I think there are roughly 3 futures we might find ourselves in in 20 years:
(1) AI is important but comparable to past technologies (like the internet)
(2) AI replaces most human labor, but humans retain control
(3) AI replaces ~all human labor, and humans lose control
What are some possible investment implications?
(1) Little reason current economic systems wouldn’t persist. You should probably invest in stocks that benefit from AI-driven productivity gains. Standard arguments for why this is hard to predict apply, broad index funds (like VT) are probably a good bet.
(2) Still seems pretty likely that our economic system would largely persist. It becomes much more important to be owning capital, as you won’t have income and UBI is uncertain and out of your control. You want to end up in the capital-owning class, not in the UBI-dependent class.
(3) Economic systems probably don’t persist. Owning capital matters much less, since it’s unclear whether/why superintelligent/super-capable AI systems would respect existing property rights. If you think this scenario is likely (as Musk seems to), it may make more sense to consume now rather than invest.
I’m operating under the following rough subjective probabilities: (1) 30%, (2) 50%, (3) 20%.
If I had more time, I’d probably think more seriously about investment strategies that outperform ‘VT’, since I do believe AI will be more transformative than what the stock market currently prices in.
Separately, I’ve spent the last ~3.5 years working full-time to hopefully reduce the chance of outcome (3), or at least make it less bad. That’s not optimal from a personal financial perspective, but I’m doing it for altruistic reasons. If you think (3) is too-likely-for-comfort and are altruistically inclined, you might want to check out https://80000hours.org/ and feel free to DM me directly.