WEBVTT - China's AI Push Meets Amazon's Cloud Surge

0:00:02.520 --> 0:00:14.920
<v Speaker 1>Bloomberg Audio Studios, Podcasts, radio News. Bloomberg Tech is live

0:00:15.040 --> 0:00:18.720
<v Speaker 1>from the heart of Silicon Valley with Ed Ludlow in

0:00:18.960 --> 0:00:19.720
<v Speaker 1>Van Francisco.

0:00:22.840 --> 0:00:24.440
<v Speaker 2>This is Bloomberg Tech coming up.

0:00:24.480 --> 0:00:27.800
<v Speaker 3>Ali Barber releases its latest Quenn model, its biggest ever,

0:00:27.960 --> 0:00:32.480
<v Speaker 3>claiming performance on par with Anthropic Plus. Amazon surpassed three

0:00:32.520 --> 0:00:35.960
<v Speaker 3>trillion dollars in market value for the first time, extending

0:00:36.000 --> 0:00:38.880
<v Speaker 3>gains after a big cloud and AI driven earnings report

0:00:38.960 --> 0:00:42.839
<v Speaker 3>last week. AWS CEO mcgarman joins us later in the

0:00:42.880 --> 0:00:47.240
<v Speaker 3>hour and how Apple's reinventing itself as a subscription giant.

0:00:47.280 --> 0:00:50.120
<v Speaker 3>Will dive into the details of the iPhone maker's Apple

0:00:50.240 --> 0:00:55.600
<v Speaker 3>Upgrade program and its next era of growth. There's an

0:00:55.640 --> 0:00:59.880
<v Speaker 3>elite club that's getting bigger. Today's big number three trillion dollars.

0:01:00.040 --> 0:01:03.000
<v Speaker 3>Amazon has reached a three trillion dollar market value for

0:01:03.080 --> 0:01:06.760
<v Speaker 3>the first time, becoming just the fifth company ever to

0:01:06.840 --> 0:01:09.679
<v Speaker 3>hit that milestone. The stock training at all time highs

0:01:09.959 --> 0:01:14.720
<v Speaker 3>following its quarterly earnings last Thursday night from Friday into

0:01:14.760 --> 0:01:18.040
<v Speaker 3>this Monday session. This is the biggest two day gain

0:01:18.120 --> 0:01:20.920
<v Speaker 3>on the stock since two thousand and nine, and the

0:01:21.000 --> 0:01:24.680
<v Speaker 3>momentum has really carried through Later in the hour, we're

0:01:24.680 --> 0:01:26.960
<v Speaker 3>going to speak to aw Seo mac Garman, and it

0:01:27.000 --> 0:01:30.800
<v Speaker 3>really was AWS that was the big driver in earnings.

0:01:30.840 --> 0:01:34.680
<v Speaker 3>The AI story, there's a lot to discuss. China's AI

0:01:34.800 --> 0:01:39.160
<v Speaker 3>race accelerates. Ali Barber has unveiled Quen three point eight Max,

0:01:39.480 --> 0:01:42.840
<v Speaker 3>its new flagship AI model, saying it can rival and

0:01:42.920 --> 0:01:45.720
<v Speaker 3>thropics best. It's the latest in a wave of Chinese

0:01:45.760 --> 0:01:50.280
<v Speaker 3>AI breakthroughs, challenging Silicon Valley's lead Bloomberg's equities reporter Henry

0:01:50.280 --> 0:01:52.480
<v Speaker 3>Wren joins us for more. Let's go over the basics,

0:01:52.520 --> 0:01:55.960
<v Speaker 3>Ali Barber sharing details of the model. The model clearly

0:01:56.120 --> 0:02:01.160
<v Speaker 3>on benchmarks, it says is high performance, and the ADRs

0:02:01.160 --> 0:02:02.760
<v Speaker 3>have reacted quite positively here.

0:02:03.960 --> 0:02:04.160
<v Speaker 1>Yeah.

0:02:04.200 --> 0:02:07.760
<v Speaker 4>Indeed, we're seeing Alibaba's ADRs training up about five percent

0:02:07.800 --> 0:02:10.919
<v Speaker 4>in US trading at this point. So the company said

0:02:10.960 --> 0:02:15.560
<v Speaker 4>that it's a new MAX three point eight models rivals

0:02:15.720 --> 0:02:19.600
<v Speaker 4>the US models, the Frontier Labs models at some benchmarks,

0:02:19.600 --> 0:02:24.200
<v Speaker 4>so for example on research reproduction, on agentic use of computers,

0:02:24.600 --> 0:02:27.280
<v Speaker 4>these are strong suits. It also says that its model

0:02:27.320 --> 0:02:30.560
<v Speaker 4>has two point four trillion terms of parameters, and that's

0:02:30.919 --> 0:02:34.360
<v Speaker 4>basically on power with the level that we've been seeing

0:02:34.400 --> 0:02:37.080
<v Speaker 4>for the Keemi K three model, which was out a

0:02:37.120 --> 0:02:40.280
<v Speaker 4>few weeks ago. Also, the company, yes also said something

0:02:40.320 --> 0:02:43.680
<v Speaker 4>about cost. Obviously per token costs, it's a bit lower

0:02:43.680 --> 0:02:45.320
<v Speaker 4>than the US frontier models as.

0:02:45.160 --> 0:02:49.040
<v Speaker 3>Well, two dollars per one million tokens. On the input side,

0:02:49.520 --> 0:02:52.799
<v Speaker 3>you have a really strong command of Chinese technology companies

0:02:52.840 --> 0:02:55.880
<v Speaker 3>from the stock perspective, Right, is Ali barbera name Henry

0:02:55.919 --> 0:02:59.560
<v Speaker 3>that trades on its AI prowess or like some of

0:02:59.560 --> 0:03:03.760
<v Speaker 3>these America can multifaceted companies a different part of its business.

0:03:04.840 --> 0:03:07.880
<v Speaker 4>Yeah, it's the AI story is really important for Ali Baba.

0:03:08.120 --> 0:03:10.320
<v Speaker 4>If we zoom out a bit today, it gains about

0:03:10.360 --> 0:03:13.560
<v Speaker 4>five percent. But if you really look at the trajectory

0:03:13.639 --> 0:03:16.200
<v Speaker 4>since June, you will see that the stock is up

0:03:16.200 --> 0:03:18.799
<v Speaker 4>about thirty five to forty percent from a June low.

0:03:18.960 --> 0:03:24.040
<v Speaker 4>And the key story here is about the more optimism

0:03:24.120 --> 0:03:27.280
<v Speaker 4>in AI. Obviously we're seeing the Quen breakthrough today, but

0:03:27.720 --> 0:03:30.560
<v Speaker 4>really last month the company has a closed store meeting

0:03:30.600 --> 0:03:33.840
<v Speaker 4>with analysts, and analysts came out more optimistic about the

0:03:33.840 --> 0:03:37.520
<v Speaker 4>company's cloud revenue growth. For example, JP Morgan is modeling

0:03:37.560 --> 0:03:40.160
<v Speaker 4>about forty five percent in revenue growth for its cloud

0:03:40.240 --> 0:03:42.640
<v Speaker 4>unit for the June quarter, so there has been an

0:03:42.680 --> 0:03:46.440
<v Speaker 4>acceleration over there, but still on the main business, on

0:03:46.480 --> 0:03:48.840
<v Speaker 4>the e commerce business, because it has been a company

0:03:48.880 --> 0:03:53.080
<v Speaker 4>known for its main shopping site in China, that part

0:03:53.120 --> 0:03:55.760
<v Speaker 4>of business is still lagging, although it's still the cash

0:03:55.760 --> 0:03:59.840
<v Speaker 4>cloud cloud, so to speak. So investors' opinions are still mixed.

0:04:00.040 --> 0:04:02.240
<v Speaker 4>That's why we were seeing sentiment has been pretty weak

0:04:02.480 --> 0:04:05.720
<v Speaker 4>into June, but since drewing some positive AI announcements has

0:04:05.760 --> 0:04:06.960
<v Speaker 4>really been boosting the stock.

0:04:07.720 --> 0:04:10.200
<v Speaker 2>Bloomberg's Henry Wren, thank you very much.

0:04:10.360 --> 0:04:13.520
<v Speaker 3>Ali Barber says its new Queen model delivers frontier AI

0:04:13.560 --> 0:04:17.000
<v Speaker 3>performance and as Henry outlined, at a fraction of the cost,

0:04:17.400 --> 0:04:21.640
<v Speaker 3>could cheaper AI accelerate adoption rather than just pressure pricing.

0:04:21.720 --> 0:04:24.760
<v Speaker 3>Joining us now is Lauren Cassidy, chief investment officer of

0:04:24.839 --> 0:04:28.440
<v Speaker 3>the Founders one hundred ETF, the ETF tracking founders in

0:04:28.480 --> 0:04:33.760
<v Speaker 3>this space meta in Zuckerberg, Musk and SpaceX. The economics

0:04:33.800 --> 0:04:36.320
<v Speaker 3>of AI is really what the Chinese story is putting

0:04:36.320 --> 0:04:38.960
<v Speaker 3>front and center. So I pose that question to you

0:04:39.120 --> 0:04:42.839
<v Speaker 3>drive adoption or lower the tokenomics here?

0:04:44.360 --> 0:04:45.560
<v Speaker 5>Yes, I do.

0:04:45.920 --> 0:04:49.360
<v Speaker 6>See this as a positive in that there's room in

0:04:49.400 --> 0:04:52.359
<v Speaker 6>the world for both open and closed models, and I

0:04:52.400 --> 0:04:55.920
<v Speaker 6>do think it will drive prices a little lower. It

0:04:55.960 --> 0:04:59.920
<v Speaker 6>isn't cheaper than anthropics haiku, but if it is fable,

0:05:00.160 --> 0:05:03.680
<v Speaker 6>fable level performance, and it's about a fit of the price,

0:05:03.920 --> 0:05:09.080
<v Speaker 6>and so I think that you know, that could put

0:05:09.120 --> 0:05:12.719
<v Speaker 6>pressure on the l elms. But overall there's still a

0:05:12.760 --> 0:05:15.760
<v Speaker 6>lot that the American llms have to offer.

0:05:16.839 --> 0:05:19.760
<v Speaker 3>What are the advantages that the US models have to

0:05:19.800 --> 0:05:20.800
<v Speaker 3>your mind, Lauren.

0:05:21.400 --> 0:05:29.120
<v Speaker 6>Trust, reliability, additional features, convenience of not having to modify

0:05:29.160 --> 0:05:33.159
<v Speaker 6>the models. So it just depends what an enterprise is

0:05:33.200 --> 0:05:34.560
<v Speaker 6>looking for, what.

0:05:36.120 --> 0:05:37.680
<v Speaker 5>Fits the best.

0:05:39.160 --> 0:05:42.400
<v Speaker 3>I'm looking at the found one ETF and again you

0:05:42.520 --> 0:05:44.440
<v Speaker 3>feel free to remind us of the thesis on the

0:05:44.480 --> 0:05:48.320
<v Speaker 3>founder side, But you don't have any of the the

0:05:48.360 --> 0:05:51.160
<v Speaker 3>A d rs for some of these big Chinese tech companies.

0:05:51.760 --> 0:05:53.080
<v Speaker 2>What are the reasons for that?

0:05:55.240 --> 0:06:01.600
<v Speaker 6>And I think that we believe that in American innovation

0:06:01.839 --> 0:06:06.480
<v Speaker 6>long term, and just as we've had waves of innovation

0:06:06.640 --> 0:06:12.640
<v Speaker 6>and greater adoption in the country of the steam engine, electricity,

0:06:12.880 --> 0:06:18.320
<v Speaker 6>modern manufacturing, computers, and the Internet, we are now on

0:06:18.360 --> 0:06:21.120
<v Speaker 6>the sixth wave, which is AI. We think it's going

0:06:21.200 --> 0:06:25.239
<v Speaker 6>to be a really big thing. We see tremendous adoption

0:06:25.320 --> 0:06:30.039
<v Speaker 6>over the next decade, and we are very exposed to

0:06:30.760 --> 0:06:31.440
<v Speaker 6>those stocks.

0:06:33.040 --> 0:06:34.160
<v Speaker 2>Let's shift gears a little bit.

