# Nvidia Goes Big on Buybacks as SpaceX Reaches Orbit

An episode of [Bloomberg Tech](https://invisiblenetwork.ai/podcast/1611767434), 28 Sep 2026, 44 min.

Audio: https://podtrac.com/pts/redirect.mp3/traffic.omny.fm/d/clips/e73c998e-6e60-432f-8610-ae210140c5b1/41764a4f-fc64-4e11-89ba-ae7c0030ab5e/5743ab89-96ff-4dbc-841d-b4d301119054/audio.mp3?utm_source=Podcast&in_playlist=9caafc41-289c-4115-995d-ae7c0030ab75

Bloomberg’s Ed Ludlow breaks down Nvidia's decision to boost the size of its share buyback plan by a record $150 billion, taking the total amount to $235 billion. Plus, a milestone for SpaceX's Starship rocket reaching orbit for the first time after the mission briefly looked at risk because an engine failed during ascent. And, Meta taps MongoDB's president and CEO to lead a new AI platform for enterprise customers that Meta CEO Mark Zuckerberg called “the next major pillar of our business.” See omnystudio.com/listener for privacy information.

## Transcript

[0:02] Speaker 1: Bloomberg Audio Studios. Podcasts.
[0:05] Speaker 2: Radio.
[0:06] Speaker 1: News.
[0:13] Speaker 3: Bloomberg Tech is live from the heart of Silicon Valley
[0:17] Speaker 3: with Ed Ludlow in San Francisco.
[0:22] Speaker 1: This is Bloomberg Tech.
[0:23] Speaker 4: Coming up, NVIDIA boosts the size of its share buyback
[0:26] Speaker 4: plan by a record $ 150 billion, taking the total amount
[0:30] Speaker 4: to $ 235 billion, plus a milestone for SpaceX's Starship rocket
[0:35] Speaker 4: reaching true orbit for the first time after a brief
[0:39] Speaker 4: scare due to an engine failure during ascent. Meta taps
[0:42] Speaker 4: MongoDB's president and CEO to lead a new AI platform
[0:46] Speaker 4: for enterprise customers it deems as the next major pillar
[0:51] Speaker 4: of its business. Let's get right to today's big number,
[0:55] Speaker 4: $ 150 billion. That's how much NVIDIA's added to its share
[0:58] Speaker 4: buyback plan, taking the total remaining authorized amount to $ 235 billion,
[1:04] Speaker 4: a massive number, reflecting CEO Jensen Wong's confidence in its
[1:08] Speaker 4: continued growth. The company expects to complete the program through
[1:12] Speaker 4: fiscal 2028, according to a statement. They're currently in the
[1:15] Speaker 4: third fiscal quarter of their fiscal year, 27. At the
[1:18] Speaker 4: same time, NVIDIA says it has a new way to
[1:21] Speaker 4: rein in rogue AI agents, introducing new security tools that
[1:25] Speaker 4: can shut them down in milliseconds. The company even going
[1:28] Speaker 4: so far as to say it could have stopped that
[1:30] Speaker 4: recent Hugging Face breach, which involved OpenAI's models. Bloomberg's Maggie
[1:34] Speaker 4: Eastland has the details, and it's kind of in two parts.
[1:37] Speaker 1: Explain them. Absolutely.
[1:40] Speaker 2: So this new product release has two layers. So first
[1:43] Speaker 2: there's Open Shell, that is an open source software program
[1:47] Speaker 2: that can run on NVIDIA's central processing units. And that's
[1:50] Speaker 2: designed so that a user can impose certain guardrails and
[1:53] Speaker 2: enforce those guardrails. Now this was first announced in March
[1:57] Speaker 2: back at their GTC conference being rolled out now in
[2:00] Speaker 2: combination with a completely new product called NVIDIA Sentry, which
[2:03] Speaker 2: is another separate system that runs on NVIDIA's data processing units.
[2:08] Speaker 2: And essentially, it's an added layer that can, like you said,
[2:11] Speaker 2: isolate an agent that's acting suspiciously, the company says, in milliseconds.
