🔮 Grading my predictions for 2022
Quantum computing, fusion, Web3, energy & climate tech, AI++
Every year, I offer my predictions for the year ahead — and at the end of every year, I review how the past predictions played out. I asked Nathan Warren, a new team member at Exponential View, to help with the assessment. Our discussion follows below.
Before we get into it, you can read my 2022 predictions here. TL;DR the main themes were the following:
Two futuristic technologies — quantum computing and nuclear fusion — will become more present in our consciousness during 2022,
We’ll reach another turning point in Web3 adoption and innovation,
The debate will shift in the wake of two 2021 occurrences: 1) the comparative fragility in European energy systems, and 2) a rethink about nuclear fission power or the role of fossil fuels, particularly natural gas, in the clean transition (note, this was before Putin’s war on Ukraine),
Big tech will remain big.
Fusion power: 1 + 1 = 3
Nathan: “At the start of the year you predicted we would see new technical milestones and increased capital flowing into Nuclear Fusion. How have you viewed the progress in the field over the year?”
Azeem: “I’ve been saved by the National Ignition Facility (NIF) because they recently announced an important scientific milestone, recording a net gain in energy generated from nuclear fusion.
But the efficiency rate is really low. It’s like they burnt a million trees in order to light one candle. It is a small step, but it's important to reach “scientific gain.”
There were other positives, such as the scientific breakthrough from the Joint European Torus earlier in the year. There was also some really meaningful commercial progress in terms of partnerships that were announced by a couple of fusion companies. Overall I feel like I got that prediction roughly right and that the direction of travel continues.
What I would say in general about this particular prediction, which was really about deep tech, is that the main change we saw was in the capital markets. The fact that there was a stock market collapse caused investors to rush to safe asset havens leaving behind growth assets that are very uncertain and take a long time to mature. This really depressed venture funding in general in 2022 which has had an effect on these deep tech firms. Fortunately for the fusion and quantum companies, most of them filled their boots over the last few years, so they’re quite well-positioned.
However, I think there’s a caveat, which is that one of the scientists in the NIF, which is a public sector facility, said this may mean we could get commercial fusion power within a few decades, which takes us back to the idea of “fusion always being 30 years away”.
On the other hand, the commercial fusion people are saying we’re making really good progress and are, in some cases, focused on delivering commercially relevant fusion sooner. And so there is this sort of bifurcation that has previously existed and is starting to become more clear on the timeline towards commercial fusion power.
Quantum computing: maybe it is, maybe it isn’t
Nathan: “Your second prediction was that quantum computing would rise in our consciousness. Do you think this happened over 2022?”
Azeem: “I think quantum firms have taken steps along their roadmaps by improving the performance metrics for the computers. There was also a growing idea that these tools could be used in industry at the noisy intermediate scale quantum computing level. The prospect of quantum computing also arrived in fields like optimisation and molecular modelling to enable new innovations in terms of simulation.
Lastly, there was a massive, beautifully written article in The New Yorker this week on quantum computing. The New Yorker is an arts and literary magazine in the US which has some of the best writing for any mainstream publication. This shows that quantum is rising in our public awareness. So I think that prediction worked out very well.”
Crypto & criminal have more in common than their first two letters
Nathan: “Your third prediction was that crypto would see a turning point in adoption and innovation in 2022. This obviously didn’t occur with the spectacular collapses of FTX and Terra. In a recent essay, you changed your view on crypto, touting that it hadn’t presented any great use cases so far. What has changed in your thinking over the last 12 months?”
Azeem: “Well, there’s definitely been a step change in adoption in the sense that loads of people have run away from crypto, and we’ve seen some incredibly innovative abuses of power with some of the key players. What was right behind my thinking was that we were seeing more consumer wallets being created. So more consumers were getting involved in crypto. When I was talking to asset allocators (e.g. pension funds) their interest in crypto and the desire to invest in crypto was definitely increasing. This enthusiasm has dampened with the flight from growth and risk assets to more stable assets like utilities, agriculture and healthcare.
