 Welcome back to Building Tomorrow, a podcast dedicated to all things tech and innovation from a libertarian perspective. I'm your host, Paul Matzko, and with me is Aaron Powell. We are returning to part two of an interview with John Aristotle Phillips, the founder of Predictit, the aptly named political prediction market, and the college student formerly known as the A-bomb kid. Yep, sometimes the truth actually is stranger than fiction. Because if teenage designer of nuclear bomb who creates a betting market for politics isn't the best plot summer you've heard for an as yet unmade Coen Brothers film, I don't know what it is. But back to the show. I know there's a relationship between, I think Aristotle was involved with some of the planning for Predictit, which is a political prediction market, but it's based in New Zealand, Alvin New Zealand University. Explain a little bit about that. Sure. That's my favorite subject. This is of more than three decades in the business of politics. This is something to which I've contributed that I am proudest of. Predictit is just the coolest thing. It's also a hell of a lot of fun. Are you guys Predictit traders? We have some here in, I think in the building. Sounds like a lot personally. You guys are definitely on Predictit. Predictit is what's called a prediction market. It's like a stock market where you buy and sell $1, winner take all shares where two people, we put two people together, one person who thinks something's going to happen, the other person who doesn't think this thing is going to happen. Let's just take Trump impeachment as an example. So right now you can buy a share, a single share of Trump to be impeached in 2019. I don't know if it's maybe selling for $0.22. That means if you put down $0.22 that says Trump will be impeached in 2019, you get a dollar if it happens. Or you can take the other side of that bet and you can put down $0.78 that says he won't be impeached in 2019 and you get the dollar if he is not impeached. So if what happens is the opposite of what you predicted, you lose whatever you paid for that share. You're limited to $850 in any market, can't buy more than $850 in a market. And what's cool about it is the price fluctuates by the second. So there are thousands and thousands of people who are holding these $1 winner take all contracts and they're watching the news and engaging in trying to figure out in this case will Trump be impeached in 2019. And as the likelihood of Trump being impeached increases or decreases, so does that $0.22 share. It goes up in value, goes down in value based upon people bidding on it 24-7, 365. So that's predicted. And it's not an original idea to Aristotle. Prediction markets have been around a while. Robin Hansen proposes them back in the 90s. Intrade was a famously fun, but also eventually shut down by the Kamai's Future Trading Commission. It was an Irish gambling site. But making predictions or people think of the stock market sometimes is gambling. There's a lot of research that goes into whether you're going to buy a share and there's a lot of thinking that goes in research that goes into whether or not you think impeachment is going to happen or what you think the margin of victory is going to be or how many Senate seats the Republicans are going to pick up or the Democrats are going to pick up. All those things factor in. And what's fascinating about predicted to me anyway is that they're so darn accurate. They're not 100%. It's not 2020 vision into the future, but you're pulling back these curtains just a little bit. You're able to see a little bit through the fog and that's what prediction markets do. And the question is why? Robin Hansen and others have been researching prediction markets for a long time. And the question really is why are markets accurate predictors of future events? How is it that is it wisdom of crowds as some people would like you to believe? Is it insider information coming into the market and causing the price to go up or go down on something like this? What is it about it? We know that markets are efficient are highly efficient at distilling billions of pieces of information. We can be talking about pork bellies or orange juice futures or the price of Tesla stock. Okay. But we're happy to talk about great trading places references. That's nice. So but they distilled billions of pieces of information and these are generally about risk and value and that distills it into odds and prices. And that's what a market does. It's what a stock market does. So this is if you think of this as a market for politics where you can make all kinds of predictions and you can put your money where your mouth is. And then of course there's this whole comments board. You know we have more than a million comments for instance on some of these markets. It's great and it's great fun. And I still think you guys are traders. You're just not saying it. So this question of accuracy. So predicted is a large portion of the data behind the website electionbettingodds.com which was the it was like it in 538 in 2016 and then again this year where the everyone in DC is obsessively hitting reload on these two sites. But one of the things about giving a percentage happening of a one-time event like will Trump or Clinton win the election. That's a one-time event. And so if predicted says that Trump has a 3% chance of winning and he wins then predicted was still right. And if predicted says that Trump has a 75% chance of winning and he wins you'd still say it was right because these are like one time you can't repeat it to see you know out of 100 times running this election he won you know either 97 or he won. Because you said because there's no way to empirically test. Right and so so when you're talking about its accuracy how are you getting to the question of its accuracy if you're always in these kind of one-off things. Well that's the old gag about Washington everybody says it's a 60% chance that something's going to happen with that way