[2:16] Speaker 4: The analogy that I've been going with this morning is
[2:18] Speaker 4: like it's putting an AI agent in a locked room
[2:21] Speaker 4: where part one controls what that room looks like and
[2:24] Speaker 4: the doors in and out of it. Part two is
[2:26] Speaker 4: a security guard that's always outside. But it's kind of
[2:30] Speaker 4: in line with Jensen Wong's school of thought that this
[2:33] Speaker 4: is an engineering issue at the heart of the AI
[2:36] Speaker 4: safety debate. Well, he's given us an AI engineering solution.
[2:41] Speaker 1: Absolutely, Ed.
[2:42] Speaker 2: That's the line I've been using. Jensen Wong, one of
[2:45] Speaker 2: the biggest beneficiaries of the AI trade, has really continually
[2:49] Speaker 2: voiced that AI progress needs to continue and that the
[2:53] Speaker 2: executives who want to slow down just need to find
[2:55] Speaker 2: a solution.
[2:55] Speaker 1: To their problem.
[2:56] Speaker 2: And now he's offering them the solution. Importantly, we don't
[2:59] Speaker 2: know whether OpenAI and Anthropic are going to adopt this
[3:02] Speaker 2: for their training run specifically, which is where they've had
[3:05] Speaker 2: the most.
[3:06] Speaker 1: Problems with these rogue agents.
[3:08] Speaker 2: But if this really can solve the issue, it could
[3:11] Speaker 2: really quiet some of these calls that we've been hearing
[3:13] Speaker 2: for a slowdown.
[3:14] Speaker 4: There is an element where each of those solutions runs
[3:18] Speaker 4: on other parts of NVIDIA's ecosystem hardware-wise. Just very quickly,
[3:23] Speaker 4: what's the NVIDIA school of thought about what this means
[3:25] Speaker 4: for the training runs that are happening in industry?
[3:30] Speaker 2: Yeah, I think the idea here is that NVIDIA has
[3:34] Speaker 2: said this potentially could have stopped the Hugging Face incident,
[3:37] Speaker 2: which was an incident where agents got out during a
[3:40] Speaker 2: training run. So if the technology works in the way
[3:44] Speaker 2: that they're saying it works, it could enable OpenAI and
[3:48] Speaker 2: Anthropic to continue with some of these training runs that
[3:50] Speaker 2: they've been more hesitant about. We saw just last Friday,
[3:53] Speaker 2: OpenAI said they were pausing training after new disclosures about
[3:57] Speaker 2: some government-related incidents and a breach of an Australian government website.
[4:01] Speaker 2: So again, this could be really a breakthrough if the
[4:03] Speaker 2: technology does work the way NVIDIA is claiming.
[4:06] Speaker 4: NVIDIA shares up 2% on multiple pieces of news, Bloomberg's
[4:11] Speaker 4: Maggie Eastland on the AI safety component of it. For more,
[4:14] Speaker 4: let's get to Erica Clower, founder and CIO of Science
[4:17] Speaker 4: and Technology Partners. And this is so interesting because this
[4:21] Speaker 4: is NVIDIA acting on their warning that this is an
[4:26] Speaker 4: engineering problem.
[4:27] Speaker 1: Here's a solution. What do you make of that? NVIDIA
[4:30] Speaker 1: just offering that to industry. So good morning, Ed.
[4:34] Speaker 5: Nice to see you. It's such an interesting time once
[4:37] Speaker 5: again for NVIDIA showing and flexing its muscle of innovation.
[4:41] Speaker 5: I think one of the things that's sometimes hard to
[4:45] Speaker 5: remember is that over half of NVIDIA's engineers are software engineers.
[4:50] Speaker 5: These are engineers who are experts, not only in terms
[4:55] Speaker 5: of software expertise, but also in cybersecurity. So it makes
[4:58] Speaker 5: total sense that the company would be stepping in to
[5:02] Speaker 5: alleviate some of the concerns around the cyber area.
[5:06] Speaker 4: What's the link, okay, between the share buyback story, this
[5:11] Speaker 4: AI agent safety story, and the other disclosures that NVIDIA
[5:15] Speaker 4: made this morning?
