The Breakdown - Bitcoin Prehistory and A Warren Buffett AI Bot

Episode Date: June 4, 2023

On this week's Long Reads Sunday, NLW reads two pieces by Byron Gilliam: Crypto's family tree Will the next Warren Buffett be an AI bot? Enjoying this content? SUBSCRIBE to the Podcast: https://pod.link/1438693620 Watch on YouTube: https://www.youtube.com/nathanielwhittemorecrypto Subscribeto the newsletter: https://breakdown.beehiiv.com/ Join the discussion: https://discord.gg/VrKRrfKCz8 Follow on Twitter: NLW: https://twitter.com/nlw Breakdown: https://twitter.com/BreakdownNLW    

Welcome back to The Breakdown with me, NLW. It's a daily podcast on macro, Bitcoin, and the big picture power shifts remaking our world. What's going on, guys? It is Sunday, June 4th, and that means it's time for Long Read Sunday. Before we get into that, however, if you are enjoying the breakdown, please go subscribe to it, give it a rating, give it a review, or if you want to dive deeper into the conversation, come join us on the Breakers Discord. You can find a link in the show notes or go to bit.ly slash breakdown pod. All right, friends, welcome back to another long read Sunday. This week, we are focusing in on the work of a radically under-heralded writer.

That is Byron Gilliam. Or Gilliam, if I pronounced it wrong, Byron, I apologize. Byron is the primary newsletter writer for Blockworks, and pretty consistently combines Bitcoin crypto and macro in a way that is extremely enjoyable to read. Today, to give you a little taste of Byron's writing, we're going to do a piece that looks backwards and a piece that looks forward. We start, of course, with the backwards. The piece is from a newsletter called Crypto's Family Tree on May 30th,

and it kicks off with a Chinese proverb. To forget one's ancestors is to be a brook without a source, a tree without a root. The good news of Crypto's long ancestry. Newsletter readers I've learned are a savvy bunch. So if I were to ask what the first blockchain was, I'm confident you'd immediately suss it out as a trick question and decline to answer. Unless that is, you've heard of surety, which offered a blockchain-based timestamping service for digital documents starting way back in 1995.

If, like everyone else, you associate blockchains with the distributed computing systems that are Bitcoin and Ethereum, you might question how surety could classify as such. Running algorithms on a decentralized set of computers would have been impractical in the dial-up era of the 90s. But that's not the only way to do it. Surety held customer documents on its own servers, but achieved trustlessness in a most analog manner, by printing its hashes once a week in the classified section of the New York Times. That admittedly does not sound very useful by today's standards. One week block times?

But I offer it here as evidence that the history of blockchain's predates Satoshi's white paper of 2008, and as a fun fact to drop into conversation at your next crypto conference. But here's an even funner fact to wow your crypto friends with. That first blockchain is itself predated by smart contracts. Crypto's family tree. You're probably like me in thinking of smart contracts as the second big step in the evolution of blockchains. Satoshi invented internet money with Bitcoin, and then Vitalik made blockchains programmable with Ethereum. But smart contracts have a prehistory of their own. In a fascinating

note, Professor Michael Monelli, co-founder of the Z-Yen think tank, offers a step-by-step guide on how proto-smart contracts were used to settle foreign exchange swaps as early as the 1980s. Counterparties would agree the terms of a swap, with contractual calculations written into a spreadsheet stamped with the hash, no different than the hash is used in Ethereum blocks. The hashed spreadsheet file was saved on a floppy disk and given to a lawyer to be held for safekeeping. When the swap was due to settle, the counterparties would convene at the lawyer's office, where the floppy disk would be retrieved from the safe. The hash on the file would be checked to ensure it was the same spreadsheet with the same

terms as originally agreed. An Oracle, e.g. That Morning's newspaper, was referenced for the prices of the relevant FX rates which were entered into the spreadsheet. The spreadsheet would calculate who owed what to whom. Settlement, I'm guessing, would then follow per normal channels. In this, the arrangement had, quote, all the characteristics of a smart contract, a transparent, immutable, and secure agreement between counterparties to exchange value according to the rules agreed.

Definitely a contract and definitely smart. Still, I wouldn't blame you for thinking that settling swaps with floppy disks is about as related to NFTs and meme coins on Ethereum as humans are to amoebas. But humans are related to amoebas, and that's a useful thing to know. Ultimately, we all come from the same place. I'd argue it's similarly useful to know that blockchains have a history that predates Bitcoin, because if crypto is its own entirely original thing, immaculately conceived in 2008, it may also be prone to immaculately disappear.

