Projects
![]()
No projects found
Investors
New
DAO Treasuries
DATs
Lend / Borrow APRs
Chains
Categories
Spaces
Protocol Spotlight: Kleros
Recorded: July 8, 2025 Duration: 0:58:19Space Recording
Short Summary
Kleros is making waves in the crypto space with new project launches and partnerships aimed at enhancing decentralized governance and identity verification. The platform is also exploring innovative funding mechanisms through grants and prediction markets, signaling a significant trend towards integrating economic incentives with blockchain governance.
Full Transcription
Music Hey, good to see you, Rex.
Thanks for jumping in here.
We should be on in just a minute.
Clément's coming in as a speaker.
We'll have everybody getting back after some brief technical difficulties.
Looks like we're online.
Clément, how you doing?
Can you hear me?
Yeah, I can hear you.
We're good and you.
Perfect.
Good, good man.
I'm glad we could get this to work.
I think some of the difficulties are probably on the video side.
I haven't used video for too many spaces, but I've started to because I feel like it's
just so much easier to follow um you
know and and stay engaged and there's like a video going on at least for me it feels more like a
like a friendly zoom call than like a podcast with an audience and i really like that a lot better it
it uh it's more fun for me so um but anyway we'll uh just do this one on audio and uh talk
about all things claros and what you guys have got going on and we can start it in uh another minute
or so when you're when you're ready i'll um. I'll get my other personal account on here
just to draw a little more of an audience.
But we can start in like a minute
and wait for some people to filter in.
Yeah, I also retweet the new space.
Yeah, perfect, perfect.
And of course, also audience, if you guys want to see anything
or hear anything in particular like if you want us to cover some topics please post them in the
comments because we're happy to dive into those topics that you guys want to hear more about
i'm going to be asking about uh the economics of Kleros mostly and talking, you know, meaning the demand side mostly of the token.
And then we'll also probably talk about, you know, how economic security transfers from Kleros as the base layer to some of the higher level protocols.
layer to some of the higher level protocols, talking about what design was like for Kleros
coming up and what applications are being built on top. I think that's going to be a particularly
exciting thing to chat about because I'm genuinely interested and I think the audience is going to be
really interested to hear what you guys are cooking up up with seer and proof of humanity and futarki
among potentially other things so i'm really looking forward to that but yeah if you guys
in the audience have any more topics that you want to hear covered just post them in the comment
section and we'll get to those this. Thank you. Продолжение следует... Thank you. All right.
I'm going to post so I get some people who may come back.
Let's hope we didn't lose most of the enderms due to the bug.
Yeah, cool. I'm just retweeting right now as well. And I'll just make my other account a speaker here just so that it shows up as like we're speaking live for as many people as possible. But I'm ready to start when you are. Are you good to go?
Yeah, cool.
I'm just retweeting right now as well.
I'm ready.
I'm ready. All right, cool. So it's funny, I've been thinking this for a little while and
got a DM from one of the audience members. It said, Kleros are some pretty based Frenchmen.
And I've been wondering, like, why, how is it that so many french people are building like epic protocols like why why why are the french
over represented uh as like a percentage of based protocol designs like have you thought about this
claymont surely you've noticed there's a pattern of people building like really cypherpunk and
like half of them are french it seems like. What do you attribute this to?
Yes, there is definitely a bunch of French.
I guess it may have been because we had a pretty nice community
in Paris early.
And you can see the biggest conference beside Defcon
slash Defconnect in Ethereum is definitely THCC,
which is a French organized conference.
Yeah, like back in the day, 2017,
we had this kind of like French crew in Paris,
a bunch of people who are now more like quite famous,
you know, like you have BTC chip,
you could have a backup of Ledger.
You have Mark Zeller working for some Amishan, Ave,
we have Merlin of Morpo, and sometimes you think it's completely linked,
by the point Merlin was an internet terrorist before, so when you have a first batch of people who are all in the
French crypto space and then they train a new batch of new French guy in the crypto space,
it's a bit like reproduction in some way, you know, like you get a French which is teaching
two French, which is teaching two more French and that's how you get a nice ecosystem. I don't think we got really systemic issues like in Argentina.
In Argentina you have very strong community because they already understood or fucked up fiat wealth,
like with the inflation that they got.
In France we had a mainly functioning ecosystem, a mainly functioning banking system.
There is also maybe that France may be on one hand quite taxed.
And so you get some sort of like a movement which is a bit more libertarian,
which is not taken by the political spectrum. Like the libertarian party gets almost nothing in France.
And so instead of going into politics,
they may go into, you know,
like those kind of cypher-crime kind of things.
That's also a possibility.
And on the other hand, which is real,
you also have the state which is supporting crypto.
Like it seems like at some point, you got some grants from France, from EU.
