The internet died a long time before AI

There's plenty of talk out there of the "dead internet theory", where most supposedly-social media content ends up being produced by AI and so eventually most of the internet is no longer "alive", but just aping the behaviour of humans to please the algorithm and get attention.

Anti-AI skeptics point to that as the downfall of AI, an ouroboros kind of situation where the AI will eat its own tail and the quality of AI output will collapse and finally, finally, finally this whole AI nightmare will be over and they'll be able to wake up and go to work doing dumb stuff that could have been automated but isn't anymore and... oh wait, I got off on a tangent. Those people are wrong anyway.

These days the bleeding edge labs are training AIs with growing amounts of synthetic data - that is, AIs are being used to generate training data for other AIs. The dead internet might come, but it won't even cause the tiniest speed bump in the rise of AI.

But what about the dead internet?

It is sad to think of the internet as populated by AI bots that mindlessly repeat winning formulae to capture the attention of the hapless folks who still wander it.

The thing is, we're already there and have been for a number of years.

Turns out, you don't need AI to have a dead internet. You just need really effective algorithms.

Most content that "does well" on social media is designed to "please the algorithm". It's not original, it's not spontaneous, it's not creative, it's just "human slop" and it's no better than "AI slop". It's just more expensive to produce.

There have been tools to automate and standardise and simplify content production for years now, and many people use them, to great success. The slop already dominates.

The internet is already dead, in the sense that it's filled with content that's not actually conveying any real human expression, it's just the same dead-eyed filler content that AI is now producing at an accelerated pace, just created by dead-eyed humans who want your clicks and your dollars more than they want to express anything about themselves or the universe they live in.

Failing the Turing Test

The question that may, or may not, that's up to you dear reader, be worth pondering is... if someone behaves in a way that's indistinguishable from a bot... are they a bot?

I say we call them bots. At least in that part of their life. If they are creating slop to please an algorithm, then in that activity they are bots, nothing more, nothing less.

It's not even an insult, just a truth. In that area of their life, they're automatons going through the motion and missing the point of being a live human. It's ok, we all do it. We're all bots in some aspects of our lives, unless we reach a Buddha-level of enlightenment where we're entirely free from all our past programming, but few people get there, even after vast sums invested in therapy, heroic doses of psychedelics, and interminable silent meditation retreats.

We're all bots sometimes, so let's be honest about it.

Being honest about it means that maybe we can start to take action to minimise those parts of our lives.

Write spontaneously instead of to please an algorithm (I'm trying to do this here on this blog, even if I don't always do it on Twitter or LinkedIn).

Just post whatever you feel like without worrying about what The Algorithm thinks of you.

After all, if you can't be free even when there is no one coercing you, what's the point?

What can you do to be more free today?