Global | Politics & Geopolitics
This is an edited transcript of our podcast episode with Azeem Azhar published on the 22 October 2021. Azeem is the creator of Exponential View, a leading platform for in-depth tech analysis, and his podcast has featured guests including Yuval Noah Harari, Tony Blair and Kate Raworth. In this podcast we discuss the four technologies driving exponential age, how tech companies defy conventional understanding, understanding US/China/Europe tech rivalry, and much more. While we have tried to make the transcript as accurate as possible, if you do notice any errors, let me know by email.
This article is only available to Macro Hive subscribers. Sign-up to receive world-class macro analysis with a daily curated newsletter, podcast, original content from award-winning researchers, cross market strategy, equity insights, trade ideas, crypto flow frameworks, academic paper summaries, explanation and analysis of market-moving events, community investor chat room, and more.
This is an edited transcript of our podcast episode with Azeem Azhar published on the 22 October 2021. Azeem is the creator of Exponential View, a leading platform for in-depth tech analysis, and his podcast has featured guests including Yuval Noah Harari, Tony Blair and Kate Raworth. In this podcast we discuss the four technologies driving exponential age, how tech companies defy conventional understanding, understanding US/China/Europe tech rivalry, and much more. While we have tried to make the transcript as accurate as possible, if you do notice any errors, let me know by email.
Introduction
Bilal Hafeez (00:01):
Welcome to Macro Hive Conversations with Bilal Hafeez. We aim to bring you the best analysis to help you successfully invest in markets from crypto to equities to bonds. For our latest analysis, visit macrohive.com.
We’re getting great feedback on our new framework for trading Bitcoin and Ethereum, so if you’re interested in crypto, make sure to read our regular analysis on the Macro Hive site. There’s a lot going on elsewhere in markets too so we published a wide range of articles this week. We have a guest article from a former portfolio manager who looks at whether rising inflation expectations will crash credit markets or not. We have another piece on the explosive rise in uranium prices, and I take a look at cinema attendance to see whether we are in a normalised world after COVID.
And then on the academic side, we summarise a new paper on how to use machine learning on text, such as articles from Wall Street Journal, and how you can use that for investing purposes. You can get access to all of this as a member of Macro Hive. This includes all our research reports, webinars, transcripts, podcasts, and more, including our member Slack room where the Macro Hive team and members discuss markets all hours of the day. It’s refreshingly different from Twitter. When you sign up to become a member at macrohive.com, the first month is free and then it’s only the cost of a few weekly cappuccinos. It’s well worth it, and many call Macro Hive the hidden gem for investors. So once again, sign up at macrohive.com.
We also have a much more in-depth product for institutional and professional investors that features all of my views, trade ideas, trading models, and much more. Hit me up on Bloomberg or email me on bilal@macrohive.com to find out more.
Now, onto this episode’s guest, Azeem Azhar. I’ve long been a fan of Azeem. He’s the creator of Exponential View, a leading platform for in-depth technology analysis. His weekly newsletter is read by over 200,000 people from around the world, and his podcast show has featured guests including Yuval Noah Harari, Tony Blair, and Kate Raworth. A member of the World Economic Forum’s Global Futures Council, Azeem contributes to publications like the Financial Times, Wired, and the MIT Technology Review. His new book is called Exponential, How Accelerating Technology Is Leaving Us Behind and What to Do. Now, onto our conversation.
So welcome, Azeem, to the podcast I’ve been looking forward to speaking to you. I’ve just finished your excellent book, and great to meet the actual author behind it.
Azeem Azhar (02:21):
Bilal, thanks for having me on the show.
Who is Azeem Azhar
Bilal Hafeez (02:22):
Now before we talk about your book, I always like to get the origin story of my guests. So it’ll be good to know in your own words, what did you study at univerisity? Was it inevitable you would end up in the tech space and become such a prolific sort of public figure in this area, and going to investing and so on?
So what’s your story there?
Azeem Azhar (02:40):
I’m going to challenge your listeners by starting even before university, even before I was born. You weren’t expecting this, were you? So in 1971, Intel releases a 4004 processor. It’s the first packaged integrated circuit. It ends up being the precursor in many ways to the technology revolution in which we’re living. I was born the year after that. So I was born in a clock speed and cadence of the development of the personal computing industry. And right now I’m holding up on our video chat, which podcast listeners can’t see my first computer, the ZX81, which is now 40 years old as of next month. And I still have it.
Bilal Hafeez (03:23):
I’ll have to get a picture of that and include it in the show notes or something. That’s excellent.
Azeem Azhar (03:27):
We can take a screenshot if you like.
Bilal Hafeez (03:28):
Yeah, I think we have it. Yeah.
Azeem Azhar (03:30):
Great. So, but I was born in Zambia in the south end of Africa, where my dad was working as an expat under contract from the British government to help a newly independent Zambia establish some economic institutions. He was an economist and an accountant and Zambia needed to build those things up as it matured into independence.