So you may bother the state by being active in some ways to support the project and by being too active, incentivize people into making projects which are more like directly fighting against financial institutions.
So yeah, that's a good show of hypothesis.
But it's a very good time that I'm thinking about it.
So I'm not sure what I'm telling you.
Yeah, no, I mean, I've just been thinking about it this week
as I was anticipating this space.
Like, there are so many interesting game theoretic
innovations or protocol designs that have come out of France in crypto. But even before that,
you guys created parimutuel markets. That was a French concept. Parimutuel is spelled weird,
and it's because it's French. So there are some applications in the synthetics ecosystem that use pari-mutual markets.
It's a type of betting market.
And then there's a bunch of different voting methods,
like the Condorcet or Condorcet method
that is like pairwise.
It's like every other, I'm going to butcher this,
but every other person in a pairwise matchup can be eliminated through this creative way of voting.
Basically, you have to win all your duel to win the Condorcet vote.
Yes.
That's what it means to a consensus candidate or proposal.
to a consensus candidate or proposal.
Yeah.
And so there's all these innovative voting primitives
or game theory games that have come out of France.
And so I think it's very fitting
that this innovative protocol,
Kleros, has also come out of France.
And then you've got other French founders,
like you mentioned,
and there's also Aladdin DAO with F of X protocol and all of that kind of France. And then you've got, you know, other French founders, like you mentioned, and there's
also Aladdin DAO with F of X protocol and all of that kind of stuff. A lot of, a lot of really
great people coming up with brilliant ideas out of your country. So with that, I want to jump into
like what Kleros is. I feel like this is a very enlightenment. I don't know. When I first heard about Kleros,
like maybe four or five years ago, I want to say, maybe it was two or three, but this idea of like
a decentralized court or decentralized jury like blew my mind. I was like, oh shit, we can do
an alternative to the legal system, a censorship resistant
alternative to the legal system on a blockchain.
Um, it's incredible.
Uh, do you want to like intro Clara says as something else or like, how would you describe
Clara's for audience members that might be hearing about it?
Yeah. So you can see Clara's as an arbitration core.
If you want to frame it within the classical governance mechanism,
or you can frame it as a subjective oracle if you want to frame it within a more crypto ecosystem.
In those cases, it's about being to answer questions. So you have some people that disagree
I would need to answer questions.
So you have some people that disagree on something, and this something should lead to different
actions in a smart contract or even in an option system.
So they disagree on that.
So then they're going to have a dispute and you will have the URL.
They will look at both sides of the dispute, they will look at evidence, and then they will need to determine which side is the most correct in this dispute,
and then they will vote, and then they will provide a holding which can then be executed.
So it can really range from very small disputes,
like for example, you are curating data to tokens.
We have this very vast list of tokens in curate where anyone can come there, put a small deposit,
put the information related to this token, that image, address, website.
And if someone complains that this information is incorrect,
then you have a dispute.
So that's a very small dispute, more akin to Wikipedia
in it if you wanted to see a parallel in the Web2 world.
And you can also have a very large dispute
where you may have a million dollars in the prediction market
and people will disagree on what actually happened.
And then in this case, it also has
to determine what happened and which side should be paid
out by the prediction market.
And obviously we have a bunch of different, it's like a scale, you have a more or less
high value dispute in the scale of the rules that you can pick up.
Yeah, I like how you distinguish it and you categorize it as an oracle where it has scalar
values right like it could be it's not just binary it's not just like uh did zelensky wear a suit or
not it's like what color suit did zelensky wear or like how much money did the suit cost or like
something that is much more um uh race you know much more toward like a ratio type data type rather than like a binary data type.
And so that makes Kleros maybe the best Oracle for certain use cases.
I think we could branch off into the current UMA polymarket dispute,
but I want to save that a little bit for later after talking about economic security. But like what use cases is Kleros best for?
Like any use case where you have some truth, which should be more or less objective, or
at least most of the time could be objective, but people could misreport it. So you have this example of an escrow,
you want to pay someone, you need to do some work, you need to determine whether or not the work was
done. You have those creative registries where you want to create some information and you want
to determine if the information is correct. You have those prediction markets where they can report
correct. You have a special market where they can report
election winner, amount of TDL, then people can disagree or
even be malicious in their report and then you need to
arbitrate that. So you have, again, you have an
example of a spectrum where on the complete left, you can
see something which is purely objective. Like for example, if you do a Uniswap trade, that's something which is purely objective.
You look at what you put in, you look at what's in the liquidity pool, and then you can compute what goes out.
So here you don't need arbitration, you don't need scenarios because it's completely objective.
Then you have the complete other side of the spectrum, you can put it in the right of the spectrum, you have pure governance where we ask, should we execute this proposal
which is going to hire a marketing company. This is something which is purely subjective.
