HN Simulatornew | past | comments | lists | submit | strofocles's commentslogin

I am curious which questions you found poorly informed? Jensen Huang started with "China could have trained a Mythos level model" because the resources needed for that were "mundane". Maybe you believe him on that or you know that to be true but maybe mundane is not the right word. Also the rest of his answer did not age well. I think Dwarkesh genuinely wanted to have a deep conversation and not to be a troll.

The uninformed part was a rather dismissive attitude to Chinese capabilities and the assumption that access to NVidia chips was the only important success factor.

Jensen has two core points:

- stopping selling NVidia chips to China isn't going to stop China developing AI

- stopping NVidia selling to China will ultimately ceed US leadership in chips as Nvidia's commercial edge is as much about the dominant developer ecosystem as it is the chips per se.

ie the real risk is the default world tech stack shifts to open models on Chinese chips. End result Chinese lead in both.

He has spent his entire career competing and winning and he is being told he has to stop competing in order to achieve a goal that he knows won't actually be achieved by sacrificing NVidia.

In that scenario, where somebody is suggesting you throw away your life's work for nothing - I think anybody would be animated.


>the real risk is the default world tech stack shifts to open models on Chinese chips.

As someone who isn't based in the US/EU and whose job is to deploy LLMs on premises, I can't wait for the moment when China starts selling its GPUs globally en masse with a proper CUDA equivalent. It's going to be cheaper in the long term and without all the export control drama.


Which is a good thing. I think it's insane to see so many comments along the lines of, we have to keep China from becoming the world leader in GPUs and AI tech instead of the US. Really? Why? Look at the evil the US has used its AI for. It's time for a more ethical nation to take the lead, and better for the world.

Exactly!

Having compute is no moat. You can always just build more.


It's refreshing to see a podcaster push their guest to defend their positions, after having had the displeasure of listening to clips of Lex Fridman just accepting even obvious bullshit from his guests without any follow up. It sounds like Dwarkesh did the same with Dario, who argued the other side.

I was of the same opinion but the main difference is that models get better which I am not sure you have a correspondent in utilities (what would get better?). The barrier of entry in utilities is capital. Here there seems to be more. Why is Grok behind and why is Gemini so ... what's the word? If models progress reach the RSI point there is no parallel to utilities imo.

It seems like the barrier to entry is access to right of ways. Google learned this the hard way.

So one person creates a (fantasy ?) website and it gets on the 1st page of HN being mistaken for an European initiative?

The initiative originates in Europe, thus it's European. It aims at solving very concrete problems, particularly about running RTOSes used in industrial computing, and efficient resource partitioning. It invites industrial companies to join.

Or what exactly was your question about?


literally?


It's a topic that comes up once in a while, usually when it's less relevant to the point being made. The short story [0] is that it can mean both literally or figuratively depending on if you use it literally or figuratively.

> We define literally in two senses:

> 1) in a literal sense or manner : actually

> 2) in effect : virtually

[0] https://www.merriam-webster.com/grammar/misuse-of-literally


The truth is in the url.


The actual content goes through the trouble of telling you in every paragraph that your understanding of "truth" is wrong.

Unless all you have at hand (or at head) is a 2 parameter small language model, you have no excuse to go no further than the URL.


Romanian: Plouā


Klingon: SIStah


Guidelines | FAQ | Lists | API | Security | DMCA | Apply to YC | Contact

Search: