Sometimes the centre of gravity in tech is very clear - everything is about PCs, or the web, or smartphones. But at other times, there are lots of things going on and none of them are The Thing, and all of them are full of questions. Of course, for some crypto people crypto is the only question and the only answer, but as we enter 2022 there are lots of areas where trillion dollar questions are wide open. These are the questions I wonder about at the moment - there are others.
Crypto is so big and potentially important, and yet so vague and so early, that we can’t even agree what to call it, and at times the noise of both irrational, religious hype and straw-man attacks can seem overwhelming. There is a set of ideas that could in principle be as central to tech as machine learning or open source, but after that, everything is a question.
The tech itself is in a period of massively increasing sophistication and complexity, as everyone builds on an open canvas and builds capability on a simple idea - early PCs or indeed the early consumer internet looked like this. But the more layers, abstractions, building blocks and primitives are created, the harder it is to know which will resolve into things normal people can use, and, paradoxically, the more likely gatekeepers become. We’re imagining the metaverse while arguing about how TCP/IP should work and whether this new ‘WWW’ thing is just a crippled Hypertext, but when can we tell our friends to ‘install Netscape’, and without giving them an op-sec lecture?
Meanwhile, what happens when the ideological fervour of decentralisation meets gravity - the dynamics and centralising forces of real products with real users? The web is radically decentralised, but centralised search and social sit on top. Open source, another crazy, religious idea, was going to transform tech, and it did! - but Office is still a huge business 25 years later. The iPhone is full of open source, and yet it isn’t open, and yet, with millions of apps and billions of downloads, in what senses and at what layers is it ‘closed’ and ‘open’? Most of my crypto questions are not ‘if?’ or ‘what?’ but ‘where?’
So - blockchains let us build distributed, trustless computers, applications and services, based on consensus, ownership and integrated incentives, but making that real involves a lot of plumbing questions and a lot of product questions that have still only just started.
We have pretty good VR devices now, but don’t know what to do with them beyond games (though we have ideas), and we can imagine AR glasses, that can put anything into the world as though it was really there, but we don’t have the optics for that today and don’t know when we will. Facebook has sold 10m of the Quest 2 at cost, but it’s far too early to know how well this is working or how big it will be.
We can certainly hope that more Moore’s Law, more engineering and more willpower can push VR into becoming the next universal device after smartphones, and we can hope that the optics for AR can work, and then we can speculate about what we might do with all this in a decade or so. Indeed, we can fill a whole whiteboard with those speculations and label it ‘metaverse’. But we don’t know. Is the future of software making everything into tangible 3D objects that merge with the world - or have we spent the last 30 years pushing for more abstraction and new kinds of tool?
The trouble is, it’s easy to make a cool concept video, but talking about this today in any kind of detail is like making detailed predictions about the mobile internet in 1999, or even 2005. It was very clear that there would be something, but we had no real idea what, and no-one thought that smartphones would replace PCs as the centre of consumer tech (or imagined that a has-been computer company from Cupertino would dominate the whole thing). AR and VR might be the next smartphone, but they might also be the next smart watch, drone, or games console - very cool but a narrow market.
Games have always been a big business, but they’ve always been a branch off the side of the tech industry. 250-350m people play console and PC games, which sounds like a lot, but is actually no bigger than Snapchat and nowhere close to the almost 5bn people using smartphones today.
Is that changing? Do games shift to more generally-appealing experiences, are smartphones unlocking a much broader audience, and do games companies look more like startups (and generate VC-type returns)? And might this merge with and reinforce VR and AR?
Perhaps. It’s a coherent thesis, with mobile game revenue now as big as PC and consoles combined and Roblox or Fortnite pointing to new creative and platform-ish possibilities. But we said that about WoW (and Second Life) too, and there’s no primary, driving mechanism (Moore’s Law, broadband) to suggest this change is inevitable.
Every big, complex, important industry has industry-specific regulation, from railways to food to aircraft to cars, and tech is becoming a regulated industry as well. But we don’t actually regulate ‘cars’ - that’s 10 or 20 different things, from congestion charges to traffic laws to seatbelts and emissions to drink driving. Regulating ‘tech’ is just as complex and full of trade-offs as regulating cars, or indeed or anything else in policy. Hence, there are lots of questions about what those rules should be, and what the trade-offs are. Very obviously, privacy rules conflict with competition rules (‘let me export data about my friends’).
