Despite all the hype around AI, artificial intelligence is not the quick fix for financial services companies as they struggle to avoid becoming mere utilities, according to Joe Lonsdale, a co-founder of Palantir and general partner at investment firm 8VC.
“AI is not the panacea,” said Lonsdale, speaking at CB Insights’ Future of Fintech conference in New York. Before AI can tackle some of the harder problems in finance it will take years if not decades to structure the data these systems will ingest and figure out how to do so for them to tackle broad financial applications.
“AI is decades away from matching human creativity,” he added.
This was a theme Lonsdale returned to several times in his chat with Robin Wigglesworth of the FT. Not every new tech trend in fintech will necessarily give legacy institutions (or upstarts) a competitive tech edge. For example, robo-advisors rely on tech that’s not especially challenging to replicate, Lonsdale argued. Much of the challenge with robo-advisory platforms is simply marketing them effectively and acquiring more users, he argued.
“Robo-advisors are less of a tech problem, and more of a brand problem,” Lonsdale said.
That said, Lonsdale did have sharp words for banks’ failure to realize that much of their tech has not kept up and that the services they offer are at risk of being further commoditized.
“It’s a competitive necessity for financial services companies to think of themselves as technology companies,” he said.
Another space Lonsdale was skeptical of is payments, where he said there are “two thousand companies” competing.
Lonsdale pointed to lending tech as a more promising area with less competitors, and mentioned one of 8VC’s investments, Blend Labs, which develops a platform that aims to streamline home mortgage originations (a $13 trillion market) by working with both lenders and borrowers.
Anand Sanwal, CEO, CB Insights: So we have a fireside chat next, which I’m really excited about. You’ve met one of the participants. So I’m gonna invite Robin Beck [SP], from the Financial Times, who you met earlier on the alternative data panel, and then I’m really pleased that we have the opportunity to have Joe Lonsdale here. Joe is a founding partner at 8VC. Also co-founder of Palantir, multi billion dollar global software analytics company, best known, I think, for its work in defense and finance.
Joe, from a VC perspective, sits on a lot of boards that are pretty relevant to the audience here. Addepar, Oscar amongst them, and then also sits on a handful of other boards. OpenGov, Illumio, Radius, Hyperloop and Wish, amongst others. So looking forward to this conversation.
Thank you guys for coming, and I’ll turn it over to you, Robin.
Robin, Reporter, Financial Times: Well, thanks so much for the introduction.
Joe, it’s great to have you here. I mean, obviously we have a lot of things. Sadly, again, I can’t take three hours. So we’re gonna try and keep it punchy. But I would like to hear, before we dive into some of the specific issues, your broad sense of, what are the big trends and themes that you’re spending a lot of time thinking about at the moment?
Joe Lonsdale, Founding Partner, 8VC: Sure. Thanks, Robin.
I mean, I guess there’s a handful of things we spend our time thinking about. At a very high level, I think the most important thing for the next decade is that there’s all these big industries, finance included, of course, that, the way they run right now is just very far off from the way they should be running if they were using technology optimally.
So I think there’s just a huge gap between how the biggest industries in America currently run, and how they will run with the best IT and with the best computer science. In general, aside from that, we’re doing a lot of things in genomic sequencing, in cancer and healthcare IT. We’re doing some things in transportation as well.
Robin: Yeah. So, transportation, that’s a very sort of boring way to say what you’re actually doing, though.
Joe: There’s a few things there that are pretty crazy. I think the Hyperloop, probably, is one of the most interesting ones.
Robin: Yeah. I mean, let’s spend a bit of time on that, just because there were recently some tests… I think, I mean, MIT, for example, presented their first little pod. I think, in August, this summer, there’s gonna be maybe some tests on what actually works.
Joe: Well, we just showed off a few weeks ago in the desert how you can accelerate to over 100 miles per hour in one second, which is really important to be able to do. The idea of the Hyperloop, as a lot of people here probably know, is you have these big steel tubes, basically, and you take most of the air out of them so it gets rid of all the friction. Once you get going, it stays going really fast.
So you just have to have a thing that accelerates it once every kind of quarter mile or so, it looks like. So you want to be able to travel about 700 miles per hour in these tubes just by using [inaudible 00:45:20] type transportation every once in awhile. So proving the transportation system works was a big part of the rest that we did.
Robin: Yeah. But what about corners, and the idea that, between San Francisco and Washington…
Joe: Well, there’s a couple of things.
Robin: …There will be some corners, right?
Joe: If you turn more than about one inch every 20 feet when you’re going 700 miles per hour, it gets very uncomfortable. In fact, everyone dies if you do more than that.
Robin: Okay, yeah.
Joe: So that’s probably an issue.
Robin: Suboptimal, basically. Yeah.
Joe: It probably would be a lot easier to do this in China. They could just knock stuff down and build stuff. Maybe Trump will do that. I don’t know. [inaudible 00:45:51] on politics.