The second thing was the spectacular amount of fraud in the ecosystem, often in ways that you wouldn’t have predicted. So for example, one former crypto billionaire was found out to have been funding one of the major independent crypto news outlets. The whole thing reminded me of Ouroboros, the mythical snake eating its tail.”
“Did you see any promising innovations over the year for crypto?”
“I think my hope is that there is a collective sense of “Oh, my God, what have we done!?” in the crypto community and that they’ll try to go back to the fundamentals: parsimoniously build real products that do useful things for customers. Some promising innovations in crypto may find use elsewhere. Consider zero-knowledge proofs, which is a way of securing communications between two parties without having to disclose what each party knows. It’s really clever and a valuable mechanism for securing all types of transactions. It’s been around for a while but crypto has helped accelerate it.
I still think that there is some possibility that we will find ourselves doing useful things with crypto. But the time it’s going to take to get there is extended and peoples’ expectations of what that will be worth have diminished. The thing is that a lot of firms have got some momentum behind them. For example, Starbucks is launching an NFT loyalty program and some banks are looking at how they can use NFTs to represent assets. So I think there is still some forward momentum but the shock of 2022 has done a lot to dampen the area. The realisation that this was a den of crime and villainy and that it is going to have to be cleaned out before ordinary folk can go back in.”
Founders fight for the climate
Nathan: “You predicted that there would be a renewed focus on climate tech over 2022. You thought that the fragility of European energy systems may force a rethink about nuclear fission power and the role of fossil fuels, such as natural gas. This did occur but perhaps not in the way you thought?”
Azeem: “Did I really say that? Wow! That is brilliant. The fact that I wrote this two months before Putin’s invasion shows just how fragile European energy security was. And that hasn’t changed. It’s obviously accentuated funding into climate tech. When looking at 2019 funding data, 6% of all venture deals were in climate tech. Three years later this increased to 25%. So, climate tech is becoming significantly more relevant. The exciting part is that you’re starting to see significant investments in quite difficult areas. For example, we have seen big investments in virtual power plants, e-trucks, carbon materials, and mining technology. This is interesting as we’ve started to see this year that it is as much a materials transition, and a metals and minerals transition as it is about decarbonisation.
There are a few other interesting dynamics to this. One was the importance of industrials & corporates getting involved in these funding rounds, quite often at early stages. We didn’t see that as much in the info tech, internet tech and social tech markets. Corporate VCs were viewed rather dismissively and rarely took early stakes in firms. This was partly because founders didn’t want them, partly because the other investors didn’t want them. But right now these climate tech companies tend to be so much more capital intensive and require so many new classes of expertise, that both founders and financial investors are more elcoming of corporate VCs at much earlier stages (even at the seed or pre-seed level). I think this is interesting because this class of capital brings expertise.
So I think that climate tech was one particular sector which really maintained and increased its momentum along the lines that I predicted. However I don’t think I would have guessed how rich and diverse the climate tech space would become over 2022.”
Big tech: bruised but not out
Nathan: “You predicted Big Tech would stay big over 2022. In terms of market cap, this did not occur. Where do you think this prediction went wrong? Or did the markets get it wrong?”
Azeem: “I think the prediction is pretty accurate. Of course, Big Tech, Meta in particular, has lost an enormous amount of market cap and they’re being undervalued by investors. But that is due to the market rotating away from growth sectors and towards more defensive ones. Inflation, rising interest rates, a consumer squeeze and recession will take its toll on racy, exotic stocks.
Google, Apple and Amazon all remain extremely powerful. One reason we know they remain extremely powerful is that Elon Musk, Epic Games, and Spotify have all been complaining about Apple, their stranglehold over the App Store and the cut it takes from transactions. These companies remain very significant intermediaries and they have a lot of power with it, although it manifests itself in very different ways.
With Apple, it’s about the iPhone and App Store control. With Google, it’s about their search, being the big traffic referrer for every business on the planet. With Amazon, there’s a whole set of different issues.
