
42:12settlement on the Eastern Seaboard of
42:15the United States the reason they're
42:17successful was they had a competitive
42:18advantage of growing tobacco there and
42:20selling it back in England and Europe
42:22and that's sort of what we've seen with
42:24space particularly in the mature Earth
42:26observation Market where they were able
42:29to put satellites up and they had a
42:30competitive Advantage because they were
42:31the only that's only way to get
42:32satellite imagery and and to handle
42:34these sort of Telecom networks they had
42:37the immediate market for that and they
42:39were able to build that in the long run
42:40those companies will likely expand this
42:42to other domains but how you build that
42:45company in the beginning is service that
42:47the second one is tracking military
42:49contracts then the first sort of uh
42:53customer for a lot of space pushing the
42:55frontier if you will is often government
42:58SpaceX is still largely derives a lot of
43:01revenue from government but looking at
43:03areas like space domain awareness which
43:05essentially is knowing what else is
43:06around you was particular specifically
43:08if it's like a very important satellite
43:10being able to see what's around you
43:12whether that be uh you know orbital
43:14debris or an enemy satellite that's very
43:17important for for space force and for
43:19other government programs also space
43:22tugs so being able to take something
43:24from a lower energy orbit and push it up
43:27higher now there's companies like
43:28Northrop which have proven this what's
43:31called life extension which is
43:32essentially taking a very expensive
43:34satellite and pushing it up a little bit
43:36more so it you know it can it can last
43:38longer in space before it falls to Earth
43:41um and then for Artemis and some of
43:43these programs going to the Moon you're
43:44going to need something to actually take
43:46it once it's in space
43:49um you know to to the Moon uh and then
43:52lastly there's a bunch of areas like
43:53tactical response that have sort of more
43:55military applications
43:57um but looking for the military and
43:59government contracts as sort of the the
44:01pathway for revenue and then when the
44:04commercial opportunities catch up to
44:05that you're going to be there and have
44:07the technology ready to go you'll be
44:09ready to ride that wave I feel like
44:12um I would be remiss if I did not cover
44:14this question as we close this out which
44:17is just you know you in your big idea
44:19kind of framed this idea of space
44:22providing the you could say that this
44:25measuring stick of power
44:27um and I think some people would say no
44:28Ryan that's still on earth right like
44:31access to land water
44:33critical materials like Cobalt that's
44:35really what is going to be the measuring
44:37stick of the different countries or
44:39entities that succeed what would you say
44:41to that I think people don't realize how
44:44important and our critical space already
44:46is to even land battles and Ukraine is a
44:48good example of this where the United
44:50States is able to provide sort of
44:52real-time intelligence of troop
44:53movements uh signal Intelligence being
44:56able to detect everything that's said
44:57every message that's that's sent
45:00um and targeting capabilities so being
45:02able to actually know where people are
45:04then also be able to send missiles to
45:06where they are all of that assets to be
45:08able to provide that infrastructure
45:09takes place in space so even even though
45:12they're they're land-based battles the
45:14infrastructure to facilitate that and to
45:15actually win wars in the future
45:18are required you know these Assets in
45:21space so these are not only important
45:22for winning but also they're one of the
45:25first targets if there was a real
45:26conflict between the United States and
45:27China or Russia they would go out for
45:29the satellites first
45:31um so so that is that is absolutely
45:32critical and also in terms of minerals
45:34mineral Discovery companies like one of
45:36our uh portfolio companies Cobalt they
45:39utilize these data sources from
45:40satellite imagery as well so even in
45:42terms of minerals they're they're highly
45:45um and more broadly looking at resources
45:50um just an insane amount of material
45:52available in asteroids and being able to
45:55actually mine some of these other
45:56celestial objects I think some of the
45:58numbers I've seen were like a small
46:00m-type asteroid might have twice as much
46:02nickel and iron that is produced on
