Completed technology to Bitcoin, and do several other amazing things as well. So, Paul Sztorc, thanks.

This talk actually is not about Drivechain, although I'm sure it'll come up in the panel at noon.

Okay, so hi, my name is Paul Sztorc. I have a relatively mainstream background.

I went to public school in Connecticut for 12 years, and then I went to a private engineering college in Ohio,

where I studied economics and statistics.

I then worked a kind of pre-doctoral research statistician job at Yale University for two years for a man named

William Nordhaus, who just won the Nobel

Prize in Economics a few months ago, or the Swedish Bank Prize if you prefer, and so my economic training is relatively mainstream.

But despite all of that, I fell in love with Bitcoin in 2012, and I published Bitcoin research on my blog

Truthcoin.info since 2014, and I presented at many Bitcoin conferences,

especially the scaling Bitcoin conferences, as well as building on Bitcoin last summer and TabConf

earlier this year.

The topic of this presentation is an idea for getting the public sector, and I put it there,

I put it there in a kind of scary red blob, to substantially inherit most of the reasonableness of the private sector.

And so this presentation is not about

replacement. It is not about replacing the public sector with the private sector via some mysterious process that no one will explain.

It is just about improving the public sector. So you have to decide how anarchist you think it is.

It's up to you, but it does involve markets. It's very Hayekian, so we'll see.

Here is the entire talk in one slide. The problem, politicians are installing bad policies and getting away with it.

Solution, make it easy to find the bad politicians and get rid of them, and

specifically, we're going to create some assets that whose value reflects the politicians' competence,

and then we'll make the pricing data available to the public, and they'll just vote for

whoever has the better numbers.

So

people might have this information available to them just before they vote, and so for each person running,

there's a column here, and it just gives forecasts about how life will be if the person's elected.

The first row guesses how much the government will cost. The second row guesses how economically prosperous the citizens will be, and

the third row just maybe would cover things like

invasion, war, health care, mental health,

gun safety, whatever. It just kind of catches all these things, and people will just look at the two numbers and decide which one they like

more. So where do these numbers come from, and why would anyone take them seriously, and that is what this talk is about.

So first, I'm going to give some background knowledge, and then second, I'm going to talk a little bit about

democracy and what I call electoral feedback, and

then I'm going to talk about these event derivatives, which are usually called prediction markets, and

fourth, I'm going to combine these ideas, and fifth, I'll explain why I think this idea actually will work, unlike other ideas that don't work.

Those last two are super short, take about 60 seconds, I hope.

So

the background knowledge that you need to have in a nutshell is that this idea is very old, and a lot of super smart people

like it, including many cypherpunks.

The second is that I've already basically implemented the idea, so it's not vaporware, and the third, there is no like scam

ICO, or any kind of like weird scamming thing in this project anywhere, which I'm going to explain.

So the first name I'm going to drop is Hal Finney. A lot of people think that Hal Finney is Satoshi Nakamoto.

He received the first Bitcoin transaction. He liked this idea a lot.

He's calling them conditional prediction markets. I'm going to call them multi-dimensional

event derivatives later.

The second name is Robin Hanson, who Ralph Merkle is name-dropping him as an important guy.

This is basically Robin Hanson's ideas just put on a blockchain, and

Ralph Merkle built most of the technology that Bitcoin uses, so

this is more name-dropping, which is that this was a paper that was put out

in order to try to convince the government to stop interfering with prediction market development and research.

It's signed by a giant list of people. All these people either like will win the Nobel Prize, or have, or should.

They're super, super well respected, and they all signed this paper,

but it's kind of going nowhere, hence the blockchain emphasis that I'm taking it sort of more

getting further away from the high road, because it's been shown to be impractical.

So the second thing is that this talk assumes that someone will solve the so-called

blockchain oracle problem, which is a technical problem. In order to make these assets, you need it to be solved. Fortunately,

I personally have solved it,

so there's nothing for you to worry about, and I wrote this paper about it a long time ago, and this paper has some pretty

cool endorsements. We've got Roger Ver up here.

