WEBVTT - Nvidia Dominates CES

0:00:00.800 --> 0:00:05.040
<v Speaker 1>From the heart of where innovation, money and power collide

0:00:05.320 --> 0:00:10.639
<v Speaker 1>in Silicon Valley and beyond. This is Bloomberg Technology with

0:00:10.760 --> 0:00:13.320
<v Speaker 1>Caroline Hyde and Ed Ludlow.

0:00:26.880 --> 0:00:29.480
<v Speaker 2>Live from New York. This is BlueBag Technology coming up.

0:00:29.640 --> 0:00:32.400
<v Speaker 2>In Vidia takes center stage at CES with new hardware,

0:00:32.440 --> 0:00:35.440
<v Speaker 2>software and services. For the stock falls, We sit down

0:00:35.479 --> 0:00:38.360
<v Speaker 2>with it's CEO, Jas and Wang. Later this hour plus,

0:00:38.479 --> 0:00:41.800
<v Speaker 2>Metad cuts its team of fact checkers, saying content moderation

0:00:41.920 --> 0:00:45.240
<v Speaker 2>went quote too far, and the US places some of

0:00:45.360 --> 0:00:49.879
<v Speaker 2>China's biggest tech giants on its military blacklist. But first

0:00:50.159 --> 0:00:52.839
<v Speaker 2>we check in on the key stop. You have to

0:00:52.880 --> 0:00:55.800
<v Speaker 2>watch today. It is the most valuable three point six

0:00:55.840 --> 0:00:58.120
<v Speaker 2>trillion dollars and it dies by four and a half percent,

0:00:58.160 --> 0:01:01.720
<v Speaker 2>in Vidia losing its pre market gains. We had touched

0:01:01.720 --> 0:01:03.440
<v Speaker 2>a new record high and now we sink the worst

0:01:03.480 --> 0:01:06.760
<v Speaker 2>days at September. Why we got a whole raft of products,

0:01:06.800 --> 0:01:09.920
<v Speaker 2>whether it's hardware, software, services, whether it's the future of

0:01:10.360 --> 0:01:14.520
<v Speaker 2>robotics from autonomous driving, from PCs and gaming. But is

0:01:14.560 --> 0:01:18.280
<v Speaker 2>it the here and now, the Blackwell program and platform

0:01:18.360 --> 0:01:20.880
<v Speaker 2>that we don't get enough detail on. Let's head out

0:01:20.920 --> 0:01:23.680
<v Speaker 2>to Las Vegas. Big story, of course, is in Vidio

0:01:23.760 --> 0:01:27.039
<v Speaker 2>and the CEO in veiling these products, we've got none

0:01:27.080 --> 0:01:31.039
<v Speaker 2>other than a key focus on the gamer, the original

0:01:31.080 --> 0:01:33.320
<v Speaker 2>customer from Jensen Wang. Have a listen.

0:01:34.480 --> 0:01:38.720
<v Speaker 3>We used g force to enable artificial intelligence, and now

0:01:38.800 --> 0:01:44.480
<v Speaker 3>artificial intelligence is revolutionizing g Force everyone. Today we're announcing

0:01:45.040 --> 0:01:49.840
<v Speaker 3>our next generation, the RTX Blackwell family.

0:01:51.720 --> 0:01:54.880
<v Speaker 2>Let's go to our gamer in residence, Ed Ludlow on

0:01:54.920 --> 0:01:58.040
<v Speaker 2>the ground at CES. There was going back to their roots,

0:01:58.080 --> 0:02:00.400
<v Speaker 2>the GPU for the gamer. There was so much much more,

0:02:00.400 --> 0:02:01.480
<v Speaker 2>but the stock doesn't like it.

0:02:04.200 --> 0:02:07.600
<v Speaker 4>Yeah, I mean, we've become so accustomed to the nvidious

0:02:07.640 --> 0:02:10.760
<v Speaker 4>story of being dominant in a market for high performance

0:02:10.840 --> 0:02:14.800
<v Speaker 4>GPUs that go into data centers and train AI models,

0:02:14.800 --> 0:02:18.200
<v Speaker 4>foundation models. But it isn't going back to its roots story, right.

0:02:18.240 --> 0:02:21.000
<v Speaker 4>It's taken the work that it's done on those AI

0:02:21.080 --> 0:02:24.320
<v Speaker 4>accelerators and put them into a factor that is more

0:02:24.360 --> 0:02:28.200
<v Speaker 4>relevant to the consumer developers in particular, and maybe the

0:02:28.240 --> 0:02:30.840
<v Speaker 4>market this morning doesn't quite understand that story. There was

0:02:30.880 --> 0:02:34.440
<v Speaker 4>plenty of big picture last night from Jensen One. Blackwell

0:02:34.480 --> 0:02:37.600
<v Speaker 4>is in full production, a kind of timeline or a

0:02:37.680 --> 0:02:41.280
<v Speaker 4>move into physical AI, which we'll get into robotics autonomous driving.

0:02:42.400 --> 0:02:46.160
<v Speaker 4>But maybe also a conversation to be had around diversifying revenue.

0:02:46.280 --> 0:02:49.280
<v Speaker 4>One hundred billion dollars of revenue is still data centerships,

0:02:49.520 --> 0:02:54.519
<v Speaker 4>but increasingly those high performance graphics cards that go into

0:02:54.520 --> 0:02:57.400
<v Speaker 4>the desktop laptops and something called project digits.

0:02:57.440 --> 0:02:58.840
<v Speaker 5>There's a lot to discuss.

0:02:58.880 --> 0:03:01.880
<v Speaker 2>That PC or on the three thousand dollars PC for

0:03:01.960 --> 0:03:04.639
<v Speaker 2>AI developers. And look, there is a macro context to

0:03:04.680 --> 0:03:07.360
<v Speaker 2>the share move today. We of course got some ism data,

0:03:07.400 --> 0:03:09.800
<v Speaker 2>we got yield spiking higher, and no wonder some of

0:03:09.800 --> 0:03:12.400
<v Speaker 2>the big tech names arounder pressure. What's interesting is some

0:03:12.440 --> 0:03:15.480
<v Speaker 2>of the suppliers, for example a Micron still leading the

0:03:15.600 --> 0:03:17.960
<v Speaker 2>charge here as its memory is going to be within

0:03:18.000 --> 0:03:20.519
<v Speaker 2>the Nvidia offering. There was a lot to be had

0:03:20.520 --> 0:03:22.040
<v Speaker 2>for some of the partnerships to play.

0:03:24.960 --> 0:03:27.520
<v Speaker 4>Yeah, and as you know, Caro, it's Bloomberg's policy not

0:03:27.560 --> 0:03:29.240
<v Speaker 4>to give questions an advance of interview, and there are

0:03:29.240 --> 0:03:30.320
<v Speaker 4>all these Nvidia.

0:03:30.040 --> 0:03:31.520
<v Speaker 6>Staff over my shoulder.

0:03:31.760 --> 0:03:35.320
<v Speaker 4>But probably the one big question is the sustainability of

0:03:35.360 --> 0:03:39.840
<v Speaker 4>the investment into AI. Those other chip names are pushing

0:03:39.920 --> 0:03:42.320
<v Speaker 4>higher because when an investment is made into a high

0:03:42.320 --> 0:03:45.600
<v Speaker 4>performance GPU, it goes in combination with the memory chips,

0:03:45.640 --> 0:03:48.960
<v Speaker 4>other CPUs and all of the telematics from those other providers.

0:03:49.200 --> 0:03:52.480
<v Speaker 4>That probably is the key question for Jensen, particularly in

0:03:52.520 --> 0:03:55.240
<v Speaker 4>a year where things run certain economically and I see

0:03:55.240 --> 0:03:57.640
<v Speaker 4>a bit of a concern about inflation in the market

0:03:57.640 --> 0:03:58.240
<v Speaker 4>this morning.

0:03:58.920 --> 0:04:00.400
<v Speaker 5>Jensen has a lot of world views.

0:04:00.440 --> 0:04:02.160
<v Speaker 4>Whether he'll talk to me about inflation, we'll have to

0:04:02.200 --> 0:04:03.000
<v Speaker 4>wait and see.

0:04:03.400 --> 0:04:06.920
<v Speaker 2>At Ludlow, we cannot wait for the conversation. We'll see

0:04:06.960 --> 0:04:10.000
<v Speaker 2>you a little bit later. Let's get an investor's take

0:04:10.080 --> 0:04:13.240
<v Speaker 2>on the here and now lack of visibility on Blackwell

0:04:13.360 --> 0:04:16.240
<v Speaker 2>or or indeed the future proofing of the business model.

0:04:16.320 --> 0:04:19.040
<v Speaker 2>Cebi's Investment Advisory Services CIO Anna Rathbun.

0:04:19.400 --> 0:04:20.280
<v Speaker 7>We tuned in.

0:04:20.240 --> 0:04:22.640
<v Speaker 2>At six thirty Pacific time as nine thirty pm New

0:04:22.720 --> 0:04:26.160
<v Speaker 2>York time. When we heard Jensen on stage, what did

0:04:26.200 --> 0:04:28.440
<v Speaker 2>you make you? In line with some of these analysts

0:04:28.440 --> 0:04:30.640
<v Speaker 2>who are out there saying, look not enough on the

0:04:30.680 --> 0:04:32.360
<v Speaker 2>Blackwell platform.