So you have the confluence of these two ideas, the computer industry on the one hand and economic institutions on the other. And you’ve just read my book, and my book is in some sense about the computer industry and in some sense about economic institutions. So the origin story I think is quite interesting because perhaps I haven’t moved very far from where I started, which is back in Zambia in fact, in 1972. And that pattern really continued throughout my career. So at my school, I was doing sciences and maths and a lot of computing as well. I was very, very lucky to get access to, for example, one of the first ARM RISC processors when it was still called Acorn RISC Machines back in the late ’80s.
But at university, I studied the social sciences, a degree called politics, philosophy, and economics. And I was the only student studying this PPE degree who, to my knowledge, audited a C++ course with a load of physicists who were learning how to program.
So this interchange between technology and the social sciences has really been a large part of my life really from the beginning. And I remember being at university and being one of the few people with a computer and using my computer to build 3D graphics that I would get printed out and sell in one of the local shops in the university town where I was studying. And that set me up to in a way be in this industry. But I graduated in 1994. I had fallen down the hole that was the internet in 1992. It’s absolutely amazing place and I recommend everyone go back to the internet of 1992. It was wonderful. But there wasn’t really a tech industry in the UK. And there wasn’t really any kind of venture capital. There weren’t startups. In fact, even as an industry, it was rather young and immature in the US. In fact, I graduated about the time that Yahoo was founded, and Netscape, which was the first browser company, was called Mosaic Communications Corporation was founded, and I was in touch with their founders Jerry Yang and Mark Andreessen way back when. It was a very, very small world and it was very hard for the world of British graduate employers to locate me.
So my first job came after 53 job rejections from banks, consultancies, accounting firms, ad agencies, newspapers. You name it, I applied to them. I didn’t get offered a job. And eventually the Guardian asked me to come in for a few days, and I did. It was mostly to help them connect a couple of computers to the internet, and I just hung around. And eventually, that meant that I’d a role in one of their sections, which I ended up doing quite a lot for, and that gave me the footing. But throughout that short journalistic career of two years at the Guardian, a couple of years at The Economist, I was still building things. And I launched some of the Guardian’s first websites, and I built The Economist’s first internet products as well. And that’s really how I got started. There wasn’t really an entrepreneurial scene in London at all. There were a couple of internet service providers and that was about it. And so, someone with my interest didn’t have that many places to go, I don’t think.
Bilal Hafeez (07:23):
And then how did you go on to set up Exponential? There’s the website, newsletter, podcast. You just published a book called Exponential. So you seem to dabble in so many different areas from what my understanding. You invest as well. You are aware of all the latest stuff that’s going on in tech. I mean, how did all of that come about?
Azeem Azhar (07:41):
Well, I think a lot of that is just about staying in the game. And many a successful long-term investor understands that holding is quite often the best strategy, and you get compounding returns. And I get compounding returns to my knowledge and my network, and my network gets richer and more complex with that compounding. And my trajectory from journalism through to where I am has often been driven by what’s interested me and what was available. Because again, the UK didn’t have a very rich tech start-up scene for a long time. I started my first company just before the dot com craze went completely crazy in the late ’90s. And then I spent a lot of time working in companies like that, young start-ups, many of which didn’t succeed, some of which did. But I also had stints as one does in large corporates. You sort of do some start-up time and you sort of sit in a large corporate for a while. And with my last start-up, which was called PeerIndex, I’d spent seven years running it. It was a really interesting space. We did a lot of very challenging work around machine learning across large amounts of data, but I’d spent seven years staring at the same problem. So when the company was acquired at the end of 2014, I just started to write a newsletter. I just started to, and sent it to 20 friends. And then within three years, it was getting to about 30 or 40,000 people and it started to take over my agenda. And it took over my agenda because the newsletter was really about two contentions I realise now. One was, we need to understand these rapidly changing, exponential technologies on the one hand. And the second is, we need to look at them not just from a tech rah rah rah, jazz hands, rubber chicken standpoint, but also from the deeper issues of how they change business models or power in society or the nature of employment or how they help tackle climate change. And because I spent a lot of time reading and engaging and being a practitioner… So for a number of years, I was a venture partner in a seed stage fund in London and I was seeing a lot of smart founders. Because I had more of a practitioner attitude and a founder attitude, but I also had this wider commercial experience, I think the way I was able to tell the story in the newsletter resonated a lot. And so a lot of companies and leadership, company leadership used to seek me out to chat to them about where things might be going. And that just reinforces the platform of what is now a newsletter, and a community, and a podcast. The book has come out, and of course it helps my early stage investing activity as well.
Bilal Hafeez (10:25):
That’s great. I mean, two questions come to mind. One, just a very simple one, what’s PeerIndex? What did that start-up do? You obviously spent a lot of time in there.
Azeem Azhar (10:33):
Well, PeerIndex has been acquired now by a company called Brandwatch, and we started way back in the day when Twitter had less than a million users and Facebook hadn’t crossed a hundred million user mark, thinking people are now putting themselves out into the market. And in any market, markets benefit from understanding who your counterpart is, just to use your financial terminology. And this was just after the financial crisis where trust in the counterparties had diminished, and you could see an evidence proof of what happens when that trust is eliminated, which was there was no liquidity and things seized up.