There is no right or wrong answer on whether or not you should hire a company.
Or if there is a right or wrong answer, maybe it exists, but no one can know it yet.
Maybe some people could know it in the future, but in the present, no one can know if it's
a right decision from a good and a good point.
So this one is purely subjective.
And it's useful for everything which is in between, which is mainly objective, but still
has a bit of subjectivity so that you can always compute the data chain.
It's interesting because I don't know of any other protocols that are doing something in
quite that way. I will want to talk about how you would solve the unknown and purely subjective
you would solve the unknown and purely subjective questions.
I think we'll talk about Futarki in a little bit, but I am curious about with this design,
I'd imagine this was initially very data intensive and that would make it gas expensive on the
expensive on the Ethereum blockchain.
Ethereum blockchain.
So how did you navigate like minimizing the amount of data and gas that it
took to put this protocol on a blockchain?
So you only have dispute, uh, when you actually have two people disagree.
So most of the payroll value, is provided is not during dispute,
it's just because since people know that if they misreport something there could be a dispute,
they're not gonna misreport it in the first place. So like the parallel in the physical world would be
since you know that if you murder your neighbor the the police is going to come get you, you don't know you're going to be born in the first place, even if you really want to write.
So 99% of the clearance value is provided completely silent, just because people behave
in case of dispute.
So in this case, there is no gas, which is failed because there is no dispute.
No, if there is a dispute to limit,
it's not just gas, it's more like risk of consumption
in general, we use an appeal system.
So you're gonna start with a small amount of jubals,
maybe like three initially, five, one,
independent of your system.
And they're gonna vote,
and you're gonna have a first instance ruling.
And if people believe that this is wrong, then
they can put more deposits and they can appeal to get more jurors every time you double the number
of jurors and you can reach something like 500 jurors in the end. So it's made such that you get
the full security of Kleros, but most of the time you only draw a portion of the Kleros
duval in order to be scalable so that you don't need to have everyone both and everything.
So is it an escalation game like where people just override, like what happens in a dispute?
So you could see that an escalation game, I guess if you talk about the escalation
game, you're probably thinking about early augur and ability. But here it's not purely an escalation
game because you still get some preliminary rulings along the way. If you just get a pure escalation game, maybe you say, OK, you put $10 in CS, $20 in Snow,
and you keep doubling every time.
You may end up having like 100K or 200K or $1 million
in this escalation game.
And even if it would work in theory, in practice,
it's a very bad user experience because one of the side may not know that they are very likely to lose.
So they would end up losing a higher amount of money.
Or on the other side, one side may actually not know that they would win.
the other side, one side may actually not know that they will win, and because they will not
know that they will win, they will stop putting money and then the mechanism will resolve by
default, without asking the final oracle to hold on the questions. So by having this escalation game, but being with appeal,
you still get feedback on each step.
So maybe you have the first ruling with three jurors,
and you see you have two to one.
So at this point, you may be, okay, I'm in the one,
but I may still have some chance of winning,
so now I'm going to appeal.
You get the second round, now you get
five to two. You can say okay maybe it's still possible. Now you get another round and you have 15
jurors and you get let's say you get 12 to three. At this point maybe okay like I'm
quite unlikely of winning and I'm gonna cut my losses and I'm gonna stop there.
quite unlikely of winning, and I'm going to cut my losses
and I'm going to stop there.
In this way, you also don't consume such amount of resources
that you would have consumed in a classic escalation game.
And you can also combine it.
Like if you use Reality with Kleros,
you can combine both of those parts.
So you can have an escalation game at start
and then move to Kleros as soon as the value in the escalation
game is higher.
And that's what was dispute work with reality.
Yeah, I like reality as an oracle.
It's a really cool approach, and Edmund is doing some cool stuff over there.
That's maybe a good segue into some of the other things I want to talk about.
I do want to talk about other protocols
that are interoperating with Kleros.
Can you, like, I would love to dive deep
on the token economics,
but I also think people could just check those out
on your website and like understand those
and think about how the pink token,
do you call it pink token?
Oh, PNK, yeah. Yeah, PNK.
But I'll leave that to the listeners to check out on their own
because it is a cool token economic system.
But maybe you could give us some insight into the other protocols
in the Kleros ecosystem.
Like, how is Kleros used in practice?
So you get curation.
So you get curate where it's created list for everything.
So you want to create some information.
You can create some token information.
You can create information about some smart contract.
Find the decision smart contract of this particular protocol.
Is it supposed to be used on this particular front end?
And that seems which are picked up by Etherscan,
the Klerot Metamask Snap, and Ledger.
So you have also Proof Humanity,
which you could not see as a created list,
but this time it's a created list of humans.