0:04:33.279 --> 0:04:36.720
<v Speaker 8>Yeah, there weren't that many details that investors were hoping for.

0:04:36.800 --> 0:04:39.159
<v Speaker 8>But if you have to talk for ninety minutes plus

0:04:39.200 --> 0:04:42.080
<v Speaker 8>almost two hours, I'm not sure if you can fill

0:04:42.160 --> 0:04:46.120
<v Speaker 8>those that timeline with a lot of details about the business.

0:04:46.160 --> 0:04:47.680
<v Speaker 7>And plus you know he's kicking.

0:04:47.480 --> 0:04:49.480
<v Speaker 8>Off a conference, you sort of want to be a

0:04:49.560 --> 0:04:53.480
<v Speaker 8>visionary and there is this a little slab of romanticism

0:04:53.520 --> 0:04:56.279
<v Speaker 8>about AI thrown in there, So you know, I think

0:04:56.320 --> 0:04:59.320
<v Speaker 8>maybe it was the wrong platform to really wait for

0:04:59.400 --> 0:05:01.919
<v Speaker 8>some of those details. But I know that the excitement

0:05:02.000 --> 0:05:04.960
<v Speaker 8>was built into the stock market certainly yesterday, so I

0:05:05.240 --> 0:05:08.200
<v Speaker 8>can understand the disappointment. But at the same time, maybe

0:05:08.880 --> 0:05:11.320
<v Speaker 8>maybe we need to wait for the earnings report.

0:05:11.480 --> 0:05:14.440
<v Speaker 2>And we're currently showing a five day shot of in video,

0:05:14.440 --> 0:05:16.960
<v Speaker 2>which gives us the context. It is ratherly eleven percent

0:05:17.000 --> 0:05:19.720
<v Speaker 2>in January. We're maybe seeing some profit being taken off

0:05:19.720 --> 0:05:21.520
<v Speaker 2>the table a little bit, hair Ona, But what did

0:05:21.560 --> 0:05:23.919
<v Speaker 2>you make of what it means for its mote, for

0:05:24.000 --> 0:05:28.440
<v Speaker 2>its ultimate leadership and technology? Does AMD do others even stand.

0:05:28.200 --> 0:05:28.880
<v Speaker 6>A chance here?

0:05:30.040 --> 0:05:32.520
<v Speaker 8>Well, you know, I think the leadership part of the

0:05:32.640 --> 0:05:37.159
<v Speaker 8>visionary talk was establishing that leadership. I think in Video

0:05:37.440 --> 0:05:39.640
<v Speaker 8>wants to be a leader and wants to continue to

0:05:39.680 --> 0:05:42.800
<v Speaker 8>be a leader. So you're taking things like you know,

0:05:42.880 --> 0:05:46.760
<v Speaker 8>self driving cars and partnership with Toyota and media attach.

0:05:46.839 --> 0:05:48.960
<v Speaker 7>Those are real things that can turn into cash flow.

0:05:49.160 --> 0:05:51.640
<v Speaker 8>But then you know, you have visionary things like humanoid

0:05:51.760 --> 0:05:56.120
<v Speaker 8>and chance gypt turning into you know, robotics, right, turning

0:05:56.120 --> 0:05:58.440
<v Speaker 8>into but robotics having a moment, right.

0:05:58.520 --> 0:05:59.840
<v Speaker 7>So these are the things that.

0:05:59.800 --> 0:06:05.279
<v Speaker 8>I think establish Nvidia, at least in rhetoric, as a

0:06:05.320 --> 0:06:09.039
<v Speaker 8>continued leader in the space. And whether or not AMDs

0:06:09.080 --> 0:06:10.760
<v Speaker 8>and other players can catch up.

0:06:11.240 --> 0:06:13.240
<v Speaker 7>I do think that Nvidia still.

0:06:13.080 --> 0:06:17.239
<v Speaker 8>Has a large, you know, sort of monopoly in the space.

0:06:17.520 --> 0:06:20.000
<v Speaker 8>The reality of catching up to Nvidia, I think is

0:06:20.040 --> 0:06:21.400
<v Speaker 8>still a difficult.

0:06:20.920 --> 0:06:24.320
<v Speaker 2>Feat and that's why i'mdphaps leaning into the PC side

0:06:24.360 --> 0:06:26.600
<v Speaker 2>of the business with its Dell announcement rather than trying

0:06:26.600 --> 0:06:29.520
<v Speaker 2>to talk up AI accelerate as in the data center part.

0:06:30.000 --> 0:06:33.160
<v Speaker 2>And to that point, are you feeling comfortable that we're

0:06:33.200 --> 0:06:37.000
<v Speaker 2>seeing diversification or at least that the demand for infrastructure

0:06:37.000 --> 0:06:39.719
<v Speaker 2>from a data center perspective is still there. Microsoft eighty

0:06:39.800 --> 0:06:41.640
<v Speaker 2>billion that they're going to spend in twenty twenty five

0:06:41.680 --> 0:06:42.680
<v Speaker 2>seems to signal.

0:06:42.400 --> 0:06:47.400
<v Speaker 8>That, yeah, I don't think this AI venture is going

0:06:47.440 --> 0:06:50.159
<v Speaker 8>to stop. And certainly, you know, likes of Microsoft and

0:06:50.200 --> 0:06:53.520
<v Speaker 8>some of these big companies are putting big dollars into it.

0:06:53.720 --> 0:06:56.320
<v Speaker 8>But we can't forget about the private side either. There

0:06:56.320 --> 0:06:58.440
<v Speaker 8>are a lot of venture companies that are going to

0:06:58.480 --> 0:07:02.400
<v Speaker 8>require as their develop of being applications for real businesses

0:07:02.400 --> 0:07:05.359
<v Speaker 8>on the user end, they're going to have to lean

0:07:05.520 --> 0:07:09.520
<v Speaker 8>into some of that infrastructure like data centers and ultimately

0:07:09.640 --> 0:07:12.600
<v Speaker 8>energy too. But that's not a conversation for today. So

0:07:12.680 --> 0:07:14.640
<v Speaker 8>I think I think the demand for this is not

0:07:14.720 --> 0:07:15.560
<v Speaker 8>going to ease up.

0:07:16.280 --> 0:07:18.920
<v Speaker 2>Can you give us your macro perspective here though as well?

0:07:19.200 --> 0:07:22.040
<v Speaker 2>We were just talking how on a macro perspective, maybe

0:07:22.080 --> 0:07:24.560
<v Speaker 2>we are seeing that feed into the stock weakness. Do

0:07:24.600 --> 0:07:28.120
<v Speaker 2>you still your mind be thinking about inflationary pressure? Does

0:07:28.160 --> 0:07:30.600
<v Speaker 2>that have any weight on a company that just is

0:07:30.640 --> 0:07:34.720
<v Speaker 2>so in focus for its innovation, not just the macro headwinds.

0:07:35.880 --> 0:07:36.240
<v Speaker 7>Yeah.

0:07:36.280 --> 0:07:38.760
<v Speaker 8>So you know, when I think about macro headwinds such

0:07:38.760 --> 0:07:41.520
<v Speaker 8>as inflation and prices, I think about who has the

0:07:41.560 --> 0:07:42.280
<v Speaker 8>pricing power?

0:07:42.480 --> 0:07:44.720
<v Speaker 7>Right, So if you're in Nvidia and you.

0:07:44.840 --> 0:07:49.040
<v Speaker 8>Have dominance in a marketplace, you have the pricing power.

0:07:49.120 --> 0:07:52.120
<v Speaker 8>So I don't think Nvidia is necessarily, you know, thinking

0:07:52.760 --> 0:07:55.600
<v Speaker 8>about whether or not they have they can pass down

0:07:56.280 --> 0:07:58.920
<v Speaker 8>any kind of inflation or price increases.

0:07:59.600 --> 0:08:01.400
<v Speaker 7>I think it's really on the consumer end.

0:08:01.440 --> 0:08:04.000
<v Speaker 8>I think it's on the end of the other max

0:08:04.040 --> 0:08:07.280
<v Speaker 8>evens that need to develop the R and D aspect

0:08:07.320 --> 0:08:09.080
<v Speaker 8>of it, they're going to have to worry about the

0:08:09.080 --> 0:08:09.920
<v Speaker 8>inflation portion.

0:08:10.560 --> 0:08:13.520
<v Speaker 2>Interesting. So for you, if you're looking at portfolio at

0:08:13.520 --> 0:08:15.360
<v Speaker 2>the start of twenty twenty five and you're thinking how

0:08:15.360 --> 0:08:17.000
<v Speaker 2>you're going to be changing up, do you make any

0:08:17.040 --> 0:08:19.840
<v Speaker 2>key decisions of the back of the latest inflation three points?

0:08:20.680 --> 0:08:23.560
<v Speaker 8>You know, our assumption has always been that it's going

0:08:23.600 --> 0:08:26.800
<v Speaker 8>to be higher for longer, both on the inflation area

0:08:26.920 --> 0:08:28.960
<v Speaker 8>as well as rates.

0:08:28.600 --> 0:08:30.600
<v Speaker 7>So our portfolio is already positioned for it.

