What if there was a crypto protocol specializing in the arbitration of chain disputes?
Imagine if, whenever the prediction markets like Polymarket have settled controversially, users had an official way of calling on a kind of neutral justice system on the chain. Or if decentralized autonomous organizations (DAO) could rely on an efficient and well informed third party to help them make decisions. Or if insurance contracts could automatically execute payments when specific events in the real world have occurred.
This is essentially what Albert Castellana Lluís and his team constitute with Genlayer, a crypto project which is marked as a decision -making system or a trusted infrastructure.
“We use a blockchain that has several AIS contact details and reach an agreement on subjective decisions, as if they were a judge,” Castellana, co-founder and CEO of Yeagerai in Coindesk in an interview. “We essentially build a global synthetic jurisdiction that has an integrated judicial system that does not sleep, it’s super cheap, and it’s super fast.”
The demand for such an arbitration project can increase in the coming years with the development of AI agents – sophisticated programs powered by artificial intelligence which are capable of performing complex tasks independently.
With regard to cryptographic markets, AI agents can be used in all kinds of ways: to exchange mecoins, arbitrate bitcoin on exchanges, monitor the security of DEFI protocols or provide information on the market through in -depth analysis, to name a few use cases. AI agents will also be able to hire other AI agents in order to complete even more complex assignments.
These agents can proliferate at an unexpected rate, said Castellana. In his opinion, most players in the cryptography market could manage a handful by the end of 2025.
“These agents, they work very quickly, they do not sleep, they do not go to prison. You don’t know where they are. Will they adopt anti-flowage rules? Said Castellana. “How can we allow rapid transactions between them?” And how can confidence occur in a world like this? ”
Thanks to its unique architecture, Genlayer could provide a solution by allowing entities – human or AI – to obtain a reliable and neutral opinion to weigh on any decision in record time. “Wherever you normally have a third composed of a bunch of humans … We replace them with a global network which provides a consensus between different AIS, a network that can make decisions in a way as correct and as impartial as possible,” said Castellana.
Synthetic court system
Genlayer does not try to compete with other blockchains such as Bitcoin, Ethereum or Solana – or even DEFI protocols such as Uniswap or Compound. The idea is rather that any existing crypto protocol can connect to Genlayer and use its infrastructure.
The Genlayer chain is fueled by Zksync, an Ethereum Layer 2 solution. Its network has 1,000 validators, each connected to a large model of language (LLM) such as the Openai Chatppt, the Bert de Google or the Meta lama.
Let’s say that a polymarket market is installed controversially. If Polymarket is connected to Genlayer, users of the prediction market have the possibility of raising the question (or, as Castellana said, to create a “transaction”) with its synthetic judicial system.
As soon as the transaction arrives, Genlayer chooses five random validators to govern it. These five validators question an LLM of their choice in order to find information on the subject at hand, then to vote on a solution. This produces a decision.
But Polymarket users, in our example, do not necessarily need to be satisfied with the decision: they can decide to appeal the decision. In this case, Genlayer chooses another set of validators – except this time, their number goes to 11. Just like before, the validators issue a decision according to the information they collect from the LLM. This decision can also be on appeal, which means that Genlayer chooses 23 validators for another decision, then 47 validators, then 95, and so on.
The idea is to rely on the theorem of the Condorcet jury, which, according to the Pitch deck of Genlayer, indicates that “when each participant is more likely than making a correct decision, the probability of a correct majority result increases considerably as the group increases.” In other words, Genlayer finds wisdom in the crowd. The more the validators are involved, the more likely they are to focus on a specific response.
“It means that we can start small and very effectively, but we can also degenerate to a point where something very, very delicate, they can always do well,” said Castellana.
The average transaction takes approximately 100 seconds to be treated, said Castellana, and the court’s decision becomes final after 30 minutes – a deadline that can be extended if several calls occur. But that means that the protocol can make a decision on major issues in a very short time, day or night, instead of going through arduous real litigation processes that can take months or even years.
Look at the incentives
Genlayer’s mission naturally raises a question: is it possible to play the system? For example, what happens if all validators select the same AI (say, Chatppt) to resolve a given proposal? Doesn’t that mean that Chatgpt will have essentially issued the decision?
Whenever you question an LLM, you generate a new seed, said Castellana, then you get a different answer. In addition to that, validators have the freedom to choose the LLM to be used according to the subject to be accomplished. If it is a relatively easy question, it may not be necessary to use an expensive LLM; On the other hand, if the question is particularly complex, the validator can opt for a better quality AI model.
Validators can even find themselves in a situation where they have the impression of having seen a certain type of question so many times that they can pre-training a small model for a specific purpose. “We believe that over time, there will just be endless new models,” said Castellana.
There is a strong incentive for validators to be on the winning side of the decision -making process, because they are financially rewarded for this – while the losing team ends up leading to costs associated with the use of calculation, without receiving rewards.
In other words, the question is not whether its validator provides a correct answer, but if it manages to store on the side of the majority.
Since the validators have no idea that other validators vote, the objective is that they use the resources necessary to provide precise information while waiting for other validators will also converge on this information – because it will happen to the same incorrect response would probably require rigorous coordination.
And if this gambit does not work, the call system is ready to get started.
“If I know that I reuse a good LLM, and I think that other people use a bad LLM and that is why I have lost, then I have a great incentive to call, because I know that with more people, there will be an incentive for them to use a better LLMS also”, because other validators will also want to win the rewards of a successful call, “said Castellana.
The system makes it difficult for validators to end, because they have only 100 seconds to make a decision, and they do not know if they will be chosen to settle specific questions. An entity should control between 33% and 50% of the network to be able to attack it, said Castellana.
Like Ethereum, Genlayer will use a native token for its financial incentives. According to Castellana. “There will be a great incentive for people to come and build more things,” he said.