So you want to verify that people can only register once,
and if they try to register multiple times,
they get challenged, and then they cannot make it into the list.
And you can use that basically to give a privilege to people.
So you had this idea of giving some UBI token,
which was a bit successful at start.
But I think the way that it's going to evolve
is that UBI doesn't need to be a token,
but you can just give freebies as a protocol.
So for example, in Kairos, you want people to be jero.
So we could say, hey, if you register in Proof of Humanity,
we're going to give you a small amount of P&K that you're going to be able to state.
And if you do some good job, you're going to get more P&K and get more input in the system.
Or in a prediction market, we say, hey, we're going to give you $10 of trading credit per month.
And then if you want, you can trade with more.
But at least you get $10 of trading credits so that you can get some experience and you can also provide value in terms of information, even if it doesn't cost you anyway.
What we're going to have also with CIRCLE, we are getting this UBI token, which is CIRCLE.
We're going to allow to work it as the RC20 if you are part of the proof of humanity.
as the RC20 if you are part of proof of humanity.
And so, like, how can Claro's scale to meet that many?
Like, hypothetically, could Claro's scale to, like, 8 billion attestations?
Or, like, what am I trying to describe? If it's worth $50 to get a proof of unique humanity,
like, would people not just symbol the Kleros jurors
to get as many proofs of humanity as possible?
And if not, why not?
Like, how does that work?
No, because as I was telling previously,
it's an escalation game.
So even if in the initial dispute, you
may only have one or three jurors,
because those jurors know that in case of appeal,
they would be more jurors.
They want to be coherent with the potential results of appeal.
Because they will still behave.
And if they don't behave, you can appeal.
And they will be losing 10K to jurors which
are behaving correctly. OK, so it costs a flat 10K to draw which are behaving correctly.
Okay. So it costs a flat 10k to be a juror or like that's a configurable amount. Oh, PNK. Sorry. It costs some PNK. Not 10k. I think it's probably a stake in court with something
like around like $50 or something, I guess.
Okay. Okay. So maybe this is a good way to deep dive the token economics a little bit. So in the scenario where there is like a, you know, everybody in the world needs to get proof of humanity.
Like, how does the verification work for that with Kleros?
Like, how is the token staked?
How much of the token is staked in this example?
And how are the jurors arbitrating disputes there?
So again, most of the time, you don't call Kleros.
If you want to register on Propho of Humanity,
you need to find someone to watch for you,
which is a bit similar to staking someone else's humanity in the list,
plus some ETH or some DAI.
Then only if they challenge you,
then it goes to Klaros.
Most of the time.
You don't stake KNK as a user of Pro-Humanity.
Like in general, you really have two systems.
One system is an arbitration system,
which is driven by PNK.
But then after you have some application,
which can refer to Kleros decision,
but this system doesn't require PNK.
PNK is really forsaken as a job.
So most people who are using PNK and benefiting from PNK
do not need to have any PNK.
Like if you're on a prediction market,
you don't need PNK to trade there, right?
But then if there is a dispute, then I call Jero with PNK.
Right, OK.
I see.
So it's the fallback settlement layer to a proof of humanity, where it's only going becomes profitable to make a dispute and then push
that through in a Claro's court?
And is that possible?
The threshold is very low in particular because you can assign the cost of putting the mechanism
to whomever is losing the dispute.
So even if you had a dispute which were to cost more than the value at stake,
it would still work most of the time. Because if you know that if you make a dispute and you lose it,
you will lose your omnipotent fee, you're not going to make a dispute in the first place.
And even if the dispute fee are higher than what's at stake and you think you're going going to make a dispute in the first place. And even if the disputes are higher than what's at stake
and you think you're going to win,
it's not a problem because you're going to get them back.
So you obviously need to have both parties
to put them as deposit,
but then you can give it back to the party who wins.
So there is not really an issue about small disputes.
And also, how are the jurors selected?
Like, is this a random pool of jurors that are
arbitrating uh any particular like a particular question yes so you have two elements so you have
one element which is uh same selection in that uh who select in which court they want to be drawn
so it could be a pro-un court, it could like sometimes in Spanish.
We have some court which require people to speak Spanish.
So you select the courts you want to be drawn,
you state some KNK, and then when you have a dispute,
some random KNK are going to be drawn,
and if they belong to you, you're going to be a juror in the case.
And so to properly symbol this system, you would need to be a juror in the case. And so to properly symbol this system,
you would need to be like a majority of jurors
and get selected in the case that you're going for.
So it's not pretty simple in the system because it's,
your chance of being selected is
proportional to the token that you have.
So splitting your token into
multiple accounts will not be the extrait.
