of course the bigger story here is that there is a LITE version of CNN on which uBlock and Privacy Badger find nothing to block!
AI breakthrough
That makes it sound like there’s something actually new here, but it seems like it’s just a Chinese company getting closer to what "Open"AI / Anthropic already have.
It’s nice that these things are open
sourceweights, but from a technological view this isn’t a breakthrough.Edit: Corrected to “open weights”
It’s nice that these things are open source
Thanks, didn’t know that!
It feels like nitpicking, but there’s a relevant difference here. If you can only access the weights (and not the training data and the learning algorithms used), it’s hard to get (say) a Chinese model to freely talk about Tiananmen Square Massacre. (Just tried it locally on Deepseek R1; it’s doable, but takes some prompt hacking.) The models may be freely available, but there’s bias and censorship baked into them.
This is huge though.
The problem AI has, especial frontier models, is that they cost BILLIONS to develop, so they need SHIT LOADS of paying customers to be profitable.
However, unlike let’s say physical invention, you almost get no advantage of your development since China can just clone your model with very little cost and then offer it at a fraction of the cost.
It’s honestly a very effective way to undermine American companies, with hardly any money investment.
Let US companies burn cash, then copy and undermine the customer base.
It’s one of the reasons people fear the AI bubble, all this money invested isn’t coming back
Dunno about getting ‘closer’ but glm-5.2 is indistiguishable from what I got before from claude and openai while being cheaper.