0:06:34.920 --> 0:06:40.279
<v Speaker 3>SpaceX has its first quarterly earnings and call Tuesday evening.

0:06:40.760 --> 0:06:44.039
<v Speaker 3>This is what I wrote published this morning. It's the

0:06:44.080 --> 0:06:48.360
<v Speaker 3>first opportunity that SpaceX investors might have to put the

0:06:48.440 --> 0:06:53.080
<v Speaker 3>question of a Tesla SpaceX merger to Elon Musk. Maybe

0:06:54.360 --> 0:06:56.520
<v Speaker 3>you and I've discussed this before from the Tesla side

0:06:56.520 --> 0:06:59.719
<v Speaker 3>of the table, but where does founder one hundred ETF

0:06:59.760 --> 0:07:01.120
<v Speaker 3>stand on that idea?

0:07:02.520 --> 0:07:08.039
<v Speaker 6>Yes, we believe that's inevitable over time. Elon Musk is

0:07:08.240 --> 0:07:13.600
<v Speaker 6>an incredible founder, and he is building the only closed

0:07:13.640 --> 0:07:17.840
<v Speaker 6>loop AI network, and over time there's going to be

0:07:18.240 --> 0:07:22.160
<v Speaker 6>more and more opportunities for Tesla and SpaceX to share.

0:07:22.560 --> 0:07:25.200
<v Speaker 6>They're going to be building Tarifab. I'm really hoping to

0:07:25.240 --> 0:07:31.240
<v Speaker 6>hear details about that build out tomorrow. They'll be sharing chips.

0:07:31.600 --> 0:07:34.760
<v Speaker 6>There's going to be a lot of opportunities for synergies

0:07:34.760 --> 0:07:37.640
<v Speaker 6>between the businesses. It probably makes sense at some point

0:07:38.040 --> 0:07:41.880
<v Speaker 6>to fold Tesla in to the greater mix as we

0:07:42.000 --> 0:07:43.680
<v Speaker 6>work towards data centers in space.

0:07:45.240 --> 0:07:48.160
<v Speaker 3>Lauren Cassidy, a founder one hundred ETF, back on the show,

0:07:48.160 --> 0:07:51.160
<v Speaker 3>thank you very much. Indeed, another top story, one of

0:07:51.160 --> 0:07:55.200
<v Speaker 3>the crypto industry's most trusted storage devices is at the

0:07:55.240 --> 0:07:59.040
<v Speaker 3>center of a major hack. Coinkite, the Canadian company behind

0:07:59.040 --> 0:08:02.160
<v Speaker 3>the cold card Hog hardware wallet, says a software floor

0:08:02.560 --> 0:08:07.440
<v Speaker 3>made some bitcoin recovery keys predictable, allowing hackers to drain

0:08:07.560 --> 0:08:10.920
<v Speaker 3>wallets that were supposed to be kept safely online. According

0:08:10.920 --> 0:08:15.000
<v Speaker 3>to Galaxy Research, more than seventeen hundred bitcoin worth about

0:08:15.160 --> 0:08:18.560
<v Speaker 3>one hundred and ten million dollars has been stolen from

0:08:18.680 --> 0:08:22.239
<v Speaker 3>roughly five thousand different wallets. Coinkite says it has released

0:08:22.280 --> 0:08:26.120
<v Speaker 3>a firmware update to fix the floor. Another stock that

0:08:26.120 --> 0:08:29.240
<v Speaker 3>we're watching because of what happened over the weekend, shares

0:08:29.280 --> 0:08:33.679
<v Speaker 3>of AMC, now modestly higher seven tens percent, had been

0:08:33.760 --> 0:08:38.320
<v Speaker 3>much higher at the open, highest ever total weekend revenue,

0:08:39.080 --> 0:08:42.559
<v Speaker 3>including records for admissions and food and beverage. By the way,

0:08:42.960 --> 0:08:46.760
<v Speaker 3>why the driven driven by the opening of Spider Man

0:08:46.920 --> 0:08:49.920
<v Speaker 3>brand New Day yet to see it, that film grossing

0:08:49.960 --> 0:09:00.280
<v Speaker 3>three hundred and fifty five million dollars domestically in the US.

0:09:00.360 --> 0:09:04.480
<v Speaker 3>Valor Atomics, the company building America's first gigas sites, secured

0:09:04.520 --> 0:09:06.920
<v Speaker 3>a billion dollars in its Series B funding round, led

0:09:06.960 --> 0:09:10.040
<v Speaker 3>by Sequoia. The company also closed a two hundred million

0:09:10.040 --> 0:09:13.840
<v Speaker 3>dollar credit facility to build the words first vertically integrated

0:09:14.080 --> 0:09:17.720
<v Speaker 3>industrial nuclear power company. Joining us now is Valo Atomic

0:09:17.840 --> 0:09:21.920
<v Speaker 3>CEO and founder Isaiah Taylor. You're also fresh off what

0:09:22.120 --> 0:09:26.439
<v Speaker 3>now seems with hindsight, a pretty key demo that was

0:09:26.480 --> 0:09:29.160
<v Speaker 3>a bit of an unlock for this round. You want

0:09:29.200 --> 0:09:31.000
<v Speaker 3>to move quickly, so let's start that. What do you

0:09:31.040 --> 0:09:33.360
<v Speaker 3>need the funding for and how you're going to deploy it.

0:09:34.160 --> 0:09:36.720
<v Speaker 7>Yeah, and thanks for having me on again. Just a

0:09:36.760 --> 0:09:39.079
<v Speaker 7>month ago, I was on the show talking about how

0:09:39.080 --> 0:09:42.440
<v Speaker 7>we became the first startup in history to make nuclear power.

0:09:42.840 --> 0:09:45.640
<v Speaker 7>We did that by powering and Nvidia Blackwell Chip, which

0:09:45.679 --> 0:09:47.880
<v Speaker 7>was an amazing event as the first time that an

0:09:47.920 --> 0:09:52.280
<v Speaker 7>advanced nuclear reactor has powered AI infrastructure directly. So you're right,

0:09:52.480 --> 0:09:54.400
<v Speaker 7>it's been an amazing couple of months. We made a

0:09:54.440 --> 0:09:56.880
<v Speaker 7>lot of progress, and really what this is about is

0:09:57.080 --> 0:09:59.480
<v Speaker 7>giving us the firepower that we need to go build

0:09:59.760 --> 0:10:03.320
<v Speaker 7>the sites. We've now demonstrated nuclear power. This is the

0:10:03.360 --> 0:10:05.840
<v Speaker 7>reactor standing behind me here, and we have to build

0:10:05.880 --> 0:10:08.040
<v Speaker 7>a lot more of these in order to make pace

0:10:08.440 --> 0:10:09.040
<v Speaker 7>with AI.

0:10:09.160 --> 0:10:10.840
<v Speaker 8>AI is building very quickly.

0:10:10.880 --> 0:10:13.120
<v Speaker 7>We need a lot of power in every direction and

0:10:13.120 --> 0:10:15.040
<v Speaker 7>this is really about scale for us. So this is

0:10:15.080 --> 0:10:16.800
<v Speaker 7>the firepower that we need to go and build.

0:10:18.200 --> 0:10:20.560
<v Speaker 3>Sequoia led the round bit. Sean McGuire is also going

0:10:20.600 --> 0:10:23.440
<v Speaker 3>to join your board. Why is that important to you?

0:10:23.520 --> 0:10:25.920
<v Speaker 3>What is it that Sean can can give you other

0:10:26.000 --> 0:10:26.880
<v Speaker 3>than just the check?

0:10:28.080 --> 0:10:30.000
<v Speaker 7>You know, it's been amazing to spend some time with

0:10:30.120 --> 0:10:32.240
<v Speaker 7>Sean over the last few months and get to know

0:10:32.400 --> 0:10:34.440
<v Speaker 7>the rest of the Sakoia team and the partnership. I

0:10:34.440 --> 0:10:36.920
<v Speaker 7>think Sean's a really, really thoughtful person. He spends a

0:10:36.960 --> 0:10:39.679
<v Speaker 7>lot of time with SpaceX and understanding that business model,

0:10:39.960 --> 0:10:41.839
<v Speaker 7>and I think one of the lessons to learn from

0:10:41.840 --> 0:10:44.360
<v Speaker 7>SpaceX is really just the power of scale and pace

0:10:44.400 --> 0:10:46.400
<v Speaker 7>of production, and that's something that we want to do

0:10:46.440 --> 0:10:47.280
<v Speaker 7>in nuclear as well.

0:10:47.360 --> 0:10:48.360
<v Speaker 8>So I think that his.

0:10:48.400 --> 0:10:50.160
<v Speaker 7>Lessons that he's to learn there are going to be

0:10:50.160 --> 0:10:52.280
<v Speaker 7>applied very well to this business, and I'm really excited

0:10:52.280 --> 0:10:53.240
<v Speaker 7>to welcome them to the board.

0:10:54.120 --> 0:10:56.199
<v Speaker 3>You know, the numbers here are really big, right, So

0:10:56.240 --> 0:10:59.320
<v Speaker 3>it's a billion dollar round six billion dollar post money valuation,

0:11:00.040 --> 0:11:02.840
<v Speaker 3>but with respect, you know, this is still early, this

0:11:02.960 --> 0:11:07.080
<v Speaker 3>is nascent, it is high risk, it is technical how

0:11:07.120 --> 0:11:10.240
<v Speaker 3>contingent was that demo that you did just on the

0:11:10.240 --> 0:11:14.559
<v Speaker 3>show a month ago in literally making the round materialize.

0:11:14.600 --> 0:11:17.720
<v Speaker 3>How much did investors want to see that this worked.

0:11:19.000 --> 0:11:21.360
<v Speaker 7>You're right, this is super early, and you know, this

0:11:21.440 --> 0:11:23.280
<v Speaker 7>is a very long journey for us. We see this

0:11:23.320 --> 0:11:26.080
<v Speaker 7>as a multi trillion dollar opportunity. We're going to go

0:11:26.120 --> 0:11:28.439
<v Speaker 7>and build the biggest energy company on Earth. So these

0:11:28.440 --> 0:11:30.440
<v Speaker 7>are the very early days for sure. You know, we

0:11:30.520 --> 0:11:32.280
<v Speaker 7>have a lot to prove still. We want to prove

0:11:32.440 --> 0:11:34.960
<v Speaker 7>pace of production. We want to go from building one

0:11:35.000 --> 0:11:38.360
<v Speaker 7>reactor in a year to six months, to once a month,

0:11:38.400 --> 0:11:40.480
<v Speaker 7>and then even faster than that. So you're right, there's

0:11:40.520 --> 0:11:42.320
<v Speaker 7>still a lot to prove. You know, I think that

0:11:42.360 --> 0:11:44.680
<v Speaker 7>this demonstration was really important. We've said for a long

0:11:44.720 --> 0:11:47.720
<v Speaker 7>time that valar Atomics is a hardware company. We're focused

0:11:47.760 --> 0:11:51.360
<v Speaker 7>on hardware execution and on full vertical integration, and that's

0:11:51.400 --> 0:11:53.080
<v Speaker 7>really what we proved here. We prove that we can

0:11:53.120 --> 0:11:55.360
<v Speaker 7>take a patch of dirt, we can build a reactor

0:11:55.400 --> 0:11:57.560
<v Speaker 7>on it, turn it on, and operate it. And that's

0:11:57.640 --> 0:11:59.720
<v Speaker 7>just one small step, but it's the first of many

0:11:59.760 --> 0:12:02.040
<v Speaker 7>to go and build the nuclear power company or the future.

0:12:03.600 --> 0:12:04.000
<v Speaker 2>Isaiah.

0:12:04.000 --> 0:12:05.480
<v Speaker 3>One of the things you and I have not yet

0:12:05.520 --> 0:12:08.560
<v Speaker 3>discussed is who the customer ends up being. You know,

0:12:08.600 --> 0:12:10.800
<v Speaker 3>you want to build the nuclear power company of the future.

0:12:10.800 --> 0:12:14.360
<v Speaker 3>Who You've explained the scale of the tech you've powered

0:12:14.360 --> 0:12:16.840
<v Speaker 3>one black Well chip, and I think that's important context

0:12:16.880 --> 0:12:18.800
<v Speaker 3>of the audience. But who do you ultimately sell it to.

