Context: AI can now program, draw, animate, create music, do math, write, and probably many more things, almost equally/equally/better than most/all humans, and at scale. Probably, soon it will close the last gaps.

Since I was young, I wanted to "leave my mark on the world". It didn't have to be a big mark - or maybe it did - but whatever. I wanted to create, or discover, something, that will be remembered or used long after I'm gone (e.g. Olympic athletes, social people, and murderers do leave some kind of impact on the world, but it wasn't what I wanted). I learned to program at around 10, I started doing my first degree (computer science) at 13, and now (early twenties) I'm doing my PhD, in... AI.

I do have some achievements I'm proud of. Around 14, I created some multiplayer video games that went somewhat popular in my school, and people literally played it quite often, around 5-10 people together (we had a lab in the school and it was forbidden to play games there, however they "only helped me debug my game"). I left this school at some point, however I continued to program games until about 18, and am still proud of each, even the ones no one besides me played.

I invited a stupid fake religion that went somewhat known/discussed in my (religious) school's year group too, that was also something.

Meanwhile, I become interested in AI (a bit before gpt2, I think). After finishing my first degree, and working a bit, I started doing my masters, and then continued to PhD. I published a few papers, some on game theory (somehow in the academy it's considered AI and is published in AI conferences), some on interpretability (not because I was concerned about AI safety, I just found it interesting). One paper got some tenth of citations and I'm also somewhat proud of it.

About half a year ago, I become a bit obsessed with playing guitar, as a hobby. Before about 2 month I created a YouTube channel where I publish music. I got some hundreds views and around 5 followers beyond people I know from real life. I'm also quite proud on that.

I'm okay with some of my knowledge becoming irrelevant; I'm ex-relogious so I'm used to it, and I'm still young enough to learn new things, even if it means the ceiling on my abilities will be lower. But AI is learning faster than me, and probably soon AI will be able to create anything I can.

I regret nothing: I'm happy that I learned to program, happy that I learned math, and I'm enjoying creating music sooo much. I did use all of these, and it was so fun. But I had so much more potential... I wanted to do so much more. I've always dreamed of taking a break for a year or two and create some really amazing PC game or something... or discovering a solution to a semi-famous math problem... or whatever. But now it's not a human thing anymore. I look at the few people I know who still program in the old style and it looks so naive, though I also envy them. But if this is my thinking, how long before I stop e.g. creating music "in the old style"? How long before I just give up on every talent I have, anything that require effort but will never give me any recognition, not anymore, and just... IDK, play video games, read books, hang out with friends... not that it's bad but I wanted so much more...

And well, I know all of these are first world problems. Except safety, which I didn't even discussed (and it's not what keeps me awake), if AI will do to medicine what it did to math, so it will probably kill us all but it could also cure cancer or whatever. But still.

So, I don't know if these personal venting fits lesswrong, and sorry that it's not very polished (not going to LLM this one...), but I figured there might be other here in similar situations, so... how do you stop yourself from going crazy, and, well, if you have any idea of something that humans alone can still create, and will probably continue to do it in the foreseeable future, I would like to hear about it.

A couple points that can maybe make you feel a little better.

Talent as part of demonstrating fitness, competing for status and mates, isn't going away just yet. People instinctively want to hang out with (and sleep with) those who prove natural talent, and instinctively mark those who use AI as pretenders. My wife and I own an art gallery and this is exactly what we see: nobody wants AI art, viewers want to find natural talents and hero-worship them as such. I also organize live music jams and the same is true there, audiences want to find natural talents and hero-worship them.

Right now AIs are quickly surpassing humans in domains where automated evaluation is possible, like games, programming, or math. But I haven't yet seen AI make interesting art, writing, or music. The challenge isn't generating a painting that looks impressionist or a fugue that sounds like Bach: all artists know that repeating is easy and won't bring you fame. The challenge is making a stylistic innovation that will resonate with society, "get trained on Bach and output the Beatles". Maybe AIs will get better at that too, but the feedback loop for now goes through society, so these tasks are much slower to train for.

I also believed that I would make some big impact on the world... and didn't... but it took me a few decades to adapt to that fact. Must be harsh when it happens overnight.

You could try to ride the wave and snatch one of the few remaining successes. For example, not what AI can write code, you could try to create a popular computer game -- figure out what needs to be done, let the AI do it, give away the alpha version for free, collect feedback from testers and let the AI implement it. Or maybe write a novel: decide the big picture (maybe after brainstorming with an AI), let AI fill in the details, run Pangram on the resulting text, and let AI fix it, etc.

Generally speaking, find those things where AI can do 90-95% of the job, but not 100% yet; or maybe even 100% but has not become common knowledge yet... and use the AI to achieve something there. Sooner or later everyone will be able to do this, but you could be among the first ones.

Another option is to ignore the rest of the world, and just try to improve yourself, as if you were the only person in the universe, with no one to compare to, but still wanting to improve. Make your goal not inventing the theory of relativity, but understanding it. Use AI as a coach. Be a renaissance cyborg .

Maybe even join all these options and go a little meta: use AI to improve your life, and then coach other people how to do it, and then write a book about it and publish it.

I'm very concerned about AI becoming uncontrollable soon and destroying our future. But unless that happens, I'm not concerned about human creativity. Whether you are creative is your decision, not a machine's.

I'm a writer, with more than 70 books published, some of them bestsellers. When I started, I knew I wouldn't be the greatest writer of all, and the competition in the book market was fierce even before AI. But I didn't care much, because writing was fun.

I wrote three complete novels before the fourth one (about an AI getting out of control, of all things) got published. Although I'm an accomplished German author by now, it still happens that books I write end up unpublished, because I always write them first and only then offer them to publishers. This may seem stupid, but it preserves the freedom and fun of writing for me. I have always made sure to never write something just to please a "target group" or (even worse) fulfill the demands of a publisher. I write because writing is fun.

Of course, I love it even more when people read and like my books. But mostly, I write because I want to write. That said, I think the demand for human creations will continue to be very different from the demand for AI-generated content. AI can create a painting that looks like a Picasso or Gauguin in seconds, but nobody will pay anything for it. The reason is simple economics 101: what is unlimited is worthless. There was only one Picasso, and he only painted a number of paintings in his life, he is also famous and important as an artist, therefore people pay millions for his works. The same logic is true for all other arts, including creating games. I believe that while there will be lots of people consuming AI content without thinking about it, there will also continue to be a market for genuine human creativity, because at least some people care who created something, and in some cases also why.

There's also the matter of live performances, which play an increasing role. People won't pay $100 to see a robot play guitar (at least not after the one-time novelty effect has worn off).

When Deep Blue beat Kasparov in 1997, some people thought "chess is dead", but we have more active chess players today than at the time. When photography was invented, portrait painters lost their jobs, but people didn't stop painting - instead, free of the necessity of painting realistically, they invented new, creative painting styles. We will not stop competing at the Olympics even when robots can beat us in all disciplines, because we care whether the athletes are human.

Of course, it may be more difficult in the future to earn money with creative content. But that will be true for everything humans do for money today, a problem that needs to be solved anyway. At least, unlike most jobs, being creative is rewarding in itself. So don't try to beat the machines in whatever you like to do. Do it because you love to do it, and share it with people because people care whether you or an AI made it.