Robin: The Trump train.
Joe: But, I mean, the key thing there is tunnels, right? So you actually need to be able to build tunnels cheaply. If you build a subway tunnel, it’s relatively expensive, because there’s a lot of air that the train has to move out of the way. But if you build a Hyperloop tunnel, it could actually be a lot smaller.
But there’s actually another angle to this that actually syncs with Fintech, which is kind of cool, which is that the way you get the Hyperloop done involves a lot of financial innovation.
Robin: So tell me a bit more. So you run prizes, right, to encourage this?
Joe: Oh, there’s different ways of doing it. The Hyperloop One that I’m working on, we’ve raised about $120 million and hired a couple hundred engineers, and we’re actually just building it more like a traditional company. But there’s a lot of open source things that are friends are doing to try to encourage innovation.
But the more interesting thing is, how do you get these things built.
So, for example, in Hong Kong [inaudible 00:26:38] their subway system, they actually used the arbitrage around changing real estate values to fund it. My favorite example that we’re looking at right now, if you look at Long Beach, there’s a port there that’s on about $200 billion of land, and they can’t hire any more people, because there’s not enough room.
So what you can do is you use the Hyperloop to move the port 100 miles into the desert, where the land costs nothing. Then it ends up having a much bigger port. Then you can free up $200 billion of real estate value. So there’s lots of ways you can use financial innovations and structures to build these things that could be interesting.
Robin: Oh, wow. That is. But, I mean, how feasible is this going to be? You can play around with the financial side of things, but…
Joe: Well, I think that’s the key thing. I think it’s quite feasible. Most of the cost is just the steel to build these big tubes, and then the variable costs of running it is just a much more efficient train. It’s better for the environment. It’s better for living near a city and getting in really fast.
But we have to figure out the real estate arbitrage to really make it worthwhile.
Robin: Yeah, because it’s gonna be hard in built up areas where it would make most sense to build a Hyperloop.
Joe: Yeah, but if you have land that’s worth nothing, then all of a sudden it’s worth a huge amount because you’re suddenly 10 minutes from San Francisco or New York, when before you were an hour and a half from San Francisco or New York, that could actually change a lot of things. So that’s pretty fun.
Robin: Do you think there would be maybe other countries that… U.S. it’s maybe harder, especially San Francisco, LA, this pretty built up area, that maybe some of the technology you find here you will be able to use in other countries.
Joe: Yeah. The biggest challenge in the U.S. is there’s this extra step, where it’s about a 5 to 10 year $10 million environmental review. We have a very strong environmental lobby in the U.S. that a lot of us respect, because it’s good for the environment. But it takes a lot loner and a lot more money to build things here because of that, so it may be tested in other places first.
Robin: Yeah, exactly.
It’s gonna be interesting.
So will I get a trip on the first one? As long as… I think that turn sounds a little bit scary.
Joe: I think our CTO wants to do the first one. I’ve told him that, if he kills himself, he wasn’t a very good CTO anyway. So.
Robin: Yeah. Exactly, yeah.
I bet he loved that joke, right? Yeah. Exactly.
Well, let’s move onto the financial technology side of things. I mean, the reason why we’re here… It would be interesting to see what you think the broad movements are.
It’s interesting. We talked a little bit about this, but payment companies. There are a lot of payment companies out there, but where are people maybe not doing as many cool things?
Joe: Yeah, sure. I’m probably unfairly biased against payment companies. I think there’s a lot of really cool innovation there too, and quite a few things I’m seeing. I worked for PayPal when I was quite young, while I was in school, and afterwards I was working with Palantir. He made me see about 2000 payments companies that wanted to meet with him. I got a little bit tired of it after awhile.
But I do think there’s areas of finance where, unlike payments, there’s almost no one working on still that are really important.
So one example I gave earlier was, a company, Blend, I work with, is working on better technology for mortgage origination. So there’s $13 trillion in mortgage backed security markets. There’s about 1000 banks, 60 really big ones, that are trying to do this origination stuff themselves, and obviously they shouldn’t be building their own systems. There should be other systems they are using.
So there’s two or three competitors in that space, which is a lot more interesting to me than a space with 2000 competitors.
Robin: Yeah. Certainly.
But let’s spend a little it… Addepar as well, one of the companies you’ve founded now, I mean, that’s in the financial [inaudible 00:49:56] technology. So that’s a platform for wealth managers to basically do everything they need to do.
Joe: Yeah, sure. So the genesis of that is that, I guess the view in Silicon Valley is you try to build and own these things that can own their own platforms, and become very large platforms. When we were looking at finance, looking at how the big banks work, they are each their own walled cities, and don’t really share data.