0:12:20.120 --> 0:12:22.280
<v Speaker 7>It's a really good question. I think that nuclear and

0:12:22.440 --> 0:12:24.280
<v Speaker 7>energy in general is a bit noisy right now. You

0:12:24.320 --> 0:12:27.160
<v Speaker 7>see a lot of deals getting signed that maybe are binding,

0:12:27.200 --> 0:12:27.720
<v Speaker 7>maybe are not.

0:12:27.920 --> 0:12:28.920
<v Speaker 8>We've been really really.

0:12:28.840 --> 0:12:31.520
<v Speaker 7>Focused and disciplined on this point that we believe this

0:12:31.600 --> 0:12:34.959
<v Speaker 7>is a hardware execution problem. Everyone needs energy, they need

0:12:34.960 --> 0:12:36.680
<v Speaker 7>it at any price, they can get it as fast

0:12:36.679 --> 0:12:39.160
<v Speaker 7>as they can get it, and so really really focused

0:12:39.160 --> 0:12:40.439
<v Speaker 7>on just executing.

0:12:40.000 --> 0:12:42.200
<v Speaker 8>In the hardware and bringing new power online.

0:12:42.400 --> 0:12:44.360
<v Speaker 7>And I think once we get that new power online,

0:12:44.440 --> 0:12:46.880
<v Speaker 7>that'll open us up to a bunch of really interesting

0:12:46.880 --> 0:12:50.520
<v Speaker 7>customers who have large balance sheets. We depend on energy

0:12:50.600 --> 0:12:52.480
<v Speaker 7>in order to build what they see is the future

0:12:52.520 --> 0:12:54.640
<v Speaker 7>as well. And so that's you know, there's a handful

0:12:54.640 --> 0:12:57.079
<v Speaker 7>of customers there that are interested in what we're doing.

0:12:57.320 --> 0:12:58.960
<v Speaker 7>But we're going to stay focused and we're going to

0:12:58.960 --> 0:12:59.880
<v Speaker 7>stay focused on hard works.

0:13:01.440 --> 0:13:03.080
<v Speaker 3>Isaiah I don't know how much you want to talk

0:13:03.240 --> 0:13:05.640
<v Speaker 3>about finance. I don't know if you're the finance guy

0:13:06.040 --> 0:13:10.320
<v Speaker 3>in truth at Valor. But an interesting credit facility Palm

0:13:10.440 --> 0:13:14.240
<v Speaker 3>Lucky's Arabor and JP Morgan. You know it's two hundred

0:13:14.240 --> 0:13:17.240
<v Speaker 3>million dollar facility. But why was that important? You really

0:13:17.280 --> 0:13:19.160
<v Speaker 3>sort of put it high up in the announcement today.

0:13:20.200 --> 0:13:21.960
<v Speaker 7>You know, it is important to us, you know, if

0:13:22.040 --> 0:13:24.400
<v Speaker 7>the long term of this business right, we don't want

0:13:24.400 --> 0:13:26.679
<v Speaker 7>to be relying entirely on equity, and we think that

0:13:26.880 --> 0:13:29.600
<v Speaker 7>power is a uniquely debt financible.

0:13:29.040 --> 0:13:31.440
<v Speaker 8>Thing, and so moving this company from.

0:13:31.320 --> 0:13:34.080
<v Speaker 7>Something that's equity financed to debt finance is a long

0:13:34.160 --> 0:13:35.920
<v Speaker 7>term goal of ours. And it's really great to be

0:13:35.960 --> 0:13:38.520
<v Speaker 7>working with partners like Airbor who led this around, and

0:13:38.640 --> 0:13:41.360
<v Speaker 7>JP Morgan Pressing Cove and the Hercules as well coming

0:13:41.360 --> 0:13:43.600
<v Speaker 7>into it, and we're really starting to demonstrate to the

0:13:43.640 --> 0:13:46.520
<v Speaker 7>market that this is a business that can take large

0:13:46.559 --> 0:13:48.719
<v Speaker 7>amounts of credit, deploy it very well.

0:13:48.720 --> 0:13:49.560
<v Speaker 8>Get the yield on that.

0:13:49.679 --> 0:13:51.240
<v Speaker 7>And this is how we're going to run the business

0:13:51.559 --> 0:13:54.120
<v Speaker 7>going forward, is that these reactors will be financed. They'll

0:13:54.160 --> 0:13:57.320
<v Speaker 7>be financed by customers by different ways of putting these

0:13:57.640 --> 0:14:00.160
<v Speaker 7>tranches together and that's really how we're going to go

0:14:00.200 --> 0:14:02.200
<v Speaker 7>and build all of this CAPEX. So we're excited to

0:14:02.200 --> 0:14:03.560
<v Speaker 7>start to demonstrate.

0:14:03.120 --> 0:14:07.840
<v Speaker 3>This with airpor Isaiah, let's end the conversation of talking

0:14:07.840 --> 0:14:12.600
<v Speaker 3>about America's attitude and progress on your industry? Is America

0:14:12.679 --> 0:14:15.760
<v Speaker 3>moving fast enough so that you can do the things

0:14:15.760 --> 0:14:18.120
<v Speaker 3>that you want to do on the timelines that you've

0:14:18.160 --> 0:14:18.600
<v Speaker 3>set out.

0:14:19.480 --> 0:14:21.600
<v Speaker 7>Look, it never feels like we're moving fast enough, right

0:14:21.600 --> 0:14:23.440
<v Speaker 7>We want to move so much faster than we are today.

0:14:23.520 --> 0:14:26.720
<v Speaker 7>China has about fourteen nuclear reactors under construction right now,

0:14:27.200 --> 0:14:29.360
<v Speaker 7>and we're just starting to break ground on our next

0:14:29.400 --> 0:14:31.800
<v Speaker 7>one here, So we need to move faster. If we

0:14:31.840 --> 0:14:33.880
<v Speaker 7>want to win on AI, we have to go faster

0:14:34.000 --> 0:14:36.960
<v Speaker 7>in all forms of power generation, but especially nuclear. I

0:14:36.960 --> 0:14:40.040
<v Speaker 7>think nuclear is uniquely well suited to win the AI race.

0:14:40.320 --> 0:14:43.360
<v Speaker 7>It's clean, it scales very well, you can build lots

0:14:43.360 --> 0:14:45.560
<v Speaker 7>of it in one place next to a large data center.

0:14:45.600 --> 0:14:47.160
<v Speaker 7>There's a lot of things that work really well for this.

0:14:47.240 --> 0:14:49.000
<v Speaker 7>But frankly, we do need to move faster, and that's

0:14:49.000 --> 0:14:50.320
<v Speaker 7>why we're so focused on the hardware.

0:14:50.400 --> 0:14:50.760
<v Speaker 5>Secution.

0:14:51.760 --> 0:14:55.280
<v Speaker 3>Valaatomics CEO and founder Isaiah Taylor back on the show,

0:14:55.320 --> 0:14:58.760
<v Speaker 3>thank you very much. Indeed, let's get to the earning story.

0:14:58.840 --> 0:15:01.560
<v Speaker 3>Stay with AI up next. Palenteer set to report its

0:15:01.600 --> 0:15:04.800
<v Speaker 3>second quarter earnings today after the closing bell. Joining US

0:15:04.840 --> 0:15:08.760
<v Speaker 3>Bloomberg's Lazett Chapman leads our coverage of Palenteer. You know,

0:15:08.800 --> 0:15:10.560
<v Speaker 3>it's so interesting we go with these periods of times

0:15:10.600 --> 0:15:13.600
<v Speaker 3>where we talk about plantic constantly, then maybe it got

0:15:13.600 --> 0:15:17.760
<v Speaker 3>a little quiet. Ultimately, the company has a lot of momentum. Right,

0:15:18.000 --> 0:15:19.600
<v Speaker 3>what are you embracing for?

0:15:19.840 --> 0:15:21.760
<v Speaker 9>Right, It's a pressure cooker quarter.

0:15:22.200 --> 0:15:22.400
<v Speaker 6>Right.

0:15:22.600 --> 0:15:28.800
<v Speaker 9>They have a history of always beating or exceeding analyst

0:15:28.880 --> 0:15:31.680
<v Speaker 9>revenue or analyst estimates for every quarter on.

0:15:31.640 --> 0:15:32.400
<v Speaker 2>Revenue, et cetera.

0:15:32.600 --> 0:15:34.800
<v Speaker 9>So this is already the expectation that they're going to

0:15:34.880 --> 0:15:37.800
<v Speaker 9>beat or exceed, and so even if they raise their guidance,

0:15:38.160 --> 0:15:40.560
<v Speaker 9>it's kind of like, so what you do this every quarter?

0:15:40.840 --> 0:15:42.480
<v Speaker 9>And so the big thing that people are going to

0:15:42.480 --> 0:15:46.680
<v Speaker 9>be watching right now is the concentration of sales, whether

0:15:46.720 --> 0:15:51.520
<v Speaker 9>it's in the US, whether it's international. This quarter they

0:15:51.520 --> 0:15:56.480
<v Speaker 9>are expecting really blockbuster growth for US commercial sales. This

0:15:56.600 --> 0:16:00.000
<v Speaker 9>has been its strongest area of growth, and next quarter

0:16:00.400 --> 0:16:04.720
<v Speaker 9>for the first time ever in its history, Analysts expect

0:16:05.160 --> 0:16:09.440
<v Speaker 9>them to exceed, have commercial growth exceed government growth. So

0:16:09.440 --> 0:16:11.120
<v Speaker 9>what they're going to be looking for this quarter is

0:16:11.440 --> 0:16:16.680
<v Speaker 9>has it continue that trajectory and how are they doing internationally,

0:16:16.760 --> 0:16:18.640
<v Speaker 9>which is something that we've talked about and we've had

0:16:18.640 --> 0:16:20.880
<v Speaker 9>Mark on and some other people on. So that's something

0:16:20.880 --> 0:16:21.960
<v Speaker 9>that they're going to be looking at.

0:16:22.000 --> 0:16:24.280
<v Speaker 3>It's a really big point because you know, in the

0:16:24.320 --> 0:16:27.400
<v Speaker 3>news headlines, Palente comes up in the context of its

0:16:27.480 --> 0:16:30.080
<v Speaker 3>work with the Pentagon or with different arms of the

0:16:30.160 --> 0:16:33.240
<v Speaker 3>US government in a defense context, but they also have

0:16:33.280 --> 0:16:35.680
<v Speaker 3>a lot of commercial case studies. So that's the moment

0:16:35.720 --> 0:16:38.880
<v Speaker 3>in time it flips to commercial. But give us the basics,

0:16:38.960 --> 0:16:41.160
<v Speaker 3>like what Palenti's business does, right, we always have to

0:16:41.160 --> 0:16:42.680
<v Speaker 3>remind us us all what is palent is?

0:16:42.800 --> 0:16:48.680
<v Speaker 9>Yeah, it is AI software that allows people to make

0:16:48.680 --> 0:16:50.880
<v Speaker 9>the best use of the information that they have to

0:16:50.920 --> 0:16:56.080
<v Speaker 9>gather data from all across different you know, sources that

0:16:56.120 --> 0:16:59.080
<v Speaker 9>they have to integrate it, make sense of it, map it,

0:16:59.400 --> 0:17:02.920
<v Speaker 9>and then throw an LLM up against it, whether it

0:17:03.000 --> 0:17:06.680
<v Speaker 9>is you know from you know, from whichever company, whether

0:17:06.680 --> 0:17:09.880
<v Speaker 9>they're agnostic exactly exactly. And so they've worked a lot

0:17:09.920 --> 0:17:14.080
<v Speaker 9>on showing different partnerships that they have with some of

0:17:14.080 --> 0:17:16.040
<v Speaker 9>the major players. That's something that they're going to talk

0:17:16.040 --> 0:17:20.040
<v Speaker 9>about today, I'm sure, along with partnerships with NVIDIAs some

0:17:20.119 --> 0:17:24.040
<v Speaker 9>of the rest. They have gotten a lot of attention

0:17:24.200 --> 0:17:29.200
<v Speaker 9>for their work with defense, but it's actually the commercial

0:17:29.640 --> 0:17:33.439
<v Speaker 9>which is their fastest growing and so we'll see how

0:17:33.440 --> 0:17:33.719
<v Speaker 9>they do.