If Satoshi's 2008 white paper was a big bang responsible for creating the entire universe of crypto, there's a chance that whole universe could be one big mistake. A detour that turns out to be nothing more than a horrendous waste of investor money and developer talent. But if crypto is just another step in the long-running digitization of everything, it's more likely to have staying power. History suggests it's the latter. Shirti's early blockchain was based on a 1991 paper, how to timestamp a digital document, written by cryptographers Stuart Haber and Scott Stronetta. On a crypto family tree that puts Shorty not far removed from Bitcoin, three of the eight papers cited by Satoshi in 2008 were authored by Haber and Stronetta. So second cousins maybe? And the ideas

from those papers can be traced back even further. In the aforementioned Z-Yen article, Professor Maynelli and his co-author recount the intellectual history of blockchains going all the way back to 1953, when hash chaining was mentioned in an internal memorandum at IBM. This, the author suggests, demonstrates that discussion of an effective blockchain may have predated Satoshi Nakamoto's Bitcoin white paper by 55 years. And that, in my opinion, is great news. All right, back to NLW here. one of the things that you sometimes hear as a reason why Bitcoin couldn't possibly be the winner in a long-term crypto space is the idea that it's somehow MySpace or Friendster, the first and inevitably weaker in a line of things that come after. One of the reasons that logic doesn't hold is exactly what Byron's pointing out, that Bitcoin is itself the culmination of a huge, huge amount of work.

Now, Byron has pointed out a couple of examples, but the 1990s and early 2000s are replete with versions of digital cash and e-ekechews. cash and all these things that just didn't quite work, or at least didn't quite get to where Bitcoin got. That's not to say, of course, that there aren't significant innovations with Bitcoin. But it is to say, as Byron points out, that it wasn't an immaculate conception. It was the culmination of a huge amount of work, creativity, and experimentation, timestamped with just a little bit of magic. Next up, we zoom forward with another Byron piece. Will the next Warren Buffett be an AI bot?

This one starts off with a quote from Stephen Pinker in 1994. The main lesson of 35 years of AI research is that the hard problems are easy, and the easy problems are hard. At the height of the dot-com bubble, the investment bank I worked at employed over 200 traders on its NASDAQ desk alone. That turned out to be the all-time high for the equities traders' jobs market, sadly, and not because the NASDAQ bubble sequentially burst, but because it was the last time the stock market was a strictly human-versus human affair.

trading algorithms were introduced shortly thereafter. Those algos exposed trading as a mostly mechanical job. Given just a little direction, machines could hit bids and lift offers many times faster than humans could ever hope to. 20 years later, that same type of institutional trading desk, doing that same amount of business requires maybe a dozen humans? And now, a new generation of AI algorithms has exposed another profession as being mostly mechanical.

Writing. I just cannot catch a break. Language, it turns out, is nothing. as complicated as we thought. AI can replicate the language function of our brains with statistics and probabilities that are not really all that complicated. The good news for writers is that while algorithms replaced nearly all of the traders, they replace none of the investors. That's because the best investors, like the best writers, are creative. In both cases, success is more a function of EQ emotional intelligence than IQ logic and reasoning. The AI machines seem to have got IQ down.

Could they get EQ too? There's no way to program. We don't know how it works in brains, so we don't know how to code it in machines. EQ, which seemed like the easy part, is actually the hard part. And IQ, which seemed like the hard part, is actually the easy part. But the fear is that EQ could become an emergent behavior of AI, that with enough computing power, machines will teach themselves how to think. There's no sign of it so far. I asked Google's barred why Nvidia shares are down this year, and it listed several reasons,

including market volatility and increased competition. It all sounded very plausible, aside from the fact that Nvidia shares are, of course, up this year by 170%. I was leading the witness because LLMs have a ready answer to every question. Warren Buffett, by contrast, would tell you, I don't know, a thousand times while he waits for one of the fat pitches that he says come along only once every five years or so. An AI investment manager would swing at every pitch, most of which I expect they would whiff on, because AI models have the same limitation as every other algorithmic model. It can only look backwards, and looking backwards is not super helpful in finance.