you're right in either way no matter what happens. So when we think of accuracy first of all the premise of the question is 100% correct in my opinion. When you say something is 20% chance that it's going to happen you know that means you know 20% of the time that thing is supposed to happen right and the other times it's not supposed to happen because you can't repeat it and all this stuff. So what how do you calculate accuracy well there's something called Brier score which is what you know people study prediction marks like it's not just being it's not just the odds are correctly cited with what actually happens but how long in advance of the event and and for how long it's okay because there's there's a virtue to being right five minutes before the event to be on the right side of a bet five minutes before the event occurs but the greater virtue is being right on the right side of that same bet six months ago when you could have bought this share when everybody else was thinking the opposite was going to happen. So you know there are there are biases that enter into it when you're you know there's long shot bias there's all kinds of bias that and it's fascinating how investors think about these things and I'm not here to tell you I figured out how why why markets are as accurate I gave you my theory but you know I think it's probably a combination of things I think in I think information flows into the market rapidly they are obsessively people are obsessively you know hitting the website trying to figure out if the odds are going to change it's what's really interesting is you follow something like the Kavanaugh nomination you know hearings or that Kavanaugh would even be picked in the first place and you can really really see the market start to tip at various points almost in advance of what's going on in terms of the testimony and and and you also you know it's not just it's not just politics of course I mean if you can predict how interesting would it be to predict you know the the the month and the year or just the year when the first human being is going to walk on mars you know all that's all that stuff is like really subject to and and again you know prediction markets and markets in general and and forecasting elections it's gone on for a long with money behind it it's gone on a long long time like hundreds of years you said that there's a was an eight hundred and fifty dollar cap on how much you invest in in a single market in here is that because of regulatory reasons and then does that would we be more accurate if we didn't have those caps so people could dedicate even more money so yes and yes both okay so there's a there's a sort of regulatory history as to why which you guys wanted to talk about I'm happy to sort of shorten it for you but basically what we did was we New Zealand University of Victoria in New Zealand had been running a prediction market and again a good one we went to them and polling is broken as everybody knows right it's expensive it's hard to do it yeah people don't want to answer questions on the phone anymore somebody's got cell phones so it's really hard to get polling to get it right and if you can harness and our interest was well wow look at these guys at intrade and they you know they they set this up and also it's just a lot of fun I mean anyway you cut it so we teamed with Victoria University and said we want to bring this to the United States and the commodities future trading commission was was the government agency to whom Victoria took the request and got what's called a no action letter and the no action letter stipulate some of these limits you're it's a balancing act in terms of the limits serve a useful function although the limits could be and probably should be higher but the limits serve a useful function they keep you know the question is well how do you keep a whale out of the market if a Russian whale wants to come in and buy a whole bunch of shares and that happened with intrade and the irony with intrade is they they relieve the whale of more than a million of his dollars before the market settled back to the to the note to what it what it would have been otherwise so markets are very efficient at taking the money away from people who rely on fake news and that's that's part of that's one of the other things about it which is it's a potential antidote to fake news if you are watching fake news or consuming fake news and you're drinking the Kool-Aid you probably should not be unpredicted because they're they're people unpredicted are going to relieve you of your money so when we talk about prediction markets and they really I mean it seems like it was with with 2016 is when they became kind of a mainstream thing for at least for people paying attention to the the political races we we say well these are more accurate than like the kind of expert predictions are terrible so the people who go on CNN and pontificate about who's going to win about horse-ray politics are terrible and the polling is not that great or I mean it's it has it has issues that you raised but if those sources of information are not that great where are the people who are betting on predicted getting their information like what information are they acting on because they're not I mean these aren't like insiders in the political campaign usually they're regular people so that's a great question so what makes I'll reframe the question okay so what makes a super forecaster because somebody at that 3% it wasn't 3% of course but to use the hypothetical whoever picked the long shot at 3% did very well six months made a killing right so our can you isolate the it gets back to our original discussion about big data and databases and the like can you isolate the characteristics that make for a great forecaster we call them super forecasters and just see what they're what they're doing right and try to figure out what how does this person I mean so for instance and those questions haven't been answered yet we have a hundred and