0:08:30.680 --> 0:08:33.360
<v Speaker 8>We never really thought that inflation could go down to

0:08:33.400 --> 0:08:35.319
<v Speaker 8>two percent it you know, I'm going to use the

0:08:35.320 --> 0:08:38.720
<v Speaker 8>word safely without causing some kind of a slowdown in

0:08:38.760 --> 0:08:42.240
<v Speaker 8>the economy. And frankly, even if it does, it's really

0:08:42.240 --> 0:08:45.160
<v Speaker 8>in the cyclical areas of the economy, and I think

0:08:45.240 --> 0:08:49.360
<v Speaker 8>tech has some immunity from that cyclicality simply because there's

0:08:49.400 --> 0:08:52.920
<v Speaker 8>just way too much momentum going into this AI venture.

0:08:53.240 --> 0:08:56.400
<v Speaker 2>And raf fan so good to have you, CIOC BIZ

0:08:56.559 --> 0:09:00.439
<v Speaker 2>Investment Advisory Services stay well. Coming up, META looks back

0:09:00.440 --> 0:09:03.520
<v Speaker 2>on fact checking across its social media platforms or on that.

0:09:03.600 --> 0:09:30.120
<v Speaker 2>Next this is Bloomberg Technology. Yet more big news auta Meta.

0:09:30.240 --> 0:09:33.520
<v Speaker 2>Today it's announced that it will end third party fact

0:09:33.559 --> 0:09:36.360
<v Speaker 2>checking on its social media platforms, saying content moderation went

0:09:36.440 --> 0:09:39.200
<v Speaker 2>too far, so instead, the company will allow users to

0:09:39.240 --> 0:09:42.079
<v Speaker 2>comment on our post's accuracy, adopting a kind of format

0:09:42.120 --> 0:09:44.840
<v Speaker 2>similar to excess community notes. This, of course follows some

0:09:45.080 --> 0:09:48.360
<v Speaker 2>notable changes to its board as well. For more bloombergs

0:09:48.400 --> 0:09:51.880
<v Speaker 2>Lin Duine joins us for more fascinating moves. It's moving

0:09:51.880 --> 0:09:54.800
<v Speaker 2>safety and content moderation teams from California to Texas and

0:09:54.880 --> 0:09:56.800
<v Speaker 2>ultimately relying on its userbase now.

0:09:57.040 --> 0:10:00.400
<v Speaker 9>It is very fascinating and obviously, as you said, further

0:10:00.520 --> 0:10:05.280
<v Speaker 9>aligns Meta to the x approach of content moderation and

0:10:05.320 --> 0:10:09.320
<v Speaker 9>fact checking, which is to sort of crowdsource. One cannot

0:10:09.559 --> 0:10:14.080
<v Speaker 9>ignore the underlying trend here. It's clear that over the

0:10:14.120 --> 0:10:19.040
<v Speaker 9>past several months, Mark Zuckerberg has tried to repair, if

0:10:19.080 --> 0:10:23.520
<v Speaker 9>you will, his relationship with president incoming President Donald Trump.

0:10:23.960 --> 0:10:26.880
<v Speaker 9>That has been a relationship that has in the past

0:10:27.000 --> 0:10:30.560
<v Speaker 9>been fairly adversarial. I don't need to remind you that

0:10:30.600 --> 0:10:35.720
<v Speaker 9>at one point Trump was suspended on Meta's platforms briefly

0:10:35.800 --> 0:10:39.800
<v Speaker 9>and has since been reinstated, and since then Zuckerberg has

0:10:39.840 --> 0:10:45.720
<v Speaker 9>personally contributed to his inauguration fund and has made efforts,

0:10:45.800 --> 0:10:48.679
<v Speaker 9>as you have said, to make appointments that would be

0:10:48.960 --> 0:10:52.760
<v Speaker 9>more conducive to a relationship with the Trump administration coming in.

0:10:52.880 --> 0:10:54.720
<v Speaker 2>I mean, it's interesting that this is basically the first

0:10:54.720 --> 0:10:58.160
<v Speaker 2>announcement coming from Jill Kaplan, who is now the chief

0:10:58.160 --> 0:11:03.080
<v Speaker 2>Global Affairs Officer as Clegg moved stage right. And we're

0:11:03.080 --> 0:11:06.280
<v Speaker 2>not only getting a more right leaning global affairs officer,

0:11:06.320 --> 0:11:09.480
<v Speaker 2>but we're also getting Dana White UFC coming to the

0:11:09.480 --> 0:11:10.360
<v Speaker 2>board of MATA.

0:11:10.400 --> 0:11:13.760
<v Speaker 9>That's right, and as you probably recall, Dana White was

0:11:13.800 --> 0:11:17.520
<v Speaker 9>a very vocal Trump supporter. He was actually featured in

0:11:17.600 --> 0:11:21.560
<v Speaker 9>Trump's first video on TikTok, another social media platform. So

0:11:22.840 --> 0:11:25.800
<v Speaker 9>a good fit because you know, Dana White and Mark

0:11:25.880 --> 0:11:29.000
<v Speaker 9>Zuckerberg have a pre existing relationship.

0:11:29.040 --> 0:11:30.720
<v Speaker 5>They for a long time.

0:11:30.840 --> 0:11:34.680
<v Speaker 9>Right, They share a very strong interest in the same sport.

0:11:35.480 --> 0:11:37.920
<v Speaker 9>So in a lot of ways, it makes sense for

0:11:38.040 --> 0:11:40.520
<v Speaker 9>Mark and for what he's trying to do with that

0:11:40.600 --> 0:11:44.480
<v Speaker 9>relationship with Trump. I think that migration for the fact

0:11:44.520 --> 0:11:48.880
<v Speaker 9>checking team from California to Texas, as you mentioned, is

0:11:48.880 --> 0:11:50.120
<v Speaker 9>a really interesting one.

0:11:50.200 --> 0:11:50.440
<v Speaker 6>Two.

0:11:50.880 --> 0:11:54.560
<v Speaker 9>Whether that's politically motivated, I don't personally know, And I

0:11:54.600 --> 0:11:56.880
<v Speaker 9>think that the bigger question is, like will there be

0:11:56.920 --> 0:11:59.559
<v Speaker 9>a broader pullback at Meta and some of the other

0:11:59.600 --> 0:12:03.480
<v Speaker 9>social media platforms with this incoming Trump administration across all

0:12:03.600 --> 0:12:06.880
<v Speaker 9>content moderation, not necessarily just fact checking, but you know,

0:12:07.520 --> 0:12:09.600
<v Speaker 9>will there be a pullback in like the monitoring of

0:12:09.760 --> 0:12:14.840
<v Speaker 9>disturbing content CSM content. I think that's the broader question.

0:12:14.640 --> 0:12:17.120
<v Speaker 2>Ask and that leads us perfectly to our next conversation.

0:12:17.200 --> 0:12:19.760
<v Speaker 2>It's as if she knows it by Magic lindaan on

0:12:19.920 --> 0:12:22.400
<v Speaker 2>All Things Meta and look as metapairs back its fact

0:12:22.480 --> 0:12:25.280
<v Speaker 2>check is there's exclusive reporting today showing that the FBI

0:12:25.320 --> 0:12:28.120
<v Speaker 2>and the Department of Homeland Security will also scale back

0:12:28.120 --> 0:12:30.920
<v Speaker 2>efforts over the past two years to disrupt violent extremists

0:12:30.960 --> 0:12:34.400
<v Speaker 2>online activities. That's according to current and former US officials

0:12:34.440 --> 0:12:37.439
<v Speaker 2>and internet radicalization experts actually fear that the trend will

0:12:37.480 --> 0:12:41.040
<v Speaker 2>accelerate under Trump. That reporting all comes fro bloombergs Jeffstone

0:12:41.280 --> 0:12:45.320
<v Speaker 2>extraordinary timing. So you're working out here that basically demotment

0:12:45.360 --> 0:12:49.520
<v Speaker 2>of honand security and indeed other agencies have not wanted

0:12:49.520 --> 0:12:52.000
<v Speaker 2>to tip off the social media companies as much.

0:12:53.040 --> 0:12:56.560
<v Speaker 10>That's right, Caroline, since the social media companies Meta being

0:12:56.600 --> 0:13:00.200
<v Speaker 10>one of them, formerly Twitter being one of them, had

0:13:00.200 --> 0:13:04.520
<v Speaker 10>a long history now of flagging false COVID related information

0:13:05.320 --> 0:13:07.800
<v Speaker 10>misinformation around the vaccines. There's been a lot of political

0:13:07.800 --> 0:13:12.400
<v Speaker 10>pushback in Washington encouraging them not to do that. The

0:13:12.400 --> 0:13:14.680
<v Speaker 10>Department of Homeland Security and the FBI have kind of

0:13:14.679 --> 0:13:17.400
<v Speaker 10>become collateral damage in that larger political fight, and they

0:13:17.440 --> 0:13:22.400
<v Speaker 10>are now incredibly reluctant and prevented in some cases by

0:13:22.480 --> 0:13:26.720
<v Speaker 10>law from dealing with social media firms to remove some

0:13:26.840 --> 0:13:30.000
<v Speaker 10>of this really incendiary violent misinformation.

0:13:30.840 --> 0:13:35.319
<v Speaker 2>Yet this comes but days after the New Year's Day

0:13:35.360 --> 0:13:38.720
<v Speaker 2>attack on a Trump hotel, what happened in New Orleans.

0:13:38.840 --> 0:13:42.360
<v Speaker 2>I mean, the idea that extremist content isn't going to

0:13:42.400 --> 0:13:47.480
<v Speaker 2>be as monitored feels pretty scary, that's right.