So I looked at what was happening in the social web. And at the time, the social networks were not walled gardens. You could look at everything publicly and you got these unfiltered feeds. They were not put through the lens of an algorithmic tuning fork, which they now are. And we thought that if we allowed people on the internet who were publishing for the first time to understand in a sense their own profile, we could use that as a sort of trust metric to improve interactions across the whole system. And we didn’t really know where that trust metric would go. There are trust metrics in things like academia, citation scores, for example. There are trust metrics obviously in financial services, credit scores, if you trust those, you know, wink, wink, nudge, nudge. But the fundamental notion was can we help you as an internet user make your experience more useful by helping you figure out who you’re talking to and helping figure out how you are presented? And the way that we did that was we looked at your public data, and you could opt into this, as you would share messages, and we would figure out the extent to which you had expertise in the topic of hedge funds or Chelsea Football Club or enterprise software. And we would figure out your influence network. We would figure out whether you were likely to be someone who could move the flow, the pattern of a story through your interactions with it.
And the first incarnation of this was in 2010, 2009 perhaps, was an algorithmic curation engine that would find the best stories that were being shared by the most trustworthy people on specific subjects. And what we discovered was that that was really exciting and lovely, but there wasn’t really any money in it. And we figured out how to scale the technology to look at larger numbers of people, again, in a as best we could a permissioned basis because that’s what we thought about. But in reality, nobody really knew where the boundaries lay, and we came up against many changes in the social networks, specifically that they became walled gardens between about 2012 and 2014.
Bilal Hafeez (13:27):
And then the other question was, why didn’t you move to Silicon Valley? Why didn’t you move to California? It sounds like there was an impediment being here in the UK, in London.
Azeem Azhar (13:36):
I had a few opportunities, and the first time I founded a company, I’d actually was holding an offer to move out to California. And an investor in London actually said to me, “Let’s just set up a company here. London’s going to be getting frantic and we’ll seed you.” And the seed was a seven-figure sum. So it was a big seed. And then what happens is I think that life slightly gets in the way. Was it that my… There was some health issue with a family member or had I just got married, and you wanted to sort of lay some foundations first, and then you end up founding the company. And what I found with PeerIndex was that there was a lot of interest in funding it in the Valley, but back in 2009, Silicon Valley VCs didn’t really invest in early stage European country companies. Moritz, Sequoia had done one deal in a company called Klarna, which has done quite well since then, but it was very, very rare. And you had to go and see them. Union Square Ventures in New York did a handful of deals in London in that period, and then didn’t for years after. And what would happen is I would go back to the Bay Area quite regularly. I mean, as often as once every three weeks, which is pretty punishing, frankly, from Cricklewood to Palo Alto, and there would be a lot of interest. But what they weren’t willing to do back then, except in these exceptional circumstances like Klarna, was actually issue the term sheet and do the deal and trust you to do the Delaware flip while you were still in London. Right? They said, “Come to the Bay Area.” And of course, with three kids that was going to be a really tall ask for a speculative move, and venture catalyst are, let’s face it, often trend followers or excited by the latest fad, and who knows if you’re still sexy when you land on the plane.
What is the ‘Exponential Age’
Bilal Hafeez (15:25):
Absolutely. And so that’s that useful context. So going on to the book now and just your work from recent years, Exponential. One thing that would be useful would be just to define some terms here. So when you say we’re living in an exponential age, what do you mean by that? Is there something specific you have in mind?
Azeem Azhar (15:41):
Yes, I have something very, very particular. And perhaps I’ll provide some context to that, which is that one of the main axioms in the book is… It’s an unsurprising one, but we should dig into it, is that that technologies will create new potentials that will shape society and industries and business. And more than that, that industries, business, and society will shape the direction of technologies. And so you see transitions in the way we live that are enabled, catalysed, and sometimes driven by technologies. And we just have to think about what mass production enabled by electricity and producing the car did for the idea of the suburb and the weekend, and full-time employment, and worker rights, and a bunch of other things. And so the notion of connecting, almost paradigmatic shift in how we organise our societies to technologies is a really, really important one. And the 20th century was largely governed by things that emerged around the technologies of the telephone and the internal combustion engine and electricity. And life had been pretty similar for quite a while despite all the change that we saw between say the 1920s and 2005. But if you look in 2009 at the world’s largest companies, the bulk of them were car companies, oil companies, electricity companies. In other words, companies of the technologies of the 1880s, generational winners. And by 2013, Exxon Mobil drops out of the top 10 American companies. And today all the large companies are based on different technology platforms. There has been some kind of transition.
So if you take the idea that technologies catalyse and shape and are shaped by societies and the norms and the frameworks that we live in, and you accept that, and then I say, “Well, we’ve now had, we’ve got evidence there’s another technology transition happening.” Well, that is a transition to a new age from that sort of 20th century Industrial Age to this new age that I call the Exponential Age.
