# Oracle’s Cloud Growth; Debate Around AI Risks

An episode of [Bloomberg Tech](https://invisiblenetwork.ai/podcast/1611767434), 11 Sep 2026, 43 min.

Audio: https://podtrac.com/pts/redirect.mp3/traffic.omny.fm/d/clips/e73c998e-6e60-432f-8610-ae210140c5b1/41764a4f-fc64-4e11-89ba-ae7c0030ab5e/f9e89c59-a563-4808-852d-b4c201496f38/audio.mp3?utm_source=Podcast&in_playlist=9caafc41-289c-4115-995d-ae7c0030ab75

Bloomberg’s Ed Ludlow breaks down Oracle's latest results and how they are helping ease concerns around the company's massive AI infrastructure buildout. Plus, the debate around AI's societal risks and the backlash against the global data center build-out grow louder. See omnystudio.com/listener for privacy information.

## Transcript

[0:02] Speaker 1: Bloomberg Audio Studios.
[0:04] Speaker 2: Podcasts. Radio.
[0:06] Speaker 3: News.
[0:13] Speaker 4: Bloomberg Tech is live from the heart of Silicon Valley
[0:17] Speaker 4: with Ed Ludlow in San Francisco.
[0:22] Speaker 5: This is Bloomberg Tech.
[0:23] Speaker 6: Coming up, Oracle's latest results help ease concerns around its
[0:27] Speaker 6: massive AI infrastructure build-out, plus the debate around AI's societal
[0:31] Speaker 6: risks grow louder, with the Pentagon's Emile Michael now pushing
[0:34] Speaker 6: back against those fears. And we'll discuss that debate, as
[0:38] Speaker 6: well as the data center build-out backlash, and much more.
[0:40] Speaker 6: with the CEO of venture capital firm Eclipse, Lior Susan.
[0:44] Speaker 6: Let's get right to Oracle. It's our main technology and
[0:47] Speaker 6: earnings story. And actually, the stock right now completely flat.
[0:50] Speaker 6: After the open, it hit a session high of more
[0:51] Speaker 6: than 8%, then went to negative territory. I want to
[0:54] Speaker 6: go through some of the big numbers. Cloud infrastructure in
[0:57] Speaker 6: the quarter, $ 7. 4 billion, a gain of 121% year on year.
[0:58] Speaker 6: The backlog, as measured by RPO, $ 664 billion. But CapEx
[1:00] Speaker 6: in that quarter that they reported... $ 28. 5 billion. Was that
[1:10] Speaker 6: the focus? They raised the full-year outlook for both revenue
[1:13] Speaker 6: and adjusted EPS. Let's get to Bloomberg's Brody Ford, who
[1:16] Speaker 6: leads our coverage of Oracle.
[1:17] Speaker 5: What's going on?
[1:19] Speaker 7: Seeing the share price gain ease this morning, I would
[1:22] Speaker 7: think it has to be the big spending in the quarter.
[1:24] Speaker 3: That's the CapEx number. Right, the CapEx in the quarter.
[1:27] Speaker 7: It came in higher than expected, and the margins went
[1:29] Speaker 7: down because cloud infrastructure is a much less profitable business
[1:33] Speaker 7: than their traditional database business.
[1:35] Speaker 3: But overall, it was a very clean quarter.
[1:38] Speaker 7: It pretty much showed that they are delivering the capacity
[1:41] Speaker 7: they promised. Revenue is ramping up. In general, these were
[1:44] Speaker 7: the figures you wanted to see if you're an investor.
[1:47] Speaker 6: I'm going to get through some of the numbers, then
[1:49] Speaker 6: we'll put the numbers to one side. And that is
[1:51] Speaker 6: debt on the balance sheet is $ 120 billion. Cash is
[1:54] Speaker 6: about $ 36 billion. So in a leverage context, they're at 3.4x, which,
[1:57] Speaker 6: I don't know, negative free cash flow is better than
[2:00] Speaker 6: expected in the quarter, right?
[2:01] Speaker 3: It's funny, right?
[2:02] Speaker 7: I mean, negative free cash flow doesn't sound good, but
[2:05] Speaker 7: it wasn't as negative as people expected.
[2:06] Speaker 6: Yes, not as bad as expected. Let's put the numbers
[2:10] Speaker 6: to one side. What is the story for Oracle? They're
[2:13] Speaker 6: trying to build data centers fast and get that capacity online.
[2:16] Speaker 6: Is it as simple as that?
[2:17] Speaker 7: Yeah, they're trying to transition from a traditional software business
[2:20] Speaker 7: to an AI data center for developer. And the question is,
[2:24] Speaker 7: is that a good business to be in? And they've
[2:26] Speaker 7: become a big kind of focus on this question of,
[2:30] Speaker 7: You know, is it worth going negative free cash flow
[2:32] Speaker 7: for a couple of years based on a couple of
[2:34] Speaker 7: huge AI contracts? And so folks are watching each quarter
[2:37] Speaker 7: very closely for any tidbits on that.
[2:40] Speaker 6: I spoke to Sam Altman, the OpenAI CEO, last week
[2:43] Speaker 6: when Astra was released. And it was very interesting to
[2:46] Speaker 6: see Oracle disclose that. one of the data central sites
[2:50] Speaker 6: that they have online was very key to the training
[2:52] Speaker 6: of Astra.
[2:52] Speaker 5: What did they say?
[2:53] Speaker 3: Very interesting.
[2:54] Speaker 7: Yeah, their flagship data center with OpenAI out in West Texas,
[2:58] Speaker 7: they said that.
[2:58] Speaker 5: This is Abilene?
[2:59] Speaker 3: In Abilene, right. They said that OpenAI's newest model was
[3:02] Speaker 3: trained there.
[3:03] Speaker 7: And it just shows that providing the infrastructure here, you know,
[3:07] Speaker 7: some might deride it as a commodity business, but it's
[3:10] Speaker 7: essential to some of the big AI labs.
[3:13] Speaker 5: Let's end on the four-year outlook.
[3:14] Speaker 6: So they're saying at least $ 90 billion of revenue... Adjusted EPS, $ 8. 10, which,
[3:21] Speaker 6: you know, that's a bit of a hike. But how
[3:24] Speaker 6: much of this will focus on the conversion of what
[3:27] Speaker 6: the remaining performance obligation backlog into revenue?
[3:31] Speaker 5: How closely does the street look at that?
[3:33] Speaker 7: Yeah, the big question is how quick can you turn
[3:35] Speaker 7: that RPO into revenue, especially in light of some of
[3:38] Speaker 7: the data center pushback that you mentioned? They've had a
[3:41] Speaker 7: couple of big sites with some tough headlines around NIMBYism.
[3:44] Speaker 5: Local pushback.
[3:45] Speaker 3: It's a recurring question is, is that going to delay projects?
[3:48] Speaker 6: Bloomberg's Brody Ford, who's had a very busy week in
[3:51] Speaker 6: the technology earnings context.
[3:52] Speaker 5: We're going to stay with Oracle.
[3:53] Speaker 6: Joining us now is Stefan Slewinski, BNP Paribas Global Head
[3:57] Speaker 6: of Software Research. He hasn't outperformed on the stock, but
[4:00] Speaker 6: actually lowered his price target from $ 290 to $ 248.
[4:02] Speaker 5: Let's start there, Stefan. Why maintain the call but lower
[4:04] Speaker 5: the price target?
[4:12] Speaker 3: Yeah, hi, Ed.
[4:13] Speaker 5: Hi, Brody.
[4:14] Speaker 8: Good to be with you today. So, as Brody was saying, look,
[4:17] Speaker 8: it was a pretty good set of results across the board.
[4:19] Speaker 8: There were some nitpicks, which we can go into. But,
[4:23] Speaker 8: you know, we did lower our price target a bit today.