0:13:47.520 --> 0:13:49.720
<v Speaker 10>I mean, there's certainly a trend where we are seeing

0:13:49.720 --> 0:13:52.760
<v Speaker 10>these extremist attacks in almost every case. You know, it

0:13:52.840 --> 0:13:54.600
<v Speaker 10>used to be a small fraction, but now in almost

0:13:54.640 --> 0:13:58.280
<v Speaker 10>every case, the perpetrators or suspected perpetrators who are carrying

0:13:58.280 --> 0:14:01.160
<v Speaker 10>out if these attacks are often taught talking about them online,

0:14:01.200 --> 0:14:03.480
<v Speaker 10>whether it be on sites like Meta in this case

0:14:03.960 --> 0:14:06.920
<v Speaker 10>or on Discord as in the case with the mass

0:14:06.920 --> 0:14:10.120
<v Speaker 10>shooter in Buffalo. The FBI and DHS are still going

0:14:10.160 --> 0:14:13.080
<v Speaker 10>to continue to investigate based on some of those leads

0:14:13.080 --> 0:14:15.080
<v Speaker 10>that they're seeing online, but they're not going to interact

0:14:15.120 --> 0:14:19.040
<v Speaker 10>with social media companies to remove that content proactively. They're

0:14:19.040 --> 0:14:21.600
<v Speaker 10>going to leave that to the social media firms.

0:14:21.880 --> 0:14:24.400
<v Speaker 2>What's interesting within your story, you really build out the

0:14:24.440 --> 0:14:28.280
<v Speaker 2>case that since so key court ruling here that as

0:14:28.280 --> 0:14:31.240
<v Speaker 2>you articulated, has pushed the DHS and the FBI, it's

0:14:31.600 --> 0:14:34.280
<v Speaker 2>less weighed in, but also there's less money going to

0:14:34.440 --> 0:14:36.760
<v Speaker 2>external third party researchers as well.

0:14:36.800 --> 0:14:38.640
<v Speaker 5>Right, that's right.

0:14:38.880 --> 0:14:42.000
<v Speaker 10>External third party researchers are a key component in this

0:14:42.080 --> 0:14:44.240
<v Speaker 10>larger ecosystem. They do a lot of the work on

0:14:44.320 --> 0:14:47.600
<v Speaker 10>behalf of social media companies to kind of proactively flag

0:14:47.800 --> 0:14:52.320
<v Speaker 10>and detect and remove some of this really again extremist,

0:14:52.440 --> 0:14:55.840
<v Speaker 10>violent material. It goes beyond a lot of misinformation. They

0:14:55.880 --> 0:14:59.360
<v Speaker 10>are now taking significantly less funding from the government because

0:15:00.360 --> 0:15:05.480
<v Speaker 10>of some of the political connotations and disputes around this. Specifically,

0:15:06.080 --> 0:15:08.240
<v Speaker 10>some of these researchers used to receive tens of millions

0:15:08.280 --> 0:15:11.120
<v Speaker 10>of dollars, whereas last year the Department of Homeland Security

0:15:11.160 --> 0:15:15.960
<v Speaker 10>gave them zero dollars for Internet related research exclusive reporting.

0:15:16.080 --> 0:15:18.120
<v Speaker 2>Jeff Stone, we thank you so much.

0:15:26.200 --> 0:15:26.400
<v Speaker 6>Now.

0:15:26.520 --> 0:15:30.480
<v Speaker 2>The US has just added China's most valuable company, Tencent,

0:15:30.640 --> 0:15:33.680
<v Speaker 2>as well as Tesla battery supply Caatl to its list

0:15:33.680 --> 0:15:37.280
<v Speaker 2>of quote Chinese military companies. It's a move which could

0:15:37.280 --> 0:15:39.640
<v Speaker 2>accelerate decoupling of the world's biggest economies.

0:15:39.760 --> 0:15:40.040
<v Speaker 5>For more.

0:15:40.040 --> 0:15:41.760
<v Speaker 2>We got to get out to Bloombergs Mike Shepard, and

0:15:41.760 --> 0:15:44.560
<v Speaker 2>this seemed to come out of nowhere yesterday, and even

0:15:44.600 --> 0:15:46.880
<v Speaker 2>Tencent itself thought it might be a mistake.

0:15:48.280 --> 0:15:50.440
<v Speaker 11>And a lot of the other companies thought so too.

0:15:50.520 --> 0:15:52.520
<v Speaker 11>They were really caught off guard, as a lot of

0:15:52.560 --> 0:15:55.840
<v Speaker 11>us here in Washington were. It showed up without fanfare

0:15:55.920 --> 0:15:59.440
<v Speaker 11>in a publication known as the Federal Register. They Pentagon

0:15:59.560 --> 0:16:02.040
<v Speaker 11>just dropped it in there. But there was a purpose

0:16:02.120 --> 0:16:05.400
<v Speaker 11>to this. It dates back to twenty twenty when President

0:16:05.440 --> 0:16:08.720
<v Speaker 11>Donald Trump was still in office and under an executive

0:16:08.880 --> 0:16:12.360
<v Speaker 11>order from him, he asked the Defense Department to look

0:16:12.400 --> 0:16:16.760
<v Speaker 11>for companies in China that had ties or actual control

0:16:16.920 --> 0:16:20.280
<v Speaker 11>by China's military. And the concern is that this nexus

0:16:20.360 --> 0:16:24.920
<v Speaker 11>between China's military and the business community in China post

0:16:25.040 --> 0:16:28.640
<v Speaker 11>a risk to global security, and as a result, we

0:16:28.680 --> 0:16:31.080
<v Speaker 11>see this listening now. It's important to note that there

0:16:31.200 --> 0:16:35.280
<v Speaker 11>is no sanction attached to it. However, there is something

0:16:35.320 --> 0:16:39.160
<v Speaker 11>that would discourage US firms from doing business with anyone

0:16:39.160 --> 0:16:40.200
<v Speaker 11>who is on this list.

0:16:40.560 --> 0:16:42.280
<v Speaker 2>I think there was initially a worry that they might

0:16:42.280 --> 0:16:45.440
<v Speaker 2>be forced to delist as well, Mike, is there any

0:16:45.560 --> 0:16:48.840
<v Speaker 2>reality in that. Certainly ten Cent came out with a

0:16:48.880 --> 0:16:51.120
<v Speaker 2>statement saying this doesn't affect our business in the here

0:16:51.200 --> 0:16:51.600
<v Speaker 2>and now.

0:16:53.000 --> 0:16:55.560
<v Speaker 11>That is a really good question, Caroly, and it was

0:16:55.640 --> 0:16:58.440
<v Speaker 11>one that we were chasing down yesterday. And as far

0:16:58.480 --> 0:17:01.400
<v Speaker 11>as we can tell, there is no immediate threat of

0:17:01.480 --> 0:17:04.719
<v Speaker 11>the listening directly as a result of this, but it

0:17:04.760 --> 0:17:09.320
<v Speaker 11>does prompt eight companies here, as I said, to revisit

0:17:09.359 --> 0:17:13.440
<v Speaker 11>and rethink relationships with ten Cent and others. And there

0:17:13.480 --> 0:17:16.840
<v Speaker 11>are now one hundred and twenty one Chinese companies on

0:17:16.920 --> 0:17:21.639
<v Speaker 11>this Pentagon list that subject them to some additional scrutiny

0:17:21.720 --> 0:17:26.040
<v Speaker 11>about their relationship with China's military, and the list also

0:17:26.080 --> 0:17:29.680
<v Speaker 11>includes CATL, which is one of the world's biggest battery makers.

0:17:29.960 --> 0:17:34.760
<v Speaker 11>They're critical to powering the electric vehicle revolution globally, but

0:17:34.800 --> 0:17:37.520
<v Speaker 11>they've also been the focus of a lot of attention,

0:17:37.800 --> 0:17:42.199
<v Speaker 11>especially from Congress, including incoming Secretary of State Mark or Rubio,

0:17:42.359 --> 0:17:44.800
<v Speaker 11>who when he was in the Senate had really zeroed

0:17:44.840 --> 0:17:48.240
<v Speaker 11>in on CATL as a potential national security risk through

0:17:48.240 --> 0:17:48.680
<v Speaker 11>the US.

0:17:48.880 --> 0:17:51.200
<v Speaker 2>So read those tea leaves for US, Mike, because on

0:17:51.200 --> 0:17:53.720
<v Speaker 2>one side, my immediate reaction is, well, Coatl is a

0:17:53.720 --> 0:17:56.479
<v Speaker 2>big supply to Tesla, and you know, Mosk isn't going

0:17:56.520 --> 0:17:59.439
<v Speaker 2>to like that. He has the air of Trump for

0:17:59.480 --> 0:18:03.159
<v Speaker 2>all some purposes, we understand. But then Marco Rubio coming in,

0:18:03.359 --> 0:18:05.560
<v Speaker 2>he doesn't much like c ATL. So what does the

0:18:05.600 --> 0:18:07.040
<v Speaker 2>next administration mean for this?

0:18:08.280 --> 0:18:10.960
<v Speaker 11>Well, the tea leaves really are tough to read here, Caroline,

0:18:11.040 --> 0:18:13.720
<v Speaker 11>and that's in part because we do see this kind.

0:18:13.600 --> 0:18:15.280
<v Speaker 10>Of division of loyalties.