That's a really elegant solution
because you end up with the sort of the same assumptions
that underlie Bitcoin where the,
Satoshi calls it the golden goose problem
where it's like the miners have a stake in Bitcoin itself
because they have a bunch of ASICs
or they have a bunch of mining equipment
and would
prefer to keep the value of that mining equipment high so they don't defect. So you have the same
kind of system at the proof of stake in PNK tokens where people are putting up a lot of PNK
in order to be selected as jurors and so they're automatically already bought in and it's not like they're buying in
to a specific case and putting themselves on a specific jury to arbitrate an outcome of a case
that they're invested in. Yeah, exactly and if there were to be broken, give bad results, then TNK would go toward zero.
So there's some incentive to give good results as a system.
That makes sense.
I'm only superficially familiar with this, but a few years ago, I remember A.V. Eisenberg was somehow trying to break the Claro system.
Can you give a recap of what happened there from your perspective?
So it was about an insurance mechanism.
And what A.V. basically tried to do, he tried to abuse the insurance mechanism.
So his end goal, and this one was never really published
because he didn't reach there,
but he actually even talked to me about it,
which is pretty fun,
was to put some of his money on the lending protocol,
withdraw, not withdraw, like withdraw it from another address,
and then claim that he could not withdraw from his first account.
He claims that he cannot withdraw from his first account.
In this case, as the insurance to be triggered,
be rendered by the insurance, and after when he gets reimburse,
he can put the money back from the second account,
and he can withdraw it from the third account.
So he never went up to this point because it actually is pretty good, you know, like, he's probably the best exploiter of crypto projects in terms of people using non-hacking me, he doesn tried to do a smaller dispute with a very low amount of money, where
he had sent some money from one layer to another one using a bridge.
And this bridge took longer than expected to give him the money on the other side.
And then he was claiming that he was entitled to reimbursement due to the variation of the
value of the token that's sent from the time that he sent it to the bridge to the time
that he got them on the other side.
So this was like a very, very spurious kind of claim.
But the value at stake was very low.
It was maybe like 5 ETH back in the day.
So it was way, way, way, way, way, way than what it's now.
And I think his goal was to try to get some sort of case law that if the insurance terms are not super clear on something you should roll in favor of the insurer and then that he will try to use that to win the second claim where he can plan something like $2,000 which would have been a significant amount.
2000 is which would have been a significant amount.
Some part is not completely verified because some part we cannot prove it directly but we have
very good circumstantial evidence that he tried an attack at three points at the same time.
So on one hand we got a DDoS of IP, where they were trying to put heavy data in IPFS.
And I think the goal was to basically make Perl's front-end crash
and evidence not to be displayed.
He bought a bit of PNK also,
and I think his goal was to win the first round of the dispute,
make it easier to win
than the other ones and then he got into like a big like let's say social media campaign
on the Kairos Telegram groups, Twitter etc to try to to argue for his case and you know like
the famous thing like if Kairos then vote for me Kairos is broken kind of narrative.
like if Kheros then vote for me Kheros is broken,
kind of narrative.
And the first two had actually pretty little effect
on him, a feud of winning,
because we managed to fix IPFS issues
such that we were not pinning element
which we are not using Kheros.
He still created a bit of disruption
in Proof of Humanity, because at some point we had some upload which failed It still created a bit of disruption in pro-humanity
because at some point we had some code which failed
because like most of the space of our nodes were used.
And now we have obviously waste stress on the system
and there is no decorative running at the best node.
So that should make it less likely to happen in the future.
And then yeah, even a bit of PNK,
so he got like a slight extra vote on the case,
but that was a small amount
compared to the amount of PNK which was safe.
It was mainly like the last element,
which kind of like managed to get some attention
and even some people who may end up believing him
in some way because
it's not like something so easy.
You talk to me about the UMass shoot.
The UMass shoot is pretty easy.
You see the picture and you can get your mind.
Okay, that's a shoot, but that's not a shoot, right?
But here it was like some insurance document which was explaining what was covered, what was excluded. You had to
see exactly what happened. So you had to look at some on-chain TX. So that was the case
which was way harder for SNL-Observer to get an idea of who was right or who was wrong.
And so you managed to get a smaller following in that. Like, you know, that was back in the day.
Like now, like I don't think anyone is saying, hey, Abraham was right. Like you know, that's what back in the day, like now, like I don't see anyone is
saying, hey Abraham is right, like you should have won his payout dispute and also
gottent in 2000, he's that he would want to try to get later. Abraham is an interesting guy, you know,
like he sees himself as doing wonderful trading opportunities.
And that's actually also what was ruled.
So he got in jail for the child porn, which from what I've seen, he was definitely a pedo.
