Searching...
19 results for “china ai”
Concept98 mentions across episodes
China
Topic5 mentions across episodes
China
Company2 mentions across episodes
China
...AI market in China is cooking. Yeah. The way I would, you know, position this is in the same week, we have Grok dropping stuff. I think ChatGPT dropped some image stuff that's going viral today. They baked in, basically, like a DALL E type image gene
42:44 / 43:47
from transcript42:44 – 43:47
it wins it by over 10%. I think it had a 15% error rate or something in the next. Like, all the best previous ones have been, like, 25 something percent. Yes. This is like someone breaking the four minute mile. Actually, in some ways, it's more impre
46:27 / 47:34
from transcript46:27 – 47:34
...I say kind of similar to how ChatGPT kicked off a movement in The US where everything had a chatbot. There's now tons of tech companies in China that are releasing very strong frontier open weight models to the point where I would say that DeepSeek i
20:08 / 21:20
from transcript20:08 – 21:20
...China's DeepSeek AI models. So let's lay it out. Nathan, can you describe what DeepSeek v three and DeepSeek r one are, how they work, how they're trained. Let's, look at the big picture, and then we'll zoom in on the details. Yeah. So DeepSeq v thre
13:28 / 14:32
from transcript13:28 – 14:32
selfishly, I'll promote a bunch of, like, Western companies. So both in The US and Europe have these, like, fully open models. So I work at Allen Institute for AI. We've been building Omo, which releases data and code and all of this. And now we have
44:13 / 45:26
from transcript44:13 – 45:26
So I ran the numbers last night, that two week training run that of six days on two GTX five eighties. If you scale, it comes out to just under five minutes Wow. On a single g v 200. Justin is making a really good point. The twenty twelve AlexNet pap
9:00 / 10:14
from transcript9:00 – 10:14
this notion of general purpose computing for their architecture for a long time. In fact, they even thought about should we relaunch our GPUs as GP GPUs, general purpose graphics processing units? And, of course, they decided not to do that, but just
49:15 / 50:32
from transcript49:15 – 50:32
I'm like, Oh, neural nets. I like neural nets. Like I remembered back to my 1990 thesis. I'm like, Oh, that sounds interesting. We should train really, really big neural nets. So that was the Which you say that, and I think it's a very interesting fi
1:05:13 / 1:06:23
from transcript1:05:13 – 1:06:23
And some of DeepSeek's earlier papers, they talk about their training data being distilled for math. I haven't I shouldn't use this word yet, but taken from Common Crawl. And that's a public access that anyone listening to this could go download data
24:41 / 25:54
from transcript24:41 – 25:54
There was a great paper in the New York Times you can go find on how the Google data centers were taking, you know, thousands and thousands of CPUs to run certain deep learning algorithms. And there was just a belief that, like, no way anybody but Go
4:53:54 / 4:55:09
from transcript4:53:54 – 4:55:09
label these images. Yeah. And if I'm remembering from our episode, basically, what happened is the AlexNet team did way better than anybody else had ever done. The complete step changed better. I think the error rate went from mislabeling images 25%
9:00 / 10:01
from transcript9:00 – 10:01
wish you were asked more in, in general? Like, you know, like you, you have such a broad scope, we've covered the hardware, we've covered the, the models research. Yeah. I mean, I think, one thing that's kind of interesting is, you know, I, I did a u
59:15 / 1:00:20
from transcript59:15 – 1:00:20
writes this seminal blog post called the unreasonable effectiveness of neural networks. And, David, I don't think we're gonna go into it on this episode, but note that recurrent neural networks are a little bit of a different thing than convolutional
26:57 / 28:00
from transcript26:57 – 28:00
...across China in the wake of DeepSeek and also a a a deluge of releases Yes. From leading Chinese companies in the tech space. I had been keeping tabs on this but I hadn't actually quite noticed how many big things had dropped. So just to give people
41:58 / 43:04
from transcript41:58 – 43:04
So coming out of the success of Franz Ox's work on Google Translate and the improvements that happened there. In, like, the late two thousands ish, 2007? Yeah. Mid to late two thousands. They keep iterating on Translate. And then once Geoff Hinton co
1:52:23 / 1:53:31
from transcript1:52:23 – 1:53:31
but 2012 was the moment that many people think was the beginning of the deep learning or birth of modern AI because a group of Toronto researchers led by professor Jeff Hinton participated in ImageNet Challenge used the ImageNet Big Data and two GPUs
18:45 / 19:49
from transcript18:45 – 19:49
Have a podcast?
Run your podcast through Clypt
Get ranked clips, hooks, and ready-to-post copy from your own episodes. Free to try.