Where do we draw the line between life and fiction?
Ari Juels is a chaired professor at Cornell Tech, a computer science faculty member at Cornell University and chief scientist at Chainlink Labs, a leader in decentralized oracle networks. He’s also just written a crypto thriller novel, The Oracle, published by Talos Press.
The novel’s protagonist is, like Juels, an expert on blockchain oracles, which are entities that connect blockchains to external systems. Juels co-authored Chainlink Labs’ 2021 white paper.
Similar to the author, the novel’s protagonist wrote a paper on rogue smart contracts, which explains why the United States Federal Bureau of Investigation (in the book) enlists his aid in thwarting a smart contract for $10,000 that has been taken out to kill a certain Professor Vigano.
This is not to say that the novel, published on Feb. 20, is a roman à clef, i.e., a book in which real people or events appear with invented names. According to Juels, it is “a warning tale about weaponized blockchain” written in the hope of preventing “rogue uses of the financial system by AI agents.”
But it’s also “a whirlwind tale of oracles, ancient and modern, vanished antiquities and conjured crypto billions, cybercriminals and digital idealists,” according to the book’s publisher.
“What happens when the crypto ideals of privacy and truth might cost human lives — especially your own?”
Last week, Cointelegraph sat down with Juels to explore why he wrote this science-based novel, what lessons or warnings it holds for both technologists and society in general, and how he sees blockchain and artificial intelligence (AI) unfolding over the next decade.
Cointelegraph: What compelled you to write a novel about a rogue smart contract, of all things?
Ari Juels: The book is actually, in some sense, a dramatization of a research paper I co-authored back in 2015. It was the first paper I wrote on smart contracts, and it was the result of my brainstorming with a colleague around the question of what smart contracts are good for. Both of us have backgrounds in security, and so we tend to think adversarially.
And the first application we came up with was crime. We realized smart contracts could be used to execute crimes. They [criminals] could benefit from the privacy features of blockchains and also take advantage of the trust models — in addition to the fact that smart contracts are, in principle, unstoppable.
CT: You don’t focus on the more familiar crypto-related crimes like the theft of private keys but something you call a “calling card” crime, a broad class of physical world crimes that can include murder and arson. In this case, a sinister party advertises through a blockchain-based smart contract for an assassin.
Juels: When I was writing the novel, I imagined that that particular scenario was something like 20 years out [i.e., before it was technically doable]. And the reason was that in order for it to become a reality, you needed an AI tool that could interpret news articles and draw keywords from those articles. Even three years ago, that seemed like a rather distant prospect.
But then, toward the end of my writing the novel, ChatGPT dropped. And suddenly, this scenario became technically feasible. That doesn’t mean that you can launch a rogue contract today — you cannot. We don’t have the infrastructure, but it means it’s now suddenly something we need to think about in the near future.
CT: On page 34 of the novel, a character named Vitalik Buterin appears. He sounds a lot like the real-life Vitalik Buterin (co-founder of the Ethereum blockchain). He provides the protagonist with an idea for thwarting the assassination — a sort of “anti-rogue-contract contract.”
Did you ever have a conversation like that with Vitalik Buterin?
Juels: That’s purely fiction. I’ve worked with Vitalik, but not on this topic.
CT: How many years away are we before a rogue smart contract of the sort described in the novel could be feasible?
Juels: It’s a question of how the infrastructure and safeguards evolve. I believe that we can evolve safeguards that will not only help prevent scenarios of this type but also make the use of AI more broadly and blockchains safer. I do think that the combination of the two technologies, particularly in the realm of smart contracts, has a lot of potential.
CT: An example of how the two technologies might be combined — for good?
Juels: There’s the possibility of making smart contracts more like real-world [legal] contracts. Now, there are still lots of problems. First, you’ll need better natural language interfaces. Large language models hallucinate, for example. That could be a serious problem. But the potential is there.
CT: In terms of reactions to the book, has anything surprised you?
Juels: The book is grounded in real science and technology and incorporates research I’ve done in the past and even some research that’s going on in my group now. One technology discussed in the book is something called a multiblock flash loan [i.e., a proof that a borrower’s contract would pay back a loan a short time later, like a ZK-proof], which relates to a project that one of my Ph. D. students is working on.
My PhD student read the book, and he said to me, you know, the book includes an application of this technology that we weren’t thinking about in our research. We should put that in our paper. And so suddenly, the dialogue between my research and the book became bidirectional. It actually became a dialogue.
CT: On the matter of blockchain-based oracles, your protagonist says on page 15:
Juels: My character has somewhat more extreme views than I do. He is a bit of an idealist, and he recognizes that for smart contracts to do anything interesting, they need to know what’s going on in the real world.
It’s not a roman à clef. There’s only one character that corresponds to a real person, and that’s one of the villains.
CT: You wrote an article recently with your Cornell University colleague Eswar Prasad where you mention an interesting blockchain use case, a “proof of reserves” protocol or app that “could help prevent debacles like FTX.” You also write in the paper about “shifting the balance back to innovation” with regard to smart contracts and blockchain technology.
Have we become overly concerned with the price of cryptocurrencies at the expense of building things?
Juels: The narrative has gotten completely distorted by the money side of things. The point we were making in the article is that what Sam Bankman-Fried did is just classic accounting fraud. There was nothing about it distinctly related to blockchain technology. Part of the problem here is that the money side of things doesn’t reflect this kind of deeper reservoir of innovation — all of the promise of the technology.
I worry that politicians are only going to see the dark side of this technology and not realize that there are really important fundamental innovations happening, not just in blockchain technology but even in tangential areas, such as applied cryptography like zero-knowledge proofs. They have advanced by leaps and bounds in the past few years, with the applications in the realm of blockchains being the impetus.
CT: Oracles aren’t just technology, your character says; they’re the “lifeblood of the smart contract revolution.” They anchor smart contracts in truth, and in a society facing existential threats to justice and truth, they “can be a force for good.” Like the ancient Oracle of Delphi?
Juels: The Oracle of Delphi played an absolutely pivotal role in the history of Greece. One of the roles that it served was in consensus forming when communities had disputes about major political decisions. Should we found a colony? Should we go to war? Should we wipe out our enemies?
Recent: Energy-efficient miners in US less likely to be impacted by Bitcoin halving
So, it served not just as a source of truth but as a source of consensus. I would love to see blockchain oracles playing a positive social role as well. And that’s a huge ambition, right?
But one of the reasons for incorporating the Oracle of Delphi into the novel is to provide a perspective on what blockchain oracles can become and the positive impact they might have if they evolve in the right direction.