The Farmer With a $50 Phone: Sergey Nazarov on Oracles, Trust, and Hybrid Smart Contracts

Lex Fridman opens with a thought experiment: live only in the digital world or only in the physical one, which would you choose? Sergey Nazarov treats it as a question about fidelity. The simulation debate, he says, grew out of philosophical papers twenty or thirty years ago that asked whether people had an intuitive attachment to something being “the real world”; *The Matrix* then carried those papers into public consciousness. His own answer is blunt. If he could inhabit a digital world at the same fidelity as this one, with immortality attached, he would take the digital world without hesitation.

Fridman immediately flags the extra clause: where did immortality come from? Nazarov’s picture of uploading is a mapping of brain patterns onto a virtual machine, and a virtual machine has no built-in lifespan. That raises a versioning problem Fridman likens to a soft fork versus a hard fork: would Sergey 2.0 be an upgrade that kills Sergey 1.0? They wander into the *Star Trek* transporter — are you killed, or transported? — and Nazarov admits he has not written a white paper on the subject. Fidelity, for him, is perception. If no dimension of experience differs from the physical world, he does not care whether the result is “real.” Metaphysics never interested him, because he could not see the practical point. If something cannot enter his perception, it does not exist for him.

Fridman recasts the simulation as an engineering problem: how hard is it to build a virtual world faithful enough that you would want to live in it? A full quantum-mechanical capture of human experience may be unnecessary; a game engine that fools the visual cortex might be enough, perhaps within a century. Nazarov is less breezy about the hardware. VR headsets already fail below ninety frames per second, which means going from one sixty-frame gaming screen to two ninety-frame screens at the eyes, a load most current machines cannot handle — and that is only two little screens, not a full hijack of every sense. On the red pill versus the blue pill he splits the question. *The Matrix* enslaves people as batteries, a moral problem at a scale he thinks almost everyone would be obliged to resist. For an individual life, going along with the system is “way more fun.” Given a binary choice at that scale of harm, he would still go against it.

Nazarov’s through-line for the rest of the conversation is a concept he calls definitive truth. It sits on a spectrum between objective truth, which philosophy sets at a standard almost no one can reach, and a story somebody invented that nobody else accepts. Commerce, he says, does not need the first and cannot live with the second. What it needs is a shared, pre-agreed definition of what will count as true for a specific agreement. If you and he write an insurance policy that pays when there is no rain, and you both accept that three weather stations coming to consensus is the trigger, then that consensus is definitive truth for the two of you. No single station gets to announce the result unilaterally.

The upgrade he wants is from “a bank or an insurer decides what happened” to “we defined in advance what truth looks like under these conditions.” Twenty nodes, thirty data sources, a stated method of consensus, a threshold of agreement: once those fire, the relationship can move forward. That standard is necessary and sufficient for contracts, and it is already a large step up from a counterparty that can rewrite the facts. Fridman asks how much of the world can be formalized this way. Nazarov’s answer is a huge amount, because two things are happening at once. Data is exploding — market prints, IoT, shipments, even likes, clicks, and views on a YouTube video — and that data is increasingly accessible to everyone.

More data, more widely available, means more definitive-truth proofs can be constructed. Feed those proofs into blockchains and smart contracts and they become tamper-proof. Combine a system that can prove things with a system that can guarantee outcomes, and you have a form of contract that, in his view, has never existed before. The rest of the episode is an argument that this combination — proof plus guaranteed settlement — is the missing piece of digital commerce, not another token narrative.

Decentralized finance, in Nazarov’s framing, is one species of what he calls a hybrid smart contract: on-chain code plus off-chain proofs that something happened. It takes the contractual relationships of global finance — lending, yield, derivatives that hedge or speculate on events — and restates them in a blockchain format. The old format is a trip to a bank, even for a hedge fund, in the hope that the product will be honored, or a deal with a counterparty in the hope that the deal will be honored. Opacity is a feature of that world. He points at 2008: many agreements, many assets, and a system so hard to see into that a boom became a bust that hit people who never participated in the boom.