You've heard it a hundred times. Past performance is not indicative of future returns. That's why, in traditional quantitative trading, the alpha comes not from a machine, but from a human with an idea telling a machine what to do. LLMs don't have ideas. They can only match patterns in the historic data. AI will discover some patterns that have hitherto gone unnoticed, but those will quickly get arbed away as every model will find the same ones.

same as when the machine started trading equities 20 years ago. Instead of extracting value from markets, the equity trading machines mostly added value by tightening spreads, adding liquidity, and making markets more efficient. I suspect they will mostly do the same in crypto, potentially with one additional benefit. The non-wisdom of crowds. It's not just investors that allocate capital. CEOs do too. Warren Buffett is, of course, a great stock picker, but what's made him the greatest investor of all time is his ability to judiciously allocate capital as CEO of Berkshire Hathaway. Berkshire employs about 380,000 people, but only about 20 of those work in the Omaha headquarters, which is where the capital is allocated, and most of the

capital is still allocated by Mr. Buffett himself. That's not only because Buffett is an investing genius, but because capital allocation needs to be centralized. The wisdom of crowds may work at the market level, but not at the company level. There's a reason why shareholders don't vote on every investment made by the companies they own. This is a problem in crypto, where everything at least aspires to be decentralized. And maybe that's an opportunity for AI. I'd argue that a mediocre AI model would likely do a better job of allocating capital than the best Dow. MakerDAO is attempting to prove me wrong, and I hope they do. Its endgame plan will put some much-needed structure to its capital allocation process by founding several sub-Dows that will allocate funds received from the

parent-dow into individual projects, which makes it sound a lot like Berkshire Hathaway. But it's missing a key element in Berkshire's success. Centralized decision-making. Holders of the maker token will vote on allocating capital to the sub-dows, and the sub-dows, with their own governance tokens, will, I think, vote on allocating to individual projects. That's too much voting for my Tradfai mind, and if it were up to me and we can be thankful that it is not,

I'd go all in on AI and let robots do the allocating. With AutoGBT, capital allocation would be just a matter of good prompt engineering. Extend loans to high-quality borrowers at fair market rates, or something. That, to me, would be an improvement on the current standards of capital allocation at many protocols, which is not a high bar. Bigger picture, it would be a way to bring centralized capital allocation to decentralized crypto. Do I think it would be as good as Warren Buffett? No, but that's an impossibly high standard. Just helping crypto do the easy things would be good enough. All right, so obviously this piece is combining two of the things that I podcast about daily,

so you'll love to see it. And I will add one little note on each side. When it comes to the AI side, I agree that I have some amount of skepticism around how far or in what ways AI-driven finance or investment is going to be an improvement. I think Byron is directionally correct that it's going to look closer to what algorithmic trading did in equities markets than to something where human investors are just replaced by robots. That said, on the way to us realizing that, I think a lot of people are going to outsource their financial decision-making to LLMs and specially purposed AIs. The optimistic case is that those dispassionate bots do a better job of trading for the average person than they would do themselves, the less optimistic

cases that they, well, don't. If you go look at chat GPT plugins right now, there are a significant number focused on finance. So you have to think that we're going to see a lot of experiments in this space in the months to come. Now, when it comes to capital allocation in crypto and more specifically within decentralized networks, the important thing to suss out here for any given situation, any given protocol or any given project, is what the goal of decentralization is. The reason the decentralization, for example, matters in democracy is not that it's an efficient way to make decisions. It's not even necessarily that it gets society to the right decisions. It's that we in democracies have determined that it matters more for everyone to have a voice

than it does to always quote-unquote get it right. Getting it wrong in the right way, in other words, is sometimes better than getting it right in the wrong way. I think something similar as at work with Dow's. If we're talking about it, just simple treasury management, sure, maybe it's the type of thing you want to outsource, whether it's to AI or traditional investment managers or some other approach entirely. But when it comes to decisions about what people want to do and how they want to allocate scarce resources, it may be that the point of many of these DAOs is to give the people in them a voice in what they do. Anyways, really good stuff to think about and another great example of Byron's writing.

If you want to go get more of that, he writes for the Blockworks Daily newsletter, their free newsletter. All you got to do is subscribe and you will get it in your inbox. And by the way, this was not sponsored by Blockworks. I just happened to like Byron's writing a lot. Anyways, guys, that is it for today's Long Read Sunday. I hope you are having a wonderful weekend. Until next time, be safe and take care of each other. Peace.