[4:24] Speaker 8: You know, we have seen a pullback in some of
[4:26] Speaker 8: the multiples in software over the last couple of days.
[4:29] Speaker 8: So from a relative standpoint, we've taken the price target down.
[4:32] Speaker 8: But we still see a lot of upside here.
[4:33] Speaker 9: This is a company that's.
[4:34] Speaker 8: Going to be growing earnings at 25%, 30% over the
[4:37] Speaker 8: coming couple of years, trading on 20 times earnings currently.
[4:41] Speaker 8: If you do give them the credit and you look
[4:43] Speaker 8: out towards 2029, 2030, you're looking at about a seven
[4:44] Speaker 8: times P.E., So if they can keep delivering, they can
[4:49] Speaker 8: keep bringing capacity online, you can keep seeing the OCI growth.
[4:52] Speaker 8: And importantly, if those margins and cash flows come through,
[4:55] Speaker 8: which is where the market's still questioning the strategy, then
[4:59] Speaker 8: the equity should perform.
[5:01] Speaker 5: Well, Stefan, let's go into those nitpicks.
[5:03] Speaker 6: We're showing that the trading of the day, we're now
[5:06] Speaker 6: down modestly four tenths of a percent. This is a
[5:09] Speaker 6: stock that early in the session was up eight and
[5:11] Speaker 6: a half percent. What's changed in the last two hours
[5:15] Speaker 6: to the mind of investors?
[5:18] Speaker 8: Yeah, I think there was even though it's a good
[5:19] Speaker 8: set of results, there was no knockout punch. I think
[5:21] Speaker 8: investors maybe were hoping for a bigger bounce. The implied
[5:24] Speaker 8: volatility showed a 10 percent move on the day of results.
[5:27] Speaker 8: And when we didn't get that, you know, I think
[5:29] Speaker 8: some people maybe took their chips off the table. Also,
[5:33] Speaker 8: you know, we're waiting for more discussion around capital requirements.
[5:38] Speaker 8: Oracle have said that they need an extra $ 20 billion
[5:41] Speaker 8: of capital this fiscal year. We think that probably comes
[5:44] Speaker 8: beginning of next calendar year. They kind of punted any
[5:47] Speaker 8: discussion around that to the investor day, which will come
[5:49] Speaker 8: at the end of October.
[5:51] Speaker 9: That'll be key.
[5:51] Speaker 8: You have a new CFO, and we think that she
[5:54] Speaker 8: will give commentary around what their capital requirements will be
[5:58] Speaker 8: beyond this fiscal year. That could be a real catalyst
[6:00] Speaker 8: for the stock, but we didn't get that last night.
[6:03] Speaker 6: I think we should talk about leverage. I was talking
[6:05] Speaker 6: about it with our guys at Bloomberg Intelligence this morning.
[6:07] Speaker 6: You know, they have $ 120 billion of debt on the
[6:09] Speaker 6: balance sheet. Cash and cash equips, $ 36 billion. A negative
[6:14] Speaker 6: free cash flow was not as bad as expected. But
[6:17] Speaker 6: everyone still seems pretty calm about that. If you compare
[6:22] Speaker 6: that rate of leverage to other hyperscalers or the average
[6:25] Speaker 6: unadjusted base in the S & P 500, how do
[6:27] Speaker 6: you feel about it? I get you're looking at this
[6:29] Speaker 6: from the equity side of the story.
[6:32] Speaker 8: Yeah, but even on the debt side, we've seen over
[6:34] Speaker 8: the last couple of weeks with the arrival of Astra
[6:36] Speaker 8: from OpenAI, that's been well received. As you pointed out,
[6:39] Speaker 8: that was trained at Abilene, Texas on Oracle infrastructure.
[6:43] Speaker 9: OpenAI hitting $ 45.
[6:44] Speaker 8: Billion of ARR. Sentiment around OpenAI has improved a bit. Therefore,
[6:48] Speaker 8: sentiment around Oracle has improved. So we've actually seen the
[6:50] Speaker 8: CDSs come in a bit, the credit come in a
[6:53] Speaker 8: bit for Oracle as well. there's been a bit more
[6:55] Speaker 8: confidence importantly the company has already said that capex will
[7:00] Speaker 8: be peaking this year and next year and therefore free
[7:02] Speaker 8: cash flow will inflect higher really in fiscal 29 so
[7:06] Speaker 8: we actually see them heading back towards free cash flow
[7:09] Speaker 8: positive If you think about it, you look at all
[7:11] Speaker 8: the hyperscalers, Microsoft, Amazon, Google, Meta, Microsoft has said they'll
[7:15] Speaker 8: stay free cash flow positive. But for the others, we're
[7:18] Speaker 8: heading down into negative free cash flow territory. For Oracle,
[7:21] Speaker 8: there's light at the end of the tunnel. We should
[7:23] Speaker 8: be inflecting back towards positive free cash flow in the
[7:26] Speaker 8: next couple of years. But again, the market wants to
[7:29] Speaker 8: see more proof before it believes it.
[7:31] Speaker 6: So you took us to tokenomics. I find this fascinating.
[7:36] Speaker 6: Astra was trained in the Abilene site. And if Astra
[7:41] Speaker 6: is a commercial success, frankly, it's run from a token
[7:44] Speaker 6: generation perspective.
[7:44] Speaker 5: For Oracle, that's good, right?
[7:48] Speaker 8: I think OpenAI with a $ 300 billion commitment to Oracle
[7:53] Speaker 8: over the next four to five years, obviously, if they're
[7:56] Speaker 8: being successful with their models and they're being successful as
[7:59] Speaker 8: a business, that makes them a good customer to have. Also,
[8:05] Speaker 8: it's a proof of concept of Oracle's infrastructure that it works. Also,
[8:10] Speaker 8: you know, we are seeing Oracle in the quarter.
[8:12] Speaker 9: They had $ 10 billion of prepayment.
[8:15] Speaker 8: So customers signing new contracts, they added $ 26.
[8:19] Speaker 9: Billion to their backlog.
[8:20] Speaker 8: And a lot of that was customers paying up front
[8:23] Speaker 8: for the CapEx. So again, that means that Oracle needs
[8:27] Speaker 8: to spend less of their own capital. They need to
[8:29] Speaker 8: raise less capital from the market in order to finance that.
[8:33] Speaker 8: It also is a validation of what they're providing. They're
[8:35] Speaker 8: not just providing balance sheet of a service. They're not
[8:37] Speaker 8: just a bank. They are providing services around the GPUs
[8:41] Speaker 8: that customers find valuable, and they're able to monetize that.
[8:46] Speaker 6: Stefan, you're Global Head of Software Research, and there was
[8:48] Speaker 6: another earnings out overnight, which is Adobe. They gave a
[8:51] Speaker 6: forecast that You know, there was a range on the street,
[8:54] Speaker 6: but it basically is disappointed. What's the Adobe story here
[8:57] Speaker 6: to your mind?
[8:59] Speaker 8: Yeah, Adobe, we're finally starting to see the cracks that
[9:02] Speaker 8: we've been expecting over the last couple of years, right?
[9:04] Speaker 8: So the RPO, the backlog, actually declined quarter of a quarter,
[9:09] Speaker 8: which is very rare for Q3. The net new ARR
[9:13] Speaker 8: was down 40% almost year over year.
[9:16] Speaker 9: The core creative apps slowed to around 8% growth. So
[9:20] Speaker 9: those concerns.
[9:21] Speaker 8: That we've been worried about for the last three years
[9:23] Speaker 8: are finally starting to come through in the forward-looking indicators.
[9:27] Speaker 9: And so this is a tough market.
[9:29] Speaker 8: Obviously, OpenAI just came out with the GPT images 2.5.
