Benedict Evans - From Winters to Summers - TOKEN2049 Singapore 2024

I was asked to come and explain bubbles to a crypto audience which I thought would be an interesting conversation um and it's kind of interesting looking at this that I first started working on crypto in 2013 I joined andreon hitz in 2014 um back when Bitcoin was still this kind of funny little curiosity um that only people like biy talked about um in fact I knew biology before he was biology he actually had a surname back then um and so now here we are 10 years later and I was asked to talk about the history of Bubbles and I thought it would be useful to refer to a kind of canonical phrase this time is different um which is what people always say in a bubble um people always say this time is different the challenge is that they're generally right um every bubble is different it always is some that something has changed something is new there's some reason why this isn't the same as it was before the bubble didn't look anything like um the Japanese bubble which didn't look anything like 1987 or 1974 um every bubble is different every bubble has something has changed and yet it can still be a bubble um and so I thought it would be interesting to talk about what that means but also talk about the trouble with saying that things are bubbles and so the first observation would be that in a bubble in a bubble everything will work if we talk about how Technologies evolve um there's generally kind of a classic scurve where to begin with every new technology um is stupid it's useless it's a toy it won't scale um it's not good for anything it's a dumb toy for Rich teenagers and then there's a period when it starts working and it's very exciting and um you want your children to get a job in it or you want to get a job in it and this is also a period when you tend to have fraud um and you tend to have Bubbles and then there's a period when it's boring um because everything works and it's become mature and all the everything's been built and so this is where smartphones are today um the new iPhone came out last week it's amazing um but nobody's very excited because it's pretty much the same as the iPhone the year before on the other hand something like VR and AR right at the beginning of this process they don't really work yet it's not clear what they're good for it's easy to dismiss them as as a toy um and so new technologies that become important tend to go through this process on the other hand there are also Technologies um um that are just out just moving from being a toy to being exciting and so this is where crypto might be this is where large language models are today however um just because that process exists doesn't mean that every technology goes through it um there's also a bunch of technologies that didn't work or didn't become huge or moved off into a niche things like 3D printing or drones um sort of five six years ago everybody bought a drone for Christmas and then 3 days later you said well I've seen the roof of my house now and it turned out that drones weren't a mass Market technology um so you can't just kind of presume that everything will follow that process and so if we think about how people tend to talk about technology we've got on new technologies we've got kind of equal and opposite fallacies on the one hand people will say it's a toy and on the other hand people will say well people always say that everyone always says it's a toy people said that the iPhone was a toy therefore this new thing will work the fact that people are saying it's a toy proves it's going to work and you don't have to think very long to realize that there's kind of a logical flaw in that there's a great quote from Carl Sean um 20 or 30 40 or 50 years ago um who said that um they laughed at Columbus they laughed at the White Brothers but they also laughed at BO the clown so just because they're laughing at you doesn't mean it's going to work on the other hand of course it's not going to be a big thing unless people are laughing at you but just to look at this point about the right Brothers there's kind of a useful distinction to draw here on the one hand this is the right brothers first aircraft and it flew 100 150 yards it carried one person um it's pretty easy to look at this and say well this is useless it won't scale but of course it did scale and five years five or six years later blo flew across the channel and people were flying hundred of miles so if you looked at this and said it's useless you're missing the point you had to understand that this was going to get much better very quickly on the other hand um most of us don't own one of these this is a bell rocket pack from about 1960 um it's amazing you can fly but you can only fly for about 30 seconds because that's how much hydrogen peroxide you can carry on your back and the only way that you can fly for a minute is to carry twice as much hydrogen peroxide and that's too heavy you can't so there was no way for this thing to SC scale Beyond flying for 30 seconds so it was very cool in 1960 and it's still very cool but there was no way that it could become any better um another kind of useful quote for thinking about this comes from a theoretical physicist called Wolf Gang Pary who was asked to look at a student's work and his colleague said I don't think it's very good and he looked at it and he said it's worse than that it's not even wrong that is to say it was a theor it was something that wasn't actually a theory it wasn't a hypothesis there was nothing that you could test there was no question you could ask that would tell you whether this is true or false and that's a challenge with a lot of people looking at early stages of technology is people make statements that don't really have any predictive value they make statements that are not even wrong and so when you look at a new thing instead of saying that won't work or people always say that you have to say well what would have to change what would have we have to believe is there a road map for this to become useful is there a roadmap for this to work um how could what do we have to believe for this to become important what do we have to believe if we think this isn't going to be important meanwhile in a bubble everything not only will everything work but everything will work right now um there's an old saying in the stock market what's the difference between early being early and being wrong and the answer is nothing um some people in the room may even have owned one of these things this is the Apple Newton from the mid 1990s this was the future but it was 10 maybe 15 years too early um and the history of tech is full of things that were the right idea in a directional sense but they were too early hence my old boss Mark andreon like to say that all of the failed ideas from the Doom bubble would