The Best Way to Prepare for AI Disruption: Invest in Tech
Convinced that AI will dominate the job market within 5–10 years, I’m going all-in on investing in tech.
“When you think you’re at the end of something, you’re really at the beginning of something else.” Seen on a garbage truck during this morning’s run.
An automated intelligent system will be better at all aspects of my job than I am in 5–10 years. This is true for all jobs (yes, including you). What does this mean?
“Learn to enjoy looking like a loser. It’s good to be underestimated.” — Simon Squibb
I’ve spent the last 18 months scrambling to figure out how to evolve my career for a post-AI world.
Said another way — the current and future AI wave will reshape the world as we know it in unprecedented ways. How do I make sure I’m surfing the wave instead of getting my face smashed in the sand?
The verdict? It’s impossible. AI will win. It’s inevitable.
My long-standing belief (and goal) has been that the way to win at this game is to own a piece of the pie by founding a tech startup. But founding a really successful startup is incredibly rare, and founding a globally leading tech company has worse odds than playing the lottery.
I’ve always considered My Job as a great fallback plan for those crap odds. If the startup thing doesn’t work out, I can always continue to sell my craft and make a decent living.
But that will soon not be true. There may be some jobs that people will continue to want a human to do even if an AI is better at the job. Software engineering is not one of those jobs.
So I’ve started to re-think how I think about this.
If you can’t beat them, join them
(None of the following is or should be taken as financial or investment advice; this is just an analysis of my thought process.)
Look at the performance of these assets over the last 10 years:
- BTC (Bitcoin): 64.86% CAGR
- SMH (VanEck Semiconductor ETF): 30.87% CAGR
- VGT (Vanguard Information Technology ETF): 22.29% CAGR
- IWY (iShares Russell Top 200 Growth ETF): 17.04% CAGR
Yes, we’ve been on a hell of a bull run. But all of the above are outperforming the bull running index:
- VFINX (Vanguard 500 Index Investor): 12.81% CAGR
Why are these assets outperforming an already performant index? Are we in an AI+Crypto tech bubble, or is something else going on?
We are probably in some kind of a bubble. $30 billion has been invested into GPU-related tech with a current ROI of $3 billion. That’s not a good return.
But I don’t think that a potential bubble is incompatible with something else going on.
In the last few decades, the “information technology (IT) sector” has accounted for the vast majority of the growth of the market as a whole.
Why do I put “IT sector” in quotes? Because it’s a different kind of sector than the energy sector, the healthcare sector, or the financial sector.
Traditional market sectors are defined by what services the company offers, while the IT sector is defined by how the company produces their services.
In terms of services, Apple is in the consumer discretionary sector (devices), the media sector (app store), the financial sector (Apple Card), and the manufacturing sector (chips), among others.
Microsoft, Amazon, Google, Meta, etc., are the same.
But these companies are considered Information Technology companies. Why?
Because they are inordinately skilled at leveraging state-of-the-art information technology (data + algorithms + compute) in whatever market they choose to enter.
And because they adeptly choose markets that are ripe for disruption by data, algorithms, and compute.
The Prediction
Why is this relevant to the AI wave / bubble?
Both “Information Technology” and “Artificial Intelligence” are composed of data, algorithms, and compute.
In fact, “Information Technology” and “Artificial Intelligence” are effectively synonyms.
Since I’m convinced AI will win all the economic marbles in the next 5 to 10 years, I will confidently make the following prediction:
The “Information Technology Sector” will become “The Market” in the next 5 to 10 years.
The Strategy
With this in mind, my current strategy for catching the AI wave is as follows:
- Continue working in whatever capacity I can for the next 5–10 years.
- Cut costs and invest aggressively in the “information technology” sector and IT leveraged assets.
Possible outcomes:
- Most likely: IT / AI eats the market. My investments grow substantially in response, and as a result, when there is no more demand for my craft, I can retire safely.
- Less likely: The bubble pops. My investments don’t return, but my job is safe.
I’ve spent the last 20 years of my life primarily investing in myself and my own capacity to do interesting things in the world, and as a consequence, putting 100% of the pressure of “making it” on my own shoulders.
I still think owning a piece of the AI pie will be critical for success in the next couple of decades, but I now think the best risk-adjusted ambitious strategy is not to go zero to one with my own startup but to go from one to 1.25 compounded annually for the next 40 years (that’s 1 to 7,523 if you’re counting).
I’m transitioning my “making it” strategy to be “invest in the organizations that are best leveraging information technology” and chill.
Is this strategy guaranteed to work? No, surely not. But does it have a better risk-reward profile than any other option I’m currently aware of? Sure does.
I was going to dive deeper into how I see the above-listed assets as high IT levers, but I’ve gone on way too long as it is, so I will leave that as an exercise for the reader.
And that’s the tickle for this week.
Be well and die later.
Miah
P.S. Don’t worry, I’ll still be working on many side hustles. But not because I think they are the best strategy for “making it.” I’ll be doing it just because it’s fun and games.
More
Asset Allocation Portfolio Analysis
- Optimizing GBTC, SMH, VGT, and IWY allocations for risk-adjusted returns: https://www.portfoliovisualizer.com/optimize-portfolio?s=y&sl=6953lC5jleaI0iA13RT6Ty. Sharpe ratio: 1.26
- Historical returns of above allocation: https://www.portfoliovisualizer.com/backtest-portfolio?s=y&sl=cQN1VJ2t6Qv0IVkX18b4U. 5-year CAGR: 38.54%.
- Monte Carlo simulation future performance: https://www.portfoliovisualizer.com/monte-carlo-simulation?s=y&sl=2Dr4xVui6oZIA3q92WubI7. 10-year median inflation-adjusted returns: 32.75% with 1.17 Sharpe ratio.
I challenge anyone to find a public asset allocation with better current metrics than this.
I plan on doing more research into which assets are going to be the biggest IT levers in the next 10 years (will semiconductors continue their dominance?). This will not be my set it and forget it allocation, but it’s a good benchmark.
The Droids are Coming
- Amazon has more robots than human workers — this is already true.
- Humanoid robots are already built, sitting in dusty warehouses, waiting for brains.
- There are projected to be as many humanoid robots as humans by 2040.
- Consider what happens when Musk merges Neuralink with Optimus.
Peter Diamandis and Cathie Wood on Longevity, AI, and Bitcoin
- Bitcoin’s likely future as cold storage for economic potential energy
- The Future of Bitcoin w/ Michael Saylor (2024) | MOONSHOTS EP #92
- Introduction to Bitcoin: what is bitcoin and why does it matter?
- “The Spanish empire collapsed because they hyperinflated their currency as a result of stealing so much gold.” — paraphrase of Michael Saylor.
- Me: Bitcoin will be the money of autonomous AI agents.
LLMs
- Current LLMs are at the state of AlphaGo before it adopted reinforcement learning. This means they are going to get A LOT more powerful.
- The most accessible technical explanation of LLMs and GPTs.
Random
- Tech layoffs are due to AI. CEOs can’t say this yet.
- The bottleneck in AI was transistors, now it’s electrical transformers, next year it will be electricity.
- The current best test for whether something is an AI is if it will say something very non-PC or NSFW (if it’s improper, it’s likely human).
- Claude 3 beats gpt4 in chatbot arena.
- The single most searched query on Google is: “what to watch”.
Ideas
- Is my job on the line? Customized reports using existing industry reports to allow a user to see the AI risk and opportunities for their specific job and tasks.