As 2021 draws to a close and Covid cases spike, it’s easy to think not much has changed. But in between pandemic waves, we’ve made major progress in science and technology. And that progress gives us clues about how the future might play out.
Host Azeem Azhar reflects on the tech trends we’ve seen this year and key shifts in industry and society that are changing cities, the labor market, and the semiconductor business.
He also discusses:
- Why semiconductor chips were hot property (and will only get hotter).
- What people got wrong about the “death of the city.”
- New media, the creator economy, and the “great resignation.”
@Azeem
@exponentialview
Further resources:
Exponential View Newsletter
Learning From 2020: Azeem’s Takeaways
AZEEM AZHAR: Hi and welcome to the Exponential View podcast. I’m your host, Azeem Azhar, and we’ve got two special editions of the podcast coming up. In the first of which, we’re going to reflect on what has been frankly quite a weird year of ups and downs and somehow stasis. In the second, we’re going to look forward to the trends from 2022 and beyond. For many of us, with COVID cases climbing and restrictions being put back in place, we may feel that life today, December, 2021, looks quite similar to December, 2020. But in between the spike of this COVID variants, there has been so much going on.
It has been quite a fantastic up and down year of developments across technology, from AI breakthroughs, to new types of chips, to record levels of funding for startups, to continued pressure on big tech firms from competitional authorities, and of course, whistleblowers. Crypto has had a crazy year of record highs and nauseous volatility. Of course, we’ve seen the arrival of a $90 million JPEG and a new multi-billion dollar industry based around non-fungible tokens or NFTs. Exponential View delivered 38 podcasts in this year, excluding this one. Of course, I published my book, Exponential or The Exponential Age, if you are in the U.S. and Canada.
So between COVID spike and COVID spike and those restrictions, a great deal has happened. So let’s turn to the retrospective of 2021. This year has been an incredible boom in the startup ecosystem and venture capital funding that supports it. Just by the numbers, prior to 2021, across the globe, we had never seen a quarter where $100 billion of venture capital was deployed into startups. In 2021, this record year, every single quarter saw more than $100 billion go into early stage companies. In one case, in the second quarter of the year, $159 billion went into startups.
So it’s pretty dramatic in terms of the growth level of interest. But where has that new capital gone? It hasn’t spread equally. So I want to pull out three little highlights. The first is what’s happened in Europe. So in 2021, around about $120 billion of venture capital went into European startups. But to put that into context, that is more than the total amount of venture capital that went into European startups in the previous four years.
So we’re starting to see investors recognize and understand the potential not just of the founding teams in this area, but also of the markets in that area, and of course, the consequences that there’s been rising valuations as well. But I do think that this starts to really put Europe more on the map than it has been previously.
There were two other areas which I think have attracted a lot more venture capital than in previous years. The second is this area of deep tech, deep technologies or hard technologies or tough technologies. So traditionally, the challenge with funding tough technologies, things that might have involved chemistry or biologics or hardware, is that they require a lot of capital and they don’t scale as quickly as consumer internet companies or software companies or mobile apps. That’s why VCs traditional, over the last 15 years or so, have vectored in towards those zero gravity software, internet-y type companies. But there’s been a realization I think that as we go through this industrial transition, what I call the exponential transition, and we build the new foundations for the next 20, 50 or 100 years, that that is going to require new types of technologies. That has drawn a new class of investor out to back deeper, tougher companies, companies that suck CO2 out of the atmosphere and turn it into jet fuel, companies building quantum computers, companies building new types of chip, companies building new battery technologies.
Of course, it’s not good enough to say, “Well, there’s an industrial transition going on and therefore you should invest.” Of course, investors have to be rational. They need to know that they’re not taking risks that are completely unmitigated. So there’s been this other thing that’s been quite helpful to unlock capital which has been the arrival of the SPAC, the special purpose acquisition companies. So these are pools of capital that will acquire startups and sort of flip them public. So what you end up doing is seeing quite young technology companies having these public listings before they’ve had commercial traction, certainly before they’ve proven longstanding business models.
The importance of the SPAC is that the SPAC sits there as a potential promise to earlier investors that we will acquire the company and there will be some kind of liquidity and some kind of exit for you. So alongside this tremendous rise in venture capital, this tremendous rise in deep tech investing have been the SPACs that have catalyzed a lot of it.
The third area that I think is really important is what has happened within climate tech, that is founders who are trying to tackle the core issue of extracting CO2 from the atmosphere or creating new industrial and business processes that use less CO2. Now, climate tech investment has grown. It’s more than doubled from last year and I think it’s a really, really fascinating area. I’m going to say more about climate tech in my forecast discussion, which is in next week’s podcast. So listen out for that.
