←back to thread

US Intel

(stratechery.com)
539 points maguay | 2 comments | | HN request time: 0.001s | source
Show context
rich_sasha ◴[] No.45027104[source]
I am torn on this.

On the one hand, I strongly agree with this article. This kind of state ownership never brings anyhing good. I don't see how this is different.

On the other, it is hard to deny how impressive the new wave of Chinese manufacturing is. No longer are they just making knock offs of Western products with stolen IP. BYD for example seems genuinely innovative, a top product. There are many other examples.

Now, these are clearly not state-ran enterprises, but equally the state is heavily involved. Or, Nvidia is concerned because China can mandate that the whole country pivots to using Chinese GPUs, seemingly with no deteiment to their AI research, while amazingly benefitting their own chip production ability.

I'm not sure how I reconcile these two.

replies(8): >>45027222 #>>45027484 #>>45028593 #>>45028676 #>>45029577 #>>45030104 #>>45030127 #>>45032850 #
margalabargala ◴[] No.45027222[source]
> On the one hand, I strongly agree with this article. This kind of state ownership never brings anyhing good. I don't see how this is different.

There's short term stable and long term stable.

Having a BDFL can, when the BDFL is genuinely concerned with the welfare of whatever they are managing, result in something much better than what would be created if designed by committee. This is equally true for software projects and nation-states. China, Singapore, Linux, Python, etc.

In the long term, having a BDFL really relies on that "B" being there, and especially when a nation state is involved, the tendency of human nature to corrupt will likely eventually take over.

Basically, while China is acting with great coordination from the now with good results, they are doomed to eventually either fall to pieces when the diktat is bad, a la "Great Leap Forward", or else transition to a more stable (less authoritarian) system.

They can only do things like:

> China can mandate that the whole country pivots to using Chinese GPUs

so many times before getting it wrong disastrously, and the longer it goes the more likely someone will get it wrong.

replies(1): >>45034889 #
woooooo ◴[] No.45034889[source]
They've lifted a literal billion people out of third world, no-indoor-plumbing poverty in the last 30 years.

And, to the point of the article, SMIC is already doing 5nm manufacturing.

Political systems are more complex than dictator/freedom, there are lots of stakeholders. The USA stakeholders tend to be short-term focused financial engineers, this is separate from whether we have checks and balances or what color tie the President wears.

replies(1): >>45043477 #
margalabargala ◴[] No.45043477[source]
Yes, they have had great success.

The comments you made about political systems indicate you don't really grasp my point.

- Governments are formed of people. People make mistakes.

- Power tends to corrupt, and the likelihood of corruption scales with the quantity of power and the quantity of people passing through the role (that is, one person may truly have best interests at heart, but will their successor? Or their successor?)

The result of the above is that when assessing long term stability of systems, problems that can happen inevitably will. If a system allows massive sweeping changes, then they will happen, and eventually an incorrect one will be made.

For example, let's take the scenario above, where Chinese companies were ordered to begin using Chinese GPUs. Did that work out? Yes. Could it have gone poorly? Yes.

Will every similar decision go as well? No. Thus, a situation where the "you must use Chinese GPUs" dictation is possible, is less stable than one where it is not.

You brought up a comparison to the US, because the US and China always get compared I guess? The US does this sort of thing more weakly, so gains less instability from this particular source of instability. Whether it's currently more unstable in other ways is left as an exercise for the reader.

replies(1): >>45046470 #
woooooo ◴[] No.45046470[source]
Im saying that people live in a culture, and that sets the parameters of what's conceivable.

You're commenting from the post-reagan consensus, which inevitably leads to Intel's current situation. But it doesn't have to be like that, America had a different consensus from 1930-1970.

(BTW, deepseek was trained on nvidia.)

Xi doesn't personally set every industrial policy in China, its a ton of people all over the country. Culture.

replies(1): >>45046769 #
1. margalabargala ◴[] No.45046769[source]
I think we're kind of talking past each other maybe.

What I'm saying reduces down to, "the scale of effect that humans are able to have on a system, will eventually lead to problems of that scale on a system".

If you can dictate that all industries within a country must do some thing, that will eventually cause problems with all industries within the country.

Culture doesn't really have anything to do with it. Whether the person making that dictation is Xi or Trump or some random bureaucrat doesn't really have anything to do with it.

replies(1): >>45047906 #
2. woooooo ◴[] No.45047906[source]
Yeah, but that is a continuum and not binary. Xi isnt personally in a room pulling levers that control everything.

There are thousands of bureaucratic orgs and corporations making their own calls, deepseek and qwen were both trained on nvidia hardware, while several chip startups and SMIC ramp up.

The nature of those calls is influenced by culture. If you have a culture of MBAs, you get Intel's actions over the last 15 years.