We also have Andrew Poelstra, who I think is the brightest technical mind in the space, and

we have also Adam Back, CEO of Blockstream, and we even have a smiley face from Peter Todd.

So it's like basically as good as it gets, I think.

This paper that I wrote spawned virtually all the projects in this area,

some of which I'm more proud of than others, but the only one I really like is this one at the top. I saw it's

oversaw its development. Personally, it's a fork of Bitcoin Core, and it was assembled by two different

Bitcoin Core developers over a period of about 24 months,

and I run the project website where I put content like papers and this presentation, and

here's some screenshots.

There we go.

Some screenshots, which hopefully by the end of the talk will kind of make some sense to you.

So we have some events here. They're like in a big list, and then we have some markets. They have prices,

and then we have like a trading view, where you're like buying and selling things. So that's just kind of to show it's not paperware.

This project could theoretically be live on the Bitcoin BTC mainnet as early as this autumn,

but I don't, you know, you never know about that type of thing. It depends on a lot of

variables, so we'll have to see about that.

So finally, I just have to mention there's no ICO and no like weird utility token going on here.

There is a second token, which leads to misunderstanding, so I'm going to explain it in about three sentences.

The second token is more

analogous to mining. So in Bitcoin, the people who transact, they don't need to mine, and

in this project, the users, people who trade in these markets and create them, they don't need to know anything about the second token,

nor do they need to own any of the second token.

So if you only own that token, if you want to go into that line of work,

same way you'd only buy an ASIC if you wanted to go into that line of work.

Okay, so now the talk can really get started actually, for real.

So the question is, why is the public sector so disappointing? Because it's really everyone that hates it.

It's not just people who come to an anarchist

conference

in Mexico. It's just kind of universally, culturally,

it's just disappointing for people. And the question is why, and the argument I'm going to put forward has nothing to do with

morality or with lofty abstractions such as freedom or whatever. It won't even have that much to do with economics.

Instead, it has more of like a physics or biochemistry flavor, because it rests on this idea of feedback.

And this is an idea of feedback in the free market, where you have

bad products being sold and

consumers don't like the products, and they buy fewer of them,

which punishes the merchant who's selling them. So a free market doesn't mean instant utopia, right?

But it does mean that things improve, and they improve basically as fast as possible.

And

so the modern democracies that most of us live in, for better or for worse, also have feedback called voting. And

it kind of goes a little like this, where if the government is too unsatisfying in some way,

it can be dismissed with a majority vote. And it's a little more complicated than that. There's like some other weird stuff

that's not really important. But basically, this is this gear that

controls the machine

very slowly. Now, the free market feedback differs from electoral feedback.

We know they're different, because one works pretty well, and the other one is always disappointing.

So what are the differences? Frequency is one difference.

The market feedback fires every time a merchant loses a sale.

So that's many millions of times per day.

The electoral feedback fires once every four or two or six years or something like that or whatever,

whatever the election is. So that's one big difference.

But I think that difference is actually relatively trivial in comparison to something that I call research.

And so you see there's actually even more feedback loops than you originally anticipated, because

when a disappointing sale is completed, the consumer suffers. And the desire to avoid suffering

stimulates the consumer's search for alternatives. And I call that search effort research, for lack of a better word.

I don't know what else to call it.

But so, okay.

Now, in both sectors, public and private, you have to do your own research,

so to speak, and you have to pay for your research. Research consumes your scarce resources, especially your time.

Important things in your life, you research a lot. Buying a house, you research it a lot.

But picking out a hotel room, on the other hand, you do not do as much research. You only do a little bit.

Cosmetic surgery, you would do lots of research. Buying a new pair of flip-flops, very little research.

So the problem with elections is that your single vote will almost never affect the outcome, because there's so many voters.

And people pretend that their vote matters in order to feel important, but really, they know it doesn't matter.

So if you master the issues and the candidates and do a lot of research,

it won't change your life in the slightest, because other people are not doing any research.

So there's never any reason for you to do any.