0:18:15.640 --> 0:18:18.320
<v Speaker 11>This is also playing out with the TikTok band that

0:18:18.400 --> 0:18:21.960
<v Speaker 11>is about to go into effect on January nineteenth unless

0:18:22.000 --> 0:18:25.000
<v Speaker 11>the company, which has a Chinese parent, is able to

0:18:25.240 --> 0:18:29.480
<v Speaker 11>persuade the Supreme Court to overturn this devest or ban law.

0:18:29.760 --> 0:18:32.639
<v Speaker 11>Now President of Like Donald Trump, has made clear he

0:18:32.720 --> 0:18:36.520
<v Speaker 11>would like to see TikTok survive as its own entity,

0:18:36.760 --> 0:18:39.639
<v Speaker 11>to exist on its own and not be banned to

0:18:39.760 --> 0:18:43.439
<v Speaker 11>somehow evade this ultimate sanction as a result of a

0:18:43.480 --> 0:18:47.200
<v Speaker 11>new law. One also that Marco Rubio have helped push

0:18:47.240 --> 0:18:50.040
<v Speaker 11>through so we will see these kinds of tensions playing

0:18:50.040 --> 0:18:53.159
<v Speaker 11>out between some of the business interests, including from Elon

0:18:53.240 --> 0:18:56.840
<v Speaker 11>Musk and others, playing out right there in the White House.

0:18:56.920 --> 0:18:59.920
<v Speaker 11>They're trying to get tough on China, yet need also

0:19:00.119 --> 0:19:02.480
<v Speaker 11>this connection to China for business themselves.

0:19:02.640 --> 0:19:04.359
<v Speaker 2>And as shocking is it might be for many that

0:19:04.480 --> 0:19:07.720
<v Speaker 2>a big gaming and social network platform Tencent might be

0:19:07.760 --> 0:19:11.399
<v Speaker 2>in any way deemed a military related organization. There is

0:19:11.440 --> 0:19:15.520
<v Speaker 2>also past tens We've seen Shaomi, for example, as a phonemaker,

0:19:15.520 --> 0:19:18.240
<v Speaker 2>but also now a carmaker is managed to get itself

0:19:18.280 --> 0:19:20.800
<v Speaker 2>off the list, right, So there's potential hope.

0:19:20.640 --> 0:19:24.680
<v Speaker 11>Here, Yes, there is. These companies do have a way

0:19:24.800 --> 0:19:27.800
<v Speaker 11>to appeal this. They can even take it to court

0:19:27.840 --> 0:19:30.879
<v Speaker 11>and contest the listing, and a number of companies like

0:19:30.920 --> 0:19:33.800
<v Speaker 11>you just mentioned show Me especially, have been able to

0:19:33.840 --> 0:19:37.640
<v Speaker 11>get themselves removed and struck from the list. However, when

0:19:37.640 --> 0:19:40.280
<v Speaker 11>we talk about ten Cent, let's turn back the clock

0:19:40.320 --> 0:19:44.119
<v Speaker 11>again to twenty twenty, when Donald Trump as president initially

0:19:44.160 --> 0:19:47.320
<v Speaker 11>tried to ban TikTok, he also went after ten cents

0:19:47.400 --> 0:19:51.639
<v Speaker 11>we chat app, which is this all inclusive communications and

0:19:51.760 --> 0:19:54.919
<v Speaker 11>payments platform that the US government had viewed as a

0:19:55.000 --> 0:19:58.639
<v Speaker 11>risk in the US by anybody who was using it

0:19:58.680 --> 0:20:01.600
<v Speaker 11>here now there is not nearly as much use of

0:20:01.720 --> 0:20:05.400
<v Speaker 11>we chair here as by TikTok, but we do see

0:20:05.440 --> 0:20:10.200
<v Speaker 11>the same sort of argument being played out with tencent

0:20:10.760 --> 0:20:13.520
<v Speaker 11>back then and perhaps the echo of it now as well.

0:20:13.960 --> 0:20:16.640
<v Speaker 2>Mike Shepard, we thank you so much on all things

0:20:16.800 --> 0:20:19.600
<v Speaker 2>US China. We can go just onto data centers now

0:20:19.680 --> 0:20:21.359
<v Speaker 2>and get back to see yes where our own ed

0:20:21.440 --> 0:20:23.920
<v Speaker 2>Ludlow is standing by with a very special guest who's

0:20:23.920 --> 0:20:26.120
<v Speaker 2>probably going to benefit from that spending in the United States.

0:20:26.320 --> 0:20:31.840
<v Speaker 4>D Yeah, Good morning from Las Vegas, Caroline and Jensen one,

0:20:32.080 --> 0:20:35.320
<v Speaker 4>CEO of Nvidia, fresh off his keynote last night on stage.

0:20:35.760 --> 0:20:36.320
<v Speaker 6>Great to see you.

0:20:36.720 --> 0:20:39.760
<v Speaker 5>Welcome to Las Vegas. Thank you every new year, familiar territory.

0:20:39.880 --> 0:20:41.439
<v Speaker 6>Congratulations on your new baby.

0:20:41.560 --> 0:20:42.480
<v Speaker 5>Thank you very much.

0:20:43.040 --> 0:20:45.480
<v Speaker 4>A lot's changed since we last spoke, actually, but of

0:20:45.600 --> 0:20:48.360
<v Speaker 4>the broad spectrum of what you announced last night. New

0:20:48.359 --> 0:20:52.879
<v Speaker 4>graphics cards, new chips actually technically in the automotive space,

0:20:52.960 --> 0:20:56.199
<v Speaker 4>products and services on the software side, which is the

0:20:56.240 --> 0:20:59.520
<v Speaker 4>single most important for in video's future.

0:20:59.600 --> 0:21:00.919
<v Speaker 6>There are important you know.

0:21:00.960 --> 0:21:01.600
<v Speaker 5>It's hard.

0:21:01.760 --> 0:21:05.760
<v Speaker 6>It's it's hard. It's hard. It's hard to pick your favorites.

0:21:05.880 --> 0:21:09.280
<v Speaker 12>You know, we announced three chips, We announced a brand

0:21:09.359 --> 0:21:13.720
<v Speaker 12>new AI, a world foundation model and first of its kind,

0:21:14.200 --> 0:21:18.840
<v Speaker 12>and we announced our work in three areas and robotics, right,

0:21:19.200 --> 0:21:22.760
<v Speaker 12>and they're all important. And you know the thing of course,

0:21:23.280 --> 0:21:28.159
<v Speaker 12>we announced a brand new Blackwell RTX and has a

0:21:28.200 --> 0:21:36.240
<v Speaker 12>new AI technology called neuro Neuroshaters, and we're combining artificial

0:21:36.280 --> 0:21:38.920
<v Speaker 12>intelligence and classical.

0:21:38.440 --> 0:21:40.880
<v Speaker 5>Computers RTX and U RTX.

0:21:40.960 --> 0:21:44.560
<v Speaker 4>Yeah, we've become so accustomed to your story being dominant

0:21:44.600 --> 0:21:48.280
<v Speaker 4>in a market for high performance GPUs server ACS data centers.

0:21:48.560 --> 0:21:51.000
<v Speaker 4>This is going back to your roots. It's in the

0:21:51.000 --> 0:21:54.520
<v Speaker 4>desktop context. There's one right there later in the year laptop.

0:21:54.920 --> 0:21:59.000
<v Speaker 4>But for a target base that is developers, a nerdy,

0:21:59.080 --> 0:22:01.879
<v Speaker 4>hardcore gamment, what's the future of that business for you?

0:22:03.720 --> 0:22:05.800
<v Speaker 6>Computer graphics is going to be here forever, forever.

0:22:06.160 --> 0:22:10.720
<v Speaker 12>And what we've done is we've fused artificial intelligence and

0:22:10.760 --> 0:22:14.960
<v Speaker 12>computer graphics together. And the images that we're generating today

0:22:15.000 --> 0:22:17.160
<v Speaker 12>wouldn't be possible if not for the fact that we're

0:22:17.240 --> 0:22:21.560
<v Speaker 12>using computer graphics to create beautiful pixels and then use

0:22:21.640 --> 0:22:27.840
<v Speaker 12>artificial intelligence to amplify that capability. For example, out of

0:22:27.840 --> 0:22:30.600
<v Speaker 12>four frames I was talking about yesterday, thirty three million

0:22:30.600 --> 0:22:31.800
<v Speaker 12>pixels to SEW and four K.

0:22:32.400 --> 0:22:33.800
<v Speaker 6>Now. Out of that thirty three.

0:22:33.600 --> 0:22:37.359
<v Speaker 12>Million pixels, two million pixels were computed. The other thirty

0:22:37.359 --> 0:22:39.680
<v Speaker 12>one million pixels were generated by AI.

0:22:39.880 --> 0:22:42.159
<v Speaker 5>In other words, the AI predicts what it thinks the

0:22:42.200 --> 0:22:42.800
<v Speaker 5>pixtels should.

0:22:42.920 --> 0:22:45.200
<v Speaker 6>That's right, Yeah, the ultimate generator of AI.

0:22:45.280 --> 0:22:47.800
<v Speaker 4>But what was interesting for me is the focus again

0:22:47.960 --> 0:22:51.159
<v Speaker 4>was away from the graphics cards, away from blackweld, undependents

0:22:51.160 --> 0:22:55.119
<v Speaker 4>and Cosmos. We probably don't have time to explain in

0:22:55.160 --> 0:22:57.880
<v Speaker 4>full detailed Cosmos, but you call it a world foundation.

