'Often something has to happen before something happens.' It's one of my favorite quotes from legendary soccer player Johan Cruijff. I was reminded of it when last week stock market analysts, a profession with the same social utility as palm readers, predicted that Nvidia could overtake Apple and Microsoft as the world's most valuable company this very year. Without analysis of why and how this is possible, it is scoreboard journalism of the worst kind. Bitcoin's huge price rise to its highest level ever, of course, is preferably hushed up by most stock market analysts because they spent years trashing crypto. Whereas the price of Bitcoin is based solely on supply and demand, with no underlying products like Nvidia. Seems a lot easier to analyze. Yet most analysts remain silent. Coincidence? No, because coincidence is logical, Cruijff also said.
In a half-baked attempt at self-analysis, I looked up what I myself have written about Apple, Microsoft and Nvidia since I started this newsletter a year ago. Along with OpenAI, maker of the revolutionary ChatGPT, they are the only companies I have discussed more than 30 times.
Nvidia is more than hardware
Looking back, it is clear that I myself struggled to understand what the advantage Nvidia has built up; for the gap with the competition is caused by much more than just the production of very fast processors, the Graphics Processing Unit (GPUs) that are the engine block of AI software.
This newsletter is too short and I lack the technical expertise to go deep into the "software shell" that Nvidia has almost secretly erected around its hardware, making it more complex than it seems for customers to achieve the same results with other vendors' equipment. But it is useful to keep in mind that Nvidia has successfully built a defensible moat around its core business, when yet another analyst swears that Google, Microsoft and Amazon will deliver similar performance to Nvidia within a few years.
Because besides designing and delivering performance in a lab, these types of AI applications have to be tested in the real world for all sorts of different types of applications, followed by optimization (I don't think it's a Mulisch-esque argument myself, but stay with me) and finally a chip manufacturer still has to know how to produce the GPUs in volume and how to support them after the initial sale. That producing, selling and supporting may be harder than designing. I am also a very good singer in principle, but in practice my dog runs away after hearing three notes.
Every day, Nvidia increases its knowledge advantage because it has been working with all these AI customers for years. While Google, Microsoft and Amazon are still on the user side. It's like being a customer at the bakery who decides one day to start a bakery himself. Then the business plan and recipes are not the parts that make the difference either: the crux is in making and selling, which is true in baking croissants and in baking computer chips.
Conclusion: Nvidia has a good chance of adding another trillion in market value within a year and dethroning first Apple and then Microsoft as the world's most valuable company. A position it will then be able to hold for some time. Once Nvidia has surpassed Apple and Microsoft in terms of company value, only then will the general public (and thus most of the media and politicians) begin to understand that a social breakthrough has occurred.
I am not claiming, as is sometimes suggested, that AI is a breakthrough comparable to the invention of the steam engine. That was the case with the invention of the personal computer combined with the Internet, which transitioned much of the world from the industrial to the digital age. There is a significant chance that mass adoption of AI applications will have the same impact on society as, at one point in time, the introduction of the assembly line. In other words, higher labor productivity coupled with fewer process jobs, fewer work hours and shorter work weeks for most desk jockeys. The question is whether average wages will be maintained, or whether higher corporate profits will prevail.
AI in the retirement home
The applications of AI extend much further than initially thought. Obviously, many administrative functions, in fact all process-based functions involving estimation based on existing data, will be replaced by AI. That stuffy insurance salesman who comes by in the evening after dinner in his ill-fitting suit with questionable tie can do a worse job of interpreting what the customer's requirements are, than an AI application that stores all current policies and all claims for the last 50 years.
It recalls the breakthrough of the World Wide Web in 1993 after the launch of the Mosaic browser as well as the period 15 years ago, after the first iPhones and Android phones were introduced. The last two innovations, smartphones and the Internet, are the enabler for today's technological revolution, because that is what we should call AI by now. Just as a mobile app was developed for everything back then, now efforts are being made to incorporate AI into everything.
A few remarkable innovations were presented at MWC in Barcelona, with AI-powered dolls for the elderly making an indelible impression on many visitors. Who thinks $3,500 for an Apple Vision Pro is not expensive, will surely buy a $1,800 Hyodol for grandma? Seems absurd, but the AI bot in the form of a six-year-old child proved more effective at reminding the elderly to take a pill, keep moving and turn off the stove in Korea.