[5:17] Speaker 5: So I think that perhaps the takeaway is that despite
[5:22] Speaker 5: all of the concerns about the amount of spending that's
[5:26] Speaker 5: going into AI infrastructure, And whether or not there's an
[5:30] Speaker 5: AI bubble, NVIDIA is saying not only are we stepping
[5:34] Speaker 5: in to buy more of our own shares and a
[5:36] Speaker 5: vote of confidence about the future of the industry, but
[5:40] Speaker 5: we also have the cash to do it. We are
[5:43] Speaker 5: generating sufficient cash flow to not only make these investments,
[5:47] Speaker 5: but also to be very shareholder friendly. So I think
[5:50] Speaker 5: it's quite a powerful message coming from NVIDIA today.
[5:54] Speaker 4: The company had also kind of disclosed that their commits
[5:57] Speaker 4: or their backlog continues to grow. Anthropic is a case
[6:01] Speaker 4: studies to date, $ 180 billion of NVIDIA hardware commitments across
[6:05] Speaker 4: its different deployments.
[6:07] Speaker 1: You're a long bias fund.
[6:08] Speaker 4: You're focused on like the absolute leading edge of what's happening.
[6:12] Speaker 4: We are through the third quarter of earnings season, right?
[6:16] Speaker 4: What did you conclude about the AI story from that
[6:20] Speaker 4: earnings season?
[6:22] Speaker 5: I think that the most important takeaway from the third
[6:25] Speaker 5: quarter was just how broad-based AI adoption has been, whether
[6:30] Speaker 5: it's a financial or a drug development company or a
[6:34] Speaker 5: consumer-facing company that's looking to improve customer service and response.
[6:40] Speaker 5: The overwhelming adoption of AI tools has been phenomenal. And
[6:46] Speaker 5: I think that that really speaks to the idea that
[6:50] Speaker 5: instead of prior projects problematic areas where capacity was built
[6:55] Speaker 5: ahead of demand, we have the flip side, which is
[6:59] Speaker 5: that we simply don't have enough compute and capacity and
[7:02] Speaker 5: power and all of this to meet the demand, which
[7:05] Speaker 5: continues to grow across many, many different industries.
[7:10] Speaker 4: The story that happened in parallel with earnings season was
[7:12] Speaker 4: the AI safety debate, pacing at the frontier, what kind
[7:16] Speaker 4: of third-party evaluation is needed to make AI development as
[7:19] Speaker 4: safe as possible. Is there any direct line to how
[7:22] Speaker 4: that impacts that supply-demand on the compute side that you've
[7:26] Speaker 4: just outlined?
[7:28] Speaker 5: I think that you bring up an excellent point, which
[7:32] Speaker 5: is that there has to be more innovation around the security,
[7:36] Speaker 5: and especially around agentic security, in order to give continued
[7:40] Speaker 5: confidence and adoption of these key technologies. So I do
[7:43] Speaker 5: think that this provides an excellent opportunity for many of
[7:46] Speaker 5: the cybersecurity software names. But I also think that this
[7:50] Speaker 5: now will become table stakes for other chip vendors, whether
[7:54] Speaker 5: it's AMD or Marvell or Broadcom. Having a security aspect
[7:59] Speaker 5: to the platform is going to be table stakes.
[8:04] Speaker 4: Erica, how much of a warning signal was the developments
[8:08] Speaker 4: around AI safety to you?
[8:12] Speaker 5: I do think that the AI safety is a profound issue.
[8:16] Speaker 5: And I think that the speed at which we are
[8:18] Speaker 5: seeing response is very reassuring. I think the other thing
[8:23] Speaker 5: that we heard a lot from many of the cyber
[8:26] Speaker 5: software companies is was that they saw a flip in
[8:29] Speaker 5: terms of trying to go out to sell their wares
[8:32] Speaker 5: to customers, to actually customers coming in to talk to
[8:35] Speaker 5: them about how they could thwart any kind of cyber threats.
[8:39] Speaker 5: So clearly this is a big issue, and what I
[8:42] Speaker 5: think is very interesting is just how fast the response
[8:45] Speaker 5: time has been.