You’ve had a lot of financial infrastructure grow up in the world in these very separated, very hard to work with ways. But then you have lots of wealth advisors and groups that have to talk to all the banks, have to talk to all the custodians, have to talk to all of the fund administrators. So it becomes a really hard data integration problem, and it becomes a really hard, frankly, operating system problem for how you can build stuff on top of this.
So what Addepar does is it’s the main software that a lot of wealth advisors use. So a lot of RIAs, now a lot of the bigger banks are starting to use it as well, and it kind of creates the financial infrastructure to get all the data in one place. Then it lets the advisors do all of the reporting of all of their work on top of it.
Robin: So what was the spark behind Addepar?
Joe: The spark was that Palantir here was the first company I was working on, and the guy who was running the financial side of Palantir with me… Palantir, about a quarter of their business is in global finance. We do a lot with the banks, and we were replacing systems with some of these banks from late 1970s, actually. This was 2006.
It’s amazing how long a lot of the technologies you all know has been around… The banks that you’ve had to build on top of over the years… The question was, how do you fix this huge mess of financial data? How do you hack into finance at the root to get thing to be more data driven?
Our insight was that the technology that powers wealth management right now is also very broken and very old, and that money sits at the base of global finances.
So if you can be the platform that powers and sits on top of a huge amount of the money, then you can actually build other things on top of that and end up making the whole thing more data driven.
Robin: So is it more individual financial advisors, or..? A big one, like a Charles Schwab, probably has its own system, and is not particularly keen on sharing…
Joe: Well, they will integrate with us. They will send advisors to us to work on it so that we can use their stuff. We sit on top of them and Fidelity, and others. Right now it’s about… I mean, it’s $500 billion investment, which is still small, I know, relative to a lot of people here. But it’s growing pretty quickly. It’s about 200 RIAs, family offices, and then a few of the big banks are getting in on it as well right now.
Robin: Oh, really? Which big banks?
Joe: I think I’m allowed to mention Morgan Stanley. I’m not sure of the other ones I’m allowed to mention. I’m always getting in trouble on this, because…
Robin: Feel free. We’re all friends here, right?
Joe: But, no. I think we’re doing some really great work with them, and there’s a handful of others as well we’re starting to work with.
Robin: Oh, that’s interesting. How do you find dealing with the banks, then?
Joe: Well the banks, I’d say, in the last 5 or 10 year, are a lot more interesting to work with for people from Silicon Valley. I think they have realized there’s a lot of things they are doing that should be utilities that they used to think of as competitive advantages. To me, that’s one of the biggest themes here in New York, in Fintech right now, is that a lot of banks used to think, “This is my competitive advantage.”
Now they think, “I’m doing this, 10 other people are doing this, I know I’m not gonna be the very best at this. It doesn’t even matter very much. How can I partner with outsiders? How can I make this into more of a utility, and just hugely cut my technology spending?”
So I think that is a really big theme for Fintech in New York.
Robin: Are they quick to do things, or do they feel slow and bureaucratic? I mean, I assume it’s slightly different time scales than what you [crosstalk]
Joe: Yes, it’s a little bit unfair. They are dealing with trillions of dollars. They are not really allowed to… I Mean, they still have to have their 6 to 12 months security reviews before they install stuff, which, to us, seems like a lifetime.
But I think there’s reasons why they go more slowly. But I think there are lots of groups within these big banks that have created different types of swat teams, and have started to move more quickly on certain things once they decide they need to.
So I’ve been generally pretty impressed with how the bigger banks are thinking of themselves as lot more as technology companies, and trying to shift that way.
Robin: That’s interesting. I mean, it’s a very trendy thing to see… I’ve seen lots of banks, and I love that lots hedge funds call themselves technology firms who just happen to manage lots of money. But is there an element of desperate trying to be the cool kid in class, that these banks are still fundamentally financial services companies that manage deposits and banks, and technology is important to them, that they have been trying to rebrand themselves. “Oh, we’re a tech company,” just feels a little bit..
Joe: I think there’s an understanding by a lot of the banks that, if they are gonna be around in 10 to 20 years, they have to figure out what their barrier to entry is in this world, and what their competitive advantages are. I think they all have a lot of data that gives them these really interesting motes. But if they don’t conceive of themselves as technology companies with how to use that data, then they are not gonna have those motes in the right way, that they are gonna be able to keep growing.
I do think it’s a competitive necessity to think of themselves that way. There probably are some of them who are just saying it because all of the other cool kids are saying it, but I think the ones that are run well are really thinking of themselves a lot more as technology groups.
Robin: Yeah. You mentioned Palantir. 25% of its revenues come from the financial services industry. I mean, what exactly are they doing for them, without giving away any sort of secrets?
Joe: No, sure. I was there the first seven years. So I’m not as intimately involved. When I was there, our first contract with a big bank was because it had bought a bunch of other banks, and it was integrating a bunch of them into it.