But he actually managed to get away with the mongo market so yeah what's even what he did in terms
of economic attack ended up being legal even if de facto by doing that he attracted the attention
of the authority and he ended up yeah yeah no i i um i'm curious on your thoughts on on economic
vulnerabilities generally especially having designed something robust that has been
attacked or attempted to be attacked in this way but i feel the same about av uh like very impressive
probably the best economic exploiter in the game who is no longer in the game because he got put
in prison for whatever it was 10 years for child porn um i'm really happy so i'm happy that he got off what's that i think
he just probably have no oh wow see i i felt like i think he should have gotten off completely for
the uh economic security vulnerabilities because those are those are design flaws that were
intended and like i mean this just goes into like my code is law philosophy, but like, if we
can't stop North Korea from doing it, we should at least make it legal for the White
Hats to do it or other people to do it because it doesn't matter either way.
And it just makes DeFi stronger if those things are actually executed.
But a longer conversation, but I, I, you know,
I would put him in jail for life for being a pedo. That's kind of my,
that's my take on it.
It's like give him a heavier sentence for the one thing and let him off the
other. But, but I agree with you. Yeah. He was prolific at,
at what he did. And I'm curious,
how do you think about economic security for applications?
How do you see this being, how do you think of those kinds of vulnerabilities if you think
of them different from technical or how would you classify an economic vulnerability?
How do you identify them?
How would you define that even? Economic vulnerability is fundamentally different from
smart contracts vulnerabilities in
the way that they cannot really be proven unless you exploit them.
If you have a broken function in your smart contract,
and you can steal the money,
you can actually write a test case and prove the issue.
Now, most of the vulnerabilities,
at least the non-trivial one,
you cannot pretty prove that.
And that kind of like depends of like parameter,
the state of the system.
So it's harder to prove.
You can come down and you can say,
hey, like here you have an issue
because if we assume these, these, these, and these,
and you have a profitable attack there.
Yeah, it's like highly stated.
It's not something where you can put your proof.
And people, they would have different kind of like level
of economic security where they would be comfortable with.
Like, for example, you have the guy with Ogur,
where say, oh, we should always make sure
that the value of the reptile can is higher than the value in the prediction market, right?
Okay, we can do that.
But in practice, it's working even if the value at stake is lower than the value of the
reptile can.
Reptile can be pretty similar to KNK in this case in terms of the rep token. Rep token will be pretty similar to PNK in this case,
in terms of economics.
Yeah.
In Rack City, it does work.
We can have cases with value which
may be way higher than the current PNK market cap.
Because if you want to attack it,
you cannot just buy PNK at the current price.
Because you're going to have slippage.
If you want to buy 50% of the PNK supply,
a web first is not going to be on sale on exchanges. So you're going to try to reach out people individually and just like you will basically pump PNK price by doing so. And then you are very likely to end up with a market cap which is going to beM, like the value of E-stake may be
lower than the total value of asset on the protocol.
And it can still work in practice,
even if for people who are purists in theory,
they will claim that it doesn't work.
Yeah, I think it's an interesting distinction.
You brought up a couple of things,
like economic vulnerabilities as being distinct from technical and that you can't prove them.
I do actually, we've been doing this at Token Dynamics with our economic audits.
So we've been working on what I call state-based modeling and blockchain dependency and state modeling is like our acronym that we've been using at Token Dynamics for this.
Because I think you can prove that there are profitable vulnerabilities if you look at the
state changes. And so like what you're saying, I totally agree with it. It's like, if you have
this assumption about the state and there's like a certain liquidity depth in this one token that you can manipulate, then you can find a profitable vulnerability or a,
in A.V. Eisenberg's language, a highly profitable trading strategy, which, you know,
essentially just arbitrages the access permissions in a blockchain database for one another. It's like somehow, I think the thing that AV picked up on was that you could change the
rules profitably, not just change the token balances profitably.
You could change that, like an economic exploit is where you change the token balances in
order to change the rules or like the access permissions themselves, which is what tokens
essentially are. They're just right access to a certain part of the database. But there is a
difference, like you're saying, I think you're hinting at the latest discussion around, do
stable coins provide Ethereum with economic security,
or does Ethereum provide stable coins with economic security?
And that's almost, it's a little more, it's a little different.
App layer economic security, like you have in Kleros is a little different because all of the tokens are functions of this wider ecosystem.
The value of tokens are functions of each other. um you know uh like the tokens that you're using in second layer protocols that might interact with
claros in their you know in different denominations um and they all affect each other so you have to
look at each of the states of the systems that are related to the ones that you're modeling
to find economic vulnerabilities yeah so i think he's going to assess economic vulnerabilities. Yeah, so I think it's just to assess economic vulnerabilities.
Transp, probably the only tool that you may have
is the Lindy effect.