work now um not literally all of them but almost all of them would kind of work now because we've got enough people we've got enough infrastructure hence something like e-commerce um has now reached sort of 15 to 20% in most developed markets more in China more in the UK um but it took 20 years to do that and there were people in 1999 who were saying it was going to happen next year um it's kind of worth pointing out incidentally this is basically the most ing chart in the tech industry because it just grew 1% every year for 20 years um then there was that kind of brief surge in the pandemic and now we've gone exactly back onto the trend line like there some kind of physical law that shapes how this works you can see something very similar in the Enterprise incidentally if you're in the tech industry then cloud is old and boring and done and that's something from the past but the cloud is still only sort of 20 to 30% of Enterprise workflows it takes a long time for big companies to change how they do things it takes a long time for consumers to change how they do things and meanwhile the old model um can hang around for a long time um IBM is still selling mainframes mainframes are still out there inside big companies doing big company things in fact IBM is selling record Mainframe capacity every year as they carry on shifting Mo's law into Banks and Airlines and insurance companies and government tax offices so they're still there it's just that they became irrelevant meanwhile if we push forward into how um how that can evolve this is a chart of smartphone sales um smart phone sales don't start with the iPhone they start arguably in the mid90s but really they start in sort of 20002 2001 with things like palm and uh blackbery and Nokia series 60 um it took a decade of people selling those devices for the sales really to pick up and when the sales did pick up it wasn't the early winners it wasn't Nokia it wasn't blackbery it wasn't Microsoft um it wasn't Palm meanwhile in 200000 um mobile operators spent something over hundred billion doar buying Spectrum to run these Services it took 10 years before anybody did anything with it and they never made any money from that Spectrum any either all the money went to Apple and Google um and so there of pulling this chart apart be careful about the early winners um this is another example of um another example of the same thing um these are P PC unit share um we've got Apple down here in the bottom and we all kind of know that Apple pioneered the market but then lost the PC market but there also something like a dozen other companies that were selling different kinds of micro computer and PC in the late '70s and early ' 80s and so Not only was Apple never the majority but the IBM PC wasn't the majority until the mid 1980s either none of those companies are still around um so be careful of looking at the early winners be careful of thinking how quickly it's going to going to work this is a chart of um market caps of the six largest companies today versus the six largest companies in the dot bubble um and there's kind of an interesting spread here because on the one hand Tech big the biggest companies today are much larger but on the other hand the EV valuations are much lower Cisco back in the spring of 2000 was at a PE 100 um and so this is really a chart about the promise of what might happen versus the reality of what does happen um and when it becomes real incidentally you can also see space Caren here in the middle um one of these is not like the other um that's a question about Tesla being valued effectively as a company that's all about the future as opposed to other large tech companies that are being valued about what they have today um meanwhile of course you look at these companies um how many of these won the internet um I don't think anybody thinks that Oracle was a big winner from the rise of the internet arguably the only company on here that is even significant in the internet is Microsoft and that's kind of an also ran in the consumer internet um and so all of the companies that were being given the highest valuations in the Doom bubble ended up not really making any money or certainly not being the biggest winners out of the Doom bubble now if you go back a little bit further this is a chart from um 1995 a diagram from 1995 of The Intern except that back then there wasn't just the internet there were lots of other networks as well um and so there were private networks in Pho and bitnet and minitel and AOL and God knows what else and meanwhile within the internet it wasn't clear how that was going to work it wasn't clear it was just going to be the web it wasn't clear that how the web was going to work it wasn't clear whether a web browser should be an editor whether this was document sharing or publishing absolutely nothing about how the internet was going to evolve was clear for 10 15 years really until the late '90s or the early 2000s and so you could have put an awful lot of money into Deli and compy serve and phonet and believed in the future of the internet and got absolutely nowhere um no one really knows anything about how the future is going to evolve it's certainly not like the last 10% which is where all the money is there to be made meanwhile in a bubble it was obvious in hindsight so we've all seen lots of bubble charts but I thought this one is kind of fun this is the market cap of Zoom versus Exxon in the pandemic um there was a brief moment when Zoom was worth more than the world's biggest oil company and even at the time that seemed kind of silly um there were a lot of reasons why Zoom didn't have a mate Zoom wasn't defensible in the time it was kind of obvious of course this is a bubble of course this is ridiculous um and you can look at something like um metaverse um obviously nobody here was ever interested in metaverse but there were other people who thought that metaverse was going to be a big deal um there may still be in another 10 years once Mark Zuckerberg has spent another 75 billion on it um nobody talks about metaverse anymore of course one should also point out um a similar chart for crypto which um was the only thing that anyone in Tech wanted to talk about for about 18 months and has now settled back to a slightly more rational level of interest um but of course you can call Bubble about anything um so these are three headlines talking about Facebook buying Instagram for a billion dollars which was obviously completely insane and and an evidence that the world had forgotten the.com bubble today Instagram is probably worth half a trillion dollarss um meanwhile and go back a little bit further when Yahoo bought flicker for 15 or 20 billion million dollars million with an M $20 million for an internet company that was obviously insane and a sign of that the bubble had returned um there's