I've been saying this for a while. Programming has been unbelievably underestimated, almost as much as art (i.e. everyone thinks they are an artist when having opinions but rarely understand anything about it).

Programming is about engaging with high level mathematical abstractions.

To the extent AI can do this high level thinking, it is going to be able to do absolutely everything else humans can. Programming (when not underestimated) is as hard as the hardest cognitive task humans can do.

▲

markasoftware 2 hours ago | parent [-]

I used to think this too, but AI capabilities over the last couple years prove it completely wrong. The skills are more specialized per domain than I thought (though there is still a lot in common between fields)

And further, surely you know some very strong programmers who are very bad at art, or math, or human relationships?

	▲	proc0 2 hours ago \| parent [-]
		Every new model is only marginally better than the last, since last year, and I'm sure they are throwing all the compute at every new model, which means it's hit diminishing returns. Even after all of the resources it takes, it still needs human guidance. At the very least the conclusion is that it's not doing what humans are doing. They might know all the programming languages and all kinds of domains but then it needs to be told basic things to not fail at simple prompts. Part of the problem is that this is very hard to talk about and describe. It requires a long conversation, probably better in real-time, to explain exactly how AIs are not really living up to the hype, and are actually introducing long-term risk by using them.