Now, of course, this is a big oversimplification,

but I think it's basically the explanation for what's going on here.

Voters usually admit that they don't know what they're doing, and I have here 71% of Americans cannot even name their congressional representative,

let alone know what they've been up to or whatever.

And why bother learning who the representative is? Because it basically makes no difference.

So at breakfast on Election Day,

voters will carefully weigh the costs and benefits of different breakfast foods that they could eat.

But when they go in to vote, they won't really know anything about the costs and benefits of their choices,

so hence this idea to kind of give them that information.

Now, an implication of this theory is that the idea of educating people in order to make them libertarians,

or make them experts in economics,

will never be able to actually solve this problem and improve society.

It's very noble and worth doing, in the same way that chemistry and music are worth doing,

and like bird-watching is like a real skill, and mastering the skill of bird-watching could be fun,

but ultimately, if persuasion were effective, then the elections themselves wouldn't be so problematic.

Friedrich Hayek would have been wrong about information diffusion,

and a centrally planned economy would be more efficient than laissez-faire.

So the research incentives are the whole problem in the first place.

So now, what should we do instead?

As I said, the plan is to have people show up on Election Day, open their cell phones, and check some prices,

and it should be as easy as checking an NFL score, or the S&P 500, or something like that.

Then they just vote for the candidate with better prices, and this is just supposed to be a huge improvement of what we do right now.

So prices of what, though? What's going on here?

This is an example of an event derivative, which is usually called a prediction market.

It's defined as an asset that pays you money if something happens.

If the thing happens, you get paid. If it doesn't happen, you do not get paid.

This is an example from a website called Intrade.

The asset was created in 2011, and it would have been worth $1 if the year 2012 had been the warmest year on record,

as reported by these NASA satellites that measure surface temperature and put it on a website where anyone can find it.

Otherwise, the asset is worth zero.

So the asset traded at $0.40 in January 2011, but by mid-2012, it was worth just $0.10,

and in December of 2012, it was worth basically nothing.

So event derivatives are similar to bets or wagers.

The difference is that with event derivatives, the prices are constantly changing,

and this variation in prices produces socially useful information.

So in a bet, the odds are fixed.

And the odds are usually just fixed at 50-50.

I bet you $20 if this happens.

And where's the fun in that?

You know, with these assets, the market price is the objective likelihood of the event that the market is about.

So as 2012 went on, it was becoming clear that it wasn't going to be the warmest year, at least according to these satellites,

and so the price collapsed over time.

The topic of prediction markets usually reminds people of election betting, which is a good start.

These are assets that pay out if the candidate wins.

This right here is an example if a candidate won the 2016 Republican presidential nomination.

So after the Iowa straw poll, Trump's price in blue plummeted because he didn't do as well as people thought,

and then it surged after he won South Carolina.

Now, is this what I suggest that we do?

A lot of people think that this is what I want us to do, but it is absolutely not at all what I want us to do.

This is lame.

It is certainly entertaining, and it makes election years much more fun, especially for me, because I like this type of thing.

But these type of markets only tell us who is probably going to win.

Instead, we want to change who wins, and we want to sabotage the incompetent candidates so that they cannot win.

And how do we do that?

This is what Hal Finney was excited about.

Although, yeah, they could never really do it the right way, but they were excited about it back in, whenever that was, 2006 or something.

Okay, so, oh yeah, sorry.

I'm going to try to explain this idea of multivariate betting in nine slides.

And hopefully after this, the whole presentation will make sense.

And the first two are super easy, and four of them are really just one slide broken into four pieces.

So, here we go.

Okay, here on the left, I have a series of coin flip events.

In this table, we're describing the behavior of the 303rd coin flip.

We don't know what will happen on that flip, but we can nonetheless fill in the probabilities in this table.

Heads and tails are equally likely, so it's 50-50.

This is just a basic statement of probability.

And this next one is also easy.

It's just a dice.

It's the same as the previous slide, but with a dice.

Since each face of the dice is equally likely to come up, they all have one sixth.

Now, here's where you need to turn on your brain, unfortunately.