2021 was the year that everybody started to care about silicon chips. In most banal way, we couldn’t get our cars for a long time because supply chain shortages had upended car manufacturers’ plans. They couldn’t get the chips they needed for parking systems and driver control systems and the entertainment systems. But there’s a bigger story to what happened within the car industry. That story was a realization that chips are going to form the fabric of the exponential transition that we’re going to be using so much more computing over the next 5, 10, 15 years than we ever have. I suggest the amount we’re going to use is going to grow even faster than has historically. So chips have started to become things that matter from a business standpoint, from a research standpoint, from an engineering standpoint and from a geopolitical standpoint. There’s something fascinating that’s happened within this industry in that we’ve noticed companies like Nvidia, which makes the graphics processing units that were first used by video gamers, now used to drive AI systems have really seen their share prices, their valuations rocket in the last couple of years.
TSMC, the company in Taiwan that actually makes chips for most of the chip companies has found itself being commercially really valuable, but also the axis of a lot of geopolitical pressure. One of my favorite stories though is the Dutch company ASML, which has become this multi-hundred billion dollar company, and it makes lasers. I think lasers are really cool, but ASML’s lasers are even cooler. They cost about 130, $140 million. They’re extreme ultraviolet lasers. They involve vaporizing tin and it’s those lasers that are required to etch the silicon wafers on which the state of the art chips are made.
So we’ve got this fascinating moment where the next 10 to 20 years means we need lots more processing power and that’s going to come from a whole array of chip companies. One indicator of how important this is, is the degree to which chip companies are now much, much more profitable. It turns out actually that for lots of their history, the profit margins on chips were not so high. The amount of global profit pools that the semiconductor industry controlled was pretty low. Actually, if you looked at sort of sector by sector, it always looked a bit like a laggard.
But just in the last 15 or 20 years, according to research by McKinsey & Company, its control of industry profit pools has come from being a laggard to being the sort of second or third most important sector anywhere in the world. So chips have found themselves coming quite sexy. Now, it’s not just the very, very big companies. It’s also that we are seeing I think sort of three really interesting other dimensions. The first is that there are new types of chip companies emerging, traditional chip companies, optical processing units and chips designed specifically for AI.
Graphcore, for example, in the UK, and Cerebras brass in the U.S. Later down the line, we’ve also got quantum computers, again, a topic I will touch on in the next podcast where I look forward. But the second interesting area is the realization by governments that being able to control one supply of chips is both a national security and an economic security issue. A couple of years ago, what governments were concerned about was the idea that chips which are inherently dual use might get in the hands of adversaries. There might be IP theft amongst other things. Then the debate started to shift and governments started to get concerned about the trustworthiness of their supply chains, the idea that with these global supply chains, kind of core semiconductors could have back doors put into them by an adversary where factories might have been located. So that became this national security consideration that was separate to the idea of perhaps military competition in dual use technologies.
Then with the arrival of COVID and the shortages that we saw, there was also this recognition that actually large parts of industry depended on these tiny little chips and you couldn’t see you where they were being made and you couldn’t control that. So what we have started to see in 2020 and ’21 is growing interest from Japan, United States, Germany, India to ask semiconductor manufacturers to build these expensive fabs in their local markets. So we have Silicon Valley, which is called Silicon Valley because of the silicon chips. In Germany, Dresden is starting to have a small cluster of chip factories. Silicon Dresden perhaps might be a phrase we’ll be using in a couple of years time.
The final area that sort of wraps all of this up is this attempt by Nvidia and SoftBank to acquire Arm, the British chip designer. Now, Arm is just an amazing company, has been building these chip designs that have become ubiquitous in mobile phones amongst other areas. Of course, it’s branching into other sectors like servers and AI chips as well. So if we look at this deal, Nvidia’s proposed over of Arm, it does encapsulate a number of the themes.
Late in 2021, the UK government asked for there to be a further review into the deal on the basis of competition concerns, which is if Nvidia controls Arm and Nvidia has this big control of the GPU industry, does that start to reduce the amount of competition in the semiconductor business in general? The second is potential national security risks. For me, that encapsulates the renewed importance of the semiconductor industry in all of our lives.
2020 in the minds of some people was of course the death of the city. Cities were rife with disease and they were expensive and urban dwellers were going to flee. Now, listeners of my podcast and readers of my newsletter will know that I think it’s far too early to call time on the city. In fact, I think that their essential components, their value as points of agglomeration for socializing and culture and research and commerce are largely unimpeachable, but we’re now a year on from that discussion coming to the end of 2021. We can reflect back and try to understand what has really had happened within cities.