[9:33] Speaker 8: Every day, it feels like there's a new competitor coming
[9:36] Speaker 8: into the market more aggressively. So on the headlines, Adobe
[9:40] Speaker 8: is still doing OK, 12% revenue growth, 42% operating margins,
[9:43] Speaker 8: trading on 12 times earnings, looks cheap. But we know
[9:46] Speaker 8: that revenue growth is going to slow here as they
[9:48] Speaker 8: really push into freemium.
[9:50] Speaker 9: So they're really prioritizing getting free users.
[9:52] Speaker 8: They now have over a billion users in total, but
[9:55] Speaker 8: that's coming at the expense of near-term ARR or near-term
[9:59] Speaker 8: revenue growth.
[10:01] Speaker 9: You have a new CEO in place.
[10:02] Speaker 8: Let's see if he'll change the strategy.
[10:04] Speaker 9: Obviously, he's not new to the company.
[10:07] Speaker 8: He's been with the company for a very long time.
[10:08] Speaker 8: And so the question is, what is going to change
[10:11] Speaker 8: the trajectory of revenue growth for Adobe?
[10:14] Speaker 6: Stefan Slawinski of BNP Paribas. I've really enjoyed our time
[10:17] Speaker 6: with you this morning covering both those earnings. Thank you
[10:19] Speaker 6: very much indeed. Now coming up, the top Pentagon official
[10:22] Speaker 6: is pushing back on fears.
[10:24] Speaker 5: Of AI doomerism.
[10:25] Speaker 6: We've got more on what is the only debate in
[10:28] Speaker 6: Silicon Valley and around the technology world right now.
[10:31] Speaker 5: Thank you. This is Bloomberg Tech. Bridgewater Associates co-CIO.
[10:46] Speaker 6: Greg Jensen is the latest to add his voice to
[10:49] Speaker 6: the growing calls for slowing down of the AI industry.
[10:53] Speaker 6: Speaking on Bloomberg's Odd Lots podcast, Jensen, who's an early
[10:56] Speaker 6: investor in both open AI and anthropic, issued a pressing
[11:00] Speaker 6: warning about the risks and consequences of the current AI boom,
[11:04] Speaker 6: likening it to the early days of the COVID pandemic
[11:07] Speaker 6: when the world took too long to react.
[11:11] Speaker 10: Unfortunately, this is like what it was like in February 2020.
[11:14] Speaker 11: Until the AI starts killing people, unfortunately, history would suggest
[11:19] Speaker 11: we're not going to do anything.
[11:21] Speaker 10: But we are going to face that. That's going to happen.
[11:23] Speaker 11: And it'd be much better if we started dealing with
[11:27] Speaker 11: it before then.
[11:30] Speaker 10: And you can do it, right? It's also not hopeless.
[11:35] Speaker 6: The debate around AI doomerism seems to be everywhere right now,
[11:38] Speaker 6: but a top Pentagon official is pushing back on the idea.
[11:41] Speaker 6: Defense Department Chief Technology Officer Emil Michael says AI harms
[11:45] Speaker 6: can be mitigated by the free market, industry-wide collaboration, and
[11:49] Speaker 6: government engagement. Bloomberg's tech editor Mike Shepard in Washington, D.C.,
[11:53] Speaker 6: joins us now. There are so many different headlines and
[11:55] Speaker 6: stories for us to pass, but what have we been
[11:59] Speaker 6: hearing from the administration on this, from President Trump and
[12:02] Speaker 6: President Trump's team about safety concerns?
[12:07] Speaker 12: Well, Ed, let's start with Emil Michael. He was asked
[12:09] Speaker 12: yesterday specifically about that high profile and now viral tweet
[12:14] Speaker 12: from the anthropic researcher who resigned earlier this week over
[12:17] Speaker 12: safety concerns. And Michael, who has led negotiations in the past,
[12:22] Speaker 12: with Anthropic over extra safeguards that the company wanted on
[12:25] Speaker 12: military use of its tools, culminating in that big blowup
[12:29] Speaker 12: that we have covered so extensively here. Michael was having
[12:32] Speaker 12: none of it on this talk about existential risks from AI.
[12:37] Speaker 12: And he really tried to play down the Doomer talk.
[12:40] Speaker 12: And we also heard from the president himself last night.
[12:44] Speaker 12: stepping off Air Force One. He was asked by our
[12:48] Speaker 12: colleague Jeff Mason, what do you make of the existential risks?
[12:52] Speaker 12: And the president said, I don't have any concerns. And
[12:55] Speaker 12: he instead pointed to his main concern with AI as
[12:59] Speaker 12: being that the U.S. needs to beat China in the
[13:02] Speaker 12: race to dominate this technology. And of course, the contrast
[13:06] Speaker 12: is that many lawmakers on Capitol Hill are seeking to
[13:10] Speaker 12: take some sort of action, ranging from a ban on superintelligence,
[13:13] Speaker 12: to even a kill switch for AI. Unclear whether we
[13:16] Speaker 12: would see actually any tangible progress on those.
[13:20] Speaker 6: I wrote to Anthropic and basically asked them to weigh
[13:23] Speaker 6: in directly with us because you have an employee that
[13:26] Speaker 6: resigned and then a current employee who amplified it, and
[13:30] Speaker 6: they haven't responded.
[13:31] Speaker 5: I just wanted to put it out there.
[13:33] Speaker 6: Let's talk about what Anthropic did say, which was they
[13:35] Speaker 6: published a report yesterday that laid out a pretty wide
[13:39] Speaker 6: range of findings, including another claim by the company that
[13:44] Speaker 6: Chinese rivals have been misusing its models publicly.
[13:47] Speaker 5: There are other security concerns. Shep, go through them.
[13:52] Speaker 12: And this is yet another layer. And of course, it
[13:54] Speaker 12: speaks to the president's desire to lead China in AI.
[13:58] Speaker 12: And what Anthropic is saying in this report, which covered
[14:00] Speaker 12: a wide range of topics. But the most interesting finding
[14:03] Speaker 12: to our view was that it was claiming that Moonshot,
[14:07] Speaker 12: Moonshot AI from China, and we all know it from
[14:10] Speaker 12: its Kimmy K3 breakthrough model from July, that Moonshot had
[14:15] Speaker 12: routed as many as 300,000 user requests through Claude, through
[14:21] Speaker 12: Anthropic's Claude model, and then tried to present the findings
[14:24] Speaker 12: as its own in a way to gain an edge
[14:27] Speaker 12: in the AI race. it used as many as 5,000
[14:31] Speaker 12: fraudulent accounts to do so, Anthropic said. And that's important
[14:35] Speaker 12: because Anthropic has really barred users in China from turning
[14:39] Speaker 12: to its technology. It sees China as a risk geopolitically.
[14:44] Speaker 12: It's something that Dario Amadei, the chief executive of Anthropic,
[14:48] Speaker 12: has articulated. And he is concerned that his company's products
[14:52] Speaker 12: not be used in China. And they have This only
[14:56] Speaker 12: adds to the claims from Anthropic and other AI companies
[15:00] Speaker 12: in the U.S. that China has resorted to unfair practices,
[15:04] Speaker 12: including distillation, where Chinese rivals have systematically extracted data from
[15:10] Speaker 12: their American rivals to build this new vanguard of chatbots.
[15:13] Speaker 12: China has rejected those claims. It's warned against U.S. retaliation
[15:18] Speaker 12: or crackdown on Chinese companies. and threatened reprisals of its own.
[15:22] Speaker 12: And of course, Ed, this all adds to tensions over
[15:25] Speaker 12: AI as President Trump prepares to meet his Chinese counterpart,
[15:30] Speaker 12: Xi Jinping, later this month.
[15:32] Speaker 5: Bloomberg's Mike Shepard, thank you very much.