[8:47] Speaker 4: We're showing NVIDIA is higher, but generally speaking, technology shares
[8:50] Speaker 4: are under pressure today. Yields are at multi-year highs. And
[8:54] Speaker 4: the headline that hit just before we came on air
[8:56] Speaker 4: is that the U.S. 10-year treasury yield has moved up
[9:00] Speaker 4: about 10 basis points, 5.27%. It's back down 5.26%. Just
[9:07] Speaker 4: how closely you're paying attention to this U.S. economy and
[9:11] Speaker 4: the actions of both fiscal policy and the Fed right now?
[9:16] Speaker 4: in the context of your investments in the technology sector?
[9:21] Speaker 5: Well, I mean, we've seen a tremendous correction in technology
[9:27] Speaker 5: shares since June. In many cases, there are companies, premier
[9:32] Speaker 5: companies that are down 30, 40, 50 percent. So I
[9:35] Speaker 5: think we're in the midst of a bottoming process against
[9:39] Speaker 5: a backdrop of fiscal tightening. And also we're heading into
[9:43] Speaker 5: midterm elections where many Concerns are being surfaced about the
[9:48] Speaker 5: pace of data center builds in the United States and
[9:51] Speaker 5: the pressure on the grid prices and also environmental concerns.
[9:56] Speaker 5: But we've been able to navigate through most of those.
[10:00] Speaker 5: My suspicion is that at the end of the day,
[10:03] Speaker 5: earnings will matter most. And we've seen extraordinary financial results
[10:07] Speaker 5: across the board for the best in class AI players.
[10:10] Speaker 5: And I think that this is probably an interesting moment
[10:13] Speaker 5: in time to really think about seeing these companies and
[10:17] Speaker 5: buying these companies at extraordinary values relative to their growth
[10:21] Speaker 5: prospects on a longer-term basis.
[10:24] Speaker 4: Erica Clower of Science and Technology Partners, great to have
[10:26] Speaker 4: you back on the show.
[10:27] Speaker 1: Thank you very much indeed.
[10:28] Speaker 4: Coming up, SpaceX was able to deploy the V3 Starlink
[10:32] Speaker 4: satellites on board, the first true orbital test flight for Starship,
[10:37] Speaker 4: but the milestone did have some challenges. We're going to
[10:40] Speaker 4: have the details next. This is Bloomberg Tech.
[10:43] Speaker 1: Thank you.
[10:52] Speaker 4: SpaceX's Starship rocket lifted off just a few hours ago.
[10:55] Speaker 1: For its 14th test, its first true orbital flight.
[10:59] Speaker 4: The Elon Musk company was able to deploy Starlink satellites
[11:03] Speaker 4: that were on board, but the trip did come with
[11:05] Speaker 4: some hiccups, which actually throughout the pre-market and into today's
[11:09] Speaker 4: session caused volatility in the shares. Starship now returning to
[11:13] Speaker 4: Earth earlier than planned. Let's get the absolute latest with
[11:17] Speaker 4: Bloomberg Space Reporter Lauren Grush. Overall successful, but where are
[11:21] Speaker 4: we at right now in this 14th test?
[11:24] Speaker 6: Right.
[11:24] Speaker 3: It wouldn't be a Starship flight if it didn't have
[11:26] Speaker 3: a little drama associated with it. So yes, it was
[11:29] Speaker 3: the first orbital mission. They were able to achieve orbit
[11:32] Speaker 3: after some confusion or consideration of whether they were able
[11:35] Speaker 3: to do so. One of Starship's Raptor engines did cut
[11:39] Speaker 3: out early during that initial climb to space. So there
[11:42] Speaker 3: was a part where they weren't sure if they were
[11:44] Speaker 3: able to do that orbit insertion, but they were able
[11:47] Speaker 3: to push ahead. as planned and successfully deploy all 26
[11:51] Speaker 3: V3 upgraded Starlink satellites on board. But they ultimately opted
[11:56] Speaker 3: to bring Starship home early. I think right about now
[11:59] Speaker 3: it's de-orbiting and coming back home. We'll probably hit the
[12:03] Speaker 3: atmosphere soon. It was supposed to be a nearly 10
[12:06] Speaker 3: hour test flight, but now it's gonna be about three
[12:09] Speaker 3: hours instead.
