There's no real incentive for the industry to coalesce around 2-3 models. I think it's very unlikely to happen.

pizzafeelsright 4 hours ago | parent | next [-]

Nobody wants to admit that while partial to one company or another, brand and all that, I highly doubt that within 1-2 years we can tell the difference between model output for 90% of our tasks.

Except for the 'vibe' of some of the models with their writing or testing style I am not certain I could tell you which model is which unless asked.

Switching from model to model at the moment has been seamless without extensive 1 to 1 bench-marking but most of my tasks are never 1 to 1.

	▲	etempleton 4 hours ago \| parent [-]
		Completely agree. And for companies, if things are that indistinguishable, it is going to come down to cost and enterprise security / privacy. Sure companies will be slower to switch, but they will. It will be a race to the bottom.

▲

arealaccount 2 hours ago | parent | prev | next [-]

It could be like the last SOTA lab to run out of money wins

▲

Zambyte 4 hours ago | parent | prev | next [-]

Training is expensive. That seems like an important incentive.

▲

etempleton 4 hours ago | parent | prev [-]

I agree. Which is why I think the bubble is going to pop in the nearish term.