[15:34] Speaker 6: As AI researchers continue to sound the alarm over the
[15:37] Speaker 6: technology's risks, Anthropic's out with a new report detailing how
[15:40] Speaker 6: its AI is being misused. And according to the company's
[15:43] Speaker 6: threat intelligence team, users from Yemen, Iran, Russia, and China
[15:48] Speaker 6: were flagged after having tapped Claude to develop missile systems,
[15:52] Speaker 6: kamikaze drone swarms, and and even biological weapons research. And
[15:56] Speaker 6: Fropik says it disrupted every operation in that report.
[16:04] Speaker 5: Let's go to OpenAI.
[16:05] Speaker 6: OpenAI is now considering slowing down the development of cutting-edge AI.
[16:09] Speaker 6: And CEO Sam Altman is urging other AI companies to
[16:13] Speaker 6: do the same. Bloomberg's Shreen Ghaffari broke the story with
[16:16] Speaker 6: Rachel Metz and joins us now. And this is about
[16:18] Speaker 6: an internal communication from Sam Altman to the OpenAI team.
[16:22] Speaker 6: And the bit I want you to explain is he
[16:25] Speaker 6: also kind of acknowledges that OpenAI can't do this in isolation.
[16:31] Speaker 9: Right.
[16:32] Speaker 4: So this was discussed in an all-hand staff meeting this week.
[16:36] Speaker 4: And the sentiment of Altman's remarks is basically that the
[16:39] Speaker 4: company is open to pacing the development of these cutting-edge
[16:45] Speaker 4: AI systems, but it can't go it alone, that it
[16:48] Speaker 4: is trying to get other companies on board.
[16:50] Speaker 1: And, you know, we will see.
[16:52] Speaker 4: How many could potentially join?
[16:56] Speaker 6: Potentially joins the big the big thing. We don't have
[16:58] Speaker 6: a sense from Anthropic specifically, right, that they have looked
[17:03] Speaker 6: at the idea of slowing down work at the frontier.
[17:08] Speaker 4: We do know that Anthropic has long been the company
[17:11] Speaker 4: that has been very vocal about the existential problem.
[17:14] Speaker 1: Risks of AI, right?
[17:15] Speaker 4: You have CEO Dario Amadei has talked about, you know,
[17:19] Speaker 4: it's actually been a sort of doomer term has been
[17:23] Speaker 4: applied to him and something that he said he's not.
[17:25] Speaker 4: But the company, I think Anthropic sort of philosophically has
[17:29] Speaker 4: certainly been open to this idea, whether in practice in
[17:33] Speaker 4: this moment they are ready to do it. And of course,
[17:35] Speaker 4: ahead of a potential public listing that could be one
[17:38] Speaker 4: of the biggest is the real question.
[17:42] Speaker 6: Shereen very quickly on on the doomerism debate and that's
[17:45] Speaker 6: probably putting it mildly from an anthropic employee that resigned
[17:50] Speaker 6: an anthropic employee who amplified that statement what is anthropic
[17:54] Speaker 6: actually said about uh everything that's being discussed largely online
[17:58] Speaker 6: at the moment.
[18:00] Speaker 4: Yeah so the company um did send a statement to
[18:02] Speaker 4: us in uh in response to the heated discussion among
[18:07] Speaker 4: staff recently. And they said, we have always been transparent
[18:10] Speaker 4: that AI will bring enormous benefits and unprecedented risks. You know,
[18:15] Speaker 4: it's a long statement, so I'm not going to read
[18:16] Speaker 4: all of it. But in summary, they sort of called
[18:18] Speaker 4: out their responsible scaling policy and how for a long time,
[18:21] Speaker 4: Anthropic has, you know, had sort of framework in place
[18:26] Speaker 4: to not release models if they feel they are overly dangerous.
[18:31] Speaker 6: Bloomberg's Shireen Ghaffari, who again broke that story on OpenAI
[18:34] Speaker 6: with Rachel Metz.
[18:35] Speaker 5: Go and read it on all Bloomberg platforms.
[18:38] Speaker 13: It's time now for Talking Tech. I'm Yahaira Anna. And
[18:40] Speaker 13: first up, Moonshot AI is targeting a big jump in
[18:44] Speaker 13: annualized revenue to $ 2 billion by the end of the
[18:49] Speaker 13: year due to the breakout success of its Kimi K3 model.
[18:53] Speaker 13: That's according to sources. It comes as Moonshot is in
[18:56] Speaker 13: the process of raising funds at a $ 50 billion valuation. Plus,
[19:01] Speaker 13: ZDOT's AI is looking to tap investors for about $ 5
[19:05] Speaker 13: billion to fund more research and development. That's according to
[19:09] Speaker 13: terms seen by Bloomberg. The firm, also known as Zipu,
[19:13] Speaker 13: has been refreshing its model with stronger coding capabilities, releasing
[19:17] Speaker 13: a new one just last month. And Micron is rewarding
[19:22] Speaker 13: its employees in Taiwan after a huge jump in profits.
[19:26] Speaker 13: Net income surged 15-fold in the third quarter, and unions
[19:30] Speaker 13: have been pushing for workers to share.
[19:33] Speaker 3: In those gains.
[19:34] Speaker 13: Now employees will get nearly $ 32, 000 in cash bonuses, plus
[19:39] Speaker 13: additional stock awards.
[19:40] Speaker 12: Ed?
[19:41] Speaker 4: Hi, Rhonda.
[19:42] Speaker 5: Thank you very much.
[19:42] Speaker 6: Apple may be late in the foldables game, but it's
[19:46] Speaker 6: hoping the wait was worth it. Bloomberg's Chris Welch got
[19:48] Speaker 6: hands-on with the new iPhone. Duo and joins us now.
[19:51] Speaker 6: And we were together at Apple Park. And then you
[19:53] Speaker 6: went down into the theater and you had time with it.
[19:57] Speaker 2: I did.
[19:57] Speaker 5: First impressions?
[19:59] Speaker 14: I'd say it just feels different. You know, I've used
[20:00] Speaker 14: all these foldables over the years. I'm sure they're late
[20:03] Speaker 14: to the game, but that's what I'm learning from Samsung,
[20:05] Speaker 14: learning from Google, learning from Apple, all these brands that
[20:08] Speaker 14: have tried. and kind of now take their own approach.
[20:10] Speaker 14: And so it just felt different. I mean, yeah, the shape,
[20:14] Speaker 14: the squat horn factor's gonna take a lot of like,
[20:17] Speaker 14: you know, it'll take time for people to adjust. It's
[20:19] Speaker 14: shorter than your iPhone Max, it's a bit wider. But
[20:22] Speaker 14: the great thing is like Apple has kind of like
[20:25] Speaker 14: just made so many changes to iOS to like the
[20:31] Speaker 14: controls on the right side. They kind of like moved
[20:35] Speaker 14: everything so your thumb can access them easily. And so
[20:38] Speaker 14: it's just like much more ergonomic than like the Fold
[20:41] Speaker 14: 8 is. And so they've learned so much from all
[20:43] Speaker 14: their competitors.
[20:44] Speaker 5: That's so interesting.
[20:45] Speaker 6: With the Galaxy Z Fold 8, I guess there's also
[20:48] Speaker 6: like the weight of it in your hand. And you
[20:52] Speaker 6: have experience personally with foldables, but also you review all
[20:55] Speaker 6: of them. What are the very simple differences with the
[20:57] Speaker 6: Fold 8 and the Duo, for example?
[21:00] Speaker 14: So the matte screen on Apple's new phone is very nice.
[21:02] Speaker 14: The nanotexture coating, that kind of like avoids.
[21:06] Speaker 14: Glare and fingerprints, and so that's like a big, and
[21:09] Speaker 14: the crease, you know, the ever-controversial crease is barely visible.
[21:12] Speaker 5: It's there if you.
