do you have a $100 mil worth of compute to train your own version of this?

I don't understand what gripe people have with open weight models and wanting it to be purely open source. the training dataset is only going to be a copyrighted set of contents you don't want to touch with a 10 feet pole. let alone have a publicly traded company host it for you, even if they internally are training on it.

▲ a2ff6eeb0 2 hours ago | parent [-]

It's just about as open source as OSX, which can be downloaded zero cost here: http://updates-http.cdn-apple.com/2019/cert/061-39476-201910...

I love that Apple made OSX open source.

▲ subygan an hour ago | parent [-]

we're calling it openweights now. what is a logical push to have a completely new paradigm be compared with traditional software binaries?

▲ a2ff6eeb0 an hour ago | parent [-]

Check the phrase I quoted. Anyways, I guess OSX is open assembly?

▲ subygan 12 minutes ago | parent | next [-]

is it not open assembly?

open weight models are better than open-assembly binaries (as you put it) because models are grown like plants, you can shape the open-weight model in a direction you want by feeding it more data and compute (aka finetuning).

which is something that is impossible in a binary.

embrace the new paradigm and it's tradeoffs. without scoffing at semantics and criticising from an armchair.

	▲	a2ff6eeb0 2 minutes ago \| parent [-]
		I'd encourage you to start referring to it that way,then.

▲ mrkeen 23 minutes ago | parent | prev [-]

Checked.

  The more open source software out there the better. And the more open weights or even over source AI stuff the better too right?