we had 150 researchers at 85 plus research you universities so they had you know economics department the marketing department the political science international relations department these these teams of researchers are digesting the anonymized data that we provide to them and trying to figure out what makes for a super forecaster how is you know is a fat cat in the beltway inside the beltway a better predictor of the outcome of the Iowa republican caucus than grandma to use her that example again who you know just reads the newspaper every day or maybe it doesn't read the newspaper maybe doesn't pay any attention to it but just you know she she has some other methodology for figuring out what's going to happen so what makes a super forecaster I don't can't answer that question for you except that they probably have really good BS detectors so this actually conversation reminds me about the debate within stock market prediction groups about whether or not you can truly beat the market right so there's the there's the whole move over the last 40 years towards index funds you know your Burton Malkiel ran the walkdown Wall Street even Warren Buffett saying that you know one of the most successful stock pickers of all time saying you shouldn't do this this is not successful over time you cannot beat the consensus of the market so there's like an there's an anti super forecaster sentiment that's been growing in stock market picking is the same thing true in political prediction markets are there folks who are skeptical of whether such a thing as a super political forecaster exists well I mean you're sitting across from somebody who's is a case study for how not to pick stocks so my 401 suffers you know so anytime I get involved every time I leave it alone it does fine when I get involved I'm sure something's going to happen pretty much bet against it so do as you say not as you do yeah so that's you know and I'm not allowed to to to uh to play on predicted because I've affiliated with Aristotle right so um so uh your question was what was so I mean do you see that I mean so you were talking about you're trying to find the super forecaster but is the super forecaster a myth a unicorn I don't know yeah okay I mean somebody's been you know people are making money right and people are losing money and over time the question is like the monkey right the monkeys throw the darts at the chart and the stock picking chart right so I wonder if so if you had something like predict it's you've got people who are logged in and making lots of bets on lots of things and I suspect that people who are really into it are betting on a lot of markets at the same time but so I wonder if you had something like that then coupled with the kind of you know almost creepy level demographic data that Facebook gathers about you if you could start like that question of then you could figure out is it the fact that grandmas in the midwest who make below a certain income are you know 40% more likely to predict the outcome of elections like you could imagine really interesting information coming out of kind of aggregating all this data that's being gathered the the analysis of what makes for you know accurate it may be that it's not grand it's not Sven Galli grandma it may be that it's that it may be that markets just function you know in a certain in a certain manner and they aggregate a lot of information I mean I don't use the that I don't use the term wisdom of crowds but that may have something to do with it you know there's there's there's some I think it's going to be fascinating to see what kind of research comes out of this just in terms of in terms of why markets are so accurate and they're not perfect as I said they're not 2020 yeah what's the I guess outside of the obvious value for investors who are participating in this at least the value for the ones who win and what's the long-term value in prediction markets in this kind of mechanism of so if we you know we get better at predicting the outcome of campaigns really far in advance does that create kind of positive externalities that's a great question I haven't been asked that question before I don't know I don't know I have to think about that one I can imagine off the top of my top of my head thinking of this being an advantage to the kind of political outsiders that we started talking about when we're talking about Aristotle where if you're not part of the two-party duopoly it's just assumed that the received wisdom is going to be you don't stand the chance but what that's not actually true the chance is rarely ever zero maybe it's only one percent or two percent or three percent but it gets you on the board a prediction market gets you on the board whereas received wisdom keeps you off the board entirely and some small yes maybe it's only a one or two percent chance but across thousands of elections across the country every year some of those one or two percent chances are more likely to pay off if we listen to prediction markets rather than to expert opinion well that that's that's possible so so there are 20 at last count between the billionaires and the and the senators and and the one of these there's 20 candidates running for the democratic nomination possible candidates this way it's probably 220 possibly there's 20 that that so we've got markets open on on them right will they run that is the market on will they get the not what when will they announce will they get the nomination will they win the chance these are all different markets that you can you can go in on and you can sort of see there's a bunch you know the two percenters right there's a whole bunch of people like clustered down there and it's shouldn't get up now that you know traditionally what happens is they wouldn't they give a good speech or have a good day or something file a good financial report and and and you know that would that would go unnoticed right unless they were anointed by the Washington press corps being the front runner or