[21:13] Speaker 14: Look very, very hard for it, but on the whole,
[21:16] Speaker 14: it's very, very hard to find.
[21:17] Speaker 6: You have covered consumer technology for some time, with respect.
[21:22] Speaker 6: Where does this duo moment kind of stack up against
[21:25] Speaker 6: releases of past generations and moments in the evolution of
[21:29] Speaker 6: consumer tech?
[21:30] Speaker 5: It feels huge for Apple.
[21:31] Speaker 14: I mean, their whole thing is not being first, but
[21:34] Speaker 14: being the best, and having full control over the software
[21:37] Speaker 14: stack and hardware. And so Samsung can only do what
[21:40] Speaker 14: Android lets Samsung do, whereas Apple can change the whole
[21:43] Speaker 14: paradigm of iOS for this device to make it feel
[21:46] Speaker 14: more intuitive, more natural, and truly fresh compared to an
[21:49] Speaker 14: iPhone Max, for example.
[21:50] Speaker 6: Just very quickly, what was the vibe of everyone that
[21:54] Speaker 6: was in the Steve Jobs Theater and also experiencing it
[21:56] Speaker 6: for the first time alongside you?
[21:57] Speaker 5: Pretty excited. Pretty excited.
[21:59] Speaker 14: Just this new product, the first huge event for John Ternes,
[22:03] Speaker 14: the new CEO. There are a bunch of famous folks
[22:06] Speaker 14: down there, influencers having fun on the floor, trying out
[22:08] Speaker 14: the Duo, testing it, just seeing what it's all about.
[22:11] Speaker 14: So I think we're in for a fun few months
[22:13] Speaker 14: for Apple fans.
[22:13] Speaker 6: Bloomberg's Chris Welch, amazing week of top reporting. I really
[22:16] Speaker 6: recommend you read Chris's review. Welcome back to Bloomberg Tech.
[22:26] Speaker 6: Another check of Oracle, which has behaved strangely in this
[22:29] Speaker 6: Friday session. It opened up and hit a session high
[22:32] Speaker 6: more than 8%. It's now higher by half a percentage point,
[22:35] Speaker 6: but had fallen on a couple of occasions into negative territory,
[22:39] Speaker 6: raising the full-year outlook for sales and for adjusted EPS.
[22:42] Speaker 6: Backlog's growing, but they're moving more quickly to bring data
[22:45] Speaker 6: center capacity online. That's been a big feature of the
[22:48] Speaker 6: show so far. zoom out it's been a jiggery week
[22:51] Speaker 6: on wall street especially this friday with the release of
[22:54] Speaker 6: the latest cpi report from the bureau of labor statistics
[22:57] Speaker 6: so next chief market strategy katherine runavera joins us now
[23:00] Speaker 6: to break everything down and and frankly now we are
[23:02] Speaker 6: thinking about inflation we are thinking about what happens at
[23:06] Speaker 6: a september fed meeting and how technology uh as part
[23:10] Speaker 6: of the equity market gets impacted what are your expectations
[23:12] Speaker 6: for that fed meeting I think.
[23:14] Speaker 2: My expectations are completely in line with consensus and market pricing,
[23:18] Speaker 2: which is for a.
[23:19] Speaker 1: Definitive hike next week.
[23:22] Speaker 2: And I think this morning's CPI data just strengthened that
[23:26] Speaker 2: consensus view.
[23:27] Speaker 1: That September, we're going to see a hike.
[23:29] Speaker 2: I think the question, Ed, is December, if we get
[23:32] Speaker 2: two hikes or if this is a.
[23:34] Speaker 9: One and done.
[23:35] Speaker 6: We're seeing a little bit of outperformance in the moment
[23:37] Speaker 6: in technology, but on the week, the NASDAQ 100 as
[23:41] Speaker 6: an example has been under pressure. And we've gone back
[23:43] Speaker 6: to the basics of higher rates discount the present value.
[23:46] Speaker 5: Of future cash flows. Old school. Is that the biggest
[23:50] Speaker 5: consideration for tech right now?
[23:53] Speaker 1: Yeah, I think it is.
[23:55] Speaker 2: It's an interesting concept because, of course, you have longer
[23:58] Speaker 2: duration equities that are going to be under pressure and
[24:01] Speaker 2: tech is one of the most vulnerable to rate increases.
[24:03] Speaker 1: But is this going to be a big rate cycle?
[24:05] Speaker 1: Probably not. It'll likely be shallow.
[24:07] Speaker 2: But the other thing that really calls my attention is
[24:09] Speaker 2: that AI is so tremendous. It's not just, you know,
[24:13] Speaker 2: the biggest driver of the equity markets. It's also now
[24:16] Speaker 2: a macro story. And so you have these countervailing winds
[24:20] Speaker 2: of AI, which is pushing stronger GDP growth. We estimate
[24:24] Speaker 2: maybe half a percentage point additive to U.S.
[24:27] Speaker 1: GDP over the past year.
[24:29] Speaker 2: That's a big number. And the Fed is trying to
[24:31] Speaker 2: kind of slow economic growth with these rate hikes. So
[24:34] Speaker 2: we have countervailing forces here, which I think shows that
[24:38] Speaker 2: the AI build-out is so important and can potentially, Ed,
[24:42] Speaker 2: mask weakness.
[24:44] Speaker 1: In other parts of the economy, including the consumer.
[24:47] Speaker 5: I don't want to.
[24:50] Speaker 6: Get to this point in a sort of superficial manner. But,
[24:53] Speaker 6: you know, every time there is a Fed meeting or
[24:55] Speaker 6: a prominent Fed speaker, the idea of whether AI shows
[24:59] Speaker 6: up in the economy or not is discussed, be that
[25:02] Speaker 6: through productivity data or even like the impact to inflation
[25:06] Speaker 6: from the spending itself.
[25:08] Speaker 5: What do you see?
[25:08] Speaker 2: I think it's generally understood by economists, and I subscribe
[25:13] Speaker 2: to this, that AI is inflationary at the outset and
[25:17] Speaker 2: then disinflationary over time.
[25:19] Speaker 1: We haven't seen that pan out.
[25:21] Speaker 9: Yes.
[25:21] Speaker 2: Yes, I do think, of course, we see prices go
[25:24] Speaker 2: up for chips and hardware and all of that stuff.
[25:26] Speaker 1: Everyone knows the story. But I think what we.
[25:28] Speaker 2: Haven't yet seen is in the offing, which is an
[25:31] Speaker 2: increase in productivity.
[25:33] Speaker 1: Now, productivity. can be very positive because it brings down inflation.
[25:38] Speaker 2: So AI over time, and I'm talking five years, 10 years,
[25:41] Speaker 2: I'm not part of the camp that says AI is
[25:43] Speaker 2: going to, I've heard one camp say.
[25:47] Speaker 5: Kill us all.
[25:47] Speaker 2: It's kind of these apocalyptic views or part of the
[25:52] Speaker 2: camp that says that it's going to take all our
[25:54] Speaker 2: jobs in six months.
[25:54] Speaker 1: I think adoption is relatively slow into the real economy.
[25:57] Speaker 2: So those productivity numbers that maybe the doves in the
[26:02] Speaker 2: FOMC are hoping for won't happen for the foreseeable future.
[26:06] Speaker 2: I think it's more of a five to 10 year
[26:08] Speaker 2: disinflationary effect. So the Fed's going to have to deal
[26:12] Speaker 2: with that in the meantime, and that means higher rates.
[26:15] Speaker 6: Catherine, throughout the program so far, we've talked about the
[26:17] Speaker 6: debate around AI doomerism, and that spans a number of
[26:21] Speaker 6: different stories and headlines, right? But in general, do the
[26:25] Speaker 6: markets treat that idea as a risk factor going forward,
[26:31] Speaker 6: the concern that AI carries risk to society?