0:17:34.240 --> 0:17:38.480
<v Speaker 3>Alex carp the CEO, is vocal. Let's say, he also

0:17:38.480 --> 0:17:41.280
<v Speaker 3>talks about the markets sometimes, but is there a clear

0:17:41.320 --> 0:17:44.080
<v Speaker 3>Palentist stock story of this year of the last year,

0:17:44.200 --> 0:17:46.280
<v Speaker 3>like how has the market reacted to them?

0:17:46.440 --> 0:17:46.640
<v Speaker 8>Right?

0:17:47.160 --> 0:17:50.800
<v Speaker 9>The market had massive expectations at the end of last year.

0:17:51.080 --> 0:17:52.040
<v Speaker 10>The stock was up.

0:17:53.080 --> 0:17:55.560
<v Speaker 9>It was almost up to two hundred or so, and

0:17:55.720 --> 0:17:58.760
<v Speaker 9>since then it's plummeted from that high. It was trading

0:17:58.800 --> 0:18:01.320
<v Speaker 9>at something like two hundred and four twenty seven x earnings.

0:18:01.320 --> 0:18:04.080
<v Speaker 9>Then now it's down to about sixty seven x earnings,

0:18:04.080 --> 0:18:07.320
<v Speaker 9>which is down from its five year average of ninety

0:18:07.359 --> 0:18:09.760
<v Speaker 9>eight x earnings. But it's still one of the most

0:18:09.760 --> 0:18:11.920
<v Speaker 9>expensive stocks on the S and P. So a lot

0:18:11.960 --> 0:18:16.679
<v Speaker 9>of a lot of investors are wondering where that disconnect

0:18:16.800 --> 0:18:19.159
<v Speaker 9>is and whether it's really worth what it said it

0:18:19.240 --> 0:18:21.360
<v Speaker 9>was going to be able to deliver.

0:18:21.320 --> 0:18:22.679
<v Speaker 3>This seems like a big one, a big one for

0:18:22.720 --> 0:18:25.359
<v Speaker 3>Market's Bloomberg's the set Chapman who covers Panenter but also

0:18:25.560 --> 0:18:28.719
<v Speaker 3>defense tech bench Capital. Thank you very much. Indeed, now

0:18:28.760 --> 0:18:31.879
<v Speaker 3>coming up, Apple's changing the way if you buyas devices,

0:18:32.160 --> 0:18:35.439
<v Speaker 3>why monthly payments could become the new normal?

0:18:35.760 --> 0:18:37.920
<v Speaker 2>Power On. This is Bloomberg Tech.

0:18:51.640 --> 0:18:54.320
<v Speaker 3>Preak check on Amazon shares at all time highs. They've

0:18:54.400 --> 0:18:57.320
<v Speaker 3>joined the three trillion dollar market cap club and this

0:18:57.359 --> 0:18:59.479
<v Speaker 3>is the biggest two day game they've posted since two

0:18:59.520 --> 0:19:02.200
<v Speaker 3>thousand and nine. After earnings later in the hour, we're

0:19:02.200 --> 0:19:05.960
<v Speaker 3>going to speak to AWSCO, maccarm and AWS, Amazon's key

0:19:06.000 --> 0:19:10.399
<v Speaker 3>cloud computing unit. Apple is making a big bet that

0:19:10.480 --> 0:19:13.280
<v Speaker 3>customers would rather pay by the month than all at once.

0:19:13.359 --> 0:19:16.159
<v Speaker 3>It's the new Apple Upgrade program, designed to make your

0:19:16.240 --> 0:19:19.880
<v Speaker 3>upgrading your device film more like a subscription. Bloomas Mark

0:19:19.920 --> 0:19:22.920
<v Speaker 3>German here with the latest power on from this weekend.

0:19:23.160 --> 0:19:25.520
<v Speaker 3>It's been really interesting to see the rollout of this.

0:19:25.920 --> 0:19:28.399
<v Speaker 3>What are you trying to capture in this this power on? Like,

0:19:28.440 --> 0:19:29.800
<v Speaker 3>what where do you think this is headed?

0:19:30.960 --> 0:19:32.960
<v Speaker 11>Yeah, the fact of the matter is is that you're

0:19:32.960 --> 0:19:35.760
<v Speaker 11>going to see a change over time from sticker prices

0:19:36.080 --> 0:19:39.359
<v Speaker 11>to subscription prices. The carriers have been trying this for years.

0:19:39.640 --> 0:19:44.560
<v Speaker 11>It's been extremely successful for the carriers in the US. Internationally, obviously,

0:19:44.600 --> 0:19:47.320
<v Speaker 11>Apple has gotten on a really good schedule with its

0:19:47.359 --> 0:19:51.680
<v Speaker 11>new new products, obviously new iPhones and Apple watches every September.

0:19:52.000 --> 0:19:56.040
<v Speaker 11>The iPad cadence has really become fall in Spring the Mac.

0:19:56.119 --> 0:19:58.879
<v Speaker 11>You know, depending on which Mac you get, you're thinking

0:19:58.920 --> 0:20:01.360
<v Speaker 11>fall or spring as well well. And so there are

0:20:01.720 --> 0:20:05.120
<v Speaker 11>upgrade cycles for consumers. Right, maybe I'm going to get

0:20:05.119 --> 0:20:06.520
<v Speaker 11>a new phone every three years, maybe I'm going to

0:20:06.520 --> 0:20:08.560
<v Speaker 11>get a new computer every five years. But making it

0:20:08.600 --> 0:20:11.600
<v Speaker 11>a subscription is going to get people into getting a

0:20:11.640 --> 0:20:12.800
<v Speaker 11>new phone more frequently.

0:20:13.280 --> 0:20:15.280
<v Speaker 10>They could give those phones then back to Apple.

0:20:15.440 --> 0:20:18.920
<v Speaker 11>Apple could resell them on secondary markets. They can reuse

0:20:18.960 --> 0:20:21.840
<v Speaker 11>the parts for new devices, right, which creates a big

0:20:21.880 --> 0:20:24.760
<v Speaker 11>cost savings and improve margins for them there too as

0:20:24.760 --> 0:20:27.680
<v Speaker 11>well as that secondary phone business. And then the consumer

0:20:27.800 --> 0:20:30.240
<v Speaker 11>is paying on a recurring schedule, so the payments to

0:20:30.280 --> 0:20:33.240
<v Speaker 11>Apple never stop and over the long term you're going

0:20:33.280 --> 0:20:36.480
<v Speaker 11>to see Apple creating far more revenue in profit because

0:20:36.480 --> 0:20:38.600
<v Speaker 11>of this program. In fact, Wall Street has been begging

0:20:38.640 --> 0:20:41.080
<v Speaker 11>Apple to do this for over half a decade. And

0:20:41.119 --> 0:20:42.240
<v Speaker 11>so it's a long time coming.

0:20:43.000 --> 0:20:45.399
<v Speaker 3>So this weekend I did my first Apple up grade.

0:20:45.400 --> 0:20:48.639
<v Speaker 3>I took my twenty twenty Macair with an M one chip,

0:20:48.960 --> 0:20:51.680
<v Speaker 3>and I took the latest Macare thirteen inches with an

0:20:51.800 --> 0:20:54.600
<v Speaker 3>M five chip, and I did the math and it's

0:20:54.640 --> 0:20:57.479
<v Speaker 3>a great deal. Basically I can trade in still and

0:20:57.560 --> 0:21:00.240
<v Speaker 3>still do the Klana Apple up grade. The thing that

0:21:00.320 --> 0:21:01.919
<v Speaker 3>jumped out is I got to wait three weeks for

0:21:01.960 --> 0:21:04.200
<v Speaker 3>that Macare. That's still an issue.

0:21:05.119 --> 0:21:07.520
<v Speaker 11>Yes, So as we wrote in power On as well

0:21:07.560 --> 0:21:10.120
<v Speaker 11>this weekend, you know, the Max Studio and the Macmini,

0:21:10.240 --> 0:21:13.480
<v Speaker 11>those are no longer the only Max with extreme supply shortages.

0:21:13.840 --> 0:21:16.000
<v Speaker 11>Now it's the most popular Mac which is of course

0:21:16.000 --> 0:21:18.439
<v Speaker 11>the MacBook Air, and new orders of the macbookare are

0:21:18.440 --> 0:21:19.240
<v Speaker 11>not arriving at.

0:21:19.119 --> 0:21:20.880
<v Speaker 10>This point until early September.

0:21:21.119 --> 0:21:24.159
<v Speaker 11>So you're seeing the supply chain shortage shift from niche

0:21:24.160 --> 0:21:26.439
<v Speaker 11>computers to the company's most popular computer.

0:21:26.520 --> 0:21:27.919
<v Speaker 10>So that's not necessarily great.

0:21:28.080 --> 0:21:30.040
<v Speaker 11>And so what they're doing now in terms of their

0:21:30.080 --> 0:21:34.320
<v Speaker 11>marketing online and in stores, they're pushing people towards the

0:21:34.359 --> 0:21:38.640
<v Speaker 11>base level M five MacBook Pro, the fourteen inch, which

0:21:38.640 --> 0:21:40.880
<v Speaker 11>they have a lot of supply of still, and they

0:21:40.880 --> 0:21:43.200
<v Speaker 11>want to sort of dwindle that down because they're launching

0:21:43.240 --> 0:21:45.720
<v Speaker 11>the new M six version in just a couple months

0:21:45.720 --> 0:21:46.040
<v Speaker 11>from now.

0:21:46.840 --> 0:21:47.919
<v Speaker 2>Mark just very quickly.

0:21:47.960 --> 0:21:50.360
<v Speaker 3>We have a story on the terminal about how Apple's

0:21:50.520 --> 0:21:52.679
<v Speaker 3>SerII stacks up against the new Chatch GPT.

0:21:53.560 --> 0:21:54.280
<v Speaker 2>What do we need to know?

0:21:55.359 --> 0:21:58.560
<v Speaker 11>Apple has a huge advantage by having SERIAI built into

0:21:58.720 --> 0:22:02.040
<v Speaker 11>every iPhone now that can run iOS twenty seven and

0:22:02.080 --> 0:22:05.960
<v Speaker 11>the latest large language models. It's competent, right. The promise

0:22:06.000 --> 0:22:08.360
<v Speaker 11>of Siri has been there for fifteen years. It's never

0:22:08.440 --> 0:22:11.840
<v Speaker 11>really worked. Now it works. It's nowhere near as good

0:22:12.000 --> 0:22:15.879
<v Speaker 11>or as capable as Chad GPT, but the floor has

0:22:15.920 --> 0:22:16.800
<v Speaker 11>improved tremendously.

0:22:17.440 --> 0:22:19.840
<v Speaker 2>Bloombergs smart German with the latest power on. Thank you

0:22:19.960 --> 0:22:20.320
<v Speaker 2>very much.

0:22:21.480 --> 0:22:24.960
<v Speaker 3>Amazon has blown past three trillion dollars in market value

0:22:25.040 --> 0:22:29.440
<v Speaker 3>after its cloud unit AWS drove big earnings performance last week.

0:22:29.640 --> 0:22:31.960
<v Speaker 3>The company says it's AI business is hit a revenue

0:22:32.040 --> 0:22:35.080
<v Speaker 3>run rate of twenty five billion dollars, and it's chip

0:22:35.119 --> 0:22:38.800
<v Speaker 3>business not the same milestone. Amazon also raised its capex

0:22:38.840 --> 0:22:41.760
<v Speaker 3>plans for this year to two hundred and twenty billion

0:22:41.760 --> 0:22:45.440
<v Speaker 3>dollars as it joins the chorus of tech companies backing

0:22:45.760 --> 0:22:50.240
<v Speaker 3>American Open Models AWS CEO Matt Garman joins US now

0:22:50.440 --> 0:22:53.439
<v Speaker 3>on Bloomberg Television and Bloomberg Radio. Matt, it's great to

0:22:53.440 --> 0:22:55.639
<v Speaker 3>have you back on the show. You know, this is

0:22:55.680 --> 0:22:58.479
<v Speaker 3>the fastest growth for AWS in nearly five years. But

0:22:58.680 --> 0:23:02.359
<v Speaker 3>to kind of capture moment in time, could we start

0:23:02.440 --> 0:23:04.520
<v Speaker 3>by talking about how much of that was driven by

0:23:04.520 --> 0:23:07.520
<v Speaker 3>the frontier labs, the big ones Open AI and Thropic,

0:23:08.320 --> 0:23:11.320
<v Speaker 3>or how much of it was something broader across the enterprise,

0:23:11.520 --> 0:23:12.440
<v Speaker 3>something bigger in AI.