46:05Earth in a single year so a country is
46:07able to actually do this at scale uh
46:09solves a lot of these material component
46:11problems and I think in the long run we
46:14likely see a lot of these once we start
46:16getting a lot of resources from space we
46:18might move other industrial very high
46:20energy intensive activities into space
46:22you don't have to worry about nuclear
46:23radiation if you have a nuclear reactor
46:25in space and you're doing other sort of
46:26CO2 emitting processes in space you
46:29might be able to do a lot of industrial
46:31stuff and then send it back down to
46:32earth not be the long-term plan next up
46:35we have Vijay hi I'm Vijay Pandey
46:37Journal partner of Bion Health at a16z
46:40and this is my biggest idea the biggest
46:42company in the world will be a consumer
46:44health tech company this may sound crazy
46:46to some but why shouldn't this be true
46:48four of the top five biggest companies
46:51in the world are consumer companies and
46:53Healthcare is one of the nation's
46:54biggest industries much larger in fact
46:56than the size of the global advertising
46:58industry in which consumer giants like
46:59Google and meta operate from that
47:02standpoint the number one slot should be
47:03the consumer health company so we see
47:05two paths to consumer health company
47:07becoming the biggest company in the
47:08world the first is a vertically
47:10integrated path of being a so-called pay
47:12fighter a combined payer and provider
47:15that eventually owns most care
47:17imagine the United Health Group but with
47:19the user interface of Apple who wouldn't
47:21choose this insurance plan and provider
47:24the second is a horizontal path of
47:26building a consumer Marketplace or
47:28infrastructure layer that enables all of
47:30their care delivery companies
47:31imagine Amazon a Visa but of healthcare
47:34we'd go as far as say that there is
47:36almost almost infinite room to improve
47:38the consumer experience in healthcare
47:40and build massive companies as a result
47:42and we expect consumer Healthcare to be
47:44front and center in 2023. all right
47:46Vijay this is a big prediction so before
47:50we jump into the nitty-gritty I want to
47:52hear from you just how big is this
47:54healthcare industry yeah actually people
47:56forget just how huge it is it's
47:59approaching 25 of US GDP it's about like
48:0322 right now is 20 in 2020 and it's just
48:07been gradually increasing it's what the
48:10healthcare crisis said we we worry about
48:12and so in that sense it's 10 times
48:14bigger the marketing budget of consumer
48:16giants like Google and meta
48:19well I'm glad you brought them up
48:20because some of these large companies
48:22like Google Apple Facebook and Amazon
48:24are pursuing Healthcare to some degree
48:26and they have so much Capital to pour
48:28into this nascent space for them but to
48:32date they have not succeeded so why do
48:35you think these super heavily
48:36capitalized companies have not come out
48:38on top yet yeah I think there's a couple
48:40different ways to think about it from
48:42purely a tech lens you can imagine the
48:44days of Google fighting Microsoft like
48:47how could something like a little
48:49startup known as Google uh sort of
48:51compete against a giant like Microsoft
48:53and it did because they were web native
48:56uh they were sort of built upon a
48:59different set of rails and with Founders
49:01that think of things differently and so
49:03even just from a tech company
49:04perspective uh it's something that's so
49:06different is hard for Giants to go after
49:09just because of the way they're built
49:10and this is basically classic disruption
49:12Theory Healthcare is also different than
49:14just Tech you can't come in with I think
49:16uh naive understanding of how Healthcare
49:18works either are on the medical side the
49:20last science side or even just how the
49:22system works and come in and make big
49:24changes and so the combination of those
49:25two things I think will make it very
49:27difficult for the existing incumbents at
49:29least on the tech side to dominate yeah
49:31so we've covered the incumbents on the
49:33tech side but there's also large
49:35Healthcare companies out there today one
49:37that you've mentioned is United Health
49:38Group which is the largest Healthcare
49:40company in the world I think it's also
49:42the eighth largest company overall what
49:44do you think stopping them from
49:45dominating in the way that you're
49:47describing well I think just like we
49:49talked about for Tech incumbents I think
