Jan Wildeboer 😷:krulorange:
Yes, you can have loud and clickbait discussions about censorship and bias in LLM/AI. Yes, you can paint it as a China v USA(I) "war".
Or you could take a step back and see that we are moving towards the next stage that happens with every new technology: Democratising and distributing. Models will be smaller and less wasteful in both training and inference. We will be able to take AI "inside" instead of having to rely on an oligopoly offering "AI as a Service".
And it will be more Open Source.
Jan Wildeboer 😷:krulorange:
@larsmb Which is why I said “will be” not “already are” …
Lars Marowsky-Brée 😷
@jwildeboer None of the models in question are actually Open Source. :-)
Best we're getting so far are open weights.
But interestingly, one of the arguments against "fully" open in the past always was "you couldn't pay for the training anyway".
If #DeepSeek is real though, at 6mil? Yes, why, indeed, *many* organizations actually could pay for it.
Jan Wildeboer 😷:krulorange:
@intrbiz Yep. Same. See https://social.wildeboer.net/@jwildeboer/113901800217019967 :)
Chris Ellis
@jwildeboer part of me thinks the AI oligopoly will do all they can to enforce their oligopoly.
GNU social JP is a social network, courtesy of GNU social JP管理人. It runs on GNU social, version 2.0.2-dev, available under the GNU Affero General Public License.
All GNU social JP content and data are available under the Creative Commons Attribution 3.0 license.