0:23:13.480 --> 0:23:16.240
<v Speaker 12>Yeah, it's actually across the board this growth that we're

0:23:16.240 --> 0:23:19.119
<v Speaker 12>seeing right now, and much of that is some of

0:23:19.119 --> 0:23:21.800
<v Speaker 12>the AI labs that are building their models on top

0:23:21.840 --> 0:23:24.639
<v Speaker 12>of AWS, but actually a lot of that growth is

0:23:24.680 --> 0:23:28.200
<v Speaker 12>spread across all of our startup and enterprise customers, where

0:23:28.280 --> 0:23:32.000
<v Speaker 12>AI is really impacting almost every single industry. And so

0:23:32.040 --> 0:23:35.439
<v Speaker 12>whether it's financial services companies or healthcare companies, or retail

0:23:35.520 --> 0:23:39.720
<v Speaker 12>companies or media companies, Really the growth is coming across

0:23:39.760 --> 0:23:43.840
<v Speaker 12>the board as companies of all different sizes in all

0:23:43.840 --> 0:23:48.280
<v Speaker 12>different industries are looking to use AI to grow their business.

0:23:48.320 --> 0:23:50.600
<v Speaker 12>And so we're seeing a lot of growth from the

0:23:51.119 --> 0:23:54.720
<v Speaker 12>frontier labs. But for AWS, we're not like some others,

0:23:54.720 --> 0:23:57.560
<v Speaker 12>maybe that's concentrated on just one or two large customers,

0:23:57.800 --> 0:24:00.560
<v Speaker 12>but it's actually growth from a really broad set of customers,

0:24:00.960 --> 0:24:01.880
<v Speaker 12>which is nice to see.

0:24:03.400 --> 0:24:03.640
<v Speaker 5>Matt.

0:24:03.640 --> 0:24:06.439
<v Speaker 3>When AWS says the AI business has a revenue run

0:24:06.520 --> 0:24:08.680
<v Speaker 3>rate of twenty five billion dollars, what does that mean?

0:24:08.800 --> 0:24:10.360
<v Speaker 2>What does the figuring compass?

0:24:11.760 --> 0:24:16.639
<v Speaker 12>Yeah, that includes both training from very large companies like

0:24:16.680 --> 0:24:20.920
<v Speaker 12>Anthropic and open Ai and other large tech companies, as

0:24:20.920 --> 0:24:22.679
<v Speaker 12>well as many startups. But it also and then a

0:24:22.680 --> 0:24:25.760
<v Speaker 12>big chunk of that is really inferenced that that broad

0:24:25.800 --> 0:24:28.920
<v Speaker 12>swath of companies are doing. And so as companies think

0:24:28.920 --> 0:24:31.880
<v Speaker 12>about how they take models in something on Amazon Bedrock

0:24:32.119 --> 0:24:34.359
<v Speaker 12>and run those models to get value out of their business,

0:24:34.440 --> 0:24:36.840
<v Speaker 12>some of them are automating processes. Many of them are

0:24:36.920 --> 0:24:40.000
<v Speaker 12>running agentic workloads that they build on AWS to further

0:24:40.400 --> 0:24:44.480
<v Speaker 12>drive efficiencies in their business or deliver new customer experiences.

0:24:44.680 --> 0:24:47.320
<v Speaker 12>We consider all of that work, whether it's agent growth,

0:24:47.359 --> 0:24:51.000
<v Speaker 12>whether it's inference, and some of that is training new models,

0:24:51.040 --> 0:24:53.800
<v Speaker 12>all of that compasses the AI business for us.

0:24:54.800 --> 0:24:57.199
<v Speaker 3>I think I've asked you this question at various points

0:24:57.200 --> 0:25:00.719
<v Speaker 3>in time since you became AWS CEO. Is there a

0:25:00.720 --> 0:25:04.679
<v Speaker 3>percentage split right now? Training versus inference that you can

0:25:04.720 --> 0:25:05.000
<v Speaker 3>give me.

0:25:06.240 --> 0:25:09.000
<v Speaker 12>Yeah, you do. And it actually keeps shifting, I would

0:25:09.000 --> 0:25:11.760
<v Speaker 12>say more and more. It keeps shifting more towards inference.

0:25:11.800 --> 0:25:14.920
<v Speaker 12>I think we still see large training clusters being used

0:25:14.960 --> 0:25:18.120
<v Speaker 12>by a number of companies, but as these as these

0:25:18.119 --> 0:25:20.560
<v Speaker 12>models get really popular and really powerful, more and more

0:25:20.600 --> 0:25:24.240
<v Speaker 12>companies are integrating that inference into their workloads. So, you know,

0:25:24.280 --> 0:25:26.280
<v Speaker 12>I don't know the exact percentage today, but it keeps

0:25:26.280 --> 0:25:28.679
<v Speaker 12>shifting more and more towards inference, and we expect that

0:25:28.760 --> 0:25:32.000
<v Speaker 12>to continue as the economics makes sense where you really

0:25:32.080 --> 0:25:35.320
<v Speaker 12>want most of that cost and spend being where value

0:25:35.359 --> 0:25:36.800
<v Speaker 12>is being created for end customers.

0:25:38.440 --> 0:25:41.760
<v Speaker 3>Amazon's overall CAPEX number for this year is very big,

0:25:42.040 --> 0:25:44.560
<v Speaker 3>two hundred and twenty billion dollars, so it's up twenty billion,

0:25:44.640 --> 0:25:49.280
<v Speaker 3>and Andy Jesse explained that's mostly AI, but it also

0:25:49.320 --> 0:25:52.680
<v Speaker 3>accounts for higher memory pricing. Right, capex higher because the

0:25:52.720 --> 0:25:57.520
<v Speaker 3>cost environment's higher. But from aws's perspective, what's the trajectory

0:25:57.600 --> 0:26:00.760
<v Speaker 3>for next year? You expect that capex will be biggest

0:26:00.760 --> 0:26:02.719
<v Speaker 3>still because I think one of the things that Amazon

0:26:02.760 --> 0:26:04.840
<v Speaker 3>is quite clear about is that even at two hundred

0:26:04.840 --> 0:26:07.520
<v Speaker 3>and twenty billion, dollars, it might not be enough to

0:26:07.560 --> 0:26:08.560
<v Speaker 3>meet current demand.

0:26:09.640 --> 0:26:12.040
<v Speaker 12>Yeah, one of the things that we're quite excited about

0:26:12.119 --> 0:26:15.600
<v Speaker 12>is just the potential business for us is just massive,

0:26:15.640 --> 0:26:18.639
<v Speaker 12>and we see this as a huge opportunity for us

0:26:18.640 --> 0:26:21.600
<v Speaker 12>to really invest and help customers take advantage of the

0:26:21.640 --> 0:26:24.520
<v Speaker 12>AI opportunity, and so we will keep investing. We think

0:26:24.520 --> 0:26:27.919
<v Speaker 12>that there's a big opportunity for us and for customers,

0:26:28.600 --> 0:26:31.320
<v Speaker 12>And as Andy kind of called out last week, we

0:26:31.400 --> 0:26:34.119
<v Speaker 12>have really great insight into what that demand is going

0:26:34.160 --> 0:26:36.359
<v Speaker 12>to be from customers, and so when we invest and

0:26:36.359 --> 0:26:39.600
<v Speaker 12>we will keep investing in kepex next year as well,

0:26:40.480 --> 0:26:43.480
<v Speaker 12>we have great insight into when that revenue is going

0:26:43.520 --> 0:26:46.239
<v Speaker 12>to land. And so it's pretty well known for us.

0:26:46.280 --> 0:26:49.760
<v Speaker 12>And as Andy mentioned, much of our capacity has already

0:26:49.800 --> 0:26:52.040
<v Speaker 12>spoken through through the end of twenty seven and even

0:26:52.040 --> 0:26:54.520
<v Speaker 12>through much of twenty eight, and so as we're investing,

0:26:54.520 --> 0:26:57.760
<v Speaker 12>we're getting five year commitments from customers. We're getting these

0:26:57.800 --> 0:27:01.080
<v Speaker 12>long term commitments, and so we're out there making investments

0:27:01.119 --> 0:27:03.040
<v Speaker 12>to keep being able to grow the business and try

0:27:03.080 --> 0:27:06.200
<v Speaker 12>to meet customer demand. But as you say, today, demands

0:27:06.240 --> 0:27:11.159
<v Speaker 12>still significantly outstriped supply, and we're trying to build and

0:27:11.280 --> 0:27:13.720
<v Speaker 12>invest to keep up with what customers are asking for.

0:27:14.480 --> 0:27:16.879
<v Speaker 3>And so we expect the CAPEX number will be bigger

0:27:16.880 --> 0:27:18.240
<v Speaker 3>next year than it is this year.

0:27:19.840 --> 0:27:22.360
<v Speaker 12>I expect us to keep investing as we see opportunities,

0:27:22.400 --> 0:27:24.800
<v Speaker 12>and some of that will see how the market continues

0:27:24.840 --> 0:27:28.480
<v Speaker 12>to grow. But right now we're excited about our investments

0:27:28.520 --> 0:27:29.919
<v Speaker 12>and we'll continue to invest.

0:27:31.359 --> 0:27:33.919
<v Speaker 3>The chip business has also got a lot of momentum.

0:27:34.080 --> 0:27:36.120
<v Speaker 3>You know, Trainium for some time has been a big

0:27:36.119 --> 0:27:38.040
<v Speaker 3>part of the strategy, and we've discussed that. I've been

0:27:38.080 --> 0:27:40.200
<v Speaker 3>able to go to Austin spend a lot of time

0:27:40.200 --> 0:27:45.960
<v Speaker 3>in Annapurna Labs to look at the server design. Let's

0:27:45.960 --> 0:27:48.119
<v Speaker 3>start by asking what when you say the chip business

0:27:48.119 --> 0:27:50.040
<v Speaker 3>has a revenue run rate in twenty five billion dollars,

0:27:50.080 --> 0:27:52.399
<v Speaker 3>what does that mean? That is the business of selling

0:27:52.480 --> 0:27:56.200
<v Speaker 3>the chips to third parties or renting out capacity based

0:27:56.240 --> 0:27:56.840
<v Speaker 3>on the chips.

0:27:57.600 --> 0:28:00.560
<v Speaker 12>Yeah, renting out capacity based on those chips. Today we

0:28:00.640 --> 0:28:03.600
<v Speaker 12>run all of our own chips inside of the AWS

0:28:03.600 --> 0:28:06.680
<v Speaker 12>cloud and that includes both Trainium chips as well as

0:28:06.680 --> 0:28:11.000
<v Speaker 12>Graviton chips or which our general purpose processors. Graviton is

0:28:11.000 --> 0:28:14.520
<v Speaker 12>incredibly popular. We've been building Graviton for many years now,

0:28:14.560 --> 0:28:17.480
<v Speaker 12>and in fact, almost all of our large customers use

0:28:17.520 --> 0:28:20.240
<v Speaker 12>Graviton as some part of their deployment, and it's a

0:28:20.359 --> 0:28:23.880
<v Speaker 12>key part of how help customers deliver extra value by

0:28:23.880 --> 0:28:28.360
<v Speaker 12>lowering their costs. Trainium is wildly popular and as we've

0:28:28.400 --> 0:28:31.120
<v Speaker 12>mentioned a couple of times now, we're largely sold out

0:28:31.119 --> 0:28:34.399
<v Speaker 12>through the end of next year for Trainium capacity and

0:28:34.480 --> 0:28:37.720
<v Speaker 12>it's really really popular. We're starting to see both large

0:28:37.720 --> 0:28:40.920
<v Speaker 12>customers as well as startups really love the cost benefits

0:28:40.920 --> 0:28:43.800
<v Speaker 12>of running on Trainium and having that choice in a

0:28:43.840 --> 0:28:47.320
<v Speaker 12>cloud where you can run on purpose built processors like

0:28:47.360 --> 0:28:51.160
<v Speaker 12>Trainium that give you really great AI performance, together with

0:28:51.480 --> 0:28:54.360
<v Speaker 12>Nvidia chips for when you need some of that processing

0:28:54.680 --> 0:28:58.360
<v Speaker 12>for GPUs as well, gives customers the right mix that

0:28:58.360 --> 0:29:01.720
<v Speaker 12>they're oftentimes looking for, and so we find that combination

0:29:01.800 --> 0:29:05.160
<v Speaker 12>to be quite powerful and our Trainium three chips are

0:29:05.200 --> 0:29:06.560
<v Speaker 12>really really popular with customers.