0:22:58.040 --> 0:23:02.240
<v Speaker 12>Model Cosmos is for or the physical world what chat

0:23:02.480 --> 0:23:05.160
<v Speaker 12>BT is for words and text.

0:23:05.800 --> 0:23:07.359
<v Speaker 6>That's the easiest way to think about that.

0:23:07.560 --> 0:23:12.760
<v Speaker 4>Okay, so model text input, but can generate synthetic data

0:23:12.800 --> 0:23:13.840
<v Speaker 4>in multiple mediums.

0:23:14.640 --> 0:23:17.960
<v Speaker 12>It understands the physical world. So for example, if I

0:23:18.040 --> 0:23:21.120
<v Speaker 12>ask it a question, if I ask it to generate

0:23:21.520 --> 0:23:25.919
<v Speaker 12>multiple futures of a car driving down the road, it

0:23:25.960 --> 0:23:29.800
<v Speaker 12>would understand the dynamics of the world, it would understand

0:23:30.440 --> 0:23:34.440
<v Speaker 12>the object permanence, it would under understand geometry and space,

0:23:34.800 --> 0:23:38.840
<v Speaker 12>and it would create a driving scenario for the car

0:23:39.560 --> 0:23:41.520
<v Speaker 12>that is plausible and so.

0:23:41.800 --> 0:23:42.960
<v Speaker 5>And you open sourced it.

0:23:43.080 --> 0:23:45.120
<v Speaker 4>So I don't really think about it as a product

0:23:45.200 --> 0:23:48.920
<v Speaker 4>or a go to market. It's more about what Cosmos enables.

0:23:49.760 --> 0:23:51.159
<v Speaker 4>Is that how we should think about it.

0:23:51.280 --> 0:23:55.320
<v Speaker 12>Yeah, Well, the automous vehicle industry and the rovice industry

0:23:55.359 --> 0:23:57.080
<v Speaker 12>is a really point to us, and we offer three

0:23:57.080 --> 0:23:58.000
<v Speaker 12>computers for them.

0:23:58.359 --> 0:24:01.240
<v Speaker 6>We offer it, of course, the training computer through DGX.

0:24:01.280 --> 0:24:04.440
<v Speaker 12>Through DGX, the robotics computer that's inside the car or

0:24:04.520 --> 0:24:07.080
<v Speaker 12>inside of robot and now we have this new computer

0:24:07.160 --> 0:24:12.760
<v Speaker 12>call Omniverse with Cosmos that is the digital twin or

0:24:12.800 --> 0:24:15.760
<v Speaker 12>the playground where these robots can learn how to be robots.

0:24:16.119 --> 0:24:18.879
<v Speaker 12>And so if we could accelerate the development of an

0:24:18.960 --> 0:24:22.959
<v Speaker 12>artificial intelligence for avs and for robotics, it brings in

0:24:23.000 --> 0:24:23.800
<v Speaker 12>a lot of business for us.

0:24:23.840 --> 0:24:26.280
<v Speaker 4>Of course, there's an academic point of tension here. If

0:24:26.320 --> 0:24:29.520
<v Speaker 4>I may Elon Musk is a customer of yours and Tesla,

0:24:30.720 --> 0:24:34.280
<v Speaker 4>their theory or practice is based on real world data.

0:24:34.080 --> 0:24:35.040
<v Speaker 5>Gathered through vision.

0:24:35.280 --> 0:24:39.520
<v Speaker 4>Yeah, does the synthetic data underpinning of Cosmos kind of

0:24:39.720 --> 0:24:41.000
<v Speaker 4>contradict that it.

0:24:40.960 --> 0:24:43.879
<v Speaker 12>Doesn't replace it augments, And so you're going to you

0:24:43.880 --> 0:24:46.320
<v Speaker 12>should collect as much world data as you can. Of course,

0:24:46.320 --> 0:24:49.320
<v Speaker 12>collecting world data is very expensive and Elon has a

0:24:49.320 --> 0:24:53.240
<v Speaker 12>great advantage because the number one his AI factory for

0:24:53.320 --> 0:24:56.280
<v Speaker 12>his cars is fantastic, has a lot of video gear

0:24:56.320 --> 0:25:01.400
<v Speaker 12>in it. His av algorithms is incredible, it's the best

0:25:01.400 --> 0:25:04.840
<v Speaker 12>in the world. And he has a very large fleet

0:25:04.880 --> 0:25:07.640
<v Speaker 12>of cars on the road that allows him to collect.

0:25:07.359 --> 0:25:07.960
<v Speaker 6>A lot of data.

0:25:08.000 --> 0:25:11.360
<v Speaker 12>And so so I think he has just a phenomenal position.

0:25:11.600 --> 0:25:13.320
<v Speaker 12>And he's been working on this for a long time,

0:25:13.359 --> 0:25:15.040
<v Speaker 12>and so he's he's going to be in a great

0:25:15.040 --> 0:25:16.639
<v Speaker 12>position to take advantage of it.

0:25:16.840 --> 0:25:18.360
<v Speaker 5>Well, may I ask you that's juncture.

0:25:18.400 --> 0:25:22.680
<v Speaker 4>He's clearly influential in this upcoming administration, but he also

0:25:22.800 --> 0:25:25.479
<v Speaker 4>positions Tesla as a leading AI and robotics company.

0:25:25.600 --> 0:25:27.760
<v Speaker 5>Yeah, how does that code for Nvidia?

0:25:27.880 --> 0:25:32.560
<v Speaker 4>Elon Musk's influence the President Electrump and also thecoming administration's

0:25:32.600 --> 0:25:34.600
<v Speaker 4>kind of attitude towards AI.

0:25:36.359 --> 0:25:40.119
<v Speaker 6>I don't know that the attitude towards AI.

0:25:40.520 --> 0:25:43.520
<v Speaker 12>I know Elon's attitude towards AI, and and he's very

0:25:43.520 --> 0:25:44.760
<v Speaker 12>optimistic about his future.

0:25:44.880 --> 0:25:47.800
<v Speaker 6>And obviously he's working on some of the most important

0:25:48.359 --> 0:25:49.200
<v Speaker 6>AI areas.

0:25:50.119 --> 0:25:57.800
<v Speaker 12>XAI is working on foundation cognitive Intelligence AI, his Tesla

0:25:57.880 --> 0:26:03.200
<v Speaker 12>is working on Thomas vehicles, and optimists were humanoid robotics.

0:26:03.400 --> 0:26:06.040
<v Speaker 12>These three areas of AI are the three most important

0:26:06.040 --> 0:26:08.879
<v Speaker 12>areas of AI, and so I think he's working on

0:26:08.960 --> 0:26:09.960
<v Speaker 12>exactly the right things.

0:26:10.119 --> 0:26:12.720
<v Speaker 4>You kind of positioned AI and video's position in the

0:26:13.160 --> 0:26:17.480
<v Speaker 4>supply chain for physical AI, robotics, autonomous driving.

0:26:17.640 --> 0:26:19.840
<v Speaker 5>Explain it a bit more, the role you see in

0:26:19.920 --> 0:26:20.440
<v Speaker 5>video play.

0:26:20.760 --> 0:26:23.359
<v Speaker 12>Well, we're a core technology company, and so we build

0:26:23.400 --> 0:26:25.160
<v Speaker 12>the foundational computing platforms.

0:26:25.880 --> 0:26:28.320
<v Speaker 6>We're also full stack, and so we developed.

0:26:27.960 --> 0:26:31.400
<v Speaker 12>The necessary algorithms and necessary AI technologies and.

0:26:31.359 --> 0:26:32.000
<v Speaker 6>Then we put it.

0:26:32.200 --> 0:26:34.240
<v Speaker 12>We put it out to the industry for them to

0:26:34.720 --> 0:26:37.639
<v Speaker 12>adopt it and turn it into in market solutions.

0:26:38.359 --> 0:26:39.719
<v Speaker 6>We're computing platform companies.

0:26:39.720 --> 0:26:41.880
<v Speaker 4>So you're on stage and you're surrounded by I think

0:26:41.920 --> 0:26:44.720
<v Speaker 4>a dozen humanoid robots, which.

0:26:44.520 --> 0:26:45.200
<v Speaker 5>Is nice for you.

0:26:45.640 --> 0:26:48.359
<v Speaker 4>When will I be here at CS and Las Vegas

0:26:48.400 --> 0:26:51.080
<v Speaker 4>and actually have there are some robots right now? In

0:26:51.440 --> 0:26:54.639
<v Speaker 4>real terms, you must have a timeline that you see

0:26:55.440 --> 0:26:58.960
<v Speaker 4>real world commercial deployment of the technology you outlined last night.

