This makes me think of the great Alfred Loomis, an amazing character if you've never heard of him before (there's a book about his life called Tuxedo Park, highly recommended). He was a very successful Wall Street operator who did a lot of electrical utility deals in the 1920s.

He was also a genius inventor who helped perfect radar during the war. Anyway, he was famous back then for having predicted the 1929 market crash and completely selling out of the stock market, thereby protecting his considerable wealth in a way that few other people managed to.

And what he said about that which has always stuck with me is that he couldn't understand why so many of his smart friends and colleagues constantly asked him when he thought the wheels would come off the bull market; they all knew it was overvalued and had to crash.

And he just thought that was a crazy thing to focus on: trying to extract the last few bucks before the music inevitably stopped. Once you were convinced a crash was coming, the only prudent thing to do was to sell out of stocks and hunker down and wait for it to happen.

That takes discipline when the market is ripping, though. Remember, in 1928, the S&P went up by a crazy ~38%! So had to be willing to forgo that gain and focus on the longer term picture.

Loomis also drew a comparison between that experience in 1928 and the period just prior to the US joining WW2, where most people "in the know" expected that the US would join the war soon, but constantly speculated about exactly when it would happen, trying to time it.

But once you're convinced that your country will soon be in a serious war, you should actually start preparing like you're already IN the war today! It's the same basic idea: that you reduce your downside risk by giving up some of the upside (in this case, a "peace dividend").

I feel the precise same way about what is coming with AI and robotics in the next few years. Whether it happens in 2028 or 2030 is actually really not that important. Once you believe that it WILL happ, you need to start changing your posture across a whole set of things.

What sort of things? Well, obviously your own job is probably the most critical thing. But also your investments. And the advice you give your kids about what they should do in life. And maybe also your consumption behaviors.

I spend a LOT of time thinking about this stuff because I'm probably 99th percentile in terms of being bullish on the transformative power of AI. I don't think it's just fancy autocomplete, I think AGI is already here and just not uniformly distributed across all tasks.

So not only do I try to predict baseline events into the future, I also try to think in terms of second-order effects. And one thing that seems clear to me is that, on the way to strong AI and universal humanoid robots, there is this narrow window or opening for outsized gains.

If you're willing to work super hard, focus on the right things, and leverage AI to the absolute hilt across every aspect of your work, you can create a lot of wealth for yourself in the next 2-3 years. This is especially true if you have some deep vertical knowledge/expertise.

And one of the basic implications of a "strong AI" future is that it further entrenches and empowers owners of capital. This implies to me that you should "pull forward" all your work efforts and focus them into the next 2 or 3 years and really crank so you can get that capital.

Anyway, if you're convinced that the strong AI future is coming and it's just a matter of timing, then take a step back and really think through the implications, and ask yourself how much of a difference it really makes from a planning perspective whether it happens a bit later.

And don't wait for it to become obvious to everyone. Everything in life is at the margin. Sure, you could buy some Bitcoin today, but the real time to buy it was before most people knew what it was and you could get it for under $1,000. Right now, most people still don't get AI.

They might think they do, they might see that it helps them do certain tasks a bit faster, but they don't really understand and haven't internalized the implications of what it can really do already if it were applied in a focused way similar to what's happening with AI coding.

What's happening in coding is about to happen to ALL knowledge work. The models can already do it, the capital has been raised with smart founders at the application layer, the tech gets better and cheaper every month, and the economics are just irresistible to employers.

And the employers that are being stupid and slow to react? They will be disrupted, very rapidly, both by their faster moving large competitors, and also by an army of startups that are scaling up revenues way faster than any startups ever have in the past. It's coming.

When you give advice to your kids or to family or friends, don't think about the impact of AI as just another factor to consider on the horizon. Explain what you actually believe, if only to really put the cards on the table, and try to help other people to see what's coming.

@doodlestein Great thread man - enjoyed this 🤝 As far as the 'moment of deluge' type thing you talk about in the first few tweets - do you envision that as being 100% tied to robotics? Or just AI in general either way?

@rektdiomedes AI agents for knowledge work, robots for physical world work (gardener, plumber, electrician, etc.).

@doodlestein Found you through this thread. Incredible work, I love how clear you laid it out in a way that many can get their heads around. Immediate follow.

@AlecTorelli Thanks, I appreciate the kind words!

@doodlestein Recently bro? This has been going on for 3 years. Or if wager since the creation of new technology in modern economies. You have doomers, dreamers and deniers. Doomers were always hacks and crazy with pdoom crap. Dreamers were crazy with AGI tomorrow and post labor economics so

@JordanSVIC Sometimes people can be jolted into action if they get scared out of their complacency. If you’re going to panic, then “panic early.”

@doodlestein great thread. my actions haven’t matched my convictions here

@benyficent Well, the good news is that it’s still pretty early. But you do have to take action soon.

@doodlestein I 100% believe this but have a hard time coming up with a good set of actions. which actions specifically are you taking? what are you changing?

@brainbunlabs I work an insane amount now. And I leverage AI to an almost ridiculous extent. And my work is increasingly oriented to AI (consulting to big companies to help them with AI automation). And I spend a lot on AI services to go faster. That kind of thing.

@doodlestein Difference is that AI will create new value, not just take it. Comparison with great depression not a good one

@razroo_chief You misunderstood the analogy then. It’s about when to act when you believe something important will happen but don’t know the exact timing. Same with entering WW2. Or the Jews who got the hell out of Germany in the 30s when things started going sideways. If you waited you lost