The whole data integration IT services challenge is something that came up a lot in government, and it is an area where we tried to prioritize a lot of this, and we’re able to do a lot better and a lot more cheaply. So it turned out it was an isomorphic problem in the financial services area.
Similarly, the next big problem we faced was actually something in mortgage backed securities, where they had a multi trillion dollar portfolio and they had tons of data sources. They were gonna pay one of the integrators half a billion dollars to integrate all of the data sources and build their system. It was gonna take nine months, and it turned out that the technology that we’d been using in the government context also could do that for them much cheaper and much more quickly.
So I think really what Palantir here did as a company is it focused on productizing [SP] IT services for systems that involve lots of analysts and lots of heterogeneous data, and two areas of the world where you have lots of these systems and lots of this spend are government finance. So it’s very natural that it would be working in both those areas.
Robin: Yeah. What about… There was a recent BuzzFeed article. They had some internal documents. I know you’re not directly involved within the company anymore, but as an interested party, it would be interesting to hear your thoughts on how you think that company is going to evolve.
I mean, it’s one of the biggest of these unicorns.
Joe: Sure. I shouldn’t speak for them. Again, even as a founder, I’m not as intimately involved.
I did write a piece in response to the BuzzFeed article. I think Peter Thiel thinks of a lot of things in terms of a zero to one and one to N [SP] framework. So Palantir here went zero to one in a few areas, and it wanted to prove it could go to zero to one in potentially a lot more areas, and then scale them.
So over the last four years it has gone to a lot of new industries in which it’s seeing if it makes sense to go from zero to one there, and it has had some very big successes in those industries, especially in healthcare type of insurance companies and the oil industries, and a few others.
There’s also been things it’s tried that it decided it was not a good fit for, and so it could focus instead on scaling the areas where it’s working. I think what the BuzzFeed article said is that there were three companies where it failed, and it was, like, “This is a really big deal. It’s failed these three companies.”
I mean, Palantir has, like, 100 big companies it works with. So the fact that they tried to make it like it was failing for not working with three companies, to me, when it’s trying a lot of things, is pretty silly, as far as I’m concerned. But, I mean, you can take data and you can tell a story. It’s a more interest in story if it’s negative.
Robin: Yeah. Journalists, you know. It’s what we do. I’ll stick up for BuzzFeed. It was an interesting article.
So, I mean, it was interesting. I was talking to a couple of CEOs of some big, big , big traditional money managers here in the U.S.
It was interesting. There has been lots of focus on the flow from active fund management… They are doing very well.. To passive ETS and [inaudible 00:57:28]
But I remember talking to one of them. He paid the point that, you know, we still have really great margins. I mean, they are coming down. But they are great. Frankly, if I was outside of the industry, I would look at the asset management industry and say, “Your margin is my business.” Who was it who said that?
Joe: I don’t remember
Robin: But have you looked at the asset management industry? Are there any particular niches that you think are more primed for a bit more disruption than others?
Joe: Sure. So Addepar actually kind of partners with a lot of the advisors, and helps them out. But I think a lot of fiduciary stuff, personally, makes a lot of sense, where you should be charging for things that are aligned. So I think that is actually amazing to me that there’s an asset management industry that’s still charging things for churning [SP] your accounts. I think that’s all shifting in the right direction, and is going a good way.
Any margins coming down there… Silicon Valley has about 20 robo advisory companies that I’ve seen over the last few years, and I think some of them are very well run, and are good companies. I haven’t invested in any of them, and I’m not that bullish on them as a space. I think it’s not a hard technology problem. I think it’s much more about your brand and who is trusted, and that sort of thing.
So I think things are shifting to more passive. I think the margins are coming down, but I think that, at the high end, if you’re managing $100 million, billion portfolios, there is a lot of human intelligence and a lot of alternatives needed. So I think that’s gonna be expensive for a long time still.
Robin: Yeah. But, so you’ve looked at these robo advisors and decided, maybe not. I mean, isn’t sometimes simple technology the best technology as well?
Joe: Oh, sure. I think it’s a good thing.
In terms of, like, if I have $10 billion to manage, would I put a few billion potentially into beta exposure that’s managed the robo advisors as cheaply as possible?
Sure. It’s not a bad idea. Would I try to build a robo advisor business as a 21st robo advisors business? Probably not. I think any of these groups, with a small team of engineers, could build a great robo advisor. I don’t think, at least the robo advisors as they exist now, are not hard technology problems. They are much more about brand than they are about technology, is my view.
So it’s just not as interesting, from a venture capital perspective, [inaudible 00:59:33]
Robin: Oh, that’s interesting. So any other niches, sub niches that you’re spending a bit of time with, without giving away tipping your hand, of course?
Joe: I mean, in the financial area, I think one really interesting trend you’re seeing in Silicon Valley is that a lot of other companies are becoming Fintech companies in order to do a better job.