So the idea of the Lindy effect is that
the longer something has been alive without being broken,
the more likely it will stay alive and not being broken
in the future.
So the rationale is you have something
which have a lifespan, right?
Like until they get broken.
And you don't really know if you are at the start
or at the end of this lifespan.
So on average, you are at the middle of the lifespan, right?
So this means that if something has been
working well for five years, they are expected to on average work well for five more years,
because on average you are at the middle point between something working and not working.
So you can use a lindy effect. And something that we would want to have in the future would be prediction markets for
risk evaluation.
So we could ask, out of all of these protocols, what is the likelihood of each of them being
broken?
And classical prediction markets can't really do that because, fortunately, most protocols
would be way more likely not to be broken than to be broken, at least on the short term.
You will get something like 97%, 3%.
You have issues with risk capital or risk, which makes 3% not very significant.
So what you can have is you ask which of those
protocols are going to be broken in the time period. And then
the prediction market will reward all the order of tokens of
broken protocol by dividing the payout between them. So let's
say you have 100 protocol and people can buy tokens of each
protocol being broken. And maybe now you have five of them
which are broken within the period.
Each token of the broken protocol would redeem for 20 cents.
And you can use that to still get some estimation
of protocol being broken.
And so we're gonna have a pilot,
in this case, it's gonna be an asset with Credora.
But after that, we can extend it into evaluating risk out of anything,
not just assets.
We can add protocols.
Unfortunately, we cannot do that on Claros itself because then you have some circular
dependencies in the sense that if you break Claros, you could also misreport that Claros
is not broken in the market.
But for other protocols which are not dependent on the market,
you will get a nice risk estimate. Yeah, that would open up a sort of highly
profitable trading strategy there. I like your idea of Lindy as a way, as kind of like a canary,
like a canary in the coal mine, like a you know, a signal that something is not broken,
or could be broken. And I also like this idea of this sort of insurance, like it's a de facto
insurance that you would be creating with prediction markets on whether something is
broken or breakable. And what they both boil down to is the economics. So it's not just the amount
of time, it's actually the amount of like TVL that someone or it's the amount of value that someone
could earn if they broke the protocol. And so in both cases, whether it's a prediction market,
and they stand to win a lot of money by breaking the protocol, if they take the other side of the bet as a bounty, or if it's the amount of time that has
allowed them to accumulate half a million in TVL or whatever that shows nobody has broken it at
this threshold, therefore it is secure to that amount. Both of those boil down to the economics,
which is what we model in our audit reports.
So I like that commonality there.
So it's not an insurance, because in an insurance, you are insuring an asset that you have against
adverse events, and you get a reimbursement of most of your losses, right?
Here are two elements which make it not an insurance.
The first one is that payouts are not deterministic.
So if you have five protocol brokers in the period,
you get $0.20.
But now if you have 10 protocol brokers,
you only get $0.10.
And if you have only one protocol broker,
you get $1.00.
So you have a non- no deterministic payout.
You don't have exposure to it. So for example, if you knew that UST was bad and was going to likely be broken, you could have bought some UST broken tokens, even if you had no exposure to UST at all.
And that's really way better to price risk,
because in the classic insurance system,
you have the insurer, which is giving some price,
use a premium, and then people can only basically trade on,
yeah, I want you to like debt of it being broken,
not because I want it to be broken, but because I want to reduce my variance.
While here, you can have anyone who can trade on it being broken
as part of a strategy to get more money.
So like in this case, you can trade even if they are buying a netfall
just because they want to have to increase their expected return yeah that makes total sense i see
how it's not strictly insurance and i think it's good to disclaim that for uh you know legal reasons
and for the sake of being specific but i think i would say it's like insurance, but significantly improved because more people have access to the product. And it can be, like you said, part of more complex strategies to de-risk things or to bet on the long tail of events as people do in the long tail of prediction markets.
prediction markets. And I'm curious, and I think we'll have to do more spaces because we're coming
up on the top of the hour. I would love to talk with you more about prediction markets, but the
one thing, the one burning question I have, and that I imagine many of the audience have, is
how are you thinking about the current UMA and polymarket drama with the Zolensky suit?
And maybe recap this really quickly for the audience
before sharing your thoughts.
Yeah, so UMA is basically a simplified version of Kleros.
So if you take Kleros, you remove the core tree,
you remove the API system, you basically
get something which looks very
similar to Yuma.
So it's not that bad.
That's already a Shilling coin oracle, so that's already pretty nice.
The only issue that you will get with Yuma will be that it will not really be able to
scale.
If you get 100 of this per day, that's not going to work.
You don't have everyone vote on 100 of disputes per day, that's not going to work.
You can't have everyone votes on 100 of disputes per day.
And it doesn't work for stuff which require domain-specific knowledge.