DeFi, he says, currently delivers three things. The first is transparency. A technical user can drill into collateral — where it is, how much, what form, how it changes block by block — because the infrastructure forces the books open. Centralized finance is “very, very good at hiding” and wrapping products in a glossy shell; escaping blockchain transparency is nearly impossible, and if someone tries, it is immediately obvious and people leave. The second is control. After Robinhood restricted trading, users discovered that the band of market conditions in which they thought they could act was much smaller than advertised. The system had been packaged to recruit their participation; underneath, it did not work the way they believed. Blockchain contracts, he argues, give people the version of the world they already assume they inhabit, which is why they do not beg for it until a Robinhood moment makes the gap painful.

The third is yield, and it is global. A dollar at a bank then paid one percent or less; the same dollar in DeFi paid seven or eight percent. Stack that against coming inflation, against systems that may fail in ways that spotlight the control problem, and Nazarov’s conclusion is almost sarcastic: the only reason DeFi is not more used is that more people do not know about it. Nobody, he says, has explained why you would take less yield, more risk, and a counterparty that can change the rules at discretion. Fridman ties the first two properties together. Transparency creates accountability; in a game-theoretic sense, if you know I will discover your deviation, you will not deviate. He extends the same hunger for authenticity and a walk-away button to social media: a big red “forget me” control, not a three-layer settings maze. Nazarov agrees the pattern is the same for banks, social networks, and every other contract that currently takes your data or your assets in exchange for features. An autonomous piece of code can offer the features without the lock-in.

The industry’s definition of a smart contract has shrunk, Nazarov says, and he wants it restored. When he started, seven or eight years before this conversation, a smart contract meant a tamper-proof digital agreement that could know things about the world, settle an outcome, and release payment. That turned out to be a tall order with three parts: the contract, the proof of what happened, and the movement of value. In practice the phrase came to mean only on-chain code. Chainlink’s 2.0 white paper — 136 pages, which Fridman found unusually readable — revives the original idea under a new name: hybrid smart contracts. On-chain code still defines the agreement, takes in tokens, and pays them out. An off-chain decentralized oracle network supplies the proof.

The counterintuitive fact is that on-chain code cannot speak to any external system. To stay tamper-proof, a blockchain is built as a walled garden that only knows highly reliable internal facts: tokens and private-key signatures. A contract cannot say “I will fetch that API,” because the fetch would import security problems and a consensus problem about what the API even said, let alone what the world is. What matters is not one API. It is a reliable form of decentralized computation that yields a definitive proof of an event. Some of Chainlink’s most used networks, Nazarov says, run well over thirty nodes and well over ten data sources on the same type of data, reach consensus, and write the result back on-chain. Without that, an agreement that can never know what happened is useless. With it, you can eventually put millions, billions, even trillions of dollars into code whose trigger is a mutually agreed, sufficiently stringent definition of truth.

An oracle node lives in both worlds. On-chain it is a contract interface, a Web3 analogue of a Web 2.0 API: request weather, request a price, ask me to send a payment. Off-chain it fetches, computes, or even generates the service. Chainlink does not run its own blockchain; Nazarov describes the goal as a decentralized meta-layer between an unreliable world and a hyper-reliable chain, sometimes generating the missing service itself. Verifiable random functions are the example: lotteries need randomness, the network produces it, nobody buys it from a data vendor. Security should scale with value at risk. A hundred-thousand-dollar beta might be fine with seven nodes and two or three sources; a thousand-node network protecting the same sum would be inefficient. As value grows, ensembles grow — fifteen, thirty-one, fifty-seven, a hundred and five. His Uber analogy is the punchline. Uber’s core code called Google Maps, Twilio, and Stripe. Without those services the founders would have had to become a geolocation company, a telecom, and a global payments firm, and they would not have. Blockchain developers, he says, are stuck in that weird scenario until oracle networks supply the services that turn tokenization into DeFi, fraud-proof gaming, global trade, and ad networks.

The applications that excite Nazarov split by geography. In developed markets he hopes transparency will blunt boom-and-bust cycles by making systemic risk visible to every participant except the tiny minority that profits from asymmetric information. Emerging markets, he thinks, will gain more, the way they gained more from the internet and from fifty-dollar Android phones. He has always had libraries; somewhere else, people without libraries now watch the same Stanford lecture on a cheap phone — a jump from zero to one. The example he keeps returning to is crop insurance, partly because it is not a tokenization story and partly because it matters more than it looks. In a rich country a farmer who gets no rain can claim a payout and argue with an insurer. Global options markets began as a way for growers to sell crops ahead of drought; they later became a casino. In an emerging market, two dry seasons can close the farm and turn the owner into a migrant worker.