49:51there's the same issue now for
49:52healthcare incumbents and that uh the
49:54winner is going to combine a deep
49:56knowledge of tech and a deep knowledge
49:58of health care and it's just as it's
50:00very difficult to sort of build that
50:02into a tech company it's very difficult
50:03to build Tech into a Healthcare Company
50:05the same way uh it's just not the way
50:07these companies are built and so it's
50:09it's the analogous problem on both sides
50:11and so something that you seem to be
50:13alluding to is the fact that we need new
50:15founder is pursuing this problem so if
50:17you were a Founder today what angle
50:19would you tackle this from I think the
50:22key thing is that the founding team
50:23ideally would bring together people that
50:26have this deep knowledge in Tech and in
50:28healthcare I think it's now sitting in
50:302023 we're in a different world where
50:32people have grown up with tech their
50:34whole lives so if you're graduating from
50:36medical school now or if you've been
50:38working the Healthcare System Tech is
50:40not something that is unfamiliar it's
50:41part of your your daily life and has
50:43been part of your daily life now for
50:45decades it's important for us to pay
50:46attention to these inflection points and
50:48so it sounds like technology has changed
50:50in a way where there's new opportunities
50:52in this space but I want to really
50:54drill into this idea that you said that
50:57there's infinite room to improve the
50:58consumer experience in healthcare and I
51:00feel like this has been true for quite
51:02some time and so perhaps in addition to
51:05some of the technical changes or the
51:07talent changes why haven't we seen many
51:09of these developments yet it's actually
51:11analogous to some areas in Tech where
51:13people from a consumer background came
51:16into Enterprise and Enterprise software
51:18started having the look and feel of a
51:21really smooth UI from you'd expect from
51:24meta or from Google or something like
51:25that and similarly it just takes a
51:28different type of founder to really
51:29understand how to create that experience
51:31on the tech side which is user interface
51:34and so on but it's also MPS is more than
51:36just about the software it's about what
51:38you can actually do and what you can
51:39accomplish and I I bet there there
51:43really isn't the same focus on user
51:46experience often in healthcare because
51:48frankly often the user the patient isn't
51:51the payer that's really I think one of
51:54the key differences for for a consumer
51:55facing Healthcare Company where finally
51:57now will be involved in maybe to some
52:00degree things will be out of pocket and
52:01that will change things but even still I
52:03think someone with consumer mindset
52:04where the patient is really Paramount
52:07and that the patient's user experience
52:09is maybe even a part of how we think
52:11about their health care that's a
52:13fundamental shift Vijay I think this is
52:14a really interesting concept but
52:17that I expect from this idea of the big
52:19biggest company in the world being in
52:21healthcare is that there shouldn't be
52:23these massive companies profiting off of
52:25patient health so what would you say to
52:28that yeah well first off I think if you
52:30think about what a human endeavor should
52:33be taking care of of of the sick is one
52:35of the most important things that we
52:36could be doing it's important to realize
52:38that a com company that's very
52:40successful at doing that is creating so
52:42much value that for it to become a large
52:45company is a very positive thing just in
52:46general but also I would add that
52:49there's different ways that companies
52:50become big and one is through rent
52:52seeking where they have some advantage
52:55that they just take it they they
52:57leverage and they don't really
52:58contribute or innovate or and this is
53:01the spirit that we see in a lot of tech
53:02companies where they actually bring huge
53:04degrees of innovation and that
53:06Innovation is really dramatically
53:08improving the patient experience
53:10surrounding a dramatically improving
53:11outcomes and as a side effect this is
53:13building the biggest company in the
53:15world that's the vision that we would
53:16love to build towards where uh the
53:18improving the human in condition
53:20improving how Human Health Care is
53:22considered to be one of the most
53:22important things we could be doing and
53:24would be in some sense represented by
53:26the fact that such a huge company could