And the reason I call it the Exponential age, and I say we enter at some point between 2013 and 2016, because it’s not like landing on the moon when we know exactly the date it happened. It sort of happens over a sort of blurry period of time. Is that the technologies that are driving this transition are not three general purpose technologies. There are really four general purpose technologies, and those technologies are improving at exponential rates, which means on a price performance basis, they improve by at least 10% on a compounding basis every year for many decades. That is you get at least 10% more in the second year than you did the previous year for the same dollar, 21% more in their third year, 32-and-a-bit percent more because it’s compounding in the fourth year, and so on. And that is the heartland idea of why we’re into the Exponential Age, what it means, and why it’s different to the age, why it could be different to the age that came before.
The four technologies driving exponential age
Bilal Hafeez (18:45):
And you mentioned general purpose technologies. You said four, I think. What are they?
Azeem Azhar (18:51):
So general purpose technology that we are most familiar with in all of this is computing. It is sort of silicon chips and beyond, and the set of derivative technologies that live above it. And then there’s a whole set of affiliated ones like storage technologies and the bandwidth technologies.
The second domain is the domain of, roughly speaking, of biology. It’s the fact that we have new science around biology from genomics to cell regulation and protein engineering where we can connect those to engineering methodologies, and we can now look at things like reading, understanding, and rewriting the gene or proteins in a kind of mechanistic way. And those technologies are improving dramatically quickly.
The third area is at a different scale, so it’s not in our bodies and tiny and squishy. It’s out there big and kind of clunky. And it’s the technologies of renewable energy and other energies in general. So solar and wind and battery storage and the associated things.
And the fourth domain is the domain of manufacturing where the technology of 3D printing or additive manufacturing, which is very niche right now, is on a dramatic price performance decline that is compounding as well.
And the thing about a general purpose technology is that it’s applicable in many parts of an industry in many different functions. It’s also applicable across many industries. And that generalisability means that there is a much more significant impact that it has in first, the economics of doing something, and then how that gets reflected in the societal mechanisms that are impacted by that economics. And it’s the generalisability I think that is really important to think about. Some of the four platforms I talk about are more generalisable than others. So computing is probably the most generalisable and manufacturing in a way less so, but it is still a broad base, broadly applicable technology.
Bilal Hafeez (20:58):
And you didn’t mention artificial intelligence and machine learning. Is that because that comes under computing, so that springs out of computing?
Azeem Azhar (21:04):
I mean, I thought quite hard about how to categorise these, and even defining a technology is quite a tricky, fraught process. And then defining the boundaries of a technology are even more fraught. And so when I look to AI and machine learning or AI and ML, I put them under the category of computing because they’re ultimately higher technologies that sit on the computing substrate.
What’s different between today and previous tech transitions
Bilal Hafeez (21:33):
That makes sense. Now, one of the things I really like about your work is that you don’t just focus exclusively on the technology side. You also think about the societal aspect, the bigger picture. And in your book, you talk about exponential gap. And in some ways, the way, I think you mentioned this in the book, technology moves exponentially and institutions move in a linear way, or there’s some way you characterised it like that, and you end up with this gap. So can you talk a bit more about this gap? And presumably the gap should narrow eventually, and is that what’s going to happen?
Azeem Azhar (22:04):
Should it? What’s interesting about the period of time we’re in now is that the rate of change is far faster than we have experienced before. And I go into that in some detail in the book. And so while technologies that diffuse slowly can still cause incredible issues, so think about the printing press which took hundreds of years to spread across Europe, and yet led to The Reformation and The Enlightenment, and many, many, many wars right across the continent, and yet the diffusion rate was often longer than lifetimes. And we have technologies now that are diffusing far faster than that. And we psychologically struggle with that pace of change because there are no real exponential processes in the wild that we normally have to contend with. We’re not selected for figuring, working with that on an evolutionary basis, but organisationally we don’t… We’ve designed things that they naturally can’t be incredibly responsive because if they were responsive, they wouldn’t be institutions, right? They’d be fads or ephemera. But they question is, what is the degree of responsiveness that they need? And they’ve often been designed for much, much more slow moving periods of history. And so what we end up seeing is a dramatic potential of the exponentiality. And I just say potentials. I’m not saying good potentials or bad potentials, but potentials. And the institutions that we rely on in our everyday, from very, very informal institutions, like when you’re playing Monopoly and you get a fine, you put it into the free parking zone. And when you land on free parking, you collect it as booty. That’s not in the rule book. That’s just an institution that we all adhere to, and everyone, if you don’t play that, that’s a bit weird, through to the convention of which side of the road you drive on, which is incredibly useful one to have, through to employment law, and companies law, and in your industry, stress tests that are mandated on large banks. Institutions are there to make life manageable and easier and to take out the headache of having to judge every single situation de novo.