Sorry about that.

But this is where it starts to get complicated, which is that this slide is just the previous two slides kind of added together.

And we have the coin flip in green along the left vertical edge of the table.

And we have the dice roll across the top horizontal edge.

And I've circled some probabilities in red.

The probability that a two will be rolled is still one sixth.

It's at the bottom.

The probability that a tail will be flipped is still one half.

Those are called marginal probabilities because they're written in the margin, so to speak.

The probability that both a two will be rolled and that a tail will be flipped simultaneously, that's shown here as one twelfth.

And okay, so with that out of the way, we can talk about relationships between events.

And I'm going to draw a contrast between these two tables here.

Both are about unknown future coin flips.

The top table considers coin flip 304 in blue across the top.

And it considers flip number 305 in red across the left side.

Each cell occurs with probability .25, which is what I have indicated here.

Now, the second table does something extremely unrealistic that is only for teaching purposes.

It plots coin flip 304 against itself.

Now, in the real world, this would never happen.

No one would do this because it's just the same thing twice.

But the point here is that if it did happen, we could write certain numbers in this table.

And the numbers we would write would have zeros for things that are impossible.

So we zero out everything that's impossible.

And we see there's a kind of clumping here.

And I'm going to explain that again with this dice, which is also the same idea here,

that if it's the same dice roll against itself, you have this kind of clumping up in a line.

If they're perfectly related, they clump up in a line.

If they're just a little related, then they only clump up a little bit.

And if they are not related at all, they don't clump.

So that's really the synthesis here is that instead of doing things that we know are related,

we check and see if other things are related.

Here I have the—we just make four—okay.

Since clumping equals relatedness, all we do is measure these two things at once,

and we look for the clumping.

And if there's clumping, that means the events are related.

So we make four different event-derivative markets, one for each of the grids here,

and we get a market price for each, and then we just look for this clumping.

And that's basically the synthesis here, which is the short part,

which is that with some arithmetic you can just add two numbers and divide them,

and you can get out of this something that's very, very easy for the layperson to understand.

They'll never have to see anything complex.

They just see a bunch of numbers that go up when the candidate is better

and go down when the candidate is much worse.

And that is how you get these numbers to show up that I mentioned at the beginning of the talk.

And so now I have two minutes left, I think, and I'm going to talk about this last part,

which is why anyone will care about this.

So why would laypeople check their phones on election day?

Well, if they're already in line, then actually this is the first Google result for election line,

and there's already people who are just on there.

These people will be glued to their phones the whole time.

So that's a small ask.

People are basically on their phone all the time.

But why would they navigate their phone to these particular prices and not just ignore them,

the same way they ignore every other piece of information that falls under this category of research that I mentioned before?

Well, basically, let's see here.

In a nutshell, the voter turnout rate is super low,

and it's so low that the people who don't vote could basically form their own political party

that would defeat both of the other parties combined.

So you really aim for these people who are just totally overwhelmed and apathetic and don't know what to do.

And I think that really describes most people,

especially when you consider that fact about 71% of Americans not even knowing who their congressional representative is,

even though everyone hates the government.

So I think that's the only way to explain all those things.

The victory margins are very narrow.

So even if you don't get every single nonvoter to form a party, you could still do very well.

And finally, you can ramp this up by betting on other stuff that people like

until they get more comfortable with this institution, and then there you go.

And that's basically—oh, yes, one final thing, and I think I have like 30 seconds left, so it's perfect, perfect timing,

which is that this idea makes rhetoric irrelevant.

So, for example, let me skip to bullet point number three.

If a politician announces that they want everyone to die and that they plan to kill everyone, that's fine,

but if the traitors don't actually believe that they will follow through or they will or can follow through with this,

then the number won't actually budge an inch because you're speculating on stuff that will happen.

The derivatives only pay if it actually happens.

So everything that politicians say can just be safely ignored.

And laypeople already know that politicians are liars, so I think they will actually warm up to this idea pretty quickly.

Okay, that's the talk.

Thank you.