I’ll give an example. One of my companies is booking tons of remodeling through contractors. So we have 1.3 million consumers a month booking contractors to do work on their homes, and there’s a few billion dollars a year being booked. Right now it’s only 1% of the market. There’s $300 billion a year booked in the U.S., and in order to get the consumers to convert better, we’re trying to figure out how to give loans ourselves, as opposed to them coming to the site, booking it and then having to go talk to their bank.
The origination itself is probably worth 50, 100 basis points.
But there’s a lot of things like that, where the thing that you used to be, like, oh, you stop and you go and you walk into the bank and you try to get a loan, is now gonna become a point online, where the technology companies themselves that own the consumer in this process are gonna be the ones that own the financial part of it to.
So I think it’s a really important trend for how these companies figure out how to partner with the financial world on things like that.
Robin: Well, the buzzwords, I’ve noticed, are very much on the financial industry… Is machine learning. But, basically, using more modern artificial intelligence techniques in everything from investing to solving problems that they have on the consumer side, [inaudible 01:01:00]
How much of this is, to return to the cool kids trying to sound with the times, [inaudible 01:01:05] machine learning is very much [inaudible 01:01:07], but do you think, is this a future we’re gonna see more things move that way?
Joe: Yeah. That one, I’m more skeptical of a lot of the efforts I’ve seen in Wall Street. So it might be a little bit more of the cool kids syndrome. I’ve had several CIOs of big institutions ping me saying they have a big machine learning project, and can I connect them to the right people for what they are doing recently.
I mean, machine learning is a tool that all the great technology companies are using now to get done certain things. When you have a bunch of data and you need to figure out what the best answer is, I think you apply it.
I mean, it’s kind of funny, right? It’s artificial intelligence, and once it works we call it machine learning, and there’s all this research on the AI front. I’m definitely bullish on what AI is making possible, but I don’t think it’s this panache or this thing that solves everything. I think the harder problem, frankly, is getting all of your data organized in an infrastructure where you could use it for the problems.
So I think the hard thing to say is, what data should we be using to solve this problem? What is that process? Then once you have all of the data organized, you can hire a single PhD from Stanford. He will do your machine learning thing for you. That’s not that hard. The hard part is getting it organized and figuring out that it’s relevant to this process.
So that’s much more of a strategy and process question to me.
Robin: But you think that it’s gonna be more, I mean, on artificial intelligence. I like that definition, that artificial intelligence is everything we haven’t been able to do that. Everything else is [inaudible 01:02:21]
But how much will we be able to do in the future? I mean, looking forward a little bit.
So I definitely have a few investments in artificial intelligence companies that are trying to push the boundary and change things. I mean, I was actually really surprised that Demis and the team at DeepMind were able to solve the Go problem.
I think that was pretty impressive. But even looking into it, it’s not really capturing, in my view, human creativity, or anything even close to that yet. So I think we’re still decades and decades away from needing to worry about being replaced.
I think the much more interesting problem that we’re focusing on kind of the next 10 or 20 years is, how do we organize the infrastructure of our companies to get the data structured and available to use it for decisions?
I think it’s not gonna be the people who are best at AI and machine learning who win. It’s gonna be the people who are best at that data infrastructure and process problem. I mean, most of the top companies in Silicon Valley, like linear regressions and things, get you 90% of the way. I’m not seeing a lot of really, really hard AI that makes the big difference, for at least the things I’m seeing.
Robin: But the approach that DeepMind used with AlphaGo, I mean, it was quite… So obviously it’s not human creativity, but when I looked at it, and, you know, I’m hardly an expert, but it did seem to be one of the closest things we’ve gotten to a bit of human intuition.
As in, it made the rules, but it was interesting that I managed to do it so quickly. I think the week before Zuckerman had talked about, we think we’ll be able to go in a few years time, and then boom. You have DeepMind out there.
It shows how quickly things are evolving. So just in the next few years, what do you think, what kind of challenges and solutions are we going to be seeing?
Joe: I think the Vision stuff is the most interesting in the next few years. So I think you’re going to be able… I mean, the way DeepMind is making the most money right now is it’s helping with YouTube search, right? So you’re already able to take videos and label things better.
I think you’re gonna be able to run that in reverse. You’re gonna be able to generate really cool 3D videos and 3D works through VR and stuff, using this AI that has gotten a lot better. So there’s gonna be some kind of fun stuff there.
I mean, obviously it has affected quant trading a lot, and there’s a lot of better ways you could do stuff like that.
But I still think, even with that, there’s a huge part of finance where you need people’s minds. You need to have the [inaudible 01:04:40] intuition. So you’re not really gonna replace people in most of these ares for a long time, is my view.
Robin: Yeah. Well, that’s good at least.
So I have to admit, I talked to a natural language processing company the other day, and seeing what they can do with automated writing was a little bit scary, as a journalist.
Yeah, yeah, exactly.
Joe: …To be negative about everything.