[26:34] Speaker 1: Absolutely. And I consider it a risk too.
[26:38] Speaker 2: I mean, at some point, all these hyperscalers and massive investment,
[26:42] Speaker 2: the Besides the MAG-7, which have had some earnings, a
[26:47] Speaker 2: lot of the earnings are.
[26:48] Speaker 1: No longer.
[26:50] Speaker 2: As productive or as strong as they could otherwise be.
[26:54] Speaker 2: But I think the market is going to demand some
[26:57] Speaker 2: additional returns here from all of this investment. So that's
[27:00] Speaker 2: the reason, in addition to the vulnerability in front of
[27:04] Speaker 2: a tightening cycle that I've been recommending recently, to our
[27:06] Speaker 2: clients to put on protective puts on technology positions, not
[27:12] Speaker 2: liquidate them, not get spooked, but put on protective puts
[27:16] Speaker 2: when complacency reigns in the market. And I measured that
[27:19] Speaker 2: with the VIX below approximately 15.
[27:22] Speaker 1: So that's worked for.
[27:23] Speaker 2: Our clients, and I think that's the way to go.
[27:25] Speaker 1: But unequivocally, there is a concern.
[27:28] Speaker 2: Of a bubble, maybe not in the equity markets, but
[27:31] Speaker 2: in the magnitude of the debt that these companies have
[27:35] Speaker 2: been accumulating.
[27:38] Speaker 5: Catherine, how do you model for the midterms?
[27:41] Speaker 6: And in this program, we talk often about data centers
[27:44] Speaker 6: being a massive consideration for the electorate. But again, as
[27:48] Speaker 6: a market strategist, is it for you?
[27:50] Speaker 2: I care about the midterms because I care about the
[27:54] Speaker 2: fiscal deficit. I think that the Fed is going to
[27:57] Speaker 2: have a tough time and has had a tough time
[28:00] Speaker 2: definitively because we're in now the sixth year of missing
[28:03] Speaker 2: the inflation target. But the Fed's gonna have a tough
[28:06] Speaker 2: time hiking rates when we have a fiscal deficit of
[28:08] Speaker 2: near 6% in an expanding economy. That's, in my view,
[28:11] Speaker 2: egregious and shows fiscal irresponsibility and profligacy unlike anything that
[28:16] Speaker 2: I've seen in my lifetime. So if we get more populism,
[28:20] Speaker 2: whether from the left, I think it probably will be
[28:22] Speaker 2: because midterms generally favor the opposition party. If we get
[28:27] Speaker 2: additional populism but from the other side, then we can
[28:32] Speaker 2: expect more spending. And that will weigh both on what
[28:37] Speaker 2: the Fed is trying to do. It'll weigh on inflation expectations.
[28:40] Speaker 2: It'll weigh on the back end of the yield curve.
[28:42] Speaker 2: And it'll also weigh on the equity markets, which I
[28:46] Speaker 2: think are facing an uphill battle with yields.
[28:49] Speaker 1: Rising at the rate they are.
[28:51] Speaker 2: The equity risk premium just pales in comparison at this point.
[28:57] Speaker 6: StoneX Chief Market Strategist, Catherine Runivera, back on the show.
[28:59] Speaker 6: Thank you very much. Warnings about the existential threats posed
[29:07] Speaker 6: by AI, while not entirely new, have gained increasing urgency
[29:11] Speaker 6: for investors. This can also be a pretty thorny issue.
[29:14] Speaker 6: How can the investors help portfolio companies grow while managing
[29:18] Speaker 6: potentially catastrophic risks? And what impact could a pullback on
[29:22] Speaker 6: AI development have on efforts to re-industrialize the United States?
[29:26] Speaker 6: We have a very wide-ranging conversation ahead with Lior Susan,
[29:29] Speaker 6: founder and CEO of venture firm Eclipse, which backs some
[29:32] Speaker 6: pretty familiar names, particularly in the worlds of physical AI,
[29:35] Speaker 6: but also across the compute stack and the supply chain.
[29:38] Speaker 6: And leora it's great to have you back on the
[29:40] Speaker 6: program i said in the last segment the the ai
[29:44] Speaker 6: doomerism debate spans many recent news stories and impacts lots
[29:49] Speaker 6: of frontier labs and others in different layers of the
[29:52] Speaker 6: stack Where's your head at with this right now?
[29:55] Speaker 15: First of all, Shana Tava, it's great to be here.
[29:57] Speaker 15: You know, I heard you talking about it in the
[30:01] Speaker 15: previous segment, and I thought to myself, when we invented
[30:04] Speaker 15: the car, naturally, car killed people.
[30:08] Speaker 5: Did we abandon cars?
[30:10] Speaker 15: No, what I think we did, we united together to
[30:13] Speaker 15: figure out the risk, the values of the cars and
[30:15] Speaker 15: moving people, moving goods, and start creating standards that allow
[30:19] Speaker 15: us to actually use car in the most safe way.
[30:22] Speaker 5: Still, by the way, unfortunately, we're losing people.
[30:25] Speaker 6: Well, so in that context, Parallel example or analogy, you know,
[30:30] Speaker 6: the seatbelt was the technology solution and regulation was done
[30:34] Speaker 6: by regulatory bodies.
[30:35] Speaker 5: Yeah, exactly. What's the equivalent here?
[30:37] Speaker 15: Yeah, I mean, I think we definitely need to have
[30:40] Speaker 15: some sort of a relationship between the private sectors and
[30:43] Speaker 15: governments of how to set some rules. regulations and how
[30:46] Speaker 15: to create a safety belt and airbags and things that
[30:50] Speaker 15: we invented for the cars for the world of AI.
[30:52] Speaker 6: The reporting, you're not directly invested in the Frontier Labs, right?
[30:56] Speaker 6: But obviously, like in the portfolio, there's exposure to them.
[30:58] Speaker 5: We work with them, yeah. Yeah, you work with them.
[31:00] Speaker 6: The proposal that was reported overnight from Sam Altman is that, well,
[31:03] Speaker 6: we could pull back at the frontier, but it would
[31:06] Speaker 6: only work if the other frontier labs agreed to do
[31:08] Speaker 6: the same.
[31:10] Speaker 15: I think I'm a big fan of the idea of
[31:13] Speaker 15: actually collaborating. Although we compete with each other constantly, I
[31:16] Speaker 15: think it's too big of a topic to not unite.
[31:19] Speaker 15: We also need to remember that if we will not
[31:21] Speaker 15: do it, our competitors, like China, are going to take
[31:26] Speaker 15: the lead. So we absolutely need to work together in
[31:28] Speaker 15: the private sectors and our leadership in Washington, D.C., to
[31:33] Speaker 15: make sure that we are aligned of how to do
[31:35] Speaker 15: it in the most safe way, but not lose the
[31:37] Speaker 15: cutting edge.
[31:38] Speaker 6: Physical AI, that could be robotics, it could be other
[31:42] Speaker 6: forms of physical AI, is probably where you.
[31:45] Speaker 5: Are most concentrated, would you say? Where's the opportunity with
[31:48] Speaker 5: that right now?
[31:49] Speaker 15: Yeah, I mean, we started 11 years ago innovating in
[31:52] Speaker 15: the physical world because we believed 85% of the world
[31:56] Speaker 15: GDP is in manufacturing and semiconductors and defense and whatnot.
[32:00] Speaker 15: And today the world's called it physical AI. We build
[32:03] Speaker 15: many companies and we invested in many companies in that area.
[32:06] Speaker 15: And it's exciting to see how AI, automation, and autonomy
[32:10] Speaker 15: is meeting each other. We are seeing the economy here
[32:13] Speaker 15: is growing faster than ever. And I think, you know,
[32:17] Speaker 15: there is a real shot to go back to grow
[32:19] Speaker 15: five plus percent, similar to what Jana is doing here
[32:22] Speaker 15: right now in the U.S.