53:28be built next up we have Julie hi
53:30everyone I'm Julie Yu General partner on
53:32the a16z bio and health Team and this is
53:34my big idea for 2023. the value-based
53:37care stack as we said somewhere in the
53:39murky middle of the adoption curve and
53:41hype cycle of value-based care or BBC we
53:45are unabashedly vbc optimists
53:47but we are also Eyes Wide Open that many
53:49value-based models haven't delivered
53:51value yet and one of the major reasons
53:53for this is that the clinical and
53:55operational models from the Legacy fee
53:58for service world have simply been
54:00transplanted into value-based paradigms
54:02resulting in the gaming of the system
54:04versus a fundamental reorientation of
54:07care models to be focused on value from
54:11and so vbc done right demands
54:14purpose-built approaches that will be
54:16built on a fundamentally different stack
54:18this emerging stack will support new
54:20entrant value-based digital Health
54:22players and a comments alike around
54:24unique requirements to deliver higher
54:26value care things like data aggregation
54:29and activation Actuarial modeling
54:31Contracting adjudication panel
54:33management Continuous Care workflow
54:36support and provider ecosystem
54:37integration around referrals chemo
54:39management and network design just to
54:43so if the first generation of the
54:45digital Health Tech stack which we wrote
54:46about in 2020 was about enabling
54:49administrative and operational
54:50efficiency for virtual first providers
54:52the next generation will be about
54:55helping providers bear risk and and
54:58enabling payers to collaborate in a more
55:00integrated fashion with their
55:02risk-bearing provider Networks
55:04we see this stack emerging in many
55:06different product phenotypes from SAS
55:08platforms to solution marketplaces to
55:11msos to service an entire range of buyer
55:15segments and levels of technical
55:16sophistication can you just break down
55:18for the listeners what the difference is
55:20between some of the value-based care
55:22that you're speaking of and the Legacy
55:24fee for service systems in healthcare
55:26of course the first thing I'll call out
55:28is that you would hope that you know all
55:30the health care that you receive is
55:32value based so it's sort of telling that
55:33we've had to concoct this term to
55:36qualify what is value-based versus not
55:38um but you know the premise of
55:39value-based care is that you know so
55:41much of the bloat of our Health Care
55:42system is the result of the sort of
55:45massively misaligned incentives due to
55:48the third party payment model that is
55:50the dominant way for how Healthcare is
55:52paid for across this four trillion
55:54dollar industry in our country every
55:55year and um it's probably actually not
55:57even accurate to call the fee for
55:59service system Legacy because it's still
56:00the dominant model today but the fee for
56:03service system is one in which as the
56:05name indicates providers essentially get
56:08paid for what they do like so for every
56:10visit that they do every test that they
56:12run every procedure they perform they
56:15are submitting this Atomic claim for
56:17each of those tasks and getting paid a
56:20fee for each of those tasks and so as
56:22you can imagine you know that kind of
56:23system incentivizes doing more stuff and
56:27um and also incentivizes kind of a focus
56:29on treating people who are sick as
56:32opposed to this value-based system in
56:34which providers are getting paid for
56:37kind of managing the health of the
56:39population of their patients and
56:42um you know versus treating it the
56:44purest form of value-based care is where
56:46providers get effectively a said Budget
56:48on an annual basis to manage their
56:51patient population and you know any
56:53overages that you have on that budget
56:55you have to eat but also any underages
56:57on that budget you get to keep a share
56:59of that savings and so again it
57:01incentivizes a different care model
57:03associated with that which is uh you
57:05want to be much more sort of proactive
57:07about getting ahead of any escalation of
57:10health issues to avoid all those
57:13catastrophic costs that might come from
57:14people really getting sick and ending up
57:16in the Ed and things like that so what
57:18you're describing sounds great it sounds
57:20like we're driving efficiency
57:24dealing with sickness before it happens
57:26and so what have been the key roadblocks
57:28to actually achieve this reality that
57:31yeah so I mean fee for service has been