Robin: Yeah, exactly. Yeah. Thanks for that. Yeah. Well, hopefully it will free me up. I’m a big fan of the man and machine marriage, putting together, ideally, the next few years, it will just be taking away my boring parts.
Robin: But, yeah. Snark. Machines aren’t good at snark yet.
Joe: They’re not.
Robin: Maybe that will be our last bastion, the kind of resistance will find itself [inaudible 01:05:27]
One of the other projects you mentioned, which I had never heard of, but it sounded fascinating a bit ] open gov, just because I’m interested in how we use their technology to change these things, especially on the public services side. Tell us a little bit more about that.
Joe: Sure. That’s another interesting area where I guess it’s kind of Fintech as well, but it’s for government. So OpenGov is, I guess, the third company that I built with a few of my friends. It’s in about a thousand cities right now in America. It’s in a lot of cities and districts, and we’re basically helping them access their systems.
The history there is a few of us tried to start a philanthropy to help government use its data better, and we did a lot of stuff with states, and got a lot of attention for showing how California compares. Certain areas are very efficient, certain areas are kind of very wasteful, and abused by special interests. A lot of cities were interested in this and said, “Oh, can you do this for us too?”
We said, “Sure. Send us your data.”
The city said, “How do we do that?”
It turns out there’s $10 billion a year in America spent by cities on these 20, 30 year old backend systems that haven’t really changed much in a long time. If you’re a mayor or a city manager, you don’t really have access to your data. If you wanted to see how were you spend on police and fire relative to 10 cities nearby us, or how we compare to other groups, you can’t do that right now. You have to pay consultants huge amounts of money to get that information.
So we decided to go in and kind of replace the financial backend in terms of how they do their budgets, how they see what they are doing, how they do all of their open data. So there’s a few different companies in this area. I think there’s been a lot of traction lately in teaching government how to use data better, how to hold themselves accountable with it.
Robin: How are they doing that, then?
Joe: Well, the key thing is really, to me, the budgeting processes. So rolling out a lot of stuff for, when you’re doing your budgeting, there might be 3000 areas you’re spending money, and you could say, “Look, this area, relative to kind of the basic model of all these cities like us, we’re doing really well. This area, we’re spending twice as much.”
Sometimes you’re spending twice as much because there’s some variable we’re not accounting for, and it makes perfect sense. This is where it’s the man, machine thing again. Sometimes you’re sending twice as much because it’s a service that’s being run by the cousin of the ex mayor, and so maybe you should look into that.
Robin: Yeah, that does happen in some [inaudible 01:07:32] accounts as well.
Do you think, is this gonna be something we should be watching in the next few years? I mean, government has been putting, whether it’s municipalities or some federal government putting more and more information online in machine readable formats… Will we be able to do a lot of things a lot better and more efficiently in the future, then?
Joe: Yeah. I think the key thing with all this open government stuff is, it’s one thing if you just put all these PDFs online, and then nothing happens because it’s so complicated. It changes maybe a little bit if there’s a really good investigative journalist who spends a lot of time and finds something, but for the most part, it just goes up, and then there’s a big nothing.
I think the key thing is actually building processes into how you’re running government, or how you’re running these organizations, that’s leveraging that data. It’s automatically doing metrics and KPIs and comparisons based on that data. So if you can create processes that are driven by technology, I think that’s how the open data stuff starts to really work its way into the system.
Robin: Yeah, that’s quite exciting. So is it a thousand cities now?
Joe: Yeah. It’s 1040. I checked with them, actually, last week. Yeah. No, it’s pretty cool. Marc Andreessen and John Chambers are on the board with us, and have been great mentors there. So it’s been interesting.
Robin: OH, yeah. That’s cool.
One thing also not Fintech related, at least not directly, but genomics. What are you doing there?
Joe: Genomics is really cool. Genomics has become a computer science problem. I’m sure a lot of people here have been watching it. It’s very funny. It used to be this obscure life sciences area, and now that you’re able to get lots and lots of really clean and good data from this space, it’s one where the very best computer science teams are kind of the ones that are driving and pushing it forward.
I mean, there’s a couple of different things there that are really important, that are probably actually of interest to everyone here who… Basically, if you’re in your 20s or 30s especially, you should probably get tested. About 2% of people in the country have a very, very high risk of cancer in a lifetime. It will be between 50% to 80% chance they’re gonna get cancer, and there’s a lot of things you could do to protect against it and prevent against it.
This would usually be a several thousand dollar test. With companies like Color Genomics, it’s now, like, a $300 test. So it’s totally crazy for everyone not to do this, because you actually can discover things that protect things.
I actually had someone very close to me who took this test, and she has a very high chance of cancer, unfortunately. So we had her whole family do it, and her mom came up as a very high chance as well, who is 60.
So the doctor said right away, “You better have your ovaries removed.” Fortunately, when they removed it, they found early stage cancer. But thank God we checked and did it.