Innovating is difficult for everyone
Most AI innovations will fail, as most attempts at innovation fail in general. Apple luckily finally stopped developing its own car and is writing off the $10 billion invested as a rounding error. The world had absolutely no need for a new electric car manufacturer. It will be interesting to see if Apple manages to incorporate AI in a useful way into the company's biggest money maker: the iPhone.
There is a lot of talk that Google has completely lost the plot in AI, but that is nonsense. In terms of product, Google has made great strides and Google Gemini was a giant leap. Only Google's corporate culture is proving a debilitating hurdle in developing breakthrough innovations, as I described the last two weeks.
David Kiferbaum left Google and wrote a truly painful account of what it's like to work in an environment where political correctness is preferred to factual correctness. Recommended reading: How Google blew up. (By the way, Kiferbaum's LinkedIn shows that he also worked for a while at Morrison & Foerster, the law firm with the most apt URL ever for a law firm: mofo.com)
Spotlight 9? Two Spotlights: Bitcoin and Ethereum
It was déjà vu, all over again, as Yogi Berra once said: Bitcoin reached a new record high and then a harsh correction followed.
Then the most frequently asked question on Whatsapp, during birthdays and after kids' soccer games was: is it too late to get into crypto?
I do not think it is too late to get in and that we are just at the beginning of mass adoption. Again, I echo the advice from 2017, which is a few hype cycles ago, from legendary investor Fred Wilson (Twitter, Tumblr, Zynga, Etsy, Coinbase etc), which he gave based on the investor's profile:
- young, aggressive risk taker - 10% of net assets in crypto
- sophisticated investor looking for a high-performing portfolio - 5% of net assets in crypto
- average investor, somewhat conservative, but with some appetite for risk - 3% of net assets in crypto
- retiree who wants to maintain portfolio value and generate income - 0% of net assets in crypto
A detailed, careful and at the same time confusing typology. After all, some elderly people may very well be able to absorb a kick where it hurts, because their house is already paid off and they have zero debt. While many young, aggressive risk-takers must sell their textbooks and become Uber drivers or start an Only Fans as they watch their memecoins evaporate.
Because I don't know any young aggressive knuckleheads who manage to limit their crypto gambling to 10% of their net worth, as Wilson advises. Around me I see young people more likely to put 90% of their money into crypto, but that may be a genetic defect in my family. More research is needed on women's investment decisions; are there still fewer women than men in crypto, or are women smarter because they are quieter about it?
How much and in which crypto?
When people ask me how much to invest in crypto, which has been a daily occurrence again since the beginning of this year, I always answer with a counter-question: can you stand to see everything you put into crypto go up in smoke? Evaporate to nothing? And just as important: will you get into a fight with your partner if you lose everything?
The couple lunatics go-getters who then remain always ask the same follow-up question, "which crypto should I buy? To that question, too, Fred Wilson was kind enough reply:
"A diverse set of crypto-assets would include Bitcoin, Ethereum, the other major layer-one blockchains (Solana, Flow, Avalanche, Polkadot, Algorand, etc.), the major DeFi protocols (Uniswap, Aave, Compound, etc.), storage protocols (Filecoin, Arweave, etc.), telecommunications protocols (such as Helium), some layer-two protocols (such as Stacks, Polygon, etc.), some gaming assets (such as Axie, Decentraland, etc.) and maybe some NFTs."
Do I follow Wilson's advice myself? No.
The difference in value
I have been convinced for years that "something enormous" will come out of blockchain innovations. Decentralization and transparency bring an intrinsic new value that cannot be realized in other ways. Unfortunately, it will take longer than I had hoped until a widely accessible application based on blockchain technology becomes available that is relevant to a large audience. Call it the ChatGPT of blockchain, that is what we are waiting for. Such an application would be of great social value.
I believe strongly in the crypto market, but I don't have the chutzpah (anymore) to think I can pick the winners. This has proven difficult with every disruptive advance in technology. The challenge is to identify potential winners early. From that follows a financial value of an innovation; incredibly fascinating, but less interesting to me personally. Before you know it, you spend hours musing about peaks, valleys and candles, without yet knowing what application it is actually about.
Obviously, every investor wants to achieve the highest possible returns with the lowest possible risk, so maybe I should create an investment portfolio myself that is trackable on a weekly basis. I would love to hear via email, LinkedIn or X if you would find such a portfolio interesting. Anyway, all tips, comments and reactions are very welcome.
Enjoy your Sunday, see you next week!