And so they’re very, very helpful, but if the fundamental basis on which the technologies operate and the types of knock-on effects they have largely starting through the companies that maximise them, but you also see it within states, then maybe we need to think about how those institutions need to change. And we’ve seen this pattern before, right? We saw again, thinking about your industry, rules change when the telegraph arrived at the end of the 18th century and you could easily arbitrage between the Boston and New York stock exchanges, right? And rules start to change. But the question is, when you have this widespread and accelerated pace of technological change, do you need your institutions to change differently? And can you find some way of narrowing the exponential gap? Or in fact, should we just live with it? And I think that that’s what I discuss in the bulk of the book.
Bilal Hafeez (25:17):
I mean, I guess one of the challenges, as you said, is the diffusion rate’s so fast now, and you have so many of these generalised technologies, are there precedents that we can fall back on and say, “Look, we can look at the late 1800s or something, or the onset of computing in the ’60s or something?” I mean, are there precedents we can learn from, or are we going to just have to make it up as we go along now?
Azeem Azhar (25:37):
There are lots of precedents, but I think there are also precedents that are a little bit dangerous. So we have to acknowledge what is the same and what is different about the environment in which we live today. And then once we’ve recognised I think not symptoms, but underlying causes and underlying processes, I think we can go back and find principles and say, “How do we apply those principles where they’re relevant to the modern day?” But there are just some examples. I just don’t think we have without really, really stretching the ecumenical or scriptural reading of our laws, find ways of applying them. I mean, a good example would be about the data that you produce in the course of your everyday interactions with the economy, and even the question of whose data is it, and should you own it? And is it something that can even be owned because ownership normally ascribes itself to rivalrous and exclusionary goods and products like physical books or water bottles. And so we lack the language in some cases to even express what the issue is. And so that’s where I think, what I try to do in the book is start to explain some of the underlying processes of the technology and how they play out as they interact in the system, and then identify where this creates this strain, this exponential gap. And then we can start to look back historically and look forward and say, “Well, what can we learn from in terms of the adjustments we need to make?”
How tech companies defy conventional understanding
Bilal Hafeez (27:11):
If we narrow the scope to say a company, again, a company’s still a big concept, but move away from government and country to company. You talk about the unlimited company, which you are hinting out there is there’s something different to what we’ve had before. Can you talk a bit more about what you mean by the unlimited company and how that fits into this Exponential Age?
Azeem Azhar (27:32):
I can, and I will do it by means of a story, in fact. So back in about five years ago, so summer of 2016, I was in a very, very posh London hotel having a very posh lunch with about 20 investors. And they were people who were personally wealthy, family office style investors. The lunch was very nice, thank you for asking. And I sat down and I talked about these very large companies. And it was at the time, Apple had a market cap in the order of $350, $400 billion. And I was explaining how there was a sort of… These companies were going to continue to expand. And I said, “Well, where do you think it’ll be in five years?” And I should caveat, I didn’t trade off this information or anything like that. I mean, I didn’t make millions of dollars on Apple, more for me. And people couldn’t conceive of it going above about a $500 billion market cap in that room. And I was pushing them, not because I thought they were right. I just wanted to see where their imagination would take them. And of course, it’s up at 2.4 trillion, I’m just looking at my screen now, today. And so who could say, look at it today and say, “Well, it couldn’t go up another sixfold in the next six years?” I mean, that’s a potentially a brave decision. You could say reversion to the mean, but which mean are we talking about?
So somehow companies have managed to defy the laws of gravity that would keep them small. And those laws of gravity were that as companies became bigger and more complex, managing them is like sort sorting through a mess of tangled headphones wires in your drawer, and it becomes harder and harder, and the company just slows down and atrophies. Well, that doesn’t seem to have happened to Microsoft or Google or any of these other Exponential Age network companies. And even some of the things that jump out of economic theory like Ronald Coase’s theory of the firm, or the idea of increasing marginal costs when you’re trying to get your inputs, don’t seem to apply to these companies because as they get bigger, they get more profitable, they grow into more markets, they generate more revenue per employee. They keep their margins high, despite all of that. And so I call that the unlimited company, and I try to explain why that comes about and explore whether it’s a problem.
Bilal Hafeez (29:52):
And is it possible for any company to potentially become an unlimited company or is this only applies to companies that somehow capture the four generalised technologies that you talked about? For example, if you are just a manufacturer of widgets, could you become this type of company or not?
Azeem Azhar (30:10):
There were many steam and railway magnates in the 1870s and 1880s, none of whom became car, telephone, or electricity magnates in the 1920s and 1930s when there was a technology transition. And so historically, the track record has not been great. The difference this time, of course, is that the managers of today have the benefits of history. They have the benefits of the innovators’ dilemma, nearly 30 years old as a text. They have the benefits of seeing what happened to the newspaper industry when it didn’t respond to the digitisation and the internet, so that they have these aspects on their side. They also have the advantage of, whether it is an advantage, as sort of a large cadre of business and technology consultants willing to help them make the transition.