57:34the dominant payment model for many
57:36decades um I think technically it was
57:38like the 1960s when the first like
57:40billing codes were implemented that led
57:42to this uh this type of proliferation
57:44and so it's fundamentally just
57:46inherently hard to like change the
57:48business model of an entire industry so
57:50I think there's just some inherent
57:52inertia that you know the entire way
57:55that our system operates has been
57:56optimized for fee for service everyone
57:58knows how to game it everyone knows how
57:59to benefit from it there's an entire
58:01like 150 billion dollar industry
58:03dedicated to just submitting and
58:05processing claims so that you optimize
58:07your Revenue flow and so you know just
58:09because so much money is on the line
58:10there's a ton of friction
58:13um against you know changing the Status
58:14Quo I liken it too like the legal
58:17industry where you know everyone's
58:19familiar with kind of the minute by
58:20minute Billing System that law firms use
58:22and how many times have we heard about
58:24some upstart you know Law Firm who's
58:26trying to move away from time-based
58:28building to fix feed building but as
58:30quickly as it comes up it goes away
58:31because there's just you know so much
58:32inertia in the system
58:34um and so therefore you know I think it
58:37would really just have to take like a
58:38top-down government mandate to force an
58:41industry to make that kind of switch and
58:43while it's not like what I would call a
58:45full out mandate we effectively have had
58:47that where we've had a number of
58:49significant government initiatives that
58:51have been put into place that create
58:53Financial incentive to get
58:55um into the value-based care game for
58:56providers and payers and take full risk
58:59as it's called Uh in the form of these
59:01you know sort of budget-based payment
59:03um we've had a number of examples of
59:04that it's been you know Medicare
59:06Advantage I think is a primary example
59:08that was started in 2003 we've had
59:11accountable care organizations or acos
59:13which were initiated as part of
59:14Obamacare in 2012 and uh some subsequent
59:17subsequent flavors of that so you know
59:19we're still in super early days I would
59:21say as far as the seeing the impact of
59:23those top-down mandates that's really
59:25what it took to kind of shake the system
59:27up and say there's a different way of
59:28doing things yeah so it sounds like some
59:30of those initiatives were over a decade
59:32old and so what is it about today in
59:352023 that you see changing whether it's
59:38through legislation through the founders
59:40who are stepping up to change things or
59:42through maybe just the step-by-step
59:44changes that are happening to kind of
59:46reorient the inertia that we've seen in
59:48the past what are you seeing that makes
59:50you hopeful about maybe some of the
59:52shifts today definitely at a baseline
59:54just the general sort of like it takes
59:56time for these things to take hold I
59:58mean it's the same way that we talk
59:59about technology shifts right like even
01:00:01you know the internet having been
01:00:03invented quote unquote on a mainstream
01:00:04basis in the 90s it's really only now
01:00:06that we're seeing some of the foods of
01:00:07that Innovation so I think there's just
01:00:09some love physics type things that are
01:00:11fundamental like that that apply here
01:00:13but um you know I think it has to go um
01:00:16without saying that certainly the
01:00:17pandemic and everything that's happened
01:00:18in the last couple of years has had a
01:00:20huge impact on frankly like the
01:00:23financial beating down of providers that
01:00:25has created this huge boarding platform
01:00:26for providers to diversify their revenue
01:00:29and just create much more resilient ways
01:00:31to survive 5 and value-based payment is
01:00:33one way to do that where you actually
01:00:34have a recurring Revenue stream versus
01:00:37these Atomic payments that rely on
01:00:40patients coming into your office and so
01:00:42that I think has been a huge Tailwind
01:00:44for for the adoption of value-based care
01:00:46models you know you also see a lot of
01:00:48decentralization of care away from
01:00:50traditional hospital and clinic-based
01:00:53um you know virtual care home-based care
01:00:55community-based care all of these things
01:00:57are now very much getting adopted
01:00:59because they are fundamentally lower
01:01:01cost you know more convenient and much
01:01:03more reliable than you know having to
01:01:05force patients to go to these