the you know the dark horse that's catching fire whatever that means right get the big mo right so but the fact of the matter is the prediction markets are are pretty adept at picking that stuff up and so it'll be really interested with this election cycle to sort of see how well they do in terms of and you know somebody may be in the front now and they may be at the back of that they may not even be a factor six months from now what do you say to I mean there are going to be some folks listening to this who are going to kind of viscerally recoil at the idea of betting on political futures I mean to some extent with with Aaron and I you're preaching to the choir when it comes to hey markets can efficiently access information it's a libertarian podcast after all but there are going to be folks who have a visceral reaction to that what do you say to the idea that this is kind of ghoulish this is unsavory that politics aren't something you should think of as a crass mechanistic financial you have a financial interest in that you should vote your heart not with your wallet or something you know I I can imagine lots of objections along those lines right so so those who do well in first of all I don't have a perfect response to the to the person who says it's you know there should be no money in politics you know that's an extension of that argument which is you know not only should you not make money but you should not put money into politics because you know you just encourage these politicians to do more of what they're doing right in any event there are lots of objections and and anything I'm involved in is never going to be pure enough for for to satisfy everybody or even many people but my my feeling is that this is in a this is one of those areas that that should be reviewed more as political speech if you decide to to put your money where your mouth is that to me is a pretty pure expression of democratic sentiment furthermore predicted does function and this is another area that's being studied right now so it's a little premature to give a to jump to this conclusion but if you ever come to the predicted parties we have you know it's debate night what they watch debates or they this election night parties or we get you know we we open a bar you have a bar which is the political bar in the state capital on election nights or nights when there are debates or something happening those places are packed and you will see a whole generation of people who really are have many of whom have not been involved in politics before but like sports fantasy or some of the other types of places where you've actually got some skin in the game you know all of a sudden there is just this avid interest in in hanging on every word in the debate because you're trying to figure out if donald trump is going to say you know crooked hillary in the tweet tomorrow so so predicted is a an engagement tool and i think one of the things that's going to come with this is we're going to discover that that especially the demographic that that does sports fantasy and does you know super bowl all this but i think you're going to see greater engagement by that that demographic of young people who otherwise you know it's a bunch of people talking about politicians a bunch of people talking about stuff is not relevant to my life and frankly i didn't i didn't see any of that tax cut but if you've got ten bucks writing on what's going to happen you're going to pay attention and it gets to another thing you just said which is you're in a market you're not supposed to vote with your heart you're supposed to vote with your head right and that's kind of what predicted does it's amazing if you guys or if your audience goes to predict it dot o r g you just put you know you can you can open an account as little as ten bucks if you put as little as ten bucks down you're going to see how it clears your head in terms of what you're making a prediction on whether you're not you know it's who the 2020 democratic nominee is going to be trump impeachment you start to think is that really worth 22 cents or should it be 23 cents maybe i'm going to put it in offer for 21 and and that's how prediction markets i think that's an element to why they're so accurate well and it produces a certain degree of another positive social externality just as say fantasy football encourages people who are previously casual consumers of football to learn more about the sport to be more engaged increases the level of knowledge about this venue the same thing it would be true in politics that the process of engaging expose and maybe expose them the new ideas right there was a certain way they were voting you know because that's what the how their family voted so they were party line voters in this way but now that they have skin in the skin in the game they're spending more and more time engaged with the issues learning about the candidate looking for their higher information who are so right about that in fact if i may just a second so people who the study shows that people who have money writing on a football game or 44 more times more likely to watch football the football games and in fact the season predicted or people who are involved in fantasy sports or betting on sports become voracious consumers of sports statistics you just have to ask espion about this right they you buy you you consume you you try to get as much of that information as you possibly can because you want to make the most accurate right what's the batting average what happens on a rainy day all that stuff the same may be true with political prediction markets is that you become real interested real fast even if you've only got 10 bucks writing on what's going to happen and the the the corollary of that is there's a reward mechanism built in as we said earlier if you're if you're making your predictions based on fake news you're you're not going to do so well i just sort of be the kind of doom and gloom guy though to counter that does it run the risk