Stepping back, though, it’s not clear how much regulation changes for how much of the tech industry. Much of what’s proposed is essentially a cost of doing business that lowers margins but also raises barriers to entry. Some narrow decisions might limit or close down entire business models (labour laws versus local delivery or Uber, say). But what more general structural changes might happen? Search, social and operating systems are natural monopolies, and you can’t really break up a natural monopoly, only regulate it. Even if Instagram was a separate company, it wouldn’t be any easier to compete with.
So, while there are obviously waves of tech regulation coming, it’s less obvious how much any of this would change our day-to-day experience of consumer technology (and most of it isn’t intended to anyway). Meanwhile, of course, the vast majority of the 3-5,000 actual tech companies founded each year in Silicon Valley alone don’t make platforms, social networks or adtech - and aren’t affected in any meaningful way by any of this.
We all agree privacy is important, and a big problem, but we don’t have any settled understanding of what that means or what a solution might look like, if we set aside bumper-sticker slogans like ‘surveillance capitalism!’ and ask what we want to be built. We don’t want to be ‘tracked’ but we quite like ‘relevance’ and ‘personalisation.’ Apple, the privacy company, built a private CSMA scanner that no-one else agreed was private. We want privacy, but don’t agree on what that means.
This matters in an immediate sense because advertising & marketing is a $1tr industry, a third of it is now online, and it’s the single most important lever for ecommerce on one hand and the growth of new brands and new competitors on the other. Most of this has been based on cookies, and privacy concerns mean that cookies are now going away, in the ‘Great Cookie Apocalypse’, but we don’t know what will replace that. The industry is trying to create ways to show ads that are both relevant and private, and there’s no barrier to that in principle. Advertisers almost never actually want or need any personal information - they just want to show diapers ads to people who have babies, not show them to people who don’t, and have some idea of which ads work better. Data isn’t oil - it’s sand. But building ‘private relevance’ is hard if we don’t agree what privacy means - tech people might argue that ‘first party’ is private and ‘on-device’ is private, but it’s not clear anyone else will go for that. Meanwhile, of course, if we don’t get an answer then by default Google, Apple, Facebook and Amazon do okay and everyone else is squeezed out - privacy conflicts with competition.
There’s a much more general issue here, though. There’s an old principle that a computer should never ask a question if it should be able to work out the answer, and the more that computers become invisible parts of our lives the more that they ‘should’ be able to work out. If I ask my Apple Glasses “I met someone from Disney last week, wearing a red shirt - what was his name?” what privacy issues arise? And when does the competition regulator force Apple to give competitors API access to that? If I use a new social network that uses a public blockchain as its source of record, what data is private and what does that mean? We live in an ever-expanding automated panopticon, and privacy questions that were always theoretical, and only worked at a small scale, now become practical at a global scale. We are trying to work out what that means, and how we change it.
The car industry is shifting to electric, and that changes a lot of what a car is - there’s an order of magnitude fewer moving parts, a very different supplier base, and much of the sophistication moves to software. We go from complex cars with simple software to simple cars with complex software.
Seen from tech, this looks a lot like the smartphone take-over of mobile phones, and there’s a lot of pattern recognition, right down to the dumb old industrial companies that think software is easy and they can just hire some developers. But it’s not yet entirely clear whether this really is disruption. An electric car is a better car but an iPhone is not a better Blackberry - it’s an entirely different thing that happens to be roughly the same size. So how much does electric really rewrite car manufacturing? Bulls think Tesla is a software company (and lots of other things), but bears think that no, it’s still a car company.
Autonomy is potentially much more profound and disruptive, and really does change what a car is - a car with no steering wheel is not really a car anymore. That raises as many questions as cars themselves did (it was much easier to predict mass car-ownership than to predict Walmart), and the tech itself remains full of questions. Can Tesla boot-strap its way through to something that works well enough? Will Waymo get there first going top-down? Are there winner-takes-all effects?