01:01:06centralized settings so
01:01:08um you know I think that those are some
01:01:10of the key obvious drivers of of seeing
01:01:13kind of the why now that's occurring
01:01:16um you know that said I think there's
01:01:17still we're still looking at depending
01:01:19on what reports you read you you were
01:01:20only at roughly about 20 percent of all
01:01:24Health Care payments being in some kind
01:01:25of risk-based model
01:01:27um so you know to use the crossing of
01:01:29the chasm framework we're probably still
01:01:31in kind of like the early adopter maybe
01:01:33early majority phase but again it comes
01:01:36back to that that concept of we're just
01:01:38still fairly early in um in the kind of
01:01:41the Fallout from some of these top-down
01:01:43government mandates and and what impact
01:01:44they can have on the industry yeah
01:01:46completely and I think to your point the
01:01:48change in inertia doesn't happen right
01:01:49away I like how you've broken down the
01:01:51healthcare Tech stack into three parts
01:01:53in the past so that includes care
01:01:54delivery services the back office admin
01:01:57which is required to run a practice and
01:01:58then the front office which interfaces
01:02:00with consumers or customers
01:02:02is there a part of that Tech stack that
01:02:04you think is really
01:02:06um particularly ripe for disruption
01:02:08today yeah yeah and just for context on
01:02:11what you mentioned you know we had
01:02:12written a couple years ago about this
01:02:14new tech stack for virtual First Care
01:02:16and you know the premise of that thesis
01:02:19um all of these new virtual care
01:02:20companies are very ill-served by
01:02:23existing I.T Solutions because those I.T
01:02:26Solutions were very much predicated on
01:02:28fee for service you know payment rails
01:02:31um sporadic facility-based patient
01:02:33encounters so their entire data model
01:02:34was predicated on kind of these you know
01:02:36again these sporadic visits and the
01:02:39workflows were basically all
01:02:40provider-centric not patient-centric and
01:02:43so we predicted that there would emerge
01:02:45this new tech stack of horizontal
01:02:47platform companies that are designed to
01:02:49deliver value-based care longitudinal
01:02:52Continuous Care models and treat the
01:02:54patient as a primary end user and so
01:02:56that did come to Bear you know the first
01:02:58wave of companies that came out started
01:03:00to build what I would call kind of the
01:03:02low hanging recruit capabilities in that
01:03:04stack so the operational infrastructure
01:03:06for billing things like Financial
01:03:07Services CRM type capabilities more kind
01:03:10of administrative workflows and so you
01:03:12know to your question I think um the
01:03:14next wave of innovation is going to be
01:03:16more on the kind of the higher levels of
01:03:19sophistication of that stack on the
01:03:21clinical side as well as on kind of the
01:03:23risk management side
01:03:25um and so what I mean by that is on the
01:03:26clinical side you know I think we're all
01:03:28waiting for the promise of things like
01:03:30Ai and ml to help supercharge the way
01:03:33that clinicians make decisions about
01:03:35patients and enable them to access in
01:03:37real time you know all of the
01:03:39intelligence that is available to us as
01:03:41Society about what works and what
01:03:42doesn't work for given patient
01:03:43population versus being beholden to you
01:03:47know the very heterogeneous distribution
01:03:48of knowledge and experience that
01:03:49inherently exists in any provider
01:03:51population so I think the clinical
01:03:53aspect of that Tech stack is one area
01:03:55that we think is very ripe for
01:03:58um and then I think the risk piece as we
01:04:01were sort of discussing earlier is you
01:04:03know providers are basically have to
01:04:04learn new skills if they're going to be
01:04:06taking risk on these kind of
01:04:07budget-based payment models they have to
01:04:09both learn how to deal with payers in a
01:04:11different way the insurance companies
01:04:12and be able to design and negotiate
01:04:14these contracts in a way that they will
01:04:17um and also again understand their
01:04:19populations in a fundamentally different
01:04:21way to avoid these you know clinical
01:04:23catastrophes that could lead to
01:04:24insolvency under these models and so I
01:04:27think that's another area where things
01:04:29like Actuarial you know capabilities
01:04:31ways to underwrite risk and proactively
01:04:35uh intervene in populations at in a
