My thesis: AI is digital abundance, crypto is digital scarcity, and the world needs both

As the next crypto bull market appears to be gathering steam, I am occasionally asked what my long-term thesis on crypto is, and how that thesis has changed since 2014.

Most lawyers don’t especially need a market thesis. If you ask a litigator, “what’s your thesis?” The response will be “people will fight, and I will bill a zillion hours.” Outside GCs and tech contracts guys such as myself are expected to be a little more opinionated. Where traders are tasked with predicting the future of prices, or VCs are tasked with predicting which companies or protocols will be winners, commercial legal advisors have to guide entrepreneurs straight through the middle of Scylla and Charybdis so that they can succeed despite the fact that the rules will likely iterate several more times, in several different places, between the date the advice is given and the date it finally becomes irrelevant. Having a thesis is table stakes for this sort of work.

One common, and incorrect, refrain I often hear – several times a week – is that you can’t found a crypto company in the U.S. and that you should instead set up as a BVI or Cayman foundation or something like that. This is, generally speaking, not correct. US laws around securities and commodities are rather agnostic as to where you set up a corporation; what they care about is where your users are, the locations of the subject matter of the transactions your startup might facilitate. If you want to sell tokens that the SEC considers securities, for example, you could rent office space in a building across the street from the SEC’s by Union Station and print ICOs all day long, just as long as you aren’t selling those tokens to Americans, and were correctly availing yourself of the appropriate safe harbors.

Put differently, our regulators are so awful that people avoid the country even though they don’t have to, and the degree to which our regulators are awful communicates to these entrepreneurs that being incorporated in America is regarded as a reputational black mark and long term business risk. This is pretty much conventional wisdom among the crypto bar at this point. Questions where practitioners tend to disagree a little more concern (1) why our regulators are so hostile, and (2) whether our regulators will continue to be hostile in the future.

The answer to (1) is actually pretty easy, as long as you’re willing to talk turkey and be unafraid of offending people. Put simply, it’s entirely political. The United States was once a free country which as recently as 1905 took less than 5% of gross national income in government tax receipts. Now, its government is a bloated, heavily-indebted, Soviet Union-sized monster, constituting more than half of national economic life, run by a bunch of borderline Marxists (which designation includes most Republicans and Democrats) (a) who have done very well shuffling paper for a living for the last 10-30 years, (b) who don’t understand technology and (c) who stand to be made obsolete by it.

These folks are currently dominant in the universities, civil service, and upper echelons of the white collar professions. Exhibit A, ticking all of these boxes: Elizabeth Warren, who is also not at all coincidentally leading the charge against crypto on the Hill. Technology is a threat to these people’s way of life, a way of life which will almost certainly end in our lifetimes. Since these folks, being in power, are also in a great position to throw wrenches in the works and slow everything down, that’s what they choose to do.

The answer to (2) is rather more speculative, but I have seen enough to be more certain than not that I’m within striking distance of what will be the eventual outcome of the great crypto experiment. My thesis is that AI is going to demonstrate crypto’s necessity and drive adoption as quickly as AI proliferates. The faster AI’s takeoff, the faster crypto’s takeoff will also be.

Critics frequently harp on the fact that cryptocurrency does not yet have product-market fit, and/or that the Securities Act of 1933 is the right long-term regulatory regime for the asset class. In 2014 this view, which I also held at that time, was correct. Today, if we assume the world will cease to exist tomorrow, it is probably also correct.

To the extent those critics are currently right based on present technology levels, it is unlikely that they will be right for much longer. Artificial intelligence is the reason why.

We are already at a point where machines and software are so advanced, so capable of portraying human voices, faces, and emotions, that, among other things, soon we will not even be able to trust our eyes when having video calls with our own loved ones or speeches from our leaders, due to so-called “deepfakes.” This is a world where authentication, metering interaction with irrevocable digital cash (a la charging for receipt of e-mail like Balaji Srinivasan’s startup 21 tried five years ago), “proof of human,” and, in particular, strong cryptography, will become exceedingly important.

One thing the machines can’t do, and won’t be able to do for the foreseeable future, is bruteforce private keys. Which is where crypto comes in. We will soon be inundated by the endless shitposting of an army of infinite generative AI models trying to wend their way into our inboxes, our feeds, our head-space. Proving who the humans are and only letting them through if they prove they’re human – or pay a price – is the only way we’ll be able to filter the spam.

If it were true that in 2009 nobody needed the double-spending problem fixed, or that in 2014 nobody needed cryptographically secure state machines with money to execute contractual obligations, or that today anyone needs their transactions encrypted and hidden from AI-powered surveillance bots run by criminals or foreign threat actors, by 2029, it is entirely possible, even probable, that everyone will.

These are functions that no stock, nor bond, nor evidence of indebtedness, nor any investment contract has ever performed, but are ones at which cryptocurrencies of various kinds routinely excel. Regulators will eventually come to realize this of their own accord or because market forces compel them to.

So that’s my thesis. Crypto didn’t have product-market fit in the past because the technology for which cryptocurrency was actually built, practical AI, didn’t exist until this year. AI drives the price of creating distraction to zero, which will drive the amount of available distraction to infinity. The more AI the world has, the more crypto it will need to keep it at bay and put a price on – and defensive wall around – the most precious and irreplaceable resource of all: our time.

Image credit Patrick Blumenthal.

Leave a Reply