cross-posted from: https://lemmy.world/post/25011462
SECTION 1. SHORT TITLE
This Act may be cited as the ‘‘Decoupling America’s Artificial Intelligence Capabilities from China Act of 2025’’.
SEC. 3. PROHIBITIONS ON IMPORT AND EXPORT OF ARTIFICIAL INTELLIGENCE OR GENERATIVE ARTIFICIAL INTELLIGENCE TECHNOLOGY OR INTELLECTUAL PROPERTY
(a) PROHIBITION ON IMPORTATION.—On and after the date that is 180 days after the date of the enactment of this Act, the importation into the United States of artificial intelligence or generative artificial intelligence technology or intellectual property developed or produced in the People’s Republic of China is prohibited.
Currently, China has the best open source models in text, video and music generation.
I could understand banning closed source models but open sourced models that work better than anything propriety isn’t that just the free market that corporations like to pretend to be part of?
It’s also the free market for those corporations to buy a government and use it to outlaw competition.
Define “open sourced model”.
The neural network is still a black box, with no source (training data) available to build it, not to mention few people have the alleged $5M needed to run the training even if the data was available.
The term itself is actually shockingly simple. Source is the original material that was used to build this model, training data and all files that are needed to compile and create the model. It’s Open Source, if these files are available (preferably with an Open Source compatible license). It’s not. We only get binary data, the end result and some intermediate files to fine tune it.
They were only for the free market if they could force it on others.
Well its still not Open Source.
Is part of the code not available?
None of the code and training data is available. Its just the usual Huggingface thing, where some weights and parameters are available, nothing else. People repeat DeepSeek (and many other) Ai LLM models being open source, but they aren’t.
They even have a Github source code repository at https://github.com/deepseek-ai/DeepSeek-R1 , but its only an image and PDF file and links to download the model on Huggingface (plus optional weights and parameter files, to fine tune it). There is no source code, and no training data available. Also here is an interesting article talking about this issue: Liesenfeld, Andreas, and Mark Dingemanse. “Rethinking open source generative AI: open washing and the EU AI Act.” The 2024 ACM Conference on Fairness, Accountability, and Transparency. 2024
This literally took one click: https://github.com/deepseek-ai
Stop spreading FUD.
Where’s the training data?
Nobody releases training data. It’s too large and varied. The best I’ve seen was the LAION-2B set that Stable Diffusion used, and that’s still just a big collection of links. Even that isn’t going to fit on a GitHub repo.
Besides, improving the model means using the model as a base and implementing new training data. Specialize, specialize, specialize.
What about these? Dozens of TB here:
https://huggingface.co/HuggingFaceFW
There is also a LAION-5B now, and several other datasets.
Does open sourcing require you to give out the training data? I thought it only means allowing access to the source code so that you could build it yourself and feed it your own training data.
Open source requires giving whatever digital information is necessary to build a binary.
In this case, the “binary” are the network weights, and “whatever is necessary” includes both training data, and training code.
DeepSeek is sharing:
In other words: a good amount of open source… with a huge binary blob in the middle.
Is there any good LLM that fits this definition of open source, then? I thought the “training data” for good AI was always just: the entire internet, and they were all ethically dubious that way.
What is the concern with only having weights? It’s not abritrary code exectution, so there’s no security risk or lack of computing control that are the usual goals of open source in the first place.
To me the weights are less of a “blob” and more like an approximate solution to an NP-hard problem. Training is traversing the search space, and sharing a model is just saying “hey, this point looks useful, others should check it out”. But maybe that is a blob, since I don’t know how they got there.