01:04:38scalable fashion is going to be another
01:04:40area where we we need um you know to see
01:04:43Innovation within this Tech stack you
01:04:46know something you touched on there is
01:04:47the focus in the past on provider-based
01:04:50care versus patient-based care do you
01:04:53see that changing as in do you see
01:04:55Healthcare becoming more of a direct to
01:04:57Consumer model or do you think that
01:04:59again using the term inertia that we've
01:05:01had in the existing system where care is
01:05:04focused on a middle layer of the
01:05:06provider do you see that changing yeah I
01:05:09mean we actually had written a couple of
01:05:10pieces recently about this notion of our
01:05:12definition of direct to consumer care I
01:05:15think the traditional
01:05:16um you know sort of uh common uh
01:05:18interpretation of that phrase in
01:05:20healthcare has been you know out of
01:05:21pocket payments where you as a patient
01:05:23are paying for a service we argue that
01:05:26the definition is much broader than that
01:05:27it's really any service in which the
01:05:30patient has a direct loyalty you know to
01:05:33an entity that is either delivering a
01:05:34service and or facilitating payments but
01:05:36not necessarily one that they're paying
01:05:37for out of pocket so in that sense uh
01:05:41um and I think it's a necessary
01:05:43component of value-based care that the
01:05:45patient be engaged in their health care
01:05:48because that is really the predominant
01:05:49Way by which we're able to catch risk
01:05:51early and so you know you hear about
01:05:53things like remote patient monitoring or
01:05:55continuous measurement where you're
01:05:57putting devices into people's homes and
01:05:58you know they're participating in the
01:06:00creation of data sets that are novel and
01:06:04also create a whole new dimension of
01:06:06visibility into how patients are doing
01:06:08and I think that's one you know very
01:06:10kind of prevailing example of how
01:06:12patients are taking more accountability
01:06:15and just more participation in these in
01:06:17these care models so so I would yeah I
01:06:19would definitely expand the definition
01:06:20of direct to Consumer beyond what we
01:06:22sort of typically think about in the
01:06:24kind of the e-commerce sense
01:06:25um to one in which patients are aware of
01:06:28you know the services that they're
01:06:29receiving and contributing you know
01:06:32novel insights and data sets to ensure
01:06:35that their providers can effectively be
01:06:37successful in these in these kind of
01:06:39value-based orientations I've also heard
01:06:41you speak about the increasing Affinity
01:06:43that consumers sometimes have with some
01:06:45of these companies that are building
01:06:46more of a relationship even if the
01:06:48consumer is not the direct payer for
01:06:51that service could you speak to maybe
01:06:53just one or two examples of companies
01:06:55that have been able to facilitate this
01:06:56successfully the first company that
01:06:58comes to mind is Firefly Health which is
01:07:00a value-based care company so they you
01:07:02know a lot of the examples that I
01:07:04mentioned earlier were in the
01:07:05government-sponsored insurance space
01:07:07Medicare Medicaid Etc one of the areas
01:07:10that we think is is extremely ripe for
01:07:12value-based care Innovation is the
01:07:13commercial insurance space so employers
01:07:16sponsored Insurance in particular in
01:07:17this case and so Firefly effectively
01:07:20sells a solution to employers that helps
01:07:22employers manage their health care
01:07:25expenses in a much more effective way
01:07:27and as all of us know insurance premiums
01:07:30continue to just Skyrocket year over
01:07:32year and especially given the macro
01:07:34environment right now
01:07:36um you know there's this is just a
01:07:37burning platform issue for CFOs and CEOs
01:07:39across the board so Firefly is kind of
01:07:41inserting itself into that equation but
01:07:43the way that they um manage their their
01:07:46patient populations is through a very
01:07:48concierge type model where you can
01:07:50literally text you know their staff on a
01:07:53real-time basis 24 7 and be able to have
01:07:57clinically meaningful encounters with
01:07:59that staff about everything from
01:08:01appointment you know schedule doing to
01:08:03Pharmacy needs to just questions about
01:08:06whether I need to take my kid into the
01:08:08clinic or not and everything in between
01:08:10and the phenomenal stat that I always
01:08:12highlight for Firefly actually I'll ask