[32:23] Speaker 5: Leveraging physical AI.
[32:25] Speaker 6: If the Frontier Labs go down the route of pulling
[32:28] Speaker 6: back development at the absolute cutting edge, how would that
[32:32] Speaker 6: impact your portfolio of companies that are working on physical AI?
[32:36] Speaker 6: They might rely on the underlying model as an example.
[32:40] Speaker 15: Yeah, I mean, I think we need to make sure
[32:44] Speaker 15: that we are a continued lead as a country in
[32:47] Speaker 15: the global race. And I think we need to do it.
[32:49] Speaker 15: And the most safe matter, and I think we need
[32:51] Speaker 15: to collaborate, and I think that was Sam's asked yesterday
[32:55] Speaker 15: that I'm completely aligned with and I'm going to sign
[32:58] Speaker 15: up for. I do think we are seeing companies in
[33:02] Speaker 15: the semiconductor space, in the energy space, in the robotic
[33:05] Speaker 15: space becoming very large businesses in our country, first time ever,
[33:10] Speaker 15: and it's just exciting to.
[33:11] Speaker 5: Be part of it.
[33:12] Speaker 6: You published some metrics around how some of your funds
[33:15] Speaker 6: have performed. You did DPI, which basically doesn't include the
[33:20] Speaker 6: paper gains. When we report on the performance of different funds,
[33:24] Speaker 6: we go with TVPI. Just explain which vintages those funds
[33:28] Speaker 6: are and where you've seen some performance, and then I
[33:32] Speaker 6: guess we could talk about valuations.
[33:33] Speaker 15: Yeah, I mean, we definitely have a few of our
[33:37] Speaker 15: vehicles crossing the double-digit TVPI that will be best in class.
[33:43] Speaker 15: That was not actually the point of the post. The
[33:45] Speaker 15: point of the post was, like, this is the best
[33:47] Speaker 15: time in the history to build. The opportunity cost is
[33:50] Speaker 15: so high. Me as an entrepreneur in this country, I
[33:53] Speaker 15: think we never had that type of an opportunity. And
[33:57] Speaker 15: I also talked about the sacrifice of being an entrepreneur
[34:00] Speaker 15: right now, of spending our weekends and our nights building
[34:03] Speaker 15: companies instead of being with our families.
[34:06] Speaker 6: There is an astonishingly well-read story on Bloomberg this morning.
[34:10] Speaker 6: I shared it with you, but it's about the idea
[34:12] Speaker 6: that of sudden wealth syndrome. So if you're a founder
[34:16] Speaker 6: and you do a seed round and within months you're
[34:19] Speaker 6: raising so much more capital and your own personal net
[34:23] Speaker 6: worth rockets How does that feel?
[34:27] Speaker 5: How does that.
[34:27] Speaker 6: Go for the founders that you back? And how active
[34:31] Speaker 6: are you in sort of counseling them on this?
[34:34] Speaker 15: Yeah, there's no school of preparing you for fast success.
[34:39] Speaker 15: I view our job as now older-ish entrepreneurs. I'm 42,
[34:44] Speaker 15: so now I'm not as my 20s when I started
[34:48] Speaker 15: my first company. is to actually work with those young
[34:51] Speaker 15: founders and explain them that even if the wealth creation
[34:54] Speaker 15: happens so fast, what matters is substance and principles. And
[34:59] Speaker 15: we need to remember that we need to do good
[35:02] Speaker 15: for others. We need to remember that we need to
[35:04] Speaker 15: build companies for the long run, kind of ignore valuation
[35:07] Speaker 15: and focusing on the substance.
[35:09] Speaker 6: Tim, let's bring that article back, just show it on
[35:12] Speaker 6: the screen for the audience. A part of what they're
[35:14] Speaker 6: discussing is the post-exit ecosystem. In your world and realm
[35:20] Speaker 6: of physical AI, you know, there's a lot of M &
[35:22] Speaker 6: A where an exit isn't necessarily an IPO or sell
[35:26] Speaker 6: in the secondaries market.
[35:27] Speaker 5: It's being acquired by a bigger entity.
[35:28] Speaker 9: Yeah.
[35:29] Speaker 6: What would you counsel a founder in that context? You know,
[35:32] Speaker 6: sometimes people want to get the return on their work.
[35:36] Speaker 12: Yeah.
[35:37] Speaker 15: And then I think, you know, my first company was
[35:39] Speaker 15: being bought in a big company.
[35:41] Speaker 5: It was a journey and I was young.
[35:43] Speaker 15: I think, you know, what we're advising those founders is
[35:48] Speaker 15: you can do a very good work post acquisition inside
[35:52] Speaker 15: a big company. You can be the next generation leader
[35:55] Speaker 15: in that business, you can have a much more resource
[35:57] Speaker 15: than you might have when you build a standalone company.
[36:00] Speaker 15: But I go back to the opportunity cost. And what
[36:04] Speaker 15: I work with founders is to make sure that they
[36:06] Speaker 15: understand the opportunity cost.
[36:07] Speaker 5: Do you mind to define opportunity cost?
[36:08] Speaker 15: Yeah, I think opportunity cost is like, you're a very
[36:10] Speaker 15: talented person, you can do many things. The opportunity cost
[36:14] Speaker 15: of not going and starting a company right now with
[36:17] Speaker 15: the resource that we have around us is very, very high.
[36:19] Speaker 15: And we want to have a market on the world.
[36:22] Speaker 15: It wants to have a mark in our society. It
[36:24] Speaker 15: wants to have a mark on our economy. So the
[36:26] Speaker 15: opportunity cost of not doing that is very high.
[36:29] Speaker 6: Just very quickly on the three vehicles that are outperforming,
[36:32] Speaker 6: or you say will be the benchmark for industry, are
[36:34] Speaker 6: there any single sectors or names that are driving that performance?
[36:38] Speaker 15: A lot of our companies, naturally in the compute, is
[36:41] Speaker 15: doing extremely well with the growth of AI. Same with energy,
[36:44] Speaker 15: same with defense, same with manufacturing. I don't know, it
[36:47] Speaker 15: was like necessarily a one event. Naturally, the Cerebras IPO
[36:51] Speaker 15: helped us a lot, and it's a phenomenal company that's
[36:54] Speaker 15: growing extremely fast, but there is a bunch of others.
[36:56] Speaker 6: So Cerebras, really interesting. That was a big moment for
[37:00] Speaker 6: the firm. You also back other compute or specialized inference platforms.
[37:04] Speaker 6: As you will have heard earlier in the show, right now,
[37:07] Speaker 6: data centers are divisive in this country. How much do
[37:12] Speaker 6: you model for that when you make an investment?
[37:16] Speaker 15: I think we did, Ed, as an entrepreneur, not a
[37:19] Speaker 15: great job explaining our society, the value of data centers.
[37:24] Speaker 5: So you agree with the Treasury Secretary and with Sam Altman? 100%.
[37:28] Speaker 15: We need to do a better job of explaining our communities,
[37:32] Speaker 15: how many jobs we are creating, the cost of power
[37:34] Speaker 15: that we can take down. We need to share with
[37:36] Speaker 15: them some of our profits of those data centers and
[37:39] Speaker 15: not let only social media turn by our adversary of
[37:43] Speaker 15: maybe influence people's mind of if data center is good
[37:47] Speaker 15: or bad. Data centers are good for our countries, but
[37:50] Speaker 15: we need to do a much better job of sharing
[37:52] Speaker 15: the profits and safety and energy with the communities.
[37:57] Speaker 5: Is it a risk factor for you?
[37:59] Speaker 6: You know, I posed the question earlier about the midterms,
[38:02] Speaker 6: but I guess in aggregate, regulation that forces the slowdown
[38:07] Speaker 6: of data center build out too.