But despite all of that, I think it’s really, really hard for companies to make the change for reasons that are really, people go get into in great, great detail, whether it’s management capability, whether it’s being able to retain the talent, whether it is being able to manage channel conflict, and many other things. Now, which ones succeed and which ones don’t, I think is going to be, depends sector to sector and company to company. So if you go into the old stage sector of farm machinery, John Deere has done quite well. And John Deere has been able to transition itself to largely a data driven company. If you look at John Deere, it’s actually called Deere & Company. In 2012, the share price hovered around $75, and today it’s at $352. So that’s not bad. That’s a 5X in five years.
Bilal Hafeez (31:56):
And this is farm, a company that historically specialise in farming equipment?
Azeem Azhar (31:59):
Combine harvesters and tractors.
Bilal Hafeez (32:01):
It’s not the most glamorous sector, but yet it’s been able to reinvent itself.
Azeem Azhar (32:06):
And what they did was they turned themselves into a technology platform. So essentially, there is a Deere app store that third-party developers who are making weather data or precision agriculture or fertilisation algorithms or whatever it happens to be, have to plug into and pay a rent. And farmers get access to these advanced features as they might through an iPhone. And you’ve seen the response in their share price. I mean, it’s… And in fact, it’s absolutely vertical in the last year. Coincidentally, since I cut that section out of my book where I wonder if there’s any connection to that.
Azeem Azhar (32:44):
So look, it is possible. And I think if you look at the car manufacturers, because they are just so big and now so committed to the transition, and because they don’t face that much incumbent competition, you can imagine the car manufacturers being able to transition towards a more electric vehicle, autonomous platform based style business model. But I think that when you look at something like that, they’ll have competition from other arenas, which is do we even use cars in the way that we used to versus other things?
Azeem Azhar (33:18):
So I think it’s a very, very difficult question that if I had to be forced to answer and say, “Will they make the change?” The answer would be, “No, they wouldn’t.” But that would only be if you forced me into it. But if I was allowed caveats, I would give you lots of caveats and say, “And this is how they can get through it,” and so on. But I think it’s very, very tough.
Why productivity has been low
Bilal Hafeez (33:39):
And one of the classic critiques of the Exponential Age or technology in general, because it depends what your starting point is for this technology shift, but Peter Thiel talked about why don’t we have flying cars? Instead, we’ve ended up with 140 characters. Or as an economist, we look at productivity as being incredibly low since the ’70s. We’ve had this computing age you could say maybe that started in the ’60s onwards. Why don’t we have like this Jetson-style world where we have flying cars and all of these sorts of things?
Azeem Azhar (34:10):
Well, I mean, the Jetson-style world is imagined without being able to understand the complexity about how society pushes and pulls and transmits needs and requirements. What we do have is an enormous growth in Reiki healers and yoga teachers and peer reviewed literature around mindfulness, which never appears in the Jetsons. So we are constructing futures, but it often they’re sort of ill thought out because they’re thought out point in time, just with a sort of technology determinism. But on the question of productivity… And thank you for asking me that question, because with the failure of every economist to answer that question, I’m bound to be the guy who knows the answer. But let me just sort of hazard a couple of thoughts. The first is that I think there is a J curve in these technologies’ utility, which is it just takes time for managers and companies to make the right investments and get the skills. And we saw that happening with typing and typewriters in the early start of the 20th century. It just took a while for managers to figure out how to change processes and typists to be trained, and then the standardisation on the QWERTY keyboard before you started to see the benefits of typing. Or likewise, the arrival of electricity, which forced a complete re-tooling. We would’ve called it a transformation of companies because power came through central drive shafts and drove massive pieces of equipment. Whereas with electricity, you could packetise the energy and put it into hand tools and lighting around individual work stations. But then you had to get all those tools and you had to train the workers, and those things take time to catch up. So there could be a delay. I think that’s one rationale.
I think the other rationale is what are we measuring? Because if we are just measuring GDP, there’s a bunch of stuff that is not included. And you can take both a free market approach to this and say, “Well, we’re not including the benefits of digital goods that are free at the point of consumption,” and that’s a problem with our measurement of GDP. We can say that there are things that are just not reflected in market prices or wrongly reflected. For example, the clean-up of pollution is represented in GDP when it shouldn’t be. And so those things will never show up in the productivity statistics as a consequence. It might just be that in all those things that we’re not measuring that free of consumption, it’s things like TikTok, which are anti-productivity, because we spend our lives scrolling rather than being, doing other things. And we’ve expanded leisure time into the workday by holding our phones.
Climate change and productivity
Azeem Azhar (36:36):
But I think there’s also potentially another reason for this, which is what productivity are we measuring? Dollars of GDP divided by number of people and over time, roughly. Okay? And let’s just say that dollars of GDP is the right thing to measure. But if we look at dollars of GDP over CO2 emitted, between 1985 and 2016, the UK got 2.5 times more CO2 productive. In other words, GDP went up about 75% and CO2 emissions declined by about 35% in that period. And so there is a productivity measure that, I’m not an economist, that exists out there that is perhaps hidden in the impact of what we’re doing. So I think there’s something to be said about looking at all of those, but in a way, part of my argument about the Exponential Age sits quite nicely with both the right and left wing critiques of GDP as a single indicator. And then the fundamental premise of your question, which is this, why has this flawed indicator divided by the number of people not moved? And it’s like, well, maybe let’s go back to the indicator.