So I think it’s really important that people… The healthcare system takes about 5 or 10 years sometimes to actually take the best practices and make them scale, and make it available to everyone else. So it actually kind of behooves everybody to be aware of what’s going on, and the fact that there are these tests that are really important, that can actually save lives, that have not quite yet become standard.
So I think the sequencing area is fascinating right now.
Robin: That’s interesting. Which companies do you have in this? What was it called?
Joe: Well, I think Color Genomics is one of the most interesting ones. So the founder, Elad Gil, he built Google Mobile, Maps and whatnot. He sold his last company to Twitter, and is a senior there… But he’s also a PhD in biology. So he’s a really cool guy, and he’s brought together a lot of our friends in the Valley to build, I think, one of the strongest teams in the space.
Another one that’s really becoming a biggest thing is called Guardant Health. Guardant Health is basically, you take a little bit of… This is an interesting area , but you take someone’s blood and you could test the cell for DNA, and basically, if they have cancer, the DNA will be there, and you can learn about it. It actually detects if you have stage two, three or four cancer as well.
So right now, because of the FDA, you have to have a doctor order it. But this is another way you could detect, actually, if you have cancer. You probably saw Bill Gates and Jeff Bezos just invested a bunch of money in something called Grail, which is also in that space. Both of these companies are on top of alumina. Alumina is the public company that does most of the sequencing. So Grail, basically, is trying to develop something that will become common, to give a little bit of blood and see if you have cancer or not.
So this stuff is all advancing pretty quickly right now. It’s really cool.
Robin: Yeah. That does sound pretty cool.
Yeah. No, I was quite struck by… I don’t know if you saw the two [inaudible 01:11:24] at the Kaggle competition with heart palpitations. I mean, just using data, you can find out a lot of things. I’m weary of being too greedy with my questions. I’m sure you guys also have some.
So I don’t know if we can have somebody that can read them out, or we can even… People can wave their hands. Something old fashioned.
Anybody have any questions?
Man: Yeah. We have some questions from the audience.
Which industries in the current world of finance do you see changing the most radically? Are there asset classes you see not existing in the future, or transactions that will look totally different in the future?
Joe: That’s a good question. I mean I think we’re already starting to see that the way that people give and get loans is gonna be abstracted out of a lot of the banks into a lot of these specialized organizations. So I think that’s one that’s accelerating. Right now it’s still really, really tiny, because the banks own that whole industry. Sometime we forget in Silicon Valley that we’re the little kids in the finance, because the amounts of money we’re playing with are really irrelevant sometimes.
But that’s one where there is this exponential growth that, I think, over the next decade, you’re gonna see a lot of the loans and credit in general not nearly as controlled by the banking system, which would be, actually, a pretty crazy shift, relative to what we had before.
So I think there’s that. I think, in general, I’ve mentioned this earlier, but a lot of the technology that runs global finance is becoming utilities and becoming tech platforms, and are gonna be spun out from the banks. In some cases, the banks will do this on purpose, and will own a piece of the resulting platform.
In some case it’s gonna be Silicon Valley companies that come in and provide this and win
So I think the way that all financial technology looks 10 years from now will be very, very different. There will be a lot of these external platforms that actually are worth more than a lot of the big banks in a lot of ways, and that a lot of different companies that didn’t used to be considered financial companies will be involved in the credit system.
Robin: Just picking up on one last bit of that, are there any particular industries where you just think they are going the way of the dodo? They are just dead. They just don’t know it yet?
Joe: Hmm. Well, I think the low end wealth advisors probably are gonna be pretty decimated. I think there’s been very strong need for good advisors for certain levels of wealth and money, but I think the technology is able to serve most of your needs for a typical middle class kind of investor, and so I do think that… Like, low end RIAs, which is a big industry in America, is probably in a lot of trouble, is one of them.
Robin: That’s interesting. I’m giving a speech at a financial advisor conference on artificial intelligence in markets. So I probably won’t tell them that half of them get…
Joe: Well, they will probably all tell you they’re not the low end ones.
Robin: …NO, exactly.
Joe: You never really offend anyone with that comment.
Robin: It’s an FD conference, so they definitely are the high end [inaudible 01:14:15]
Any more questions?
I’m happy to ask more. But I just felt…
Man: Yeah. We have a couple more questions.
Man: As someone who has been very involved in data, are there data sets in the world that you think are underutilized, or data sets you wish existed that don’t currently?
Joe: That’s a great question. There’s a few of the really… I Generally am not excited about quant funds. I always try to convince my friends, don’t go to a quant fund – help us build a real business. That sort of debate. But there’s actually a few really smart quant funds that are using data sets to profit in ways that have surprised me and impressed me.
So what happens is, a lot of the most talented ones will approach our various companies working in industries. So Silicon Valley right now, we have a lot of these vertical platforms forming and growing that are becoming very important in different parts of healthcare, different parts of government, different parts of… Like I mentioned the contracting one.