[38:09] Speaker 15: I think regulation generally, it's a risk for the economy.
[38:13] Speaker 15: And that's why we need and the private sector to
[38:17] Speaker 15: work faster and spend our calories with a different administration
[38:21] Speaker 15: to work with them rather than go and block by them.
[38:24] Speaker 15: And I feel much better in the last two years
[38:27] Speaker 15: about some of the activities that I'm seeing between the
[38:29] Speaker 15: private sector and the governments of how to actually set
[38:32] Speaker 15: regulation right, because if set wrong, it will slow down
[38:35] Speaker 15: the economy.
[38:36] Speaker 6: Let's end just very quickly on venture as an asset class.
[38:39] Speaker 6: You know, you've raised funds over the last year. You've
[38:42] Speaker 6: hit a pretty large critical mass. Is it easy to
[38:45] Speaker 6: go out and convince LPs to raise new funds right now?
[38:48] Speaker 15: I mean, I think LPs see the opportunities. I think,
[38:53] Speaker 15: you know, the physical industry is having a moment of
[38:57] Speaker 15: transforming from analog to a digital. As a result of that,
[39:00] Speaker 15: you're seeing in IPOs like SpaceX and Cerebras. And as
[39:04] Speaker 15: a result of that, LPs want to have more access
[39:06] Speaker 15: to those sectors.
[39:07] Speaker 6: Leo Susan, founder and CEO of venture capital firm Eclipse,
[39:10] Speaker 6: back on Bloomberg Tech.
[39:10] Speaker 5: Thank you very much. Good to see you. Indeed. Now,
[39:13] Speaker 5: coming up, we're going to stick with the AI build-out.
[39:15] Speaker 6: Microsoft plans to triple its compute power in the next
[39:18] Speaker 6: few years. Despite the increasing public pushback against data centers,
[39:22] Speaker 6: Bloomberg broke that story, and Brody Ford is back with
[39:24] Speaker 6: us next.
[39:25] Speaker 5: This is Bloomberg Tech.
[39:39] Speaker 6: Microsoft plans to more than triple its data center capacity
[39:42] Speaker 6: in an effort to overcome a computing shortage that's forced
[39:45] Speaker 6: the company to turn away some AI business and some
[39:48] Speaker 6: cloud business. Bloomberg's Brody Ford broke the story of Matt
[39:50] Speaker 6: Day and is back with us. Let's get into the
[39:54] Speaker 6: details that we're reporting, all according to sources. But this
[39:57] Speaker 6: is a massive ramp up, or the plan at least.
[39:59] Speaker 3: It's a big ramp up, right?
[40:00] Speaker 7: And the big question for a lot of people, whether
[40:03] Speaker 7: in Wall Street or in the industry, is just, How
[40:06] Speaker 7: big are these data center fleets and how big are
[40:08] Speaker 7: they going to get? This current building boom, how long
[40:11] Speaker 7: does it last? How long do we get ramped up?
[40:14] Speaker 7: And this is the first time that I think we've
[40:16] Speaker 7: had a pretty good sense of where a company thinks
[40:19] Speaker 7: it's going. More than tripling a data center fleet through
[40:22] Speaker 7: the coming years, hitting as much as 38 gigawatts.
[40:26] Speaker 9: Wow.
[40:27] Speaker 7: That's a big ramp up. And this is a plan
[40:30] Speaker 7: for multiple years out. And of course, plans change based
[40:33] Speaker 7: on a number of things, delivery, demand signals. But this
[40:37] Speaker 7: is the trajectory that one of the biggest data center
[40:39] Speaker 7: developers in the world is on.
[40:42] Speaker 6: What did Microsoft say about our reporting? Did they respond
[40:44] Speaker 6: to requests or comment?
[40:45] Speaker 3: Initially, no.
[40:47] Speaker 7: Later, they said the numbers were not accurate without giving
[40:49] Speaker 7: a whole lot more. But, you know, as we said,
[40:52] Speaker 7: this is a plan for multiple years out. Companies tend
[40:56] Speaker 7: not to give us much more than a year or
[41:00] Speaker 7: so out. And so it doesn't surprise me that we
[41:02] Speaker 7: don't hear a lot from the company.
[41:04] Speaker 6: Just for transparency, let's go back to the idea that
[41:08] Speaker 6: because they're compute constrained, they've turned away some AI business.
[41:12] Speaker 6: How has that shown up?
[41:15] Speaker 14: Right.
[41:15] Speaker 7: This is something that Microsoft, Amazon, Google, everyone complains about,
[41:19] Speaker 7: that if it weren't for the supply constraints as data centers,
[41:23] Speaker 7: they'd be able to juice their cloud revenue much more.
[41:25] Speaker 3: And so that's why you see this building.
[41:27] Speaker 7: Frenzy trying to capture this demand while it's there. We
[41:30] Speaker 7: saw Timu, for example, the Chinese e-commerce company, big customer
[41:34] Speaker 7: of Microsoft Azure, they couldn't get more capacity, so who'd
[41:37] Speaker 7: they go to?
[41:38] Speaker 3: They went to Oracle.
[41:39] Speaker 7: And this is the kind of situation that Microsoft is
[41:41] Speaker 7: really trying to avoid.
[41:43] Speaker 6: So that brings us to the competitive landscape. Is Microsoft
[41:46] Speaker 6: good at building data centers?
[41:49] Speaker 3: Seems that way. You know, I mean, they have a
[41:51] Speaker 3: very big fleet.
[41:52] Speaker 7: Today, they have 12 gigawatts across the world, two of
[41:56] Speaker 7: that specifically for AI.
[41:57] Speaker 3: And that's big.
[41:59] Speaker 7: I mean, we don't have the precise estimates for everybody,
[42:02] Speaker 7: but likely that's bigger than anybody but Amazon Web Services.
[42:05] Speaker 6: One thing people will forget, because before AI, there was
[42:08] Speaker 6: cloud computing, and cloud computing was for running any given
[42:11] Speaker 6: software and for hosting and for storage. Like, how does
[42:14] Speaker 6: Microsoft manage that, AI versus non-AI?
[42:17] Speaker 12: Yeah.
[42:17] Speaker 7: It's a really good point because a lot of people
[42:19] Speaker 7: out there seem to think that GPU means AI workload.
[42:23] Speaker 7: And the message that Microsoft has been trying to give
[42:25] Speaker 7: is that in the agentic era, a lot of what
[42:29] Speaker 7: you think of as AI workloads is being done off CPUs, right?
[42:32] Speaker 3: You're inferencing.
[42:33] Speaker 7: Agents are going and accessing databases and creating records. And
[42:36] Speaker 7: that in this era, you need maybe more CPUs than
[42:40] Speaker 7: you expected even a year or two ago.
[42:42] Speaker 6: Bloomberg's Brodie Ford with a must read on Microsoft and
[42:46] Speaker 6: its plans to triple compute capacity. That does it for
[42:49] Speaker 6: this edition of Bloomberg Tech. Let's take a final look
[42:52] Speaker 6: at technology markets. NASDAQ 100 higher on the day, basically
[42:56] Speaker 6: flat on the week. The MAG7 slightly outperforming. And Oracle, again,
[42:59] Speaker 6: is flat, but it's been up as much as 8.5%
[43:02] Speaker 6: post earnings and seen a little bit of chop. What
[43:04] Speaker 6: a week it's been. Recap on the podcast. You know
[43:07] Speaker 6: where to find it. From San Francisco, have a great weekend, everyone.
[43:10] Speaker 6: This is Bloomberg Tech.
[43:12] Speaker 1: Thank you.

Shows are listed from Apple Podcasts' charts and directory, and Podcast Index where it is used. Every show and episode belongs to its publisher and plays from the publisher's own feed; this site hosts none of them.