that if i have bet a bunch of money on a slate of candidates say that when i then go to the polls the person i'm going to vote for is not going to be the one who i think will govern the best has the best principles is the best the job but rather the person who i have predicted is going to win and so therefore my slate of candidates will pay off which would seem to be i mean it's it's good for my the my income but it's not necessarily good for democracy if that's the reason we're picking candidates that you may have a perfectly valid point i don't know if i i talk to people who who's bemoan the fact that they put money on the opposite of what they're going to do because they edge their bets so there's a that way i'm not going to be disappointed when my candidate doesn't win right so there's that thing going on how much of it is what you describe which is probably a lot and how much is what i just anecdotally describe to you i don't know but that's one of those things i mean wouldn't it be just interesting to know look our our who's the better who's the better predictor of the outcome of a democratic or republican primary campaign is it somebody that's you know a democrat or republican or independent or i don't know the answer that we're going to know the answers to some of those questions pretty soon though i still like the theory it's a urs fengali grandma in the midwest i think clearly every time um so while we move on for sake of time to your original title back before you were a ceo back before you ran for congress you were known as the a bomb kid right so and this is uh a quite a literal appellation you were uh as i understand that an undergraduate princeton and you designed an atomic bomb something that you know most undergraduates do usually buy their sophomore year i was on academic probation at the time so that's uh you know they had to do something um my thesis advisor was the eminent uh physicist freeman dyson that's one of those interesting fun facts to no one tell so freeman was told me when i set out to do this this is a junior uh thesis is what they have so you do these in your junior you pick a topic and you dig into it and you ignore your other classes and you're supposed to do so that's that's what this that's what this project was um you know freeman informed me from the first day when i asked to be my thesis advisor he wasn't going to be able to help me with any of this uh because he had security clearances and i frankly nobody thought i was going to succeed anyway uh you know it at coming close but yes that's that uh that happened and uh it was it was quite a roller coaster after that um it was uh you know the point was to see how difficult it would be for a terrorist group or a uh criminal organization or even a third world country to build a bomb which i didn't do i've designed it right and the the premise much of the debate at the time about proliferation it's eerily similar to debate now but much of the debate about proliferation of nuclear weapons was that you didn't need to tighten up security around plutonium and uranium the wrong ingredients for uh atomic bomb because it would take a manhattan project to design and then build a bomb and therefore we should be lax in exporting technology and materials around the world because places like pakistan would never be able to come up with an atomic bomb it was this sort of this hubris which proved not to be the case and and as a as a result of a lot of work by people who are very concerned about proliferation and been working the issue for a long long time especially you know in washington democrats and republicans um uh a um um some of the there were curbs placed on on the spread of some of the technology mostly plutonium reprocessing facilities this is a whole great subject for another podcast yeah i'm happy to talk about it uh you know this is like i had a bit roll in this whole thing but it was sort of a punctuation mark i guess for you know for the proliferation debate because once once i did what i did then it sort of took ripped the rug out from underneath the argument that you're not the secure physical material less so i have to ask you said you you designed one but you didn't build one yeah but could you have uh the questions could i have without killing myself uh sure yeah well there was there was that kid who built one in his shed the boy scout in michigan who built one in his potting shed in the 90s may have yeah irradiated himself pretty severely uh you know it would not have been uh a trivial matter uh if you had part of the part of the the unfortunate part about the whole thing is that the whole subject is that um even if you you know the the atomic bombs that were that were used over uh Hiroshima Nagasaki only consumed a fraction of the physical material so they were not what we would call efficient and efficient design a lot of the work so you know in a lot of the work around atomic nuclear devices is to make it and make it efficient so they can be smaller so they can go on the top of the warhead that sort of thing or unfortunately in a suitcase somewhere so the the the and and failure if you create a dirty bomb right if that's your intention you set out to create a dirty bomb you don't really you know you're you're you're striving to do something different than create an efficient uh device so uh there's you know there's still very good reason to uh to tighten keep security tight on these materials and there you know there's that personally i think we we were in a much more precarious situation especially with the with the collapse of the Soviet Union um it was an example of at the time it wasn't perfect it was an example of you know bipartisan support uh and international support to get these to to stop these from these weapons from falling into the wrong hands and and how successful how successful we were there's evidence of that because we're we're here today but you know i don't it sounds like something betting market on that's right yeah i don't think