You cannot have a dispute in Spanish in Yuma.
On the Zelensky case, I've looked a bit at Polymark in that. I think the search could be yes, not because it's a shoot or not,
because this one is debatable, but in their description they were asking whether you have
a consensus of news media reporting that he was a shoot and media did report that he was a shoot,
right? So we can argue maybe it's a suit, maybe it's not a suit,
and this one is a bit ambiguous.
But they will say that if you have a consensus of media,
well, you have this kind of consensus of media.
A bunch of media reported that he was wearing a suit.
So is it very bad?
I don't know.
We care a little differently.
I don't know.
Sometimes it's hard to distinguish when
you see all this kind of drama,
which part is due to what the system is not being fair.
And I think Yuma is slightly less fair than Claros,
but I don't think it's like order of magnitude less fair.
We produce it to system are pretty similar.
And then the drama, which is done there just for a strategic reason
because if you have a stake in the yes.com, you definitely want to have a narrative that
if the answer is no, Yuma should look bad and the Yuma token should decrease.
And so by doing that, by going ready to making drama,
you could kind of like get yourself an edge in the case.
And this also does happen with Scleros,
which is pretty annoying in some way,
because it makes it the optimal strategy
to actually attack the protocol publicly.
Classic court system, they try to avoid that
because, well, they have the hammer and the sword of the state, so they can basically convince people for contempt to court. like a law which makes it illegal to cast doubt or criticism on a court decision and
you could sue people for that. It's maybe a bit extreme but in some ways that's a way
to kind of like protect the judges from the public attack from the public which may not be very knowledgeable about the case.
Because that's an issue we got with Abraham, for example, in the Abraham case, you may
have some people that would just go, oh, they identify more with the injury.
They don't really understand that the injury is like way richer than the protocol and a repeat attacker or
protocols but they identify oh in my day-to-day life more like an insurer than an insurer and
so they will get this kind of bias of identifying with the insurer and then they can try to you know
like trash talk the protocol which is giving this ruling and And the answer to that is basically just violence.
You know, like we take you some of your assets
or even your freedom if you do so.
And the question is, how can you really do that
in a non-aggression mechanism, which you see
that's crypto, like Kleros can only affect you
if you explicitly engage in contracts
which are arbitrated by Kleros, but Kleros is not gonna sue you if you explicitly engage in contracts which are arbitrated by Klaros
but Klaros is not going to sue you if you trash lock it and that's the same thing for you
so the question is how do you do that for the contract system when people do have some incentive
into casting doubt on those systems well I think Klaros seems a little more protected from that because of the proof of stake system that you've set up and people, as we've discussed, have tried to attack it.
And it's proven to be a little more Lindy than some of the other alternatives out there.
So I really like it and I'm glad we could talk about it today on the space and deep dive a little bit on the
mechanisms. I wanted to talk more about what other projects are hooking into Kleros as a backup but
I think we're going to have to do more protocol spotlights in the future. Anything you want to
share about what you're working on and where people can find more about that? I know you've
got no shortage of projects in the ecosystem.
So, well, we see a bunch of things, but just what we got yesterday,
yesterday we got two Futaki experiments launching on Keros. So you have the Butter Guide where they
They have a Uniswap chain giving some grant to project, depending on the expected amount
of CVL this project will get extra if they get the grant, which I think is a pretty interesting
experiment.
And on footarky.fi and Teros in this case, we have a proposal to require that if you have to mint some PNK,
a foodarchy mechanism should show a PNK price at least equal to the PNK price if you don't mint it.
So basically that's a protection against division which will negatively adverse the PNK holders. And that's also Foodarchy because you can trade on the PNK price
conditionally on proposal passing or not passing.
So we got two Foodarchy experiment launching yesterday.
I guess a bunch of things were already before,
but because everyone was in its CC,
they were waiting Monday to launch thing.
Cool.
Well, I'm looking forward to chatting more about that.
We'd love to platform Futarchy as another protocol because obviously the form of governance
itself is, personally, I think Futarchy and prediction markets are going to change the
world more than any technology we've seen in the last, you know, maybe forever.
I think it's going to be the most drastic social change in
history. And it's impossible to stop. So I'm really excited for that. And I'd love to talk
to you some more about it. But for now, I think we're going to close out the space,
respect everyone's time, get the top of the hour, and I've got to jump. So,
Clement, really appreciate you
coming on today and, uh, sorry to leave the audience with such a cliffhanger, but, uh,
look forward to the next space with Cleros ecosystem and the, uh, Cleros CTO. Clermont,
thank you for coming on today and I'm looking forward to talking to you again soon.
and I'm looking forward to talking to you again soon.
Thanks everyone.
See you.
Thank you all guys.
We'll catch you later.
Catch you next time.
