Insights from Coinbase co-founder Fred Ehrsam
I host a podcast which takes deep dives into science, tech, and sociological topics. I do this via interviews with world-class experts who have the patience to engage in truly unhurried discussions of their fields. My episodes are untethered from the headlines, as they’re meant to resonate with future as well as present-day listeners, ideally over a span of years.
Occasionally, though, things happen to line up with current events. And was that ever the case this week — as today’s episode is about cryptocurrencies, which have been repeatedly dashing price records, even as Bitcoin debuted on the futures markets this past Sunday.
My guest is Coinbase co-founder Fred Ehrsam. With over 13 million users, Coinbase stores more cryptocurrency than anyone else, and the total monthly volume of trades between currencies on its platform stretches into the double-digit billions. You can find our interview right here — or you can hear it on your smartphone by typing “After On Podcast” into the search field of your podcast app.
Before we sat down, Fred and I put significant thought into structuring our conversation as both a rigorous introduction to cryptocurrencies (the first half hour or so), as well as an advanced discussion of the more wild and speculative things which could emerge from all this. Our prep, the interview itself, and the research I did in connection with it was a great education, and I’d like to share some of the more intriguing things that emerged from it.
Fred discusses the financial incentives which radiate through crypto ecosystems in more nuanced ways than I’d previously encountered. Those enjoyed by cryptocurrency miners, of course, are obvious. In exchange for running the servers that secure the networks, maintain their ledgers, and process transactions, miners get transaction fees from users, and also collect the new coins which enter money supplies at carefully-metered rates (if you’re new to this, Fred and I discuss miners in detail at timestamp 21:10 in the interview).
What I’d never considered are the massive incentives facing the developers of open cryptocurrency software. The social and psychic rewards of traditional open source are certainly there for them, along with a certain communitarian ethic. But before Fred mentioned it, I hadn’t considered that most of these folks are also deeply bought into the currencies they work so hard on.
In retrospect, this seems head-smackingly obvious. If you pour your evenings and weekends into building something, of course you’re going to use the thing you’re building. And if the thing you’re building is money, you’ll inevitably buy at least some of the currency you’re creating. You hopefully won’t go into debt betting on it (please don’t!). But if you were an early developer on Bitcoin, Ethereum, or another successful project, a minor bet would be enough to make you hugely vested in its ongoing success today.
The diffuse rewards of open source have long created amazing things without this financial dimension. Linux and Wikipedia should rank with the Great Pyramids in the pantheon of human wonders (and as a former Cairo resident, I don’t say this lightly). Now couple that power with hard economics — itself the world’s mightiest organizing force — and you really have something. While there are many reasons why BitCoin is worth $300 billion, a giant one is that it’s the super-sexy open source project that paid.
Fred astounded me by saying that though hundreds have made real contributions to the codebase, BitCoin’s truly core developers number perhaps fifteen. Brilliant newcomers would no doubt be welcome to join their ranks and make major contributions. But that would mean joining a project whose old-timers have all banked fortunes, which your own voluntary efforts could never hope to replicate. This surely helps explain the explosion in new currencies. Because unlike the situation in other open source areas, being established and well-known will reduce, not enhance, a project’s allure to developers in crypto.
Much as I love the ethos of Mozilla and Burning Man, this doesn’t bother me, because I also love Google and the Rolling Stones. Magnificent things — even great art — can be made by folks who are deeply interested in making money. So by extending incentives in subtle new ways, the blockchain and crypto might just create emergent wonders, which the world would not otherwise see.
Tech has always aligned incentives more deliberately and aggressively than any other industry, to its great benefit. Facebook achieves things that Clorox and Ford can’t dream of in part because it can make its employees stupidly rich. And here, it’s worth noting that when a cryptocurrency really succeeds, the upside radiates beyond its developers to include much of its user base (as anyone who got into BitCoin before, say, last week will attest. For now).
And what if Facebook not only cut its employees in on the deal, but its users as well? Of course we can’t know, because corporate structures don’t support such things. Fred — whose perspective on this extends clear back to the Dutch East India Company (timestamp 53:37), and indeed far beyond that — enjoys speculating on how blockchain incentives and governance might enable entirely new methods of organization. We discuss the notion of an Uber-like entity that’s blockchain-based — perhaps collectively owned by its drivers, or even its users (timestamp 1:03:02), and indeed at least one entity is chasing this very dream.
One of Fred’s more fascinating topics is blockchain-based AI (timestamp 1:21:54). The precept is that AI’s lifeblood is great masses of data, for training and honing its output. And however much data Google has, the rest of us combined surely have more. A blockchain-based AI, which takes data contributions from one and all, and rewards data donors when their contributions are used, might therefore become far mightier than any proprietary AI.
I find this notion riveting. But a huge countervailing force will be the vast relative efficiency of proprietary networks over decentralized ones that securely leverage the contributions of unknown (and therefore untrusted) contributors. Bitcoin’s founding genius lies in enabling the latter type of network, by cracking the famous “Byzantine general’s problem” (timestamp 23:30). But this comes at great cost, in terms of the compute cycles demanded by its underlying cryptography and “proof of work” requirements.
How great a cost? Well, the carbon footprint of Bitcoin’s infrastructure is said to rival that of Denmark. And we can safely assume the Visa card network needs far less juice than six million Europeans in chilly latitudes. Yet Visa can handle 47,000 transactions per second, vs. a single-digit number for Bitcoin. And as AI is orders of magnitude more computationally demanding that ledger transactions, an open, blockchain-based AI will have to overcome tremendous headwinds.
That said, Fred’s points about the immense rallying power of financial incentives are spot-on, and armies of brilliant programmers are just starting to explore the blockchain’s potential as a computing platform. Perhaps in the end it will be a horserace between the power of incentives vs. the efficiencies of centralized networks?
If these subjects interest you, you should enjoy our interview — particularly if you’re seeking a rigorous primer on cryptocurrency basics, which we start out with (experts may want to jump to the 30-minute mark, or thereabouts). Topics covered in prior episodes of the podcast include neuroscience and consciousness, drones, augmented reality, Medium itself (an interview with Ev Williams), Fermi’s Paradox, and more. I’ll add that my guests have included Chris Anderson (who runs the TED organization), legendary tech observer and participant Tim O’Reilly, and the ever-controversial Sam Harris.