01:08:15you this question Steph how many how so
01:08:17you you know all of us kind of typically
01:08:18see your primary care doctor maybe once
01:08:22um how many annual encounters do you
01:08:25think clinically meaningful encounters
01:08:27do you think Firefly has on average with
01:08:29its patient population well now I feel
01:08:31like it's high because I feel like it's
01:08:33more than one but um this is per patient
01:08:39um I'm gonna guess take a wild guess
01:08:45um 45 clinically meaningful interactions
01:08:48that they have on an annual basis with
01:08:49their patient population these are you
01:08:51know commercially insured right so these
01:08:52are people who have full-time jobs where
01:08:54they're receiving health care benefits
01:08:56so typically you're speaking about a
01:08:58much more healthy population than you
01:08:59would see in like the Medicare or
01:09:00Medicaid populations and so because of
01:09:03that context is even more phenomenal
01:09:04that they're getting that level of
01:09:05Engagement but it's um you know they
01:09:08they use this text messing modality they
01:09:10do proactive Outreach to their patients
01:09:12they also provide remote monitoring
01:09:14devices they will send providers into
01:09:16the home when there needs to be kind of
01:09:18Hands-On on the body to do some kind of
01:09:21assessment or treatment so um because
01:09:23they're taking this multimodal uh
01:09:25approach to delivering primary care
01:09:27they're able to earn the right you know
01:09:29to have these these very frequent touch
01:09:32points with their member population
01:09:33which not only helps with kind of just
01:09:36consumer experience frankly and uh the
01:09:38development of loyalty towards them but
01:09:40also obviously from a clinical lens that
01:09:42allowed their clinicians to really be
01:09:43proactive about highlighting you know
01:09:46opportunities to intervene early in
01:09:48their care Journey so that that to me is
01:09:50a kind of a primary example of the way
01:09:52that you can Implement using technology
01:09:54in really systematic ways and not just
01:09:56relying on throwing bodies at the
01:09:57problem as traditional providers have
01:09:58done and you know which obviously is
01:10:00inherently unscalable to be able to
01:10:04um you know provide a delightful
01:10:05experience to the patient while also
01:10:07delivering clinical outcomes and lower
01:10:09that's insane 45 versus as I think you
01:10:13said less than one encounter for most
01:10:15people and I assume that the price point
01:10:19of that is is much lower as well through
01:10:21the facilitation of technology and
01:10:23that's exactly right the Insight here is
01:10:25that only because of the technology that
01:10:28now exists to be able to facilitate
01:10:30multimodal care meaning virtual care
01:10:32home-based care care in your community
01:10:34as well as you know when needed um you
01:10:37know steering you towards in-person care
01:10:39because they're able to diversify the
01:10:42sites of care that across which they're
01:10:44able to deliver their services the the
01:10:47cost structure of their Clinic versus a
01:10:49traditional Clinic is is a fraction
01:10:52right so that ultimately I would call
01:10:54out is also another key driver and key
01:10:56unlock for making value-based care work
01:10:59is the ability to fundamentally reduce
01:11:01the cost structure of care delivery and
01:11:04that I would you know I would go as far
01:11:05as to say it's not even possible to do
01:11:07that without the technology that we now
01:11:09have today a whether it be mobile
01:11:10whether it be these connected devices
01:11:13Etc and even just frankly technology
01:11:15adoption amongst providers which did not
01:11:17exist until very recently for better for
01:11:19worse and their ability to communicate
01:11:21directly with each other as they're
01:11:22managing collectively these patients in
01:11:24these in these budget-based risk models
01:11:27so um you know that that is absolutely a
01:11:30fundamental part of the unlock that you
01:11:31know Firefly and many other companies
01:11:33have been able to take advantage of I
01:11:34hope you enjoyed this two-part series
01:11:36and if you're looking for the full list
01:11:37of 40 plus ideas you can head over to
01:11:39a16c.com and you'll find it on the home
01:11:41page see you next time
01:11:45thanks for listening to the a16z podcast
01:11:47if you like this episode don't forget to
01:11:49subscribe here on YouTube to get our
01:11:51exclusive video content we'll see you