Because we need to have the theory of mind to recognize and understand what other agents want. This is crucial for survival. Our own experience enables this skill.

He doesn't need to evaluate all of them to know their shit or hit 😉

This reminds me of the "philosophical zombie" arguments by David Chalmers on conscious experience. For decades I thought the p-zombie was a silly concept, but here we are...

This is all to say that I did not fully understand how you came to your conclusion from reading that paper (if I interpreted your words correctly that is).

In my definition of Computable most simulations are only useful when we compute irreducible rule systems, like rule 30, for example. I would further propose that our universe itself is a vast but finite state space.

Your conclusion from the paper you review confuses me: "It suggests that most artificial life systems and simulations, which rely on computable rules and finite state spaces, are inherently incapable of strong OEE."

I read your post because your terminology 'computability' confused me. I think it means something like 'reducible computation', eg finding an algebraic solution.

Once you have a good code template that it can follow, the quality experience can drastically shift: less hand-holding required, faster iteration times. And you're right of course, good engineering skills are still very much required to sculpt the final result to something really decent.

Yeah but, even statistically, this mistake is an outlier in c code base training data...the number of basic C introductory courses etc. you get my point.

This is a human mistake, an LLM will not make such a mistake unless it does not know it's writing C.

1 such bot is harmless, but thousands of them and you can influence the overall sentiment and overall narrative. That's useless in a dead internet, but I don't think that's the end goal of these actors.

Dead internet is a meme, but the truth behind these bots is far more sinister. These bots are being selected like Darwinian evolution: their fitness function is simply to be not unmasked as a bot and build credit points.

Unfounded confidence on both sides of the stick is the only interesting observation to be made.

Replace ChatGPT in your arguments with 'Computers' and you know how well this will age.

Only that it's not the story. Little detail.