Tricky to make predictions based on costs or quality when the costlier parts, training and inference, are totally disconnected from reality thanks to VC money and when total cost of ownership over time is radically different.

▲

IshKebab 2 days ago | parent [-]

They aren't totally disconnected from reality. There are plenty of providers who don't have a ton of VC money selling tokens at basically cost price.

▲

simianwords 2 days ago | parent [-]

There's a misconception going on that somehow the inference prices are so costly that eventually we would see some market correction. This is false.

	▲	nullsanity 2 days ago \| parent [-]
		[dead]