If you're z.ai in 2025 with $10 million to spend on GPUs, what's better:
1) buying B100s by the palette with 15 years of stable work on CUDA, or
2) some rinky-dink Huawei chip that's 3x slower, no software framework and crashes 50% of the time?
It's a no-brainer. If there were no export controls, Nvidia would be taking all that oxygen out of the room. Chinese chip efforts would be put back by at least 5 years, and probably more like 10.
Sure, they might get there eventually but Nvidia would still have that lead when they do.
> If you're z.ai in 2025 with $10 million to spend on GPUs, what's better:
> 1) buying B100s by the palette with 15 years of stable work on CUDA, or 2) some rinky-dink Huawei chip that's 3x slower, no software framework and crashes 50% of the time?
Whichever one the CCP representative on your board tells you to do. You don't get a choice on things like this.
There's no way having the option of buying chips from Nvidia would have hurt China. You can't bait them into acting against their own interests by giving them stuff. If the harm to the domestic chip industry would really outweigh the benefits, the CCP would issue the order and no-one would buy the Nvidia chips and that would be that.
> Whichever one the CCP representative on your board tells you to do. You don't get a choice on things like this.
Clearly untrue, because many companies are using the gimped H20s and the full H200s (both legally imported or smuggled) despite official government policy being silicon-self-sufficiency. It's a bit more complex than "CCP issues order and everyone immediately falls into line".
> There's no way having the option of buying chips from Nvidia would have hurt China.
I'm not saying it "hurt China", I'm saying there's a fine line to tread between supporting Chinese AI models now, at the expense of the Chinese silicon industry in 20 years' time. The US export ban forced the issue and took the alternative off the table.
"You can't bait them into acting against their own interests by giving them stuff"
Generally this does work. China is pretty resistant to it as countries go.
See how difficult it is for the US Chip market to get off the ground, because it is so easy to just buy from other countries. A government directive to "Build It", only goes so far. So the Chinese Gov directing its companies to build chips, would have headwinds if Nvidia chips were available.
Yes it is Chinas MO. But it probably possible to influence their timing and/or degree. I do not thing they have anywhere near replaced western/Taiwan x86/ARM CPUs, for example?
I think even if NVIDIA were founded recently, it still wouldn't have such a share structure. The companies that do have dual class shares are software companies that can grow significantly without raising a ton of capital. NVIDIA being a hardware company, that is a lot harder to do even today.
The affair itself is not the problem. The problem is if his wife doesn't like it, that secret is worth billions of dollars to keep. What would Jensen do to keep it secret?
The problem is SCOTUS has already ruled software is protected under the First Amendment, and the landmark cases that establish this precedent applies perfectly to model files.
Weights aren't really software though? Any more than a png is software. And God knows you can put things in a png that are not covered by the first amendment.
Encryption algorithms require substitution-boxes or other data for the algorithm to work correctly. Changing the data separately from the algorithm fundamentally alters the correctness of the algorithm.
LLM algorithms also require model data for the algorithm to work correctly. Similar to cryptography, changing the data separately from the algorithm fundamentally alters the correctness of the algorithm.
This precedent has already been perfectly established thanks to Phil Zimmermann and Daniel J. Bernstein.
You mean like in Harrington v. Purdue Pharma L.P. (2024) where SCOTUS struck down Purdue Pharma’s multi-billion-dollar bankruptcy reorganization plan? Being cynical and demoralized (read: a doomer retard) is what the oligarchs wish.
I think whether or not there is an economic impact from AI is really independent of Anthropic and OpenAI.
If there is an economic boost, it will come from broad productivity improvements and deflation of costs for services. If OpenAI or Anthropic can capitalize on that or not is a separate question.
What exactly do you propose they do? Do you think they are holding back supply? Should they sell at less than market price? That's just handing the profits to scalpers/middle men.
I'm sure NVIDIA, AMD, Intel, Qualcomm, etc. would sell more hardware if they could, but they can't.
I can run a linear regression and get the same result every time or I can go to chat.com and get A/B tested upon. yes these are cherry picked examples. they're also the ones that apply to me personally. for a higher level version of this argument please see my response to the other responder.
Wait, I thought we were discussing loss of employment of engineers? Whether the AI companies continue their exponential growth curve is another question entirely. Making up for the (thus far hypothetical) loss of engineering roles doesn’t require the AI companies to even stay in business, let alone grow exponentially.
reply