I somewhat disagree. we still need humans to provide direction and context and expanded the concept of... well everything. But we will certainly need less of them. AI will afford the ability to more quickly disseminate state-of-the-art. As soon as a breakthrough is discovered, you no longer need to read 300 whitepapers to hopefully stumble upon it, AI will identify relevancy quicker.

AI didnt break through stavier-nokes until a human set the initial direction. That's not going to change.

▲

famouswaffles 4 hours ago | parent | next [-]

>AI didnt break through stavier-nokes until a human set the initial direction. That's not going to change.

That's not what happened. There was no 'initial direction' to set because OpenAI didn't know whatever tristan et al were working on or what direction they took. They heard rumors thta navier stokes was solved and set their best model on it.

▲

pixl97 8 hours ago | parent | prev | next [-]

>AI didnt break through stavier-nokes until a human set the initial direction. That's not going to change.

Why isn't this going to change?

Right now it's easy to see why don't set AI lose on every problem, someone human or AI has to delegate rare resources between competing interests. But if we look at general compute it was no different in the past. In 1980 you had to ask for permission to get CPU compute time. In 2026 you run it on your own computer, or maybe pay for it on Amazon. The constraints are much different.

If we keep pumping out chips and increasing efficieny someone will just make the LLM into an agentic loop (build the harness in) and set it lose on problems.

▲

johnsmith1840 2 hours ago | parent | next [-]

Model collapse

If you solve it then yes humans are toast mental usefulness wise

	▲	int_19h an hour ago \| parent \| next [-]
		Models have been trained on synthetic inputs for over a year now. Collapse only happens in very specific circumstances, it's not a limit in practice.
	▲	ChickeNES 2 hours ago \| parent \| prev [-]
		Model collapse as an insurmountable wall is a fiction dreamed up by Luddites, actual frontier labs understand the failure mode and avoid it.

▲

SpicyLemonZest 8 hours ago | parent | prev [-]

It’s like asking why we don’t just run all the programs instead of making users decide what to run. There’s an infinite number of problems that could be solved, and the question of which ones are important to solve is a question about us and what we want.

▲

BalinKing 2 hours ago | parent | prev [-]

(minor nitpick, but it’s “Navier–Stokes”)