I think we are still seeing how this is going to play out. But there were a few things that I would like to kind of nuance. The first is that the city didn’t die by any means. We are still, of course, seeing people move from certain cities to other cities. But if you actually look at the real estate prices and how they have shifted over the last few years, there is a really interesting trend. If you look at low and mid-density cities, prices are going up. If you look at central business districts, so sort of typical skyscrapery homes of the commercial, prices have really collapsed. In fact, 10% off their Feb, 2020 levels by the end of 2021.
So that does paint a picture of the pressure of certain city designs. I think that that’s the important thing for us to bear in mind. The important thing is that we’ve had technologies that have eliminated distance many, many times before. The arrival of the car and the telephone at the tail-end of the 19th century could have brought with it the death of the city. I mean, those are technologies of distance. But in fact, what we saw was the arrival of the first 10 million person cities, New York in the 1930s, built off the back of the very technologies that enabled life at a distance.
The other thing to bear in mind is that we are still an urbanizing species. When we look at the data about people fleeing cities, we might be looking at places that aren’t very big by city terms. I mean, San Francisco is just not a big city. London in reality is not a big city. The biggest cities in the world are and will become the Karachis and the Dar es Salaams and the Lagoses and the Shanghais and the Mexico cities, Asia and Africa and other parts of the world that are getting richer, that are getting more complex knowledge driven economies. In those cases, the story will be that cities will remain these crucibles of innovation, creativity, and importantly, places where the ravages of climate change can best be addressed for large numbers of people.
2021 also brought to the fore questions about what work was and what was it to be an employee or to do work. We saw the great resignation take hold certainly in the United States. We saw all sorts of labor shortage problems emerge. This is happening alongside the renewal of the idea of the creator economy, the idea that given the tools of technology, we can all become economically productive and we can reach markets that are wider than just a few people in our neighborhood, or even within our own city. Platforms like Upwork allow anybody to do web design or email management or data-based analytics or copywriting for any client anywhere in the world.
So you have this kind of confluence of this exogenous shock hitting the labor markets, which was the sort of COVID lockdown and the great awakening I think of workers alongside these underlying technologies that are getting better. The kind of fundamental idea that has been around for as long as the internet has really been around the idea that anyone can connect to this network and can start to participate.
So it did feel that in 2020 to 2021, we saw this turning point around what’s been called the creator economy. There were a couple of really big funding rounds, Clubhouse, the online audio service raised money at a $4 billion valuation. Substack, which hosts newsletters, raised money at 500, 600, $700 million valuation. There was even data coming out about how much money individual creators were making and it was starting to look more significant.
Stripe, the online payments company did a really extensive survey of the creator economy and they reckoned that they tracked by their definition about 668,000 different creators who had in aggregate received about $10 billion in payments. Those payments coming either in the form of subscriptions or purchases of particular products and sort of premium products. What I had noticed was that there are many, many more platforms and tools and products available for individual creators than there were 10 years ago, let alone 20 years ago.
There’s also been a shift in worker expectations about what work might look like. As I talked to entrepreneurs with my investor hat on, what I’m also seeing is more and more entrepreneurs trying to look at this market and saying, “Well, does YouTube really work for a creator who is a yoga teacher, or who teaches a woodworking skill, or who is in helping cooking? Does it really work as a platform or do we need more specific platforms to meet those needs?” So, ’21 I think showed that there was a demand on both the creator side and a demand on the consumer side for these more artisanal approaches to making life work.
I think that’s pretty exciting. Now, what we also know, though, is that it’s not all sunshine and roses. There was a really interesting data that demonstrated that the creator economy was a highly, highly skewed economy. The phrase was the creator economy needs a middle class. There were a small number of incredibly successful streamers on Twitch or YouTubers, and a long, long tail of people for whom their creator economy earnings might have helped pay for a hobby, but certainly wouldn’t have paid for the mortgage.
My summary of the early experiments of 2021 and the growth of an ecosystem of tools from small companies and startups through to bigger firms like Stripe supporting creators, the idea that the social networks who have so heavily relied on creator content would start to directly pay them and pay them more cements the idea that creator economy will and can emerge. For historical reasons, I think one of the most interesting things written about this is Kevin Kelly’s essay, a 1,000 True Fans. If you haven’t read that essay, it’s now almost a couple of decades old, I think, I would recommend you have a read, of course, put it into its own appropriate historical context.