Understanding US/China/Europe tech rivalry
Bilal Hafeez (37:46):
And you mentioned TikTok, obviously a Chinese company. It seems like we’ve got splinternets. We’ve got this different tech empires. How do you think about the different architectures we have? You have China with its own tech ecosystem. We have the US, and then we have Europe, which is somewhere in the middle that seems to be the sort of regulator in chief. That seems to be its big thing. How do you see things from a geopolitical perspective?
Azeem Azhar (38:11):
Well, I think to such a fascinating subject and there are two slightly orthogonal frames that we need to take. So let’s look at the first frame, which is the one that you provided. And the idea of the splinternet and the localisation of technologies, specifically the localisation of the governance and control of those technologies. Now I would suggest that we are recording this in September, and in the last two months, China has probably taken the lead as the regulator in chief, right, across the PBOC and the CAC and other government bodies who have really come in and said the growing power of these Exponential Age companies is now deleterious to the functioning of the state and we are going to do something about it. And I think in the slightly anodyne way that I’ve described what the Chinese did, that sounds quite understandable in a funny sort of way, right? We might not like the patterns of democratic accountability within China. We might not like the way the decisions are taken. We might not even like the specific decisions, but the framing of the job of any state. I mean, this goes back to Thomas Hobbs and Leviathan, and is to essentially maintain itself for the benefit of its, protection of its citizens. If it perceives a threat, it needs to tackle that threat. So many, many nuances around China that we can dig into in a second conversation. But that’s what they’ve done, and they are now showing other models out there. But of course you’ve got the European Union, which has its own strong ideas. Very, very pro rights, human rights, pro citizen. But China’s privacy policies and personal data privacy policies relative not to the state, but to private actors, are pretty stringent as well. And you have India saying, “You need to have your own… If you’re going to serve Indians, you’ve got to have the data centres in India where they’re within reach of a judicial warrant.” You’ve got the Ethiopians saying, “We want to run our own social networks for Ethiopians within Ethiopia.” And then of course, you’ve got the US, which I think is trying to figure out the hard realities of how you tackle regulation in the Exponential Age when you’ve been bound on the one hand by amendments to the US Constitution that are a couple hundred years old, and on the other hand, by 50 years of political dogma that says government shouldn’t regulate anything. And I think that creates a bit of a tension for them. So that’s how… And then of course you’ve got the Russias and the people who are mischievous actors in of all this.
I think it’s a really, really fascinating period and it’s one of the critical exponential gaps because we got here because the technologies of the semiconductor industry and the internet walked hand in hand, where now I’m introducing a third person into this marriage. Walked hand in hand down the wedding aisle with both the technologies and the policies of globalisation. And they were closely connected, and governance was essentially almost by accident, less than by design, aligned to US technology commercial interests up until the mid ’90s and late ’90s. And so it’s not that we had global rules prior to the splinternet emerging. It’s that our rules had largely fallen out of, certainly on the internet, a vision that happened to be aligned with American interest. But in reality, those initial architects of the internet were not thinking in those terms. They were not thinking in national interest terms. They were thinking in protocol terms, in technology scalability terms, but it didn’t have passing interest to those particular questions. It just happened to line up with the political dogma of the time.
So that is my sort of abbreviated answer to the first framing of the question that you asked.
Importance of big tech in dealing with cyber risk
Bilal Hafeez (42:16):
And then you said there’s another way to frame it as well.
Azeem Azhar (42:18):
So the other way to look at it is that, I mean, there’s never been a moment where the nation-state is too small for certain types of problems it has to solve. It’s also too big for many others. And that moment is when you look at things like climate change or pandemics or cyber risk. And if you think about cyber risk as something that’s emerged directly out of the Exponential Age, but when a cyber threat manifests itself, no military can stop it. Not the Danish military, not the British military, not the American military, not even Space Force can stop it. You have to rely on Akamai or Cloudflare or Google or Microsoft or Apple because those core infrastructures, the domain name system of the internet, the content caching, the operating system on the is it public, is it private interface of our smartphones that we all carry, are all run and maintained by private companies that while notionally are largely in America, end up taking actions globally. And they have the capabilities and they have the reach.
So when the Ukrainian bank link host was attacked by a threat group affiliated with the Russian security services, the NotPetya virus spread around the globe and brought companies like Mondelez and Merck to its knees, we’re dependent on the tech companies, not on national militaries to stop that threat, to seal that threat off, and then to vaccinate and eliminate that threat. And so there is a new geopolitical actor on the stage, and we have to have, be delighted that they are willing to act like this and delighted, in particular companies like Microsoft step up and try to create global compacts around this.