Those farmers can now read Wikipedia on a phone, but they are still decades from an insurance company setting up shop, because the local legal system cannot support the product and courts cannot unwind fraud. A smart contract that sees satellite or weather-station data for that geography lets the farmer lock value in and receive an automatic payout on the same phone. He leapfrogs a government that will not provide legal infrastructure, leapfrogs insurers who may price-gouge and often fail to pay, and lands in guaranteed outcomes that Nazarov says are as good as, and in many cases better than, what farmers elsewhere already have. The same pattern extends to inflation-resistant accounts, selling goods on a global market without a middleman who owns every trust relationship, and other backend problems rich listeners do not feel. Fridman traces the ripple down the supply chain: if any producer can formalize agreements, raise reliability, and prove that goods were sourced ethically in a way that cannot be gamed, buying power, supplier power, and how consumer goods are made all move.

Politics is the further horizon. Nazarov imagines a candidate who promises to redistribute oil revenue writing that promise into a smart contract, solid on a chain, to be switched on after the election. Fridman reaches for a digital constitution — the U.S. founding documents recast in the twenty-first century — and even for formalizing military conflict, wondering whether Afghanistan or Iraq might have resolved faster with honesty, transparency, and control baked into the terms. Nazarov has not thought much about war contracts; the idea is a little scary, and it is not in the white paper. The three ingredients he will stand on are simpler. You need trusted code, agreement about how the world should work, and a system that can tell the code what happened. Where those hold, any set of agreements can in theory be formalized. Fridman adds a computer scientist’s point: programming languages force a clarity natural language does not. There will not be a shelf of books about what one line meant. Writing human agreements in code is not only a technical or financial idea. It is a philosophical one.

Fridman asks where non-human intelligence fits once agreements live in digital space. Nazarov is not an AI expert and says so; he looks at the field through the only lens he has, which is trust. Everyone saw *The Terminator* in the nineties. At some point, wherever we are on that curve, people may want AI to run important parts of life and also be a little afraid of how much it can do. His answer is straightforward. Limit the system with private keys and blockchains. Create guardrails and firm walls it cannot pass, assuming encryption still holds and breaking encryption is not that model’s specialty. Shipping, defense, the financial system: if the machine is inventing things no one would have thought of in a hundred years and feels a little too unbelievable, bake in conditions, written in the unambiguous language of a contract, beyond which it may never act. Supervision, in this picture, is not “is it smarter than us.” It is “will the encryption hold.”

Would AIs write contracts with one another, so that most agreements eventually run between non-human parties? Nazarov suspects they would prefer the deterministic, guaranteed format. An AI cannot pick up the phone, call Dave at the bank, soothe him through a derivatives trade, and manage the social cues. Language-and-law ambiguity is a bad fit; a high-quality model would notice the risk. Mapping today’s natural-language contracts onto smart contracts is, for many industries, less hard than it sounds. Plenty of legal agreements already point at a computing system — e-signatures, a repository, a settlement engine. You swap the engine. The harder limit is data. You will not get a hybrid contract about whether someone painted a house the right blue unless a drone with a color tool exists, and if demand is high enough, that feed will be created. Insurers already fly drones over construction sites to count hard hats and void policies when the percentage is too low. They already pay IoT firms to put sensors in customers’ infrastructure so a policy can be priced. Where there is a data source that can settle a dispute for both sides, Nazarov sees no reason not to use autonomous code.

Privacy inside those contracts is itself a trust issue. A driver might not want an insurer to see a high-resolution feed; a smart contract can evaluate the data against the policy without the company ever touching it. Search-engine optimization was an early example from his own work: clients would not pay until rankings moved, firms would not work until they were paid, so bitcoin sat in a contract and released when a keyword hit a Google rank. Machine-learning data markets are the same pattern. You do not know my data’s quality, so you do not know what to pay; I cannot show you the data without giving it away; you do not want me to see the algorithm that scores it. An impartial on-chain agent plus an oracle network samples, scores, and prices. I never see your algorithm, you never see my full data, neither of us controls the judge, and we can even forbid uses we dislike. Stay in this industry long enough, he says, and you end up with one useful question: what trust issue is this solving? Answer it case by case and the relevance of blockchains stops being a slogan.