Now, we’ll get to the last trend. I hope as a loyal listener you will allow me to talk about my book. Of course, 2021 I had my first book published. It was a labor of love and also just a lot of labor. But I learned a lot. I learned a lot about the publishing industry. I learned a lot about how reviews and rankings work in eCommerce. I think they’re really relevant to helping us understand what happens to products, in particular cultural products, in this new digital age. So it was a moment of bringing together this old industry, the publishing industry with the ideas we discuss here. It has been a fascinating experience.
I’ll reflect on three things, right? The first is that writing is thinking. So it’s an amazing process to sit down and write a book. It’s a hard process and you’re distilling a lot of material, a lot of ideas. One of the things that I learned in writing the book was how far in history I would go back and how so many of the ideas that are important in framing the way we see the world today are things that have been thought about in the academy across many different disciplines back for many, many decades. Sometimes that gets lost in that constant stream of news flow in which we find ourselves.
I did find the publishing industry, frankly, to be very empathetic and very attuned to the process. Quite often, people look at it from the outside and say, “Well, that looks like a really old industry. It moves really, really slowly.” My experience of it is that the nature of the medium, the nature of the output is such that it’s hard to do it at tweet speed. So actually I think there’s what we would say in the entrepreneurial game a product market fit in the way that publishers work in order to help authors produce their books. So that has been an amazing experience.
I’ll talk about a couple of other things. Of course, the vast majority of books in the UK and in the U.S. are sold through Amazon. So Amazon ends up being the gatekeeper, the control point of the book. Yes, there are external taste makers that matter a lot. They matter emotionally and they also matter for driving sales. Those are the reviews in the top tier publications and amongst the bloggers and across Twitter. But at some point, reviews and ratings on Amazon matter a lot. I think as many authors will know, the first few days after your book being published, you spend a lot of time hitting reload on the Amazon rankings pages.
I want to reflect on a few of those learnings. The first is that there’s not a lot of filtering that goes on. My book actually got a one star review from someone a month before I had submitted the final edit to the publishers. Because essentially, anyone can go off and look at forthcoming books and say, “I’m going to give this book a one star review or a five star review.” It’s quite hard to go back into Goodreads and sort of pull those reviews out. The question is, does that really matter?
Well, it turns out that getting reviews on Amazon matters a great deal. There is some academic research that shows that the number of five star reviews that you get in the first few days of a product being available is strongly correlated with the final level of sales. So for any author, anyone producing a new product they’re selling through Amazon, you’ve got to figure out how do you get your five star reviews in upfront. That talks a little bit to the power that Amazon now has on determining what books are successful.
I noticed, for example, that when there are orders that go out on Amazon, you’ll see a spike in your ranking, but orders only count to your ranking for a couple of days. They are like a short-lived radioactive isotope with a short half-life. So the orders that you see on a Monday that might give you a spike don’t count very much towards your ranking by the Wednesday, certainly not by the Friday. So all of that creates a kind of a dynamic where one needs to think a little bit about how you are going to create visibility within Amazon.
Because the rankings aren’t actually exclusively an ego thing. They are also about discoverability. They’re also about getting the Amazon label that says best-seller in whatever topic. I was really lucky that I would get the best-seller in geopolitics, international economics, technology and engineering. When people see that orange label, they’re much more likely to go ahead and buy the book, which then reinforces the ranking. That is something that we’re familiar with, that these ranked, algorithmically-ranked lists that you see on Amazon or Twitter or elsewhere have a certain dynamic of self-reinforcement.
There’s a lot of nuances in there. I may one day write a blog post about all the other nuances I discovered. But that takes me to my third observation of what it was to write a book. I was really lucky that I have a group of people who read my essays every week and people who listen to this podcast and who are willing to support me and who regularly give me feedback and hold me to account. But it does mean that having an audience makes a really big difference when you produce a book.
Today, a large part of getting a book to a publisher, getting a book into Amazon, getting it to sell ends up being about those other things that as a creator, to hark back to the previous trend, you can bring to the party. Now, what I would say is I loved writing the book. I got a lot of intrinsic pleasure from doing so. It was incredibly hard to do and I was also really happy by the critical reception that it got. Now, it’s too late if you’re listening to this to get the book as a Christmas present for a loved one, but it’s still not too late to pick up a copy perhaps to help them see in 2022.
So that wraps up my five retrospective trends for 2021. I know that it’s been another strange year for most of us. So I want to wish all of you a very happy and successful 2022. Before I finally sign off, many, many thanks also to the brilliant Exponential View team, Fred Casella, Joseph Dana, Ilan Goodman, Mischa Frankl-Duval, Emily Judson, Sanjana Varghese and Chantal Smith who have helped bring you what I hope has been a brilliant year’s worth of insights and interests. Thank you.