0:29:06.640 --> 0:29:10.800
<v Speaker 3>Right now, we're live on Bloomberg Television and Bloomberg Radio.

0:29:10.920 --> 0:29:11.880
<v Speaker 2>This is Bloomberg Tech.

0:29:11.920 --> 0:29:14.560
<v Speaker 3>In conversation with Matt garm and, the CEO of AWS,

0:29:15.280 --> 0:29:18.880
<v Speaker 3>explain that the difference in economics between capacity based on

0:29:18.960 --> 0:29:23.200
<v Speaker 3>Trainium and capacity based on Blackwell, for example, like what

0:29:23.360 --> 0:29:26.680
<v Speaker 3>is the selling point of going with Amazon Silicon.

0:29:28.560 --> 0:29:29.320
<v Speaker 2>Yeah, we think that.

0:29:29.800 --> 0:29:32.240
<v Speaker 12>Look, we think customers want choice, and so we offer

0:29:33.040 --> 0:29:36.360
<v Speaker 12>great offerings for both products. Some customers really like to

0:29:36.360 --> 0:29:39.760
<v Speaker 12>build on in Vidia GPUs and those are fantastic, and

0:29:39.800 --> 0:29:43.120
<v Speaker 12>we're one of Nvidia's absolute largest customers in the world.

0:29:43.320 --> 0:29:45.440
<v Speaker 12>We've been offering in Vidia GPUs in the cloud for

0:29:45.760 --> 0:29:50.760
<v Speaker 12>over a decade now, and AWS is the most scalable

0:29:51.080 --> 0:29:54.880
<v Speaker 12>and secure and stable place to run in Vidia servers anywhere,

0:29:55.520 --> 0:29:57.840
<v Speaker 12>and so we're very excited about that business. But when

0:29:57.880 --> 0:30:00.400
<v Speaker 12>you add Trainium, we're able to lower the ca for

0:30:00.480 --> 0:30:03.680
<v Speaker 12>many workloads, and so customers like to have that opportunity

0:30:03.680 --> 0:30:06.760
<v Speaker 12>where they can take these really massively scaled AI workloads

0:30:07.040 --> 0:30:09.600
<v Speaker 12>or inference lart workloads where we can tune them because

0:30:09.600 --> 0:30:12.760
<v Speaker 12>we kind of control that whole stack. We control the

0:30:12.800 --> 0:30:15.960
<v Speaker 12>training chips, the data centers, and the inference stack and

0:30:16.000 --> 0:30:18.880
<v Speaker 12>can really tune that whole stack. And that helps customers

0:30:18.920 --> 0:30:22.280
<v Speaker 12>lower their costs and improve performance. And we're really starting

0:30:22.280 --> 0:30:26.480
<v Speaker 12>to see that flywheel go where model providers of all

0:30:26.520 --> 0:30:30.240
<v Speaker 12>types and customers love that they get that value and

0:30:30.320 --> 0:30:33.680
<v Speaker 12>great performance. And we think that that's a real flywheel

0:30:33.680 --> 0:30:35.640
<v Speaker 12>that we can keep turning and are quite excited about

0:30:35.920 --> 0:30:37.200
<v Speaker 12>continuing to grow that business.

0:30:38.080 --> 0:30:39.840
<v Speaker 2>Could you make it tangible for me? What is the

0:30:39.840 --> 0:30:40.479
<v Speaker 2>cost saving?

0:30:42.120 --> 0:30:44.480
<v Speaker 12>It's workload by workloads, so there's not like a specific

0:30:44.520 --> 0:30:48.840
<v Speaker 12>cost savings, but for workloads where we see this optimization happen,

0:30:49.120 --> 0:30:52.680
<v Speaker 12>customers can oftentimes save twenty thirty percent off of their

0:30:52.680 --> 0:30:54.720
<v Speaker 12>inference costs when they run that on Trainum.

0:30:54.760 --> 0:30:55.160
<v Speaker 2>Interesting.

0:30:56.240 --> 0:30:59.760
<v Speaker 3>Would you consider selling the chips outright as opposed to

0:31:00.240 --> 0:31:02.240
<v Speaker 3>capacity based on them.

0:31:02.360 --> 0:31:04.600
<v Speaker 12>Yeah, we've mentioned that we might consider that in the future.

0:31:04.680 --> 0:31:06.480
<v Speaker 12>Right now, we have so much demand inside of the

0:31:06.520 --> 0:31:09.960
<v Speaker 12>AWS cloud that we really love that business, and so

0:31:10.040 --> 0:31:12.719
<v Speaker 12>that's the place we're focusing right now. But it's an

0:31:12.720 --> 0:31:15.720
<v Speaker 12>interesting opportunity and it's something that we would definitely consider

0:31:15.760 --> 0:31:16.240
<v Speaker 12>in the future.

0:31:17.240 --> 0:31:20.800
<v Speaker 3>The other big development was AWS signing the open Weights letter.

0:31:21.640 --> 0:31:25.600
<v Speaker 3>You yourself communicated that Why was it important to you,

0:31:25.680 --> 0:31:27.960
<v Speaker 3>Matt to sign that and participate in that?

0:31:29.160 --> 0:31:29.400
<v Speaker 5>Yeah?

0:31:29.800 --> 0:31:31.959
<v Speaker 12>I think that Again, a lot of this boils down

0:31:32.000 --> 0:31:35.960
<v Speaker 12>to customers wanting choice, and the big frontier labs have

0:31:36.040 --> 0:31:39.000
<v Speaker 12>awesome models today, whether it's open AI running on bedrock,

0:31:39.240 --> 0:31:42.400
<v Speaker 12>whether it's Anthropic running on bedrock, Customers really love using

0:31:42.440 --> 0:31:45.800
<v Speaker 12>those models. But they also like being able to customize models,

0:31:45.840 --> 0:31:49.560
<v Speaker 12>and so having a broad ecosystem of open weights models,

0:31:50.120 --> 0:31:53.720
<v Speaker 12>I think is incredibly important for innovation. It's important for

0:31:53.760 --> 0:31:56.200
<v Speaker 12>our customers, and so you think about things like the

0:31:56.280 --> 0:31:59.640
<v Speaker 12>Nematron models from Nvideo, or you think about Kimmi three

0:31:59.800 --> 0:32:02.400
<v Speaker 12>or some of the Chinese open weights models. Those really

0:32:02.440 --> 0:32:05.240
<v Speaker 12>help customers be able to customize to their own data

0:32:05.880 --> 0:32:08.720
<v Speaker 12>and really be able to innovate. And so we think

0:32:08.720 --> 0:32:12.400
<v Speaker 12>that it's important to not overlegislate there and give that

0:32:12.480 --> 0:32:15.480
<v Speaker 12>flexibility for customers to use whichever models they find to

0:32:15.520 --> 0:32:17.920
<v Speaker 12>be the best fit. And that's actually why we really

0:32:17.920 --> 0:32:20.200
<v Speaker 12>focus on having all of those available inside of Bedrock,

0:32:20.240 --> 0:32:22.400
<v Speaker 12>so that customers can choose which ones they want to use.

0:32:22.920 --> 0:32:25.720
<v Speaker 3>Not to over legislate. I mean, I had a conversation

0:32:25.920 --> 0:32:29.240
<v Speaker 3>with Jensen Wong the day the letter was posted about

0:32:29.680 --> 0:32:31.840
<v Speaker 3>why then, what is the rationale? It does seem like

0:32:31.880 --> 0:32:35.760
<v Speaker 3>the concern is that the US government overly regulates open models.

0:32:35.840 --> 0:32:37.560
<v Speaker 3>Was that sort of a motivation for you?

0:32:39.360 --> 0:32:41.280
<v Speaker 12>Yeah, I think that's the concern, and I think you

0:32:41.400 --> 0:32:43.160
<v Speaker 12>just want to make sure that we kind of have

0:32:43.240 --> 0:32:45.640
<v Speaker 12>an even playing field, and it doesn't mean that there's

0:32:45.680 --> 0:32:47.480
<v Speaker 12>none by the way, I think there should probably be

0:32:47.920 --> 0:32:50.600
<v Speaker 12>the same level of oversight of both frontier models as

0:32:50.600 --> 0:32:52.240
<v Speaker 12>well as open weights models, and I think we should

0:32:52.280 --> 0:32:55.560
<v Speaker 12>have a consistent framework, but kind of making sure that

0:32:56.560 --> 0:32:59.560
<v Speaker 12>folks realize how important that is to what customers and

0:32:59.600 --> 0:33:01.920
<v Speaker 12>are into is out there building on and that those

0:33:01.960 --> 0:33:03.640
<v Speaker 12>open weights are a key building block of that, and

0:33:03.720 --> 0:33:06.800
<v Speaker 12>we want to make sure that, however legislation lands, it

0:33:06.880 --> 0:33:08.280
<v Speaker 12>lands evenly across all of those.

0:33:09.640 --> 0:33:12.160
<v Speaker 3>Matt a question from our Bloomberg Tech audience for you

0:33:12.320 --> 0:33:16.320
<v Speaker 3>is aws's attitude towards Kimmy K three. The open weight

0:33:16.440 --> 0:33:20.240
<v Speaker 3>was released touly twenty seventh. You know, given the platform's

0:33:20.240 --> 0:33:22.280
<v Speaker 3>AWS offers, what is the plan there.

0:33:23.800 --> 0:33:25.080
<v Speaker 5>Yeah, it's a great model.

0:33:25.120 --> 0:33:28.640
<v Speaker 12>The team there has done a really nice job and

0:33:28.640 --> 0:33:32.400
<v Speaker 12>we see really great performance there, and we'll continue to

0:33:32.400 --> 0:33:36.400
<v Speaker 12>support in bedrock offering the Kimmi models. I think the

0:33:36.480 --> 0:33:38.800
<v Speaker 12>interesting thing is also as you see many of these

0:33:38.800 --> 0:33:41.760
<v Speaker 12>open weight companies, they're starting to think about how they

0:33:41.800 --> 0:33:43.880
<v Speaker 12>make money in these models as well, and so they're

0:33:43.920 --> 0:33:47.120
<v Speaker 12>starting to introduce licensing around some of these open weights

0:33:47.120 --> 0:33:50.040
<v Speaker 12>models as when customers want to run them in a

0:33:50.040 --> 0:33:52.720
<v Speaker 12>cloud environment, and so I anticipate that there's going to

0:33:52.760 --> 0:33:54.920
<v Speaker 12>be a blending of some of these models too, where

0:33:55.120 --> 0:33:57.720
<v Speaker 12>these companies won't continue to spend lots of money and

0:33:57.720 --> 0:34:00.720
<v Speaker 12>then just offer their IP up to the world. They're

0:34:00.760 --> 0:34:02.640
<v Speaker 12>going to be licensing models. They're going to have ways

0:34:02.680 --> 0:34:04.520
<v Speaker 12>of making money on those as well. And you're starting

0:34:04.560 --> 0:34:06.920
<v Speaker 12>to see that with some of the open weights models

0:34:06.960 --> 0:34:10.279
<v Speaker 12>actually having some licensing around them. And so we'll keep

0:34:10.480 --> 0:34:14.560
<v Speaker 12>supporting these in Bedrock and working with those companies in

0:34:14.640 --> 0:34:17.759
<v Speaker 12>order to offer these in the best possible way.

0:34:18.760 --> 0:34:21.120
<v Speaker 3>That's why I'd like to end the conversation the business

0:34:21.160 --> 0:34:25.200
<v Speaker 3>case and economic opportunity for open models from aws's perspective.