[12:11] Speaker 1: Exactly.
[12:11] Speaker 4: So at the bottom of the spacecraft part of Starship,
[12:14] Speaker 4: there are six Raptors, one of them misbehaved in the ascent.
[12:18] Speaker 4: They did reach orbital altitude, 275 kilometers. And there'll be
[12:23] Speaker 4: people watching, Lauren, that are like, hold on, I've seen
[12:26] Speaker 4: Starship in space. Like, what's the difference? So I think
[12:29] Speaker 4: get into that. It's important that Starship could reach that
[12:34] Speaker 4: orbital altitude of 275.
[12:37] Speaker 7: Yeah.
[12:38] Speaker 3: Up until now, all of Starship's test flights have been suborbital,
[12:41] Speaker 3: meaning it's been on a trajectory where no matter what,
[12:44] Speaker 3: the gravity of Earth would take it back just so that,
[12:48] Speaker 3: you know, if there were any issues with Starship on board,
[12:50] Speaker 3: that it would come back and fall in a place
[12:53] Speaker 3: that wouldn't have any kind of issues for people or
[12:56] Speaker 3: uninvolved property. And so going to orbit is a big
[12:58] Speaker 3: decision because of, you know, if there was an issue
[13:01] Speaker 3: with the engines, they want to make sure that they
[13:03] Speaker 3: can be able to take Starship out of orbit safely,
[13:06] Speaker 3: which is hopefully what they're doing right now. But yeah,
[13:09] Speaker 3: it's a big moment. And, you know, we'll see what
[13:11] Speaker 3: the FAA says and what SpaceX says following the data
[13:15] Speaker 3: review after this, but it could potentially open the door
[13:18] Speaker 3: for more orbital flights to come, more Starlink deployments to come.
[13:22] Speaker 3: And as we know, that's a big moment for SpaceX.
[13:26] Speaker 1: Bloomberg's Lauren Grush, thank you very much.
[13:28] Speaker 4: Let's talk to someone who's worked on Starship. Scott Morton
[13:31] Speaker 4: spent nearly a decade at SpaceX as an engineering manager
[13:35] Speaker 4: for Starship, focused on software before leaving to found Revel,
[13:39] Speaker 4: which builds software to help companies design, test, operate complex machines.
[13:43] Speaker 4: I think we'd agree, Scott, that Starship is a complex,
[13:47] Speaker 4: a multifaceted machine. But just your reaction, you've been inside
[13:51] Speaker 4: the beast, right? You understand how important this milestone was
[13:56] Speaker 4: for Test Flight 14.
[13:59] Speaker 7: Yeah, I mean, I think, I mean, overall, this was
[14:02] Speaker 7: a monumental step forward, like actually going to orbit. I
[14:11] Speaker 7: You know, they did have a hiccup with one of
[14:13] Speaker 7: the engines, but, you know, they deployed Starlink satellites in orbit.
[14:19] Speaker 7: And then they've also, it sounds like they've successfully completed
[14:22] Speaker 7: the first deorbit burn. So I think overall, this is like,
[14:26] Speaker 7: I mean, if I were internal at SpaceX still, I
[14:29] Speaker 7: would be seeing this as a massive success.
[14:33] Speaker 4: When you, the time you'd left SpaceX, you were the
[14:36] Speaker 4: software manager for the Starship program, right? We talk a
[14:40] Speaker 4: lot about the iterations on the hardware, the heat shield
[14:42] Speaker 4: in particular, but generation to generation, those hardware changes they make.
[14:47] Speaker 4: But software is absolutely critical because the launch system and
[14:52] Speaker 4: the spacecraft itself is essentially autonomous, right?
[14:55] Speaker 1: Explain that part.
