Really interesting setup, Peteris! You also just made me realize that I actually disagree with Bryan Johnson's conclusion, but I do agree somewhat with his premise, which for me is usually the opposite; I agree with conclusions, but I'm upset when people get their premises wrong. As for how I'm dealing with uncertainty: 1. Creating control: I could have built Kiwi on X Y Z cloud platform. I could have made it dependent on this or that VC, etc. I could have also used the fiat money that I now spent on building the site to invest in Ether or the S&P500. Yet, I find myself uniquely well-positioned to take on uncertainty and have control over it, so I have invested heavily in myself. I love having choices and being independent. I think this is super important for me to deal with uncertainty. It will slow growth though. 2. Trust in the Lindy effect model of the future: I'm not exactly sure where he wrote this, but Taleb has a chapter — maybe it was in Anti - fragile—called "How to predict the future," which has become my go-to model for how the world works. I trust in this model, and it helps me to navigate. 3. Become really good at something, then don't plan: I love it when a plan works. Sometimes, I'm planning a weekend with friends and family, and when a plan works, I end up very happy. But I generally, in life and work, try to be very open-ended regarding plans. For work, for example, I do believe somewhat in determinism and that if I work hard enough, I will get some place some time. Yet, I don't plan to become the top CEO or German chancellor or whatever. During COVID or other scary moments in my life, I asked myself: How can I hedge an uncertain future? To me, the answer always has been that if I find joy in trying to become the best possible person, personally and in my career as a software dev, then in each individual moment, I never have to be fearful about my future because the accumulation of my decisions will always make sure that I end up in places that make me stronger or more proficient or sought-after. Although I gotta say, I have had some doubts about that recently with regard to family planning and high-risk startup work. Btw. I think I kinda do all the prongs which you have mentioned: - look for scarcity in creating network effects with software: that's Kiwi - focus on short-term exploitation: I try to make Kiwi work NOW - Make bets on companies that'll drive change: I've built Kiwi (which I believe in, obviously) on Ethereum, Farcaster, ENS, etc.
Embracing uncertainty is a great attitude I think in approaching life and work. I am trying to do this for years and it has brought me a lot of good things and joy. Also it is quite relaxing. Instead of saying “I want to become … in the future.” - you can say “I don’t know what I can become.” It leaves room for discovery and truly creating something new. You can only think of stuff you already know. So if you focus on something specific predetermined, you are likely to miss opportunities along the way. Better to give yourself over to the journey and see where it takes you!
@tim - absolutely believe that patient bootstrapping (especially when activities are only focused on bottlenecks) is one of the best ways of thriving in times of uncertainty @0xC304 - like this a lot, although I've always debated in my head whether more unique moments lie in depth (spending 10 years on something so you access insights that no other human has accessed) or breadth (traveling 70+ countries).
Interesting post, I think these 3 methods make perfect sense. My take on LLMs, in particular, is that - before ASI - it will be just a tool. Today, it's an army of interns & junior employees. With GPT-5, it might become an army of senior employees. And with GPT-6, maybe staff-level employees. So it's still up to us to decide what they should do, which market to choose, what goals to follow, how to build what we want to build, and so on. And just like startups can have great teams and tons of cash and burn it all, the same goes for LLMs/AI. The difference is that the costs should be lower, so it might be easier to iterate.
@mac - completely agree on LLM. It's a weird world where anyone has an infinite org below them but one that we can still reason about :).