In this episode, Sunny Ray talks with Emin Gun Sirer about his path from a tech obsessed childhood in Istanbul to becoming a pioneer in distributed systems. Sirer explains how his early work at Cornell on the Karma project, a proof of work based minting scheme, predated Satoshi Nakamoto's Bitcoin by six years, though he credits Satoshi's genius in pairing proof of work with a probabilistic consensus mechanism. The conversation moves through Sirer's contributions to Bitcoin security, his role advising Ethereum during the DAO crisis, and his eventual focus on building Avalanche, a consensus protocol he considers the biggest breakthrough in distributed systems since Satoshi's white paper. Sirer details why Bitcoin excels as a settlement layer but cannot scale for broader use cases, why Ethereum lacks native asset awareness, and how Avalanche's subnet architecture allows enterprises to tokenize real world assets like real estate with legal compliance and flexible governance built in.
The professor who built proof of work years before Bitcoin explains why Avalanche, not Bitcoin or Ethereum, is built for tokenizing the real world.
Where does your story begin, in terms of your worldview and background?
I was born in Istanbul and grew up in a developing country where nothing was built to perfection, which instilled in me a love of building perfect systems. Computers felt like a universe you could get exactly right. I wanted to build self organizing systems with strong guarantees, which led me to Princeton on scholarship and eventually into operating systems and distributed systems research.
How did the idea for Karma come about at Cornell around 2001?
Peer to peer file sharing was booming but everyone wanted to download and nobody wanted to upload. I thought the simplest fix was inventing magic internet money you earn just by proving you did something useful, then spend to download. My suggestion was to mint that money using proof of work. Karma, built with a couple of brilliant students, was likely the first proof of work minting scheme, about six years before Satoshi.
Why didn't you continue pursuing Karma after building it?
Mentors, older professors, told me it was technically exciting but a dead end for funding, especially since this was right after 9/11 when everyone worried about terrorist financing. Talking about virtual internet currency just would not go anywhere for a young assistant professor, so I abandoned it. Satoshi came later, after the 2008 financial crisis, with far better timing and one key innovation I hadn't made.
What was Satoshi's real breakthrough compared to earlier academic consensus work?
Academics like Lamport and Liskov built classical consensus protocols trying to replicate a small closed parliament with one hundred percent guarantees. Satoshi redefined the problem as probabilistic, realizing that something happening with 99.99999 percent probability is effectively identical to certainty given our imperfect computers. That shift let him combine proof of work minting with consensus itself, creating an open system anyone could join, which was the real innovation.
What are Bitcoin's limitations if you want more than a store of value?
Bitcoin is perfect as electronic gold and a settlement layer, but it doesn't scale in participants, transactions per second, or latency. If you want a medium of exchange or a platform for others to issue digital assets, you need different infrastructure entirely. Bitcoin is playing its own game very well, but that is no longer the game I'm interested in solving.
How does Ethereum fall short for representing real world assets?
Ethereum is a wonderful general computational platform, but it has no native concept of an asset besides ETH. ERC20 tokens are invisible to miners, who just execute code without knowing its value. That means the system can't provide special handling, storage, or value added services for specific asset classes, so something like tokenized real estate gets treated no differently than any other transaction.
What makes Avalanche's design different from one coin, one virtual machine systems?
Avalanche has a groundbreaking scalable consensus protocol, but also a flexible foundation where assets are explicit to every layer, can be coded in any language, run on any suitable virtual machine, and have network participants chosen by the asset creator through subnets. Governance also lets token holders vote on changing economic parameters, unlike fixed schedules baked permanently into chains like Bitcoin.
Why would an enterprise choose Avalanche over private permissioned blockchains?
Almost every Fortune 500 company has tried private permissioned blockchains and abandoned them as dead ends. Avalanche offers a continuum between fully private and fully public blockchains through subnets, so you can require specific resources, restrict participation, or even mandate compliance with a jurisdiction's laws for a given asset, something you simply cannot do on Ethereum or a purely private chain.
Building something daring? Sunny talks to founders like this every day. Fifteen minutes to see if your story belongs on the stage.
Claim your pre-interview