Originality Is Dead

Think of an account you follow on Twitter. Someone whose tweets you actually read and don't just scroll past. Now take the person away. No name, no face, nobody behind it, just the text showing up in your feed.

Is it still good?

Take a second with it, because I think both answers are kind of weird.

If yes, you only care about the content. That means the creator is replaceable, and good AI content would hit exactly the same.

If no, the person is what you care about, and the tweets were never really the point.

The paradox

So why do you follow someone? Because their content is good. And why is the content good? Because they're the one saying it.

The creation matters because of the creator, and the creator matters because of what they create. It's a loop and there's nothing at the bottom of it. If you follow it all the way down, it kind of looks like humans don't create any value at all.

I was stuck on this for a while.

Someone has done it before

I've thought for a long time that humans aren't really creating anything new. Pretty much any idea I have, some philosopher probably wrote down a thousand years ago. This one too. Every thought you've ever had, some Greek guy had it first.

For years I thought this was just my problem, that I was kind of stupid or unoriginal. At my first job I kept thinking that the work we were doing had been done a billion times already. Tons of other people were building almost the exact same product with slight differences and a different logo. So why not just copy them and skip their mistakes? It started to feel like we weren't doing anything useful, like most jobs are basically daycare. You're not creating new value or new knowledge. You're repeating patterns so you don't get bored.

It's the same with everything I do. My first thought is always who has done this before, and someone always has, and it messes with my head. At a hackathon my teammates get ambitious and excited about some idea, like recursive reinforcement learning on SQL queries, and I'm sitting there thinking that for sure someone more skilled than us has already done it, or thought about it and skipped it. Every single time that turns out to be true.

This is also why I have a really hard time with startups. There's no way your idea is so original that nobody has thought of it, so what's the point of trying to find one? And in most cases it isn't even about the idea. Everyone is doing the same thing, you differentiate with some slight execution differences and contacts, and then luck basically decides who wins.

Maybe I'm extremely nihilistic.

Why I'm AI pilled

I'm extremely AI pilled right now, and I think this is the reason. The instant I saw the first output of an LLM, I understood it, because that's how I had always seen human work. Basically everything we do can be done by predicting the most likely next word out of our mouths.

A lot of people can't accept that an algorithm that predicts words is able to produce basically all of human knowledge. For me it was obvious. I had already accepted that 99% of us aren't creating anything. We repeat what we hear, so of course it could be automated.

I think this is where the existential dread comes from. If you thought your life was original and that you were creating value, AI is telling you that you weren't. There's no value added by humans, sorry!

That's the loop from before with nothing at the bottom of it.

Your friends aren't original either

But think about your friend group. Right now there are probably thousands of other friend groups having the exact same conversation as yours, with the same jokes and the same takes on the same news. Nobody cares, and that's the point.

You don't talk to your friends because they have original thoughts. You talk to them because they're your friends. The conversation is worth something because of who's in it.

Following a creator is the same thing from further away. People don't really follow content. They follow a guy, and the content is kind of just proof that the guy is still alive.

It's the same with this post. The idea is old, so the value I'm bringing can't be the idea. What's actually happening in your head while you read this is that you're imagining me explaining it. Some guy called Viljami going on about the idea of ideas. That's the only new thing here.

So the loop does have a bottom, and it's the person.

This is also the one thing AI doesn't have. An LLM can write the same tweet a person would, sometimes a better one, but there's nobody behind it. It can be useful. It can answer your question or write your code. But it's hard to enjoy it the way you enjoy someone. It can have the ideas, because they were never really ours anyway. It can't be the one saying them.

Accept the delusion

For years I thought, what's the point of writing? For sure someone has had these exact thoughts already, so isn't it a waste of time to communicate them again in a less researched and less well written way?

Now I understand that the point is not to be the best or the most original. The point is kind of just to talk to your friends.

I think my hackathon teammates had it right the whole time. You have to be in a bit of a delusion that you're creating value, and sometimes you accidentally do create something interesting.

So there's no value added by humans, and that's fine. I think accepting it frees you. Accept the delusion, don't dread it.