0:26:59.520 --> 0:27:03.600
<v Speaker 12>It depends on use case, I would say first use. Yeah,

0:27:03.640 --> 0:27:05.840
<v Speaker 12>the first use case will probably be in manufacturing. You know,

0:27:05.880 --> 0:27:11.720
<v Speaker 12>there's estimates tends, if not one hundred million jobs that

0:27:11.760 --> 0:27:14.840
<v Speaker 12>are workers that are the world is short of workers

0:27:15.560 --> 0:27:23.000
<v Speaker 12>and aging population, declining, declining birth rates, and so I

0:27:23.040 --> 0:27:27.240
<v Speaker 12>think the world needs a lot more workers. Robotics is

0:27:27.240 --> 0:27:30.520
<v Speaker 12>one of the best ways for us to supplement all

0:27:30.560 --> 0:27:33.639
<v Speaker 12>of that and help companies recover the lost revenues on

0:27:33.680 --> 0:27:38.000
<v Speaker 12>the one hand and drive productivity which reduces inflation for

0:27:38.040 --> 0:27:40.479
<v Speaker 12>the world on the other hand. And so I think

0:27:40.560 --> 0:27:42.919
<v Speaker 12>robotics is going to be very important to that in

0:27:42.960 --> 0:27:46.840
<v Speaker 12>different different areas. You could have probably deploy into manufacturing

0:27:46.840 --> 0:27:49.120
<v Speaker 12>first because they obviously need it most.

0:27:49.119 --> 0:27:52.560
<v Speaker 4>Which you see is a ginormous potential market addressable lots.

0:27:52.560 --> 0:27:55.720
<v Speaker 6>It's a fifty trillion dollar industry that wants to.

0:27:55.600 --> 0:27:58.080
<v Speaker 4>Grow, and it thinks to grow. Sorry to interrupt you, Jensen.

0:27:58.080 --> 0:28:00.200
<v Speaker 4>I think that something that wool Street struggling with this morning,

0:28:00.440 --> 0:28:05.000
<v Speaker 4>among many things, is dg X. They understand the investment

0:28:05.040 --> 0:28:08.680
<v Speaker 4>there that trains the foundation models in terms of Nvidia's

0:28:08.680 --> 0:28:13.960
<v Speaker 4>business model what you outlined last night, Cosmos, and then

0:28:14.160 --> 0:28:17.960
<v Speaker 4>later on on the inference side, how physical AI helps

0:28:18.000 --> 0:28:19.640
<v Speaker 4>guare your business from the drives.

0:28:19.720 --> 0:28:21.560
<v Speaker 5>D GX growth just simple as that.

0:28:21.680 --> 0:28:23.400
<v Speaker 12>Yeah, as simple as that, I think if you if

0:28:23.440 --> 0:28:25.760
<v Speaker 12>you just look at simply like that, we have three computers,

0:28:26.400 --> 0:28:29.920
<v Speaker 12>and two of the computers d GX and Omniverse drives

0:28:29.920 --> 0:28:32.800
<v Speaker 12>an enormous amount of data that is necessary to train

0:28:32.880 --> 0:28:36.440
<v Speaker 12>the AI models right, and so Omniverse creates the data

0:28:36.760 --> 0:28:39.680
<v Speaker 12>that we then use to train AI models. The training

0:28:39.760 --> 0:28:43.680
<v Speaker 12>is what drives DGX sales. And the more robots that

0:28:43.720 --> 0:28:46.400
<v Speaker 12>are that are available, the more data we can create,

0:28:46.440 --> 0:28:46.880
<v Speaker 12>the more.

0:28:46.760 --> 0:28:48.720
<v Speaker 6>AI models we have to we have to go train.

0:28:49.400 --> 0:28:53.920
<v Speaker 12>That's that cycle is ultimately what we're striving for. All

0:28:53.960 --> 0:28:56.760
<v Speaker 12>of that drives consumption from data center growth.

0:28:56.840 --> 0:29:00.360
<v Speaker 4>There is a surge in AI spending data center growth.

0:29:01.600 --> 0:29:03.640
<v Speaker 4>Some of our audience are a bit concerned about how

0:29:03.680 --> 0:29:07.160
<v Speaker 4>sustainable that is, short, medium, and long term.

0:29:07.520 --> 0:29:11.840
<v Speaker 12>Well, at the limit, Artificial intelligence is a single most

0:29:11.880 --> 0:29:15.520
<v Speaker 12>important technology force of our time, and it's really about

0:29:16.040 --> 0:29:18.440
<v Speaker 12>we're at the beginning of that and in the future,

0:29:18.960 --> 0:29:23.040
<v Speaker 12>every single data center will be driven by AI and

0:29:23.080 --> 0:29:25.640
<v Speaker 12>the type of computing that we build today. And so

0:29:25.680 --> 0:29:28.640
<v Speaker 12>if you look at the world today, we're about a

0:29:28.800 --> 0:29:32.080
<v Speaker 12>year and a half into the remodeling, if you will,

0:29:32.920 --> 0:29:36.720
<v Speaker 12>the modernization, the reinvention of computing, and so I think

0:29:36.760 --> 0:29:38.840
<v Speaker 12>that over the next several years you're going to see

0:29:39.200 --> 0:29:40.760
<v Speaker 12>the transition from the old.

0:29:40.560 --> 0:29:44.080
<v Speaker 6>Way of doing computing, general purpose computing. There's a new way.

0:29:43.920 --> 0:29:47.320
<v Speaker 12>Of doing computing, artificial intelligence and accelerated computing, and so

0:29:47.360 --> 0:29:48.640
<v Speaker 12>we have a lot of growth to go do.

0:29:49.160 --> 0:29:52.080
<v Speaker 4>Let's go back to your and nvidious routes, and therein

0:29:52.160 --> 0:29:56.120
<v Speaker 4>lies the complication, right and the story accelerated computing. You

0:29:56.160 --> 0:30:00.240
<v Speaker 4>spent four years redefining the computer for me, and we

0:30:00.320 --> 0:30:05.120
<v Speaker 4>get rtx Blackwell single black Well in that form factor.

0:30:05.520 --> 0:30:09.800
<v Speaker 4>The target audience is hardcore gamers but also developers, and

0:30:09.800 --> 0:30:12.120
<v Speaker 4>I wondered if you could give me any early examples

0:30:12.520 --> 0:30:15.600
<v Speaker 4>or evidence of how you see the gaming industry adopting

0:30:15.640 --> 0:30:16.440
<v Speaker 4>your technology.

0:30:17.640 --> 0:30:23.560
<v Speaker 12>Well, AI is going to reinvigorate the video game industry.

0:30:23.880 --> 0:30:26.680
<v Speaker 12>On the one hand, for developers, it's going to reduce

0:30:26.680 --> 0:30:29.920
<v Speaker 12>the cost of creating the content. On the other hand,

0:30:30.200 --> 0:30:32.360
<v Speaker 12>all of the characters that are in the games are

0:30:32.400 --> 0:30:34.960
<v Speaker 12>going to be smart characters in the future, so every

0:30:34.960 --> 0:30:36.800
<v Speaker 12>time you interact with them, they're going to be interacting

0:30:36.840 --> 0:30:39.080
<v Speaker 12>with you in a much more intelligent way. And so

0:30:39.160 --> 0:30:41.479
<v Speaker 12>the games are going to be more interesting, the characters

0:30:41.520 --> 0:30:45.240
<v Speaker 12>are going to be more interesting. The content development cost

0:30:45.360 --> 0:30:47.320
<v Speaker 12>is going to decline, and that's going to be really

0:30:47.320 --> 0:30:49.600
<v Speaker 12>great for the industry, and so I think that the

0:30:49.640 --> 0:30:53.520
<v Speaker 12>future is really bright for video games and these virtual

0:30:53.560 --> 0:30:57.240
<v Speaker 12>worlds and artificial intelligence is going to really reinvigorate it.

0:30:57.760 --> 0:31:00.920
<v Speaker 5>Project digits, may I pick it up? Oh yeah, yeah,

0:31:01.080 --> 0:31:06.680
<v Speaker 5>three thousand dollars digits? Yeah, a supercomputer costing.

0:31:06.520 --> 0:31:09.320
<v Speaker 12>Exactly How could you imagine you're just sitting there just

0:31:09.360 --> 0:31:10.920
<v Speaker 12>like that, You're working on your PC?

0:31:11.480 --> 0:31:14.880
<v Speaker 5>Well, here, why would I need one of these? Probably

0:31:14.920 --> 0:31:15.240
<v Speaker 5>not me?

0:31:15.920 --> 0:31:18.959
<v Speaker 4>But how big is the addressable market for this? What

0:31:19.080 --> 0:31:20.280
<v Speaker 4>is the addressable market?

0:31:20.400 --> 0:31:24.880
<v Speaker 12>There are thirty million software developers. They're probably something along

0:31:24.880 --> 0:31:29.400
<v Speaker 12>the lines of ten million designers around the world. Probably

0:31:29.440 --> 0:31:33.720
<v Speaker 12>another twenty million creative artists. Hard to say exactly how

0:31:33.720 --> 0:31:36.920
<v Speaker 12>many students. I'm going to guess probably a couple one

0:31:37.000 --> 0:31:39.400
<v Speaker 12>hundred million students around the world.

0:31:40.040 --> 0:31:41.560
<v Speaker 6>Everybody is going to have to board it.

0:31:41.680 --> 0:31:45.160
<v Speaker 12>Everybody's going to have to Well, they can afford computers,

0:31:45.640 --> 0:31:47.760
<v Speaker 12>and so here's if they can afford computers and they

0:31:47.800 --> 0:31:51.840
<v Speaker 12>would like to have a companion that helps them do AI,

0:31:52.840 --> 0:31:53.800
<v Speaker 12>this is the way to do it.

0:31:53.800 --> 0:31:54.960
<v Speaker 5>Can I just clarify something?

0:31:55.000 --> 0:31:55.560
<v Speaker 6>Yeah? On it?

0:31:55.600 --> 0:31:59.760
<v Speaker 4>I think you said on stage mac os, Linux and Windows.