0:34:25.640 --> 0:34:28.160
<v Speaker 3>So if the model makers themselves would like to make money,

0:34:28.360 --> 0:34:31.080
<v Speaker 3>you know, how does AWS see that going in your

0:34:31.160 --> 0:34:35.000
<v Speaker 3>favor to make money from wide use of open models?

0:34:35.080 --> 0:34:38.560
<v Speaker 12>Yeah, yeah, I mean we basically that is exactly what AWS.

0:34:38.640 --> 0:34:41.640
<v Speaker 12>It's a great platform for companies to come and offer

0:34:41.719 --> 0:34:46.160
<v Speaker 12>their IP and their capabilities to the world, and so everybody,

0:34:46.200 --> 0:34:49.640
<v Speaker 12>whether they're startups, whether they're governments, whether they're large enterprises,

0:34:49.680 --> 0:34:52.759
<v Speaker 12>everybody can build on top of AWS. And if they

0:34:52.800 --> 0:34:56.560
<v Speaker 12>have choice and they have access to these different models,

0:34:56.800 --> 0:34:58.920
<v Speaker 12>it allows companies to be able to come and monetize

0:34:58.960 --> 0:35:01.600
<v Speaker 12>them and sell when they have you to customers to

0:35:01.640 --> 0:35:03.600
<v Speaker 12>be able to offer value when they think that they

0:35:03.600 --> 0:35:06.040
<v Speaker 12>can go at lower prices or better performance. And so

0:35:06.120 --> 0:35:09.640
<v Speaker 12>having that open platform allows everybody to have that competitive chance.

0:35:10.120 --> 0:35:13.160
<v Speaker 12>And AWS is a fantastic partner for all these model

0:35:13.160 --> 0:35:15.440
<v Speaker 12>providers to get access to the broad set of customers.

0:35:15.480 --> 0:35:19.080
<v Speaker 12>And so it's a great ecosystem where model providers benefit

0:35:19.160 --> 0:35:21.640
<v Speaker 12>and customers benefit by having all these in the same place.

0:35:22.719 --> 0:35:26.959
<v Speaker 3>Matt Garman, COOFAWS, Amazon Web Services back on Bloomberg Tech,

0:35:27.000 --> 0:35:31.960
<v Speaker 3>thank you very much. Now, coming up, home battery startup

0:35:32.040 --> 0:35:35.799
<v Speaker 3>base Power has a new valuation thirteen billion dollars. We're

0:35:35.800 --> 0:35:38.560
<v Speaker 3>going to talk to CEO Zach Dell about its latest

0:35:38.560 --> 0:35:41.359
<v Speaker 3>funding round and the launch of its US built.

0:35:41.400 --> 0:35:44.879
<v Speaker 2>Home powering battery. That's next. This is Bloomberg Tech.

0:35:58.160 --> 0:36:01.760
<v Speaker 3>Rising demand for electricity in US is fueling investor interest

0:36:02.160 --> 0:36:05.480
<v Speaker 3>in backing battery startups, among them base Power, which just

0:36:05.560 --> 0:36:08.480
<v Speaker 3>closed a one billion dollar Series D round at a

0:36:08.560 --> 0:36:12.440
<v Speaker 3>thirteen billion dollar post money valuation. Base Power CEO and

0:36:12.440 --> 0:36:15.439
<v Speaker 3>co founder Zach Dell joins US now in Bloomberg Tech.

0:36:15.880 --> 0:36:19.719
<v Speaker 3>A big round, right, and you know, inevitably part of

0:36:19.800 --> 0:36:22.960
<v Speaker 3>needing the funds is to scale to get moving on

0:36:23.200 --> 0:36:25.799
<v Speaker 3>the operations side. But let's start there, Like, what is

0:36:25.800 --> 0:36:29.600
<v Speaker 3>it you want to do in the first instance, Zac, Yeah, Well.

0:36:29.440 --> 0:36:32.560
<v Speaker 13>We really want to scale energy infrastructure to meet rising

0:36:32.560 --> 0:36:36.359
<v Speaker 13>demand for power. So we build a hardware product that

0:36:36.680 --> 0:36:39.680
<v Speaker 13>we install on homes, and we leverage a home battery

0:36:39.760 --> 0:36:42.640
<v Speaker 13>that we design and manufacture ourselves, and we leverage those

0:36:42.640 --> 0:36:45.080
<v Speaker 13>batteries to support the grid in times of high demand.

0:36:45.320 --> 0:36:47.879
<v Speaker 5>So we're adding to basilely to the system to meet this.

0:36:48.400 --> 0:36:52.920
<v Speaker 13>Rising demand from AI electrification, of heavy industry, of transportation.

0:36:53.280 --> 0:36:56.560
<v Speaker 13>We're adding capacity the system through distributed batteries. The grid

0:36:56.600 --> 0:36:59.040
<v Speaker 13>is able to use those batteries when demand is high,

0:36:59.040 --> 0:37:01.680
<v Speaker 13>and the homeowners that are have those batteries in sold

0:37:01.719 --> 0:37:03.520
<v Speaker 13>on their home are able to leverage those batteries when

0:37:03.520 --> 0:37:04.000
<v Speaker 13>the grid.

0:37:03.840 --> 0:37:06.080
<v Speaker 5>Goes out, Zach.

0:37:06.320 --> 0:37:10.320
<v Speaker 3>This is a residential product right at a basic level.

0:37:11.280 --> 0:37:15.000
<v Speaker 3>How tied is your fate and future to the adoption

0:37:15.040 --> 0:37:17.680
<v Speaker 3>of solar in this country?

0:37:18.360 --> 0:37:19.439
<v Speaker 5>Not really at all.

0:37:19.880 --> 0:37:23.800
<v Speaker 13>Solar is a generation technology much like wind or gas

0:37:23.920 --> 0:37:26.719
<v Speaker 13>or coal, and batteries are really mork into a.

0:37:26.640 --> 0:37:28.520
<v Speaker 5>Transmission and distribution technology.

0:37:28.640 --> 0:37:30.680
<v Speaker 13>So the way to think about it is that polls

0:37:30.680 --> 0:37:34.000
<v Speaker 13>and wires move energy through space and batteries move energy

0:37:34.120 --> 0:37:37.080
<v Speaker 13>through time, and so the battery is really a utilization

0:37:37.600 --> 0:37:40.600
<v Speaker 13>tool that is helping increase the capacity factor of the

0:37:40.640 --> 0:37:43.960
<v Speaker 13>system and bring down costs for everybody.

0:37:44.520 --> 0:37:47.920
<v Speaker 3>So I ask you that, because the energy or electricity

0:37:48.000 --> 0:37:52.359
<v Speaker 3>generation isn't the problem here, what is the problem that

0:37:52.400 --> 0:37:53.960
<v Speaker 3>the base power is trying to solve for?

0:37:55.280 --> 0:37:57.760
<v Speaker 13>Yeah, we've got a utilization problem on the grid today.

0:37:57.800 --> 0:38:00.720
<v Speaker 13>So the grid is really built for peak demand, which

0:38:00.760 --> 0:38:04.600
<v Speaker 13>means that most of the time we're not utilizing all

0:38:04.600 --> 0:38:07.600
<v Speaker 13>of the infrastructure, which ends up being very expensive for

0:38:07.760 --> 0:38:10.920
<v Speaker 13>rate payers or consumers of electricity. So by deploying batteries

0:38:11.280 --> 0:38:13.840
<v Speaker 13>to the edge of the grid, you're able to increase

0:38:13.840 --> 0:38:17.680
<v Speaker 13>the utilization, increase the capacity factor, and thereby lower costs

0:38:17.760 --> 0:38:19.040
<v Speaker 13>for consumers.

0:38:20.239 --> 0:38:20.680
<v Speaker 2>At home.

0:38:20.800 --> 0:38:23.919
<v Speaker 3>I have a Tesla power Will three. It does rely

0:38:24.040 --> 0:38:27.640
<v Speaker 3>on solar to charge it during the daytime. But let's

0:38:27.640 --> 0:38:29.759
<v Speaker 3>talk about that as a comparison or a case study.

0:38:29.800 --> 0:38:33.280
<v Speaker 3>At the technology level, the base call is better.

0:38:33.360 --> 0:38:34.360
<v Speaker 2>How is different.

0:38:34.480 --> 0:38:37.920
<v Speaker 5>How it's really designed to be a grid resource.

0:38:37.960 --> 0:38:41.240
<v Speaker 13>So it's almost three maybe over three times the size,

0:38:41.360 --> 0:38:45.279
<v Speaker 13>it's installed much faster, it's actually owned and operated by

0:38:45.360 --> 0:38:48.240
<v Speaker 13>us the company based power and used as a grid resource.

0:38:48.400 --> 0:38:50.040
<v Speaker 13>So when you go buy a home battery or a

0:38:50.040 --> 0:38:52.839
<v Speaker 13>home generator, it's owned by the homeowner and it's really

0:38:52.920 --> 0:38:56.319
<v Speaker 13>only there when the grid goes out. Our asset is

0:38:56.360 --> 0:38:59.960
<v Speaker 13>really a piece of grid infrastructure. And so in partnership

0:39:00.120 --> 0:39:03.279
<v Speaker 13>some of our utility partners, those batteries are used as

0:39:03.280 --> 0:39:06.759
<v Speaker 13>a grid resource by the utility in times of high

0:39:06.760 --> 0:39:09.520
<v Speaker 13>demand like I mentioned earlier, and then in the deregulated

0:39:09.560 --> 0:39:12.160
<v Speaker 13>markets in Texas and Illinois where we operate, we use

0:39:12.200 --> 0:39:15.200
<v Speaker 13>those batteries as a whole seal power asset. So we're

0:39:15.239 --> 0:39:18.400
<v Speaker 13>really operating a fleet of infrastructure. Whereas those companies that

0:39:18.440 --> 0:39:20.919
<v Speaker 13>sell batteries or generators, they're just trying to sell high

0:39:20.920 --> 0:39:23.680
<v Speaker 13>margin hardware outfront. And our view is that batteries are

0:39:23.719 --> 0:39:26.840
<v Speaker 13>really an infrastructure asset, and they're better sold and built

0:39:26.960 --> 0:39:29.880
<v Speaker 13>kind of as such, right, and not really as a

0:39:29.960 --> 0:39:30.800
<v Speaker 13>high margin product.

0:39:31.560 --> 0:39:31.719
<v Speaker 2>Zach.

0:39:31.719 --> 0:39:34.160
<v Speaker 3>I also feel like this is a Texas story from

0:39:34.160 --> 0:39:36.840
<v Speaker 3>the manufacturing standpoint, right, you know this is going to

0:39:36.920 --> 0:39:41.040
<v Speaker 3>be an American built technology. What were the challenges in

0:39:41.160 --> 0:39:45.960
<v Speaker 3>establishing Texas is the base, the supply chain, making this

0:39:46.239 --> 0:39:48.680
<v Speaker 3>an economically viable plan.

0:39:50.040 --> 0:39:53.000
<v Speaker 13>Well, Texas is a fantastic place to build companies and

0:39:53.080 --> 0:39:57.160
<v Speaker 13>also to establish manufacturing. And we see actually huge advantages

0:39:57.200 --> 0:40:00.839
<v Speaker 13>in co locating engineering and manufacturing because the team that

0:40:00.920 --> 0:40:03.560
<v Speaker 13>designs the product needs to be able to be very

0:40:03.560 --> 0:40:05.200
<v Speaker 13>close to the team that builds the product. And so

0:40:05.520 --> 0:40:09.040
<v Speaker 13>by coupling engineering and manufacturing and having that done by

0:40:09.080 --> 0:40:12.319
<v Speaker 13>one team, one company, and one location, you get a

0:40:12.320 --> 0:40:15.319
<v Speaker 13>bunch of efficiencies. It is much easier to iterate and

0:40:15.360 --> 0:40:16.280
<v Speaker 13>you can move a lot faster.

0:40:17.520 --> 0:40:19.800
<v Speaker 2>Should we talk about some of your investors in the round?

0:40:20.560 --> 0:40:23.640
<v Speaker 3>What was the importance of assembling that group beyond just

0:40:23.760 --> 0:40:26.600
<v Speaker 3>the checks that came in. What are you hoping to

0:40:27.280 --> 0:40:31.240
<v Speaker 3>leverage from their experiences and their expertise. We're showing ribit

0:40:31.520 --> 0:40:34.359
<v Speaker 3>Co two valor Ultimate Sense Capital D one.