we'll avoid that one so that that um thesis made you something of a household name lots of talk show circuit appearances uh you're the first 24 year old not involved in fronting a rock band to get like a cover article in Rolling Stone as far as i as far as i know um and you you kind of leveraged some of that fame to run for congress we talked about those campaigns in Connecticut in 80 and 82 so what made you decide to switch from being the you know the front the the face of a campaign and then the next year after that second failed bid is when you start Aristotle right like so what's going on in your head as you go from you know young and i think by the standards of the time maybe not by the standards of 2016 2018 you were considered fairly liberal on a lot of policies in 1982 like how do you decide to go from that to running a bipartisan political campaign platform so so it was those experiences so first of all you know my brother and i were broke so we had to we needed a job and so the the thing we knew how to do was to run a campaign which didn't succeed in the general election but did succeed in the primary election campaign and and so it was it was a um we loved the campaign the the democratic process itself it's like it's fascinating we use how people come together how they try to articulate their point of view it was highly inefficient computers were just the brand new thing and and and we felt based on my my what my the program my brother had cobbled together for the campaign that this is an area where you know some there's room for growth and there's also room a lot of room for improvement it was just too hard to run a political campaign in those days if you didn't you know even if you had the blessing of the party it was just really really hard and it shouldn't be that hard it should be easy to you know to get your point of view across and and and then let the voters make a decide it's a one-day sale you let the voters make a decide as to how much merit your ideas have and that's the basis on which i know it sounds you know naive but the fact of the matter is that is pretty much what aerostyle is dedicated itself to you know in this intervening period of time is to try to make it easier for points of view which may not be mainstream or mainstream to be heard by voters so that they because ultimately voters make the right choice you know in all these elections we were involved in ultimately they they weigh you know their own personal circumstances but also what they think is best for the their community or their country and their neighbors and and that's how they that's how they cast their ballot one last thing i think unless you have something erin as i was reading the the rolling stone interview actually there was this this is the section where you talked i mean this is 1980 so it's been a while you talked about admiring like the great figures in history like joe de vark alexander the great napoleon you wanted to be like the ancient greeks who tried to make a substantive contribution in one's lifetime as a way of leaving a legacy that outlived you so let's talk about like i want to be preserving civilization rather than than you know destroying civilization um you also mentioned like a childhood dream of being an astronaut in all this so young political relative political radical for the time trying to get into politics in the late 70s early 80s these big dreams and visions and actually while reading that i thought of some another politician new gingrich someone who's also obsessed with space travel loves history it might sound like a bit more of a uh a uh i don't know if that's a compliment given the trajectory of his career but it's interesting to me that there might be something about your generation who are coming of age in the 70s so it's post new left post 1960s they have you had a big vision but also saw a lot of really systematic institutional problems in america and wanted to do something about it albeit i assume your solutions in 1980 were rather different from new gingrich's but some of that same energy i what that puts in mind be in mind now i think for the my final question would the you know 1980 24 year old liberal wonder kid who invented the atomic bomb and then ran as a radical democrat for congress like would he be happy with you now the 60 some-year-old you know oh i'm sure i'm sure grossly disappointed uh i was you know my my viewpoints when i ran for congress were would today be considered pretty much mainstream or conservative i mean that's the irony of it right medicinal marijuana and then so i you know i ended up uh look what i what i do look i think what i what i what i'm involved in what my company's involved in things like predicted they're there i think these are these are important sort of endeavors right and and i've been very lucky they're also i want to stress how much fun it yeah it is to be doing what i do i really enjoy being the CEO of a company and i really enjoyed it gives me the opportunity to stuff like predicted or look how fascinating it is to be able to go to a foreign country and be able to help manage a political campaign try to get your arms around what's going on and how i told you earlier on it's invigorating it's inspiring you see people waiting in the sun for hours to vote in an election where you know you're trying to make sure that the votes are all counted etc it's just it's great so i have i have no regrets about that i don't think i would have for the record i don't think i would have made a very good congressman so i don't think so either and that's a compliment all right yeah yeah and i do think the uh the ancient greeks would approve so i think here 19 version be happy uh and that's all for today and until next week be well building tomorrow is produced by test terrible if you enjoy our show please rate review and subscribe to us on itunes or wherever you get your podcasts to learn about building tomorrow or to discover other great podcasts visit us on the web at libertarianism.org