Transparency and privacy pull in opposite directions. Blockchains are public; they are not private. Nazarov expects encrypted on-chain smart contracts eventually, but he thinks they will take years, run inefficiently, and come with a pile of nuances. The interim move is to place part of the computation in a decentralized oracle network. Chainlink presented a paper at Stanford in February 2020 called Mixicles on exactly this. You control the committee of nodes, and you can require trusted execution environments so that even those nodes do not know what they are computing.

The paper’s example is a binary option. Payout happens on-chain, but an observer of the chain cannot reverse-engineer the contract’s outcome. The result is computed in the oracle network; a switch then decides who receives the money. Users still have on-chain guarantees: if the network finds a given outcome, the relevant party is paid, and there is still a place to lock value. Privacy, Nazarov argues, is not a nice-to-have. Many contracts will never be born as hybrid on-chain agreements without it. That is why oracle networks keep trying to ship services — privacy-preserving computation among them — that let developers compose products they are otherwise blocked from shipping.

He reaches for the internet’s own history. E-commerce was painful while credit cards traveled over public HTTP. HTTPS arrived, cards could move on an encrypted channel, and buying things got easy in a way that was probably unexpected at the time. Privacy services in hybrid contracts play the same role. Until on-chain encryption is ready, the decentralized oracle network is the place that feature lives, and without it a substantial share of real contracts will stay off this infrastructure.

Nazarov splits the world into two camps. One is brand-based, paper-guarantee trust: a bank logo, an insurance logo, the biggest building in town, marble arches larger than the church. Letters of credit from branded entities really did make early global trade possible, and that is better than nothing. It is still a paper agreement with a legal framework. When Robinhood has to change, it changes, and you cannot do much about it. Banks did not build beautiful buildings because they enjoy spending money on stone. They built them to manufacture confidence. Gold-rush cities that needed banking still have some of the most beautiful buildings in the United States for that reason. Until now, a bad experience with one marble building meant walking down the block to another color of marble and another paper contract.

The other camp is mathematical: encryption-secured, decentralized-infrastructure-powered agreements. The mechanism is consensus. Independent nodes, provably independent computers, reach the same conclusion impartially, verify it, and produce a final global state — address A had one bitcoin, now address B does. When things go wrong, the contrast stops being theoretical. Counterparties turn out not to have to pay you. Trading is halted. ATMs in Greece, a few years before this recording, dispensed sixty-six euros a day whether you were a person or a business. Centralized finance works well until it does not; then you discover what was under the contract. Users, he thinks, will migrate a portion of what they own into systems where thousands of machines, not one institution, guarantee unfettered access. A farmer evaluating crop insurance can look at Ethereum’s hash power, look at an oracle network of fifteen or twenty-five nodes drawing five weather stations, and decide whether that ensemble is enough. Each successful payout writes immutable proof back onto both the contract and the network. Reputation becomes a body of settled value, not a logo above a train terminal.

He does not think banks and insurers will all die. The internet is the script: some adopt fast, some slow, some never. The real pivot happens on an earnings call. A competitor launches a blockchain bond or a smart-contract derivatives platform, prints ten billion in revenue or volume, and someone asks what your plan is. By the next call there is a plan; by the one after that, the question is when the plan ships. Compensation structures are built around winning and losing business. Fridman is skeptical that old banks can revolutionize themselves — even their basic online experiences are clumsy — and wonders whether a Square or a Cash App will do to banking what insiders will not. Nazarov knows the first-principles troublemakers inside those firms. They are getting listened to more, because when the core line of business is finally hit, they are the only people with a coherent explanation of why.

Asked to value bitcoin, then around sixty thousand dollars, Nazarov declines to play analyst. He spends his time building systems. What he sees in bitcoin is a conception: the first true non-governmental fiat money, possible because it is decentralized. Government fiat is, in his telling, a roughly forty-year experiment. Historically, anyone who tried to issue a competing fiat inside a country was shut down at once. Bitcoin could not be shut down that way. How powerful is non-governmental fiat? As powerful as the ideas people ascribe to it. Internet money — what is that worth? The money of countries that lack a good currency — what is that worth? Portable digital gold with built-in scarcity — what is that worth? All three at once? He has not done the analysis. It might be worth a lot. What matters is how society’s conception evolves, and that depends on current events, not on him, Fridman, or the creator. A world where people keep ten or twenty percent of wealth in non-governmental fiat as a hedge is a different set of zeros.