The one interesting case is Meta. Zuck has gone off and bet the multi-billion dollar house on a product (VR/metaverse) that nobody knows if anybody wants, no one knows how to build and no one knows when it’ll be ready. This decision occurred at the time when his core products were being attrited by Snapchat, TikTok and regulatory pressure. I think this is going to be a business school case study of one of the single most disastrous unforced errors in business history.
This was hard to predict since you can’t predict that someone would take a massive amount of risk. Now Meta shares have fallen by over 60% since the start of 2022. It’s hard to imagine them crawling back out of that. I could be wrong because they have brilliant, talented people, and Zuck’s a talented individual, so there is a chance they can turn it around. Either way it’s still going to be a business school case study.”
Metaverse: I still don’t understand the why behind the metaverse
Nathan: “You also made predictions that the metaverse was starting to gain steam in the public consciousness. You said that this was against your intuition but were swayed by VR headsets outselling the XBox video game console. How did you see the metaverse progress in 2022?”
Azeem: “Microsoft’s attempted acquisition of Activision Blizzard (a games company) is a strong statement that immersive experiences are going to be important. Immersive experiences like Fortnite, Roblox or Minecraft are still growing rapidly.
On the device side, the number of Oculus devices sold is small, and it’s hard to track. It’ll be interesting to see how well the Oculus Quest does over Christmas since the reviews have not been very good. Again, you can’t forecast unforced errors by companies, they may still sell quite a lot.
Apple is slowly but surely pushing their augmented reality system. It’s got a new name for its mixed reality operating system, xrOS. There will be more new high-end devices coming out which will have LiDAR. We are also seeing augmented reality increasingly becoming part of Android devices.
On the software side, one of the big issues is how well you can generate assets that can live in a 3D world and that can be used for overlays. Graphic assets of 2D and 3D video images are getting better and better, and the cost of generating them is coming down further and further.
So I think that some of the core components of this are starting to get into place but I still think that we are going to struggle with what that use case is.
This is especially so with Meta. What strikes me as odd with their approach is that they’re trying to start with a complete ecosystem. Whereas what I think you’ll see is that these things build in layers over time. Apple understands this.
This can be seen with the iPhone. One of my top apps on my phone is my parking app, Just Park. This app didn’t exist at the iPhone’s launch. You could only get it on the iPhone after years of market development. People start to understand the device and developers realise you can do something with it. Meta seems to want to do the whole thing at once and I don’t think that will work.
It will take time, and an ecosystem growth is organic. You can’t force it to arrive quicker by turning up the temperature.”
Nathan: “AI has had a breakthrough year with the release of ChatGPT, how have you viewed its overall development in 2022?”
Azeem: “Overall, it was a really important year. You’ve had a lot of work being done on large language models which had appeared a few years earlier. They’ve started to get more and more performant. And we have, of course, ended the year with ChatGPT. However, all the problems that emerge from generative AI are becoming very apparent.
On the research side, we also saw really interesting avenues, perhaps not as much as we would like though. We saw AI in protein folding, we saw AI in tensor multiplication, we saw AI systems win at games of negotiation such as Diplomacy and games of incomplete information like Stratego. These are some really important research milestones that show we’re just starting to tap this technology.
The third leg of that stool is how companies are using these technologies. I think it’s fair to say that companies rushed into what they could do with AI a bit too quickly over the last three or four years. This year they started to understand what they can do and they’ve been able to make the internal changes they need to be able to make use of these types of products. McKinsey’s annual survey of enterprise AI showed that the froth had come out of the enterprise AI, to be replaced by wider and deeper adoption across businesses.
Overall we’ve had a really interesting year for AI across all of those different pillars, core science, development and commercialisation.
🎄 Merry Christmas to all who are celebrating! In a week, we will be back with my 2023 predictions. In the meantime, we invite you to share your predictions for next year with us. Fill out this form, and we may feature your forecast in a special edition of EV.