That’s a great example. So the contracting one happens to know the rate of remodeling going on, and then the contracting activity going on in lots of different regions in America, and there’s been a company that built indices based on this, and used it a a really important input. So I think, in general, data from the startup world, in terms of what it says about the economy… Another really good example is there’s 20 million SMBs, approximately, in America, and there’s a lot of data coming from about… The different startup platforms about what these SMBs are doing, what they are purchasing, what their economic activity is in real time, and a lot of stuff is starting to be used by the very best quant funds to profit.
That’s something that I think is very clever. It’s not something I personally spend as much time on, but I think it’s definitely really valuable data.
Robin: Why are you skeptical of some of the quant funds? I mean, below the top tier ones.
Joe: Well, I think that’s the problem. I don’t think you’re adding that. Maybe I’m wrong here, because they make money. But I don’t think you’re adding that much value to the world, beyond there being a few of them.I don’t know. Maybe making markets more efficient is really good, and we need a lot of smart people doing that.
But at some point, being part of building and creating a business that’s solving a real need, to me, seems like a better use of a really smart person’s time, and just more arbitrageurs of the market, I guess, is my advice.
Man: We have time for one more question. But where do you see the opportunities in insurance, given you investments in Oscar and Ladder [SP]?
Joe: Oh, cool. Ladder is an early stage life insurance company that’s really cool.
Yeah. Insurance is also a really interesting area. I think a lot of the big insurance companies here are aware that they could be using data to do their job a lot better, whether it’s life insurance or health insurance. Health insurance is interesting. It’s very regulated, so you can’t use data to do better stuff with individuals. But you can for groups.
You also can for cutting your costs. So, I mean, I think there’s a ton more to do in health insurance. It’s a trillion dollar market. I am very bullish on Austria. I think it’s gonna be very big. But hopefully there’s more stuff you see there that transforms it.
I think in property and casualty insurance, that’s obviously a very data driven thing. So the best run PNC one is the [inaudible 01:17:14] they combined, right? So they have by far lower loss ratios than other companies, and I got to know some of the people there. I think they are a very well run company, but there’s still a ton of data they could be using that they are not.
So you still have these really big insurance companies that are not really up to date in terms of the latest data sets and data platforms. There’s lots of room for distribution there, and I think it’s become an area that a lot of people in Silicon Valley are focusing on. So you should expect to see a lot more new insurance companies, as well as insurance technologies.
Robin: Yeah, that’s interesting.
Do you have a question?
We live in a very controlled world [inaudible 01:17:46] in the way we’re so used to [inaudible 01:17:48] and what you see with tech is it goes across boundaries. If you start [inaudible 01:17:54] out of that system, and insurance starts to be something else, [inaudible 01:17:58] like [inaudible 01:17:59], do you think the tension between national governments and private enterprises is going to increase, because they [inaudible 01:19:07]
Joe: I think that’s a great question. It’s something we talk about a lot with some of my friends in the Valley. I probably have kind of strange biases with this. Personally, my view is that the global center banks are all kind of collaborating to do something that they think is good right now that’s actually very destructive, in my view, in terms of how they are all basically in control of this credit system.
It’s kind of like everyone is on drugs, basically, is my view of it. They just keep giving us more and more cocaine, and eventually you’re gonna have to have a hangover. It’s gonna be a mess. So there’s a lot being written and discussed about this.
So my view is that it’s a really good thing that, because we have all of this new stuff happening outside of the purview of the traditional credit mechanisms and government control, you’re gonna basically take back the control from these crazy neo-Keynesians that are running the whole global economy right now.
But, yes. That is the question. There’s this power system that’s been developed over 100 years, and there’s all this stuff that’s coming up that’s gonna be a lot harder for them, and they are gonna have to regulate in totally different ways. I’m hoping they kind of ignore that for 10 or 20 years, because maybe then we can take back the control from them.
But, yeah. I think that’s a great question.
Robin: Do you think you will take back control?
Joe: I think by we, it’s actually the bottom up system. So I think the greatest wisdom of the last 200 years is the extent to which bottom up things, empirically with markets, can create wealth and prosperity better for everyone, and that whenever you do things top down, you end up having cronyism and you end up breaking stuff.
So my view is that it’s really good for the world overall to be able to get around this kind of control mechanisms that have been put in place for well meaning but stupid reasons.
So I think you’re gonna see a shift more out of their control in a way that’s positive, in my view.
Robin: So you think a president Trump would put [inaudible 01:19:45] in charge of the federal reserve?
Joe: I’m not gonna comment on that one.
Robin: Go on. Okay. Well, I like to keep things trim.
So one more question? Maybe short one?
No? Well, great. Okay.
Well, ending just ahead of time. Thanks so much.
Joe: Thanks, Robin.