0:32:00.080 --> 0:32:02.240
<v Speaker 5>No, No, the location said just Linux, No.

0:32:03.600 --> 0:32:07.800
<v Speaker 12>Whatever computer you use, you're literally enjoying how it's going

0:32:07.840 --> 0:32:10.520
<v Speaker 12>to be used. It's sitting right there and you'll just

0:32:10.520 --> 0:32:12.880
<v Speaker 12>connect to what wirelessly like it's at your personal cloud.

0:32:13.640 --> 0:32:15.959
<v Speaker 4>I promised the audience I'd garify that because very excited.

0:32:16.040 --> 0:32:19.320
<v Speaker 4>Oh yeah, we're running short on time. President Electrump has

0:32:19.360 --> 0:32:23.680
<v Speaker 4>been speaking during the course of conversation. How imperative is

0:32:23.720 --> 0:32:25.960
<v Speaker 4>it that you go to mar A Lago and meet

0:32:25.960 --> 0:32:29.680
<v Speaker 4>with him? If Nvidia is America's leading AI company, and will.

0:32:29.480 --> 0:32:31.400
<v Speaker 6>You I would be delighted to go see him. Have

0:32:31.480 --> 0:32:32.240
<v Speaker 6>you been invited?

0:32:32.880 --> 0:32:34.800
<v Speaker 12>Not yet, but I would be delighted to go see

0:32:34.840 --> 0:32:40.520
<v Speaker 12>him and congratulate him and do everything we can to

0:32:40.520 --> 0:32:42.240
<v Speaker 12>help this administration succeed.

0:32:42.600 --> 0:32:44.520
<v Speaker 4>A lot of what you outlined on stage last night

0:32:44.560 --> 0:32:47.920
<v Speaker 4>in the remophysical AI. You know, I saw x Pong,

0:32:48.000 --> 0:32:51.800
<v Speaker 4>for example, in the autonomous driving context that's happening in China,

0:32:51.920 --> 0:32:53.520
<v Speaker 4>like they are doing a lot on robotics.

0:32:53.720 --> 0:32:55.520
<v Speaker 5>So I'm going to ask you about tariffs.

0:32:56.080 --> 0:32:59.000
<v Speaker 4>You know, it's likely this coming administration will be as

0:32:59.040 --> 0:33:02.520
<v Speaker 4>restrictive on technology export and tarifs will be a function.

0:33:02.640 --> 0:33:04.160
<v Speaker 5>How have you prepared for that? Jensen?

0:33:07.080 --> 0:33:11.920
<v Speaker 12>Whatever the administration ultimately decides, will give them as much

0:33:12.400 --> 0:33:16.040
<v Speaker 12>insight as we can from our perspective, and I'm sure

0:33:16.040 --> 0:33:18.680
<v Speaker 12>that the administration will make the right moves that's the

0:33:18.720 --> 0:33:19.880
<v Speaker 12>best interest of our country.

0:33:20.520 --> 0:33:21.400
<v Speaker 5>Twenty twenty five.

0:33:21.800 --> 0:33:23.520
<v Speaker 4>I kind of started the conversation by saying, a lot

0:33:23.520 --> 0:33:26.880
<v Speaker 4>has changed since we last spoke in the summer. You

0:33:27.000 --> 0:33:29.800
<v Speaker 4>said that the age of general robotics is just around

0:33:29.800 --> 0:33:33.440
<v Speaker 4>the corner. Is twenty twenty five the year of general

0:33:33.520 --> 0:33:35.720
<v Speaker 4>robotics or is that a little premature.

0:33:35.200 --> 0:33:36.720
<v Speaker 6>To your mind?

0:33:36.960 --> 0:33:40.680
<v Speaker 12>The development is going gangbusters, as you can see all

0:33:40.680 --> 0:33:42.600
<v Speaker 12>the different robots that are going to be around here,

0:33:43.080 --> 0:33:47.520
<v Speaker 12>and the enabling technology necessary for general robotics is coming together.

0:33:47.600 --> 0:33:50.719
<v Speaker 12>All the pieces are coming together. The industry still has

0:33:50.760 --> 0:33:54.120
<v Speaker 12>a lot of engineering to do. If you looked long term,

0:33:54.480 --> 0:33:57.720
<v Speaker 12>you know, pick your horizon in ten or twenty years,

0:33:58.040 --> 0:34:01.200
<v Speaker 12>the number of robots that's to be on Earth that's

0:34:01.240 --> 0:34:03.880
<v Speaker 12>going to be measured and probably tens of nine hundreds

0:34:03.920 --> 0:34:08.080
<v Speaker 12>and not potentially billions of robots, and so those days

0:34:08.120 --> 0:34:11.680
<v Speaker 12>are clearly coming. Is it going to happen in the

0:34:11.719 --> 0:34:13.520
<v Speaker 12>next couple of years or the next five years, hard

0:34:13.520 --> 0:34:16.279
<v Speaker 12>to say, but the development of robotics is going to

0:34:16.280 --> 0:34:19.240
<v Speaker 12>be all over the world now and we're seeing startup companies,

0:34:19.320 --> 0:34:24.040
<v Speaker 12>large companies, industrial companies, consumer electronics companies all getting involved

0:34:24.040 --> 0:34:26.520
<v Speaker 12>in the future of robotics. And our offering to the

0:34:26.560 --> 0:34:30.000
<v Speaker 12>industry is a three computer system. And so whether they're

0:34:30.640 --> 0:34:33.560
<v Speaker 12>developing the robot, training the robot, we have DJX systems

0:34:33.560 --> 0:34:36.360
<v Speaker 12>for them and DJX clouds for them. If they're simulating

0:34:36.400 --> 0:34:38.960
<v Speaker 12>the robots, we have omnivers for them, and if they

0:34:39.080 --> 0:34:39.759
<v Speaker 12>want to deploy them.

0:34:39.800 --> 0:34:41.759
<v Speaker 6>When they're ready to deploy the robots, we have.

0:34:43.560 --> 0:34:47.160
<v Speaker 12>Little computers that basically is the computer brain of the

0:34:47.280 --> 0:34:49.720
<v Speaker 12>robot that they can put inside the robot.

0:34:50.000 --> 0:34:50.920
<v Speaker 6>And so we'll work with.

0:34:50.880 --> 0:34:54.000
<v Speaker 12>The industry across the board from the development of the

0:34:54.080 --> 0:34:57.040
<v Speaker 12>robots to the deployment of the robot. And we have

0:34:57.120 --> 0:35:00.399
<v Speaker 12>computer systems for them, algorithms for them, ais for them,

0:35:00.760 --> 0:35:03.400
<v Speaker 12>and we'll partner with the industry and make this feature happen.

0:35:03.320 --> 0:35:04.120
<v Speaker 6>Very very quick.

0:35:04.560 --> 0:35:07.560
<v Speaker 4>Which line of business grows fast? Is this year, data center,

0:35:07.760 --> 0:35:08.759
<v Speaker 4>gaming or other.

0:35:09.719 --> 0:35:12.239
<v Speaker 6>They're all going to grow fast. I think gaming is

0:35:12.239 --> 0:35:13.080
<v Speaker 6>continuing to grow.

0:35:13.800 --> 0:35:18.000
<v Speaker 12>Our automas vehicle business is already on its way to

0:35:18.040 --> 0:35:20.799
<v Speaker 12>be a five billion dollar business this year, and so

0:35:21.360 --> 0:35:24.880
<v Speaker 12>right and sometimes right run rate, and so the autonomous

0:35:25.000 --> 0:35:27.760
<v Speaker 12>vehicle business is just starting to get off the ground,

0:35:28.040 --> 0:35:30.279
<v Speaker 12>and that tells you something about how we address it.

0:35:30.320 --> 0:35:32.560
<v Speaker 12>And the reason for that is we get the benefit

0:35:32.760 --> 0:35:35.759
<v Speaker 12>from the beginning of the development of the AIS all

0:35:35.800 --> 0:35:38.799
<v Speaker 12>the way to the deployment of the cars. Because a

0:35:38.840 --> 0:35:43.000
<v Speaker 12>car company needs two factories, a car factory that builds

0:35:43.000 --> 0:35:44.960
<v Speaker 12>the cars and an AI factory that builds the AI

0:35:45.040 --> 0:35:47.919
<v Speaker 12>sport the cars, and both of these, both of these

0:35:48.239 --> 0:35:49.640
<v Speaker 12>areas we could participate.

0:35:49.719 --> 0:35:51.600
<v Speaker 6>And so I think it's going to be a very

0:35:51.640 --> 0:35:52.160
<v Speaker 6>large business.

0:35:52.160 --> 0:35:54.279
<v Speaker 4>Send I've made you late for your next appointment. Yeah,

0:35:54.280 --> 0:35:56.400
<v Speaker 4>it's crazy, literally grateful a few times. It's good to

0:35:56.400 --> 0:35:59.200
<v Speaker 4>see you as well. Jensen one, the Nvidia CEO.

0:35:59.040 --> 0:36:05.080
<v Speaker 2>Caroline, what a conversation, ed Ludlow live from Las Vegas.

0:36:05.120 --> 0:36:07.080
<v Speaker 2>So good to have you back. We'll let you go

0:36:07.120 --> 0:36:09.319
<v Speaker 2>and return to paternity leave from New York.

0:36:09.840 --> 0:36:11.040
<v Speaker 6>This is blue Bag technology.