0:40:35.480 --> 0:40:35.680
<v Speaker 5>Yeah.

0:40:35.719 --> 0:40:39.040
<v Speaker 13>I mean, we're building the company to compound over many decades,

0:40:39.080 --> 0:40:41.920
<v Speaker 13>and we want to establish a cap table that is

0:40:41.960 --> 0:40:44.640
<v Speaker 13>aligned with that kind of long term thinking. So these

0:40:44.640 --> 0:40:47.360
<v Speaker 13>investors who joined our capitable today and who've been investing

0:40:47.360 --> 0:40:49.480
<v Speaker 13>with us now for a couple of rounds, they see

0:40:49.520 --> 0:40:53.359
<v Speaker 13>the opportunity to scale the business across the US internationally

0:40:54.400 --> 0:40:56.440
<v Speaker 13>and do that over a long period of time. And

0:40:56.480 --> 0:41:00.879
<v Speaker 13>they add a lot of interesting insights perspectives, and they're

0:41:00.880 --> 0:41:03.799
<v Speaker 13>really are thought partners and our operating partners day to day.

0:41:03.800 --> 0:41:06.600
<v Speaker 5>So we're super lucky to have a whole.

0:41:06.360 --> 0:41:09.000
<v Speaker 13>Host of new investors joining the cap table and then

0:41:09.239 --> 0:41:12.320
<v Speaker 13>basically all of our majorgitcing investors reinvesting in the company.

0:41:13.920 --> 0:41:16.520
<v Speaker 3>Earlier in the program, we had Valor Atomic CEO Isiah

0:41:16.560 --> 0:41:19.840
<v Speaker 3>Taylor on you know, different part of the energy stack,

0:41:19.960 --> 0:41:24.120
<v Speaker 3>but we talked about how is America doing in this domain?

0:41:24.680 --> 0:41:24.839
<v Speaker 8>Right?

0:41:24.880 --> 0:41:27.880
<v Speaker 3>We talk a lot about China, how quickly China moves

0:41:28.200 --> 0:41:31.600
<v Speaker 3>I guess at the nation level and at the federal level,

0:41:31.680 --> 0:41:34.520
<v Speaker 3>are you getting everything you need, Zack to be able

0:41:34.560 --> 0:41:35.879
<v Speaker 3>to execute in the way you want?

0:41:35.920 --> 0:41:37.280
<v Speaker 2>Get the regulatory support?

0:41:38.640 --> 0:41:41.840
<v Speaker 13>Yeah, Look, I think there's a lot of energy being applied,

0:41:42.280 --> 0:41:45.239
<v Speaker 13>no pun intended to the space today, whether that's from

0:41:45.239 --> 0:41:49.560
<v Speaker 13>the regulators or policy makers, but also entrepreneurs like like

0:41:49.680 --> 0:41:51.680
<v Speaker 13>you know, our team at Base and the folks you

0:41:51.719 --> 0:41:53.759
<v Speaker 13>mentioned earlier, some of the other companies entering the space.

0:41:53.800 --> 0:41:56.000
<v Speaker 13>I mean, when we started Base three years ago, it

0:41:56.080 --> 0:41:58.680
<v Speaker 13>was kind of unique to start an energy technology company.

0:41:58.719 --> 0:42:00.400
<v Speaker 13>Either were not a lot of those out there, and

0:42:00.440 --> 0:42:02.000
<v Speaker 13>now they're very common and you see them pop up

0:42:02.000 --> 0:42:03.800
<v Speaker 13>all the time, and we love that. We think that

0:42:04.160 --> 0:42:07.399
<v Speaker 13>the industry needs more great entrepreneurs and teams solving hard

0:42:07.400 --> 0:42:10.480
<v Speaker 13>problems and helping us scale our capacity to consume electricity

0:42:10.480 --> 0:42:12.640
<v Speaker 13>in this country. So I think we're in a really

0:42:12.680 --> 0:42:15.799
<v Speaker 13>good spot and we've just got to keep executing relentlessly

0:42:16.160 --> 0:42:18.640
<v Speaker 13>and building great solutions to some of these hard problems

0:42:18.640 --> 0:42:19.520
<v Speaker 13>in the energy industry.

0:42:20.719 --> 0:42:23.360
<v Speaker 3>You're building across the electric stack, and I kind of

0:42:23.400 --> 0:42:25.960
<v Speaker 3>like read a lot about you and think you're pitching

0:42:25.960 --> 0:42:29.920
<v Speaker 3>yourselves as like a next generation utility almost, And I

0:42:29.960 --> 0:42:31.560
<v Speaker 3>think we should end the conversation with like, what is

0:42:31.560 --> 0:42:35.160
<v Speaker 3>your relationship like with the existing utilities with the rest

0:42:35.200 --> 0:42:37.400
<v Speaker 3>of the grid, because you're all kind of trying to

0:42:37.400 --> 0:42:39.040
<v Speaker 3>move together in the same direction.

0:42:40.160 --> 0:42:42.040
<v Speaker 13>Yeah, I wouldn't think of us really as a next

0:42:42.080 --> 0:42:44.560
<v Speaker 13>generation utility. I would think of us as a modern

0:42:44.680 --> 0:42:48.120
<v Speaker 13>energy technology company that is a provider of choice to

0:42:48.160 --> 0:42:51.000
<v Speaker 13>the utilities to scale their infrastructure. So we work very

0:42:51.000 --> 0:42:54.120
<v Speaker 13>closely with utilities with all kinds of utilities, with municipal

0:42:54.200 --> 0:42:57.440
<v Speaker 13>utilities like Austin Energy, with co ops like Coserve and Gveck,

0:42:57.760 --> 0:43:00.600
<v Speaker 13>with Investorr owned utilities like all Pass Electric and we

0:43:00.719 --> 0:43:03.839
<v Speaker 13>build them technology solutions so they can scale their infrastructure

0:43:04.120 --> 0:43:06.839
<v Speaker 13>and meet the moment, meet the rising demand that's coming

0:43:06.880 --> 0:43:09.440
<v Speaker 13>from AI and all these things that you discuss all

0:43:09.440 --> 0:43:13.960
<v Speaker 13>the time. There's an incredible inflection happening and energy consumption

0:43:14.000 --> 0:43:17.160
<v Speaker 13>in America, and it's never been more important for the

0:43:17.239 --> 0:43:21.440
<v Speaker 13>utilities to kind of step up and adapt, and we

0:43:21.520 --> 0:43:23.400
<v Speaker 13>are really helping them do that. And it's our honor

0:43:23.440 --> 0:43:25.560
<v Speaker 13>to work with the utilities. They're incredible partners and they

0:43:25.560 --> 0:43:28.759
<v Speaker 13>have a huge responsibility in this coming decade, and we're

0:43:28.800 --> 0:43:30.520
<v Speaker 13>really excited to partner with them.

0:43:30.880 --> 0:43:33.440
<v Speaker 3>Zach very quickly, what's it like trying to hire right

0:43:33.480 --> 0:43:34.680
<v Speaker 3>now the talent you need.

0:43:36.200 --> 0:43:40.480
<v Speaker 13>Look, I think that we have an incredibly mission driven

0:43:40.640 --> 0:43:43.920
<v Speaker 13>culture and company, and there's a lot of people out there,

0:43:44.000 --> 0:43:47.280
<v Speaker 13>Hopefully some of them are watching the show that feels

0:43:47.320 --> 0:43:53.240
<v Speaker 13>strongly that's scaling our capacity consumer electricity and really driving

0:43:53.280 --> 0:43:56.880
<v Speaker 13>towards energy abundance on the path to human prosperity is

0:43:56.880 --> 0:43:59.200
<v Speaker 13>one of the most uplifting and exciting missions that you

0:43:59.200 --> 0:44:01.360
<v Speaker 13>can work on. And so we've had the great fortune

0:44:01.360 --> 0:44:03.400
<v Speaker 13>of having tons of the best engineers in the world

0:44:03.760 --> 0:44:05.520
<v Speaker 13>come to base and say, Hey, this is the mission

0:44:05.520 --> 0:44:07.759
<v Speaker 13>I've got to be a part of and so we've

0:44:07.760 --> 0:44:10.840
<v Speaker 13>had a really great time, frankly building out an incredible

0:44:10.840 --> 0:44:14.680
<v Speaker 13>team Dan and Austin of just very mission aligned individuals,

0:44:14.920 --> 0:44:17.480
<v Speaker 13>and we're excited to continue to grow that team as

0:44:17.520 --> 0:44:18.600
<v Speaker 13>we conin to scale a business.

0:44:19.360 --> 0:44:21.799
<v Speaker 3>Zach Dell, co founder and CEO of Based Power, thank

0:44:21.840 --> 0:44:23.799
<v Speaker 3>you very much. In tea, there's a lot of other

0:44:23.840 --> 0:44:26.360
<v Speaker 3>tech headlines. It's time for talking tech and first start

0:44:26.640 --> 0:44:30.440
<v Speaker 3>South Korean AI chit designer DPEX secured funding it roughly

0:44:30.520 --> 0:44:34.080
<v Speaker 3>four times its previous valuation, buffing the firm to two

0:44:34.080 --> 0:44:37.560
<v Speaker 3>point two billion dollars. The news comes months after the startup,

0:44:37.600 --> 0:44:41.520
<v Speaker 3>which developed semiconductor architecture tailored for AI tasks, said it

0:44:41.560 --> 0:44:45.359
<v Speaker 3>secured over thirty mass production contracts from customers across eight

0:44:45.440 --> 0:44:50.360
<v Speaker 3>countries plus. According to a PwC survey the thousand finance executives,

0:44:50.640 --> 0:44:52.640
<v Speaker 3>eighty six percent of his ex said they thought AI

0:44:52.680 --> 0:44:56.320
<v Speaker 3>training was more valuable than an MBA for new hires.

0:44:56.520 --> 0:44:59.040
<v Speaker 3>Ninety one percent of those execs also said they were

0:44:59.040 --> 0:45:03.920
<v Speaker 3>willing to pay according increasing compensation for AI literate employees,

0:45:04.120 --> 0:45:07.080
<v Speaker 3>pointing to a growing urgency for firms to adopt AI

0:45:07.200 --> 0:45:10.640
<v Speaker 3>tech and Spider Man Brand New Day was the biggest

0:45:10.680 --> 0:45:14.080
<v Speaker 3>debut in Hollywood history, with the movie opening to three

0:45:14.200 --> 0:45:17.040
<v Speaker 3>hundred and sixty million dollars in the US and Canada.

0:45:17.360 --> 0:45:20.720
<v Speaker 3>Was a smashing success in China too, hauling the country

0:45:20.760 --> 0:45:24.080
<v Speaker 3>out of its post pandemic Hollywood slump and contributing to

0:45:24.120 --> 0:45:27.760
<v Speaker 3>a global box office debut of nine hundred twenty seven

0:45:28.000 --> 0:45:32.160
<v Speaker 3>million dollars. That does it for this edition of Bloomberg Tech.

0:45:32.239 --> 0:45:34.000
<v Speaker 3>These are the earnings that we're looking at this week.

0:45:34.160 --> 0:45:38.080
<v Speaker 3>Earnings continues. We've got Palenteer SpaceX first ever earnings report

0:45:38.120 --> 0:45:41.319
<v Speaker 3>and call as a public company, AMD a big one

0:45:41.320 --> 0:45:44.720
<v Speaker 3>in the AI infrastructure space, and then later a focus

0:45:44.840 --> 0:45:49.600
<v Speaker 3>on entertainment, the economy Paramount, Disney Uber all on Wednesday,

0:45:49.800 --> 0:45:53.319
<v Speaker 3>Sorry Disney and Uber on Wednesday Recap Everything on the.

0:45:53.320 --> 0:45:55.919
<v Speaker 2>Podcast was a pretty fast start to the week.

0:45:56.040 --> 0:45:58.400
<v Speaker 3>You know where to find it on the terminal, on

0:45:58.440 --> 0:46:02.799
<v Speaker 3>all Bloomberg platforms and online on Apple, Spotify, and iHeart.

0:46:02.880 --> 0:46:04.680
<v Speaker 2>Have a great week. This is Bloomberg Tech.