The conception has already changed once. When he entered the industry, bitcoin was micro-payments: cheap fees, machine-to-machine, useful in emerging markets. Price rose, the story evolved. Governments can steer the public’s idea of their money with a spokesperson, a budget, a plan, a promise of safety. Bitcoin has no such office, so a hole remains that everyone fills. Who is Satoshi Nakamoto? Nazarov does not know. If he had to guess, a group, some of whom may no longer be alive. He is grateful either way, and he thinks the identity should not matter: it should have no bearing on the system’s function, value, use, or security. Whoever it is was wise not to step forward. Fridman asks the trap question: if you were, would you tell me? Nazarov laughs, says he is not, and adds that if he were he probably still would not. Relinquishing that attention, Fridman says, is in the same family as Washington giving up power.

On bitcoin maximalists he is calm. Results speak for themselves. Bitcoin drove the industry; he grew up in bitcoin meetups among libertarians who shared his views, and anyone this early came through that ecosystem. The protocol’s focus on its own ledger, not colored coins, not eighty-byte OP_RETURN, not contractual outcomes, is a design decision with the benefits of focus and the costs of a chain that will not hold other people’s contracts. People are building systems that anchor security in bitcoin; that path will eventually need a deeper conversation with the protocol itself. Ethereum, by contrast, is “completely complementary.” Chainlink does not want to run smart contracts; it wants to let them go further, and an oracle network is worth far less without a mission-critical contract that needs its data. Before Ethereum, writing a new contract meant asking protocol developers to bake logic into the chain itself, a process of months or more than a year that almost nobody would endure. Scriptable contracts, auditable by others, in a matter of weeks, is the leap he still thanks Vitalik Buterin and the core developers for.

Chainlink’s design problem, Nazarov says, is the reverse of a blockchain’s. A chain picks a block size, a set of transaction types, a contract language, a fee structure, and then keeps composing blocks forever. That focus buys security, reliability, tamper-resistance, and transparency, at the cost of a feature-poor environment: no API calls, often no native randomness, no privacy-preserving compute. Oracle networks have to face all the world’s data and all the use cases that have trust problems — insurance, lending, ad networks, data sales — so variability is enormous. On-chain, flexibility is the enemy. Off-chain, flexibility is the goal. The job is a framework that lets people compose the network they need, not a five-year wait for someone else to build it.

A DeFi contract locking a million dollars might start with seven nodes and three data sources. At a billion it should move to fifteen or thirty-one nodes, five or seven sources, circuit breakers, and extra checks the network must also reach consensus on. Chainlink already runs networks that secure well over a billion dollars, he says, and they keep adding sources and features. Crop insurance wants a different consensus method and a different aggregation than a lending market; gaming might want NFTs tied to real-world sports. The stack on offer is templates, node-operator sets, pre-integrated sources, a reputation system, and a way to scale the ensemble as value scales. Increasingly, one hybrid contract calls several oracle networks at once: price data, proof of reserve, randomness. More advanced contracts compose more networks, which he describes as a virtuous cycle: more hybrid contracts on Ethereum and other chains means more real-world outcomes, which is what he thinks the industry actually signed up for.

For developers, the on-ramp he recommends is the Chainlink YouTube tutorials, upcoming hackathons that run over weeks rather than days, and the write-ups of past winners. The hello-world is probably DeFi, because testnets already have rich data feeds and many existing contracts to copy. Prior hackathons produced a Tesla-rental contract that used the car’s API and a payment on-chain, and a cross-chain project called D-Bridge that used oracle networks to confirm data on different chains. There is even a gov-tech category. The explicit goal is to stop the industry from being only about bitcoins, tokens, or NFTs — “the snow coming off the top of the iceberg” — and start it being about global finance, insurance, and trade. Fridman then needles him for praising bitcoin and Ethereum while skipping “one of the most profound pieces of technology,” Dogecoin. Nazarov treats it as closer to a social experiment, interesting mainly for how a technology rides fun and memes. He has not done the analysis on a trip to the moon. Fridman defends the leftover humor in a culture that bitcoin maximalism has made very serious, and calls fun a force that bends experienced time the way mass bends spacetime.

