941–950 of 1001 posts
Earlier quoted context omitted.
I think that’s unfair unless you give specific examples and clear evidence he’s wrong. I disagree with PG on economics and politics, but much of his writing on that is subjective.
“Clear evidence he’s wrong” Good science is dead, isn’t it?
Earlier quoted context omitted.
> DeepSeek is still a big model that requires a lot of resources to run I can run the largest model at 4 tokens per second on a 64GB card. Smaller models are _faster_ than Phi-4. I've just switched to it for my local inference.
Isn't the largest model still like 130GB after heavy quantization[1] and 4 tok/s borderline unusable for interactive sessions with those long outputs? [1] https://unsloth.ai/blog/deepseekr1-dynamic
I run my models in an agentic framework with fast models that can ask slower models or APIs when needed. It works perfectly, 60 percent of the time lol.
Earlier quoted context omitted.
I think that’s unfair unless you give specific examples and clear evidence he’s wrong. I disagree with PG on economics and politics, but much of his writing on that is subjective.
He recently said that evil people can’t survive long as founders of tech companies because they need smart people to work for them and smart people can work anywhere. There are lots of other examples. Especially read his recent tweets/essays that aren’t about his area of expertise.
We might also want to delve into what "evil" means here. Presumably he's talking about evil in the way that they treat their employees?
As sibling says, that's his core expertise, so not quite relevant.
Here’s a take I haven’t seen yet: If training and inference just got 40x more efficient, but OpenAI and co. still have the same compute resources, once they’ve baked in all the DeepSeek improvements, we’re about to find out very quickly whether 40x the compute delivers 40x the performance / output quality, or if output quality has ceased to be compute-bound.
I don't really understand why the stock market has decided this affects nvidia's stock price though.
90% of the comments in this thread make it clear that knowing about technology does not in any way qualify someone to think correctly about markets and equity valuations.
An experienced person on how is challenging to implement this in the Tech world: https://news.ycombinator.com/item?id=42658998
Basically, if you look at only the stories, you get a dire picture -- like, "how will they make payroll????"
However, if you look at the actual share price in context, you see . . . a not very interesting event.
Look here:
https://www.macrotrends.net/stocks/charts/NVDA/nvidia/pe-rat...
The graph definitely shows a share price adjustment, but it by no means erases the run-up Nvidia has enjoyed in the last 2 years. All that happened was the stock dropped back to it price of about the middle of last summer.
This may be old news to most of you, but: in markets we also track a figure called the P/E ratio, which is the ratio of the price of the company vs. its profits. Old-school manufacturing firms -- what used to be "blue chip" stocks -- would be in the 18-22 range here. Apple which absolutely PRINTS cash, is at a high-flying 38. NVidia's P/E is still 58.8.
The tl;dr here is that NVidia is still valued very, very highly. They're still what, the 3rd most valuable company in the world, with an enviable position in chipmaking. They still make money hand over fist. The weird part is that their firm is SO valuable that they could take a $600B haircut and it be almost a non-event.
Earlier quoted context omitted.
Isn't the largest model still like 130GB after heavy quantization[1] and 4 tok/s borderline unusable for interactive sessions with those long outputs? [1] https://unsloth.ai/blog/deepseekr1-dynamic
I told it to skip all reasoning and explanations and output just the code. It complied, saving a lot of time)
Fascinating. I think Meta is a big winner from this - they still control the content and now mining it for value has been proven even cheaper by DeepSeek.
What kind of quality of content can Meta theoretically mine with all the vast data that they collect? They have a good advertising and a content algorithms but does it approximate general intelligence? And the content you see on Facebook, Instagram, Whatsapp, that's just a lot of average content that might actually be great for AGI.
Earlier quoted context omitted.
One thing you’re missing is that there’s nothing that says the value must correct. There are at least two very good reasons it might not: Nvidia now has huge amounts of money to invest in developing new technologies, exploring other ideas, and the other is that very little of the stock market is about the actual value of the company itself, but speculation. If people think it will go up, they buy it, reducing supply,…
Eventually people will sell their stock to invest in some business that is actually growing or giving proportional dividends. Of course, that "eventually" there is holding a way too much load. And it's very likely this won't happen in a time the US government is printing lots of money and distributing it to rich investors. But that second one has to stop eventually too.