But equally, they’re not democratically accountable and their ultimate responsibility is first and foremost to their shareholders, however that’s articulated. And so you end up in this interesting, challenging position, which is who governs those decisions, where is their seat at the table, and what are the rules by which this escalation happens, and who gets to pull the strings, and which nations get the hotline to the headquarters to stop that happening? And increasingly what we’ve seen, I mean, this started very early on, of course, with the Germans saying in the late ’90s, early 2000s to eBay and other auction sites, “You can’t host Nazi memorabilia on your site.” We then get to the point with the Chinese and the Great Firewall and more and more takedowns. And then eventually very recently, the Russians saying that the Navalny app should not be present on the Apple or Google app stores. So it’s a really complicated problem that intersects much needed, good intent and great capability from technology platforms spilling over into things that we might feel uncomfortable with, and then the fundamental question about how do you make power accountable?
Best investment advice Azeem has received
Bilal Hafeez (45:14):
We could talk on and on and on, and I urge listeners to read the book to get more of your insights. But I just want to round off with a few more personal questions. One is, this is something I ask all my guests, what’s the best piece of investment advice you personally have received?
Azeem Azhar (45:33):
You don’t have to do it. So a pass is as good as getting into it. And so don’t ever feel that pressure of I’ve got to chase this and I’ve got to get into it now. And I can’t stick to that, but it does sit with me that sometimes you should just let it be and either hold cash for a bit longer or leave your position where it is. So itchy trigger fingers, not very welcome.
Productivity hack
Bilal Hafeez (45:52):
That’s a good piece of advice. The other one was, I do notice that in your newsletter you seem to digest huge amounts of information. Do you have a system for that, filtering system or some kind of productivity hacks to make yourself so productive?
Azeem Azhar (46:04):
I’ve been at the mercy of the software start-ups in this area. So I’ve had to continually move as they’ve gone out of business or been acquired and been shut down. So when you ask me this question today, we don’t have a great system because the companies in question have been sort of swallowed up and we don’t really have the tools that we used to have. The thing that does help though, is I’ve been looking at this for a long time. And I know that it takes me about 18 months to start to feel a sense of confidence in a subject. It’s at the end of the 18 months that I have a sense of what I don’t know and all the questions I need to ask, whereas I have none of that knowledge when I start. And then I can start to compound my knowledge after that.
But the second thing, and this is the absolute beauty of the internet is that you can stand on the shoulders of many giants. So I follow a lot of interesting people on Twitter and I try to follow discerning experts. And initially, I don’t know who the discerning experts are. I just know the noisy person in CAR T therapeutics or in space satellites. And then over time, I start to figure out who is discerning, and I start to follow them and I drop the other people. And that gives me this wide network of smart people. And effectively Twitter and a couple of other things allow me to sidle up and eavesdrop on the more interesting and cleverer conversations that are going on.
Books that influenced Azeem
Technological Revolutions and Financial Capital (Perez) and Letters to a Young Poet (Rilke)
Bilal Hafeez (47:25):
And finally, what books have really influenced you in your work?
Azeem Azhar (47:29):
In this particular piece of work, I read a hell of a lot of books, but I think two were really relevant. One is Carlota Perez’s Technological Revolutions and Financial Capital, which she wrote in about 2000 and is an absolutely brilliant analysis of how technologies emerge into an economy and how they create cycles of boom and bust, and then through a process of eruption take us to golden ages. And I think that there are some extensions we need to do of it given the nature of exponential technologies and also given the nature of the exogenous threat of climate change and how the economy will respond to that. And then the second book is something that I think about quite a lot. It’s a book by a poet called Rainer Maria Rilke called Letters to a Young Poet written about a hundred years ago. And what’s interesting about this is that he’s, Rilke’s talking to this poet about process, about the art of writing, about how he writes, and why he writes, and how he forms the words, and the method and the process he goes through. And I think that a lot of writing a book I have learnt now, which I didn’t know two years ago, is about the process and you choose the form you take, and that instructs the way in which you create your narrative and construct your arguments.
Keep up with Azeem
Bilal Hafeez (48:47):
That’s great. Both books sound really, really good. So with that, listeners who want to follow your work, what’s the best way for them to do that?
Azeem Azhar (48:53):
You can find me on Twitter, which is @azeem. In the UK, that’s A-Z-E-E-M, and for American listeners, A-Z-E-E-M. And you should also sign up to my newsletter, which you can find at exponentialview.co or just type in Exponential View into Google. And of course, I’d love you to buy the book. It’s called the Exponential Age in America and Canada, and Exponential in the UK and the rest of the world. And all the book sellers have it. You just go to their website and pop in that title or my name or a combination of both and you should find it.
Bilal Hafeez (49:23):
Great. I’ll include links to everything you just mentioned there. So with that, big thanks. I learned a lot in this conversation and it’s been great having you on.
Azeem Azhar (49:30):
My pleasure. Thanks so much.
Bilal Hafeez (49:31):
Thanks for listening to the episode. Please subscribe to the podcast show on Apple, Spotify, or wherever you listen to podcasts. Leave a five star rating and a nice comment and let other people know about the show. We’ll be super grateful. Sign up to become a member of Macro Hive and macrohive.com. We’ll be back soon, so tune in then.