Asked for books and films that shaped him, Nazarov starts with Plato, especially the *Protagoras* and the *Gorgias*, read in his early teens. The dialogues showed him what the assessment of information looks like: how an idea can be taken in either direction, and how external or internal conversation can burn off fuzzy thinking until a more accurate picture of reality remains. Playing those dialogues in his head taught him how to tell when someone has no clue — a useful skill, because many people are excellent at announcing how the world works, how you should spend your time, your money, and your attention. He still does not understand why high schools and universities lack a course in evaluating information so you are not captured by the picture other people need you to have. After Plato he read Plutarch’s *Lives* and a run of biographies, on the theory that an idea which has survived two thousand years of rebuttal is worth more than one he invented ten minutes ago.

The biography that stuck recently is Vanderbilt’s. Light on formal education, strong on steamboats and on building monopolies without much concern for whether the means were right, Vanderbilt’s fate hung on a Supreme Court fight over state-granted steamboat licenses. Had those monopolies been upheld, he would have been in serious trouble; they were not. The ruling unseated an old landed gentry in the 1830s and 1840s, and he had seen it coming. The detail Nazarov likes is that the ferries in New York Harbor now run as a public good. A capitalist who hammered one industry ended up making something so important it left the category of private enterprise. That is his definition of success in crypto: whether you can make a Linux, an HTTP, an HTTPS — a system that lives a long time as a public good, even if it began as a capitalist idea. He feels lucky to be working on that kind of body of work, with a team and a community that care about a hybrid-smart-contract world, because bitcoin originally attracted him as a picture of how the world could work, not as a coin.

On screen, the shows that keep rising in his estimation are *Star Trek: The Next Generation*, *Voyager*, and *Deep Space Nine* — not the original series, whose effects he cannot take seriously, and not the newest shows, which he finds a little strange. He now watches them as a sketch of utopia: Maslow’s lower rungs already satisfied, nobody hunting a paycheck, every episode a formula to save a planet or an ecosystem. Technology can take civilization toward that picture or toward a dystopia; those middle *Star Trek* years are the former. Fridman, from the AI corner of public discourse, argues that fear-mongering about how everything goes wrong has been overdone. You have to think about failure, in moderation. You also have to design toward a better world, because it is hard to build anything while you are afraid. Nazarov agrees that optimism is better for design and creativity than pessimism.

Advice for the young, Nazarov says, is not one career path. It is a framework about time. Obligations rise and free time falls: a job, a partner, children. Young people assume the drop is ten percent. He thinks it is ninety, and the remaining ten percent will be spent recovering for work. Borrowing Jeff Bezos’s regret-minimization frame, he shortens the horizon. Do not ask what you will regret at eighty. Ask what you would feel worst about if, over the next ten or twelve years, choice itself were cut by ninety percent. If you will never travel after twenty-five, how does that feel? If you will never start a company after marriage and children, how does that feel? Whatever produces the most fear and disgust is what you should do, and you should not listen to anyone else. His colder addendum: if you are really smart you will make it anyway, and a lot of the pressure on you is other people’s fear; if you are not, you are in trouble anyway. Either way, stop agonizing and spend the window on what you would most regret skipping.

He is sharper about the social clock. At twenty-five, peers start comparing achievements; at thirty, families, companies, and PhDs raise the pressure until even the fun thing is no longer fun. Living in a hostel at thirty stops being normal. If your friends are wasting scarce time, get new ones. Do not assume your capacity to get a broad education will improve. Capitalism is good at fashioning people into narrow tools. If you want philosophy, Italy, a band, a painting practice, do it now. Startups get harder with age because they do not generate income for years — evictions and ramen are real — while a household depends on you. Start the company when nobody is eating from your paycheck, when sleeping on the floor can still feel like a great time. Do not take three years of consulting for “experience.” Experience is doing the thing, failing, and doing it again. Mentorship is luck; most people never get it.

Fridman closes on meaning. Nazarov will not answer the cosmic version. He shrinks it to a prior choice: do you want to live in society? He tried the woods, learned a great deal, and got a lot of clarity. Some people leave; some immigrate to another society. Staying means accepting minimums, compromising values a little, and then facing a second question that confuses him when people dodge it: if you chose to stay, what are you doing here? People who accept the deal and refuse to do anything should, in his view, go somewhere the rules fit. For those who want to leave something behind, the question becomes which body of work you will be able to name when you look back. His own answer is unusually clear. He and Fridman are building toward economic fairness and transparency. He thinks he is in a position to enable it, with a real chance of succeeding, and that the impact on many people would be substantial. Looking back on that body of work is something he would be proud of. Fridman agrees that hybrid smart contracts could push the world further than he can currently imagine.