But more importantly, we don’t know when, how or where any of this will work. There was a period of euphoria a few years ago when AVs looked imminent, but it may now be that autonomy is like the old joke that AI is anything that doesn’t work yet. ‘Full’ autonomy may be as many decades away as ‘general AI’ (indeed it might require general AI!) but we’ll get all sorts of much more limited automation in the meantime.
A billion smartphone users leapfrog the second half of the 20th century and go direct to apps in everything from grocery shopping to consumer finance, generating a torrent of new ideas, business models and applications, most of which none of the rest of us can use, test or really understand first-hand, but all of which sound very interesting. And yet, this sits behind a firewall, in a totally different market structure, run by a deeply autocratic state that, in the last year, has kneecapped some of its biggest champions. How many questions does that pose? How many would you like?
I’m not a China analyst, and I try to leave most China tech discussions to those who can actually use the products and read the language, but three sets of questions seem important from outside. First, Chinese consumer tech has become a well of ideas for other people, especially other ecommerce and social media companies in other parts of the world, to go and copy. That probably isn’t going to change. Second, can the companies themselves go global? By default no (WeChat), but sometimes yes, with the right approach (TikTok), and does the current crackdown make Chinese companies look more aggressively for less regulated foreign opportunities, or focus on their home market? (Either way the ideas and models certainly go global.) What’s the next Shein, which may now be the biggest fast fashion retailer in the USA? And third, if China really does becomes uninvestable, where else does that capital go?
I’m not a China analyst, and I’m certainly not a macro analyst, but tech has been in a macro moment for a while. On one hand the pandemic (the ‘Covid Rotation’) has crystallised and accelerated the realisation of how big and central tech has become - where winning once meant a $100m or $1bn company, now it means $100bn or $1tr. On the other, the scale of company creation and opportunity is accelerated by a decade of low interest rates and a torrent of capital looking for tech returns - most of which are now in private markets, without public markets governance.
So, Tesla’s valuation is driven to $1tr by unprecedented options trading. A16Z now has over $20bn under management, and 300 people (10x a typical venture firm), but Tiger Global probably did 300 deals in 2021. Welcome to the $100m seed round. It’s different this time, of course, but then it’s always different, until the music stops and you find out who was good and who was lucky, and people have been wondering about that for a while now. Ironically, for a brief moment people thought that Covid might be the external shock that would stop the music and instead it accelerated things. But if history teaches us anything, it’s that something will happen. Check your email retention policies.
Most of the questions I’ve discussed so far are about the future of tech, but a central theme in the trends presentation I published last month was all the ways that industries much bigger than tech are being disrupted now by things that tech was excited about 10, 15 or 20 years ago. Across brands, consumer goods, advertising & marketing, TV, retail and ecommerce, all the old value chains break up, all the cards are thrown in the air, and no-one knows where they’ll land. Old gatekeepers and toll gates go away and new ones emerge. US pay TV subs are down by a third, Amazon’s GMV has probably passed Walmart - and yet Shopify is now 45% of the size of Amazon Marketplace.
These are big industries - global retail is over $20tr - and they’re being disrupted by tech ideas from a generation ago (“people might buy stuff online!”), but most of the questions are not tech questions. Netflix is enabled by tech, but all the questions are TV questions. The internet enables ecommerce, but the margin trade-offs of BOPIS are retail questions (and note that half of US and UK ecommerce comes from physical retailers). Does the transformation of marketing and distribution mean we’ll have far more brands (supermarkets drove a 10x increase in grocery SKUs), or far fewer? How many D2C companies will there be, and how big can they get? What happens as LVMH goes online? Bernard Arnault is richer than Jeff Bezos - which of them should I ask about Shein?
A decade ago, Uber and Airbnb represented ‘software eating the world’ - companies that used software to change the nature of a product. Airbnb doesn’t sell software to hotel companies, but changes what a hotel is. This is happening over and over again in every other industry. But Netflix or Shein might be a much more general trend - that tech changes the playing field but then the game is played by that industry.
Benedict Evans is a Venture Partner at Mosaic Ventures and previously a partner at a16z. You can read more from Benedict here, or subscribe to his newsletter.