a letter to my past self

dear jason,

because of artificial general intelligence – ai that can surpass humans at almost all cognitive tasks – the next few years are going to be wild. it's no longer a small crowd of internet weirdos who believe this. it's some of the richest, most powerful people in the world. and you can, with this knowledge, at least be slightly more prepared.

the first thing i'd say is understand what's coming. i'd say one of the most considered takes is this series by rudolf (although the ai futures project will come out with another one too).

i think you're in a position to materially change the outcome of how this goes. so, i'd roughly put your options as:

make sure the transition goes well.

make money.

and, let's be real, if you're just entering high school, both of these are hard, but the latter is more achievable. the first thing to do is to tell whatever sources of wealth you have access to, your parents, probably, to get themselves a small bit of exposure to the potential economic growth that's happening. this means nvidia/microsoft/google. but even just being in index funds, s&p500 + nasdaq, is sufficient. probably 10% of your portfolio is sufficient. just make sure it's not sitting in a bank. go do it. now. it'll take one hour, tops.

at this age, you probably won't have a job, at least not the ones that you're thinking. automation will hit the most junior positions first. there will be things that we need super high trust on, or we'd prefer humans, like politics or high-level decision-making, but the realistic timelines for those are hard.

i'd spend the minimum amount of time possible doing school work (which, as you can probably already tell from existing ai systems) is almost wholly useless – try not to fail out of class though. instead, spend your time building. it's so incredibly easy. you can do it right now. download cursor, shell out $20/month for claude.ai professional plan, and then start building and deploying webapps. any questions? ask claude. no ideas? try and motivate yourself by looking around you and seeing what jobs are maximally automatable, and ideally some niche (don't focus on education). build end-to-end tools. get on twitter. publish them on a site. github pages + vercel is free. you don't even need to know how to code – just practice the art of copy-pasting. and once you're comfortable, start building businesses.

the end idea is to make a startup to capture the many many billions that are, and will be, sloshing around in the market. you can pull it off. other teenagers have done more impressive things.

the central takeaway that you should have is to move yourself up the decision-making pyramid as fast as possible – you want to be a ceo, a president, a manager, not a software engineer, a policy analyst, an employee. these are much harder to automate just because of the nature of the technology itself. humans are able to think naturally over long horizons. ais are not.

jason