a classic quote here that's kind of apocryphal that economists have predicted nine of the last 15 recessions um so it's very easy to call Bubble about everything it's very easy to say this is a bubble just as it's very easy to say that it's not a bubble so how can you think about this when you're in it when by definition you kind of can't know and so I think there's a few axioms that one can think about first of all be careful of people drawing straight lines on log scale charts um some of you may remember a lot of log scale straight lines on log scale charts around um crypto a couple of years ago that kind of analysis tends not to end very well in fact you should probably be careful about straight straight lines on charts in general um this is nvidia's quarterly Revenue um now I'm not a semis analyst um I'm also not a data center analyst um but I am kind of reminded of the line from The Godfather when Michael Kone says that if history's taught us anything it's you can kill anybody and I think the technology industry equivalent of this is probably that if history has taught us anything it's it semiconductors are cyclical and commodity technology gets cheap and of course that new technologies tend to lead to Bubbles um one should also be careful about politics um this is a famous document written by John Perry Balor who is um part of the Grateful Dead um who wrote um governments of the industrial World You Weary Giants of Flesh in steel I come from cyberspace you have no sovereignty Where We Gather there are quite a lot of people in the crypto industry who thought that um National governments had no sovereignty and none of the laws apply to them and as you heard on the previous panel um turns out that real world laws do apply to crypto just as they applied to anything else so be careful of politics be careful of radical politics be careful of thinking that New Politics will change how Technologies work um there is an observation that you know truly transformative Technologies very often come with crazy politics because you kind of have to be crazy to believe that you're going to change the world like that and so that's how a lot of early Silicon Valley happened it's also how the early internet happened as we see here it's what happened with open source um and indeed a lot of early crypto came with crazy politics but you kind of have to cautious about believing that um you have to be cautious of people saying that the old rules don't apply and just as being careful about politics you also kind of have to be careful about religions um so open source people were convinced that micro that open source was going to destroy Microsoft that commercial software was evil and that nobody was going to buy software anymore and here we are 25 years later and the iPhone is full of Open Source software but it's not open but it has millions of apps and billions of downloads so what does it mean to say that it's not open open and this is kind of the messy reality of how amazingly exciting new ideas get translated into products that billions of people use um the new ideas happen but at different levels of the stack and maybe the consumer knows maybe the consumer doesn't know maybe it's important at this level but not important at that level arm is a open standard but it's not open um iOS is full of Open Source software but it's not open but it has more software installed than any platform in history so is it open or closed well it kind of depends what you mean and depends what level of abstraction you're thinking about um finally of course you should be aware of being in a bubble yourself this is a quote from John Peele who's a famous British DJ and he told a story that he said to a friend at a record company I can't believe that album wasn't a hit I everybody I knew bought a copy and the record exact said no you knew everybody who bought a copy and I think we can kind of all think of examples of cases where people believe that everyone else thinks the same as them and will make the same product decisions as them and the same purchase decisions as them and this is kind of a recurring challenge particularly in the technology industry we've seen this with everything from guies to command lines to smartphones to open source um be careful of believing that your own product preferences will automatically become a universal preference that everybody will want else will want to follow um you should also incidentally be aware of the fear of missing out which is of course the greatest sign of a bubble ever um most people will remember here Sequoia put up um a very long essay talking about how wonderful s bankman freed was about 3 days before he was arrested um and then they deleted it from their website so here we've got a screen grab from web archive because nothing is ever forgotten forever um fomo is probably the greatest ever signal that you're in a bubble because the moment the moment that you actually know you're in a bubble is when you say to yourself well gee I don't really understand that thing and I'm not quite sure why the price is doing what it is but all these other idiots seem to be getting really rich so maybe I should have a goto that is pretty much always the signal that you're 3 to six months away from the top of the bubble um the challenge of course at that is that you can call a bubble for years before it pops um hence another important quote from um JM kins markets can remain irrational longer than you can remain solvent um calling a market is a very specific kind of skill and it's very tra and trading is very different from investing and that takes me to um one more chart um this is the annual return on the S&P um back since 1930 um in generality in the long term the market tends to go up but every now and then it doesn't um every time incidentally that I go and look at statistics from 1929 1930 1931 I generally think I've made a mistake in my spreadsheet as you make the chart and you think wait no construction equipment sales can't have gone down 90% in 1931 so oh yes they did newspaper sales can't have gone down by three quarters in 1932 oh yes they did um that's kind of what you're seeing on the left side of this chart um but of course over the long term the market goes up um it just wobbles a bit in the meantime as long as you're sure that you believe in the growth of the US economy um and hence I'll leave you with one final quote um from the what the man who was the richest man on Earth at the time um John D Rockefeller um the story is that an eon a reporter a newspaper reporter asked him what he thought would happen to the stock market and he said uh it will fluctuate um which has generally been a good way to think about what's going to happen to the stock market in the short term in the long term of course um it tends to do something else and with that I will say thank you