[14:58] Speaker 7: Yeah, I mean, it's actually autonomous, I'd say, even further
[15:01] Speaker 7: out on the launch pad getting ready for flight. But yeah,
[15:08] Speaker 7: if you can plan for everything that you want to
[15:10] Speaker 7: do with the vehicle and code all that in software,
[15:13] Speaker 7: it's a much easier kind of thing to reason through
[15:17] Speaker 7: than having a lot of manual commanding. As much as
[15:19] Speaker 7: you can automate as you possibly can, it's definitely the
[15:21] Speaker 7: best approach. And even like the scenario that happened today,
[15:24] Speaker 7: this was definitely fully planned for. There was software already
[15:29] Speaker 7: ready to go for all these types of contingencies.
[15:33] Speaker 4: I am familiar with over-the-air software updates for my Tesla vehicle, right?
[15:39] Speaker 4: And that gives, with existing hardware, new capabilities to the car,
[15:44] Speaker 4: improves performance. How does that work in the context of
[15:48] Speaker 4: the most powerful rocket and spacecraft launch system that humankind
[15:53] Speaker 4: has developed?
[15:54] Speaker 1: Does it work the same way?
[15:57] Speaker 7: At least at my time at SpaceX, we could not
[16:00] Speaker 7: kind of do over-the-air updates. I'm not sure where they
[16:02] Speaker 7: are today on that. There are a lot of things
[16:04] Speaker 7: that you can tweak, parameters and things that you can update.
[16:09] Speaker 7: But generally, you want to have the software be quite
[16:12] Speaker 7: locked down and not be kind of making changes, especially
[16:14] Speaker 7: not on flights. I assume long-term they'll have some type
[16:19] Speaker 7: of capability for this when you have vehicles on orbit
[16:21] Speaker 7: for very long durations. But at least in my time
[16:26] Speaker 7: on the program, you want to have the software very
[16:29] Speaker 7: locked down. There's a lot that you need to make
[16:30] Speaker 7: sure that you're operating within the bounds of what is
[16:35] Speaker 7: expected and not going too far off of those plans
[16:41] Speaker 7: in real time.
[16:44] Speaker 4: When people kind of question whether or not a mission
[16:47] Speaker 4: is successful against its original goals, one of the things
[16:49] Speaker 4: that SpaceX always talks about is the data that is
[16:52] Speaker 4: generated from whatever the outcome of the mission. How do
[16:56] Speaker 4: the software teams take the data and then use it?
[16:59] Speaker 4: Like if there was a sort of iterative process on software,
[17:04] Speaker 4: how meaningful is the data from the mission that's gathered?
[17:08] Speaker 1: I mean, it's massive.
[17:09] Speaker 7: They're going to spend weeks and weeks, probably actually months
[17:11] Speaker 7: on this one. going through every single system, all the
[17:15] Speaker 7: software that was run, literally point by point, in some cases,
[17:19] Speaker 7: like telemetry point by point, to make sure everything performed
[17:23] Speaker 7: as expected. So yeah, the data is actually, I'd say,
[17:26] Speaker 7: the most valuable thing coming out of this flight. I
[17:29] Speaker 7: was on the program from the very beginning. I think
[17:33] Speaker 7: I was Probably the first software engineer down in Boca Chica, Texas,
[17:36] Speaker 7: not in a Starbase. And the fact that it has
[17:39] Speaker 7: now gone to orbit is just incredibly exciting for me.
[17:42] Speaker 7: But yeah, the most valuable thing coming out of this,
[17:46] Speaker 7: all the data that they're going to be going through,
[17:47] Speaker 7: I'm sure they've learned a tremendous amount.
[17:49] Speaker 1: From this flight.
[17:51] Speaker 4: Take me back again to your time there. Prior to
[17:57] Speaker 4: SpaceX's IPO, the prospectus made it crystal clear that all
[18:00] Speaker 4: of those kind of out of this world but future
[18:03] Speaker 4: business lines, they're all predicated on Starship working, being fully reusable,
[18:07] Speaker 4: rapidly reusable.
[18:09] Speaker 1: But when the.
[18:09] Speaker 4: Program started, what was the kind of vision for Starship?
[18:13] Speaker 4: What were you all talking to each other about, hoping
[18:15] Speaker 4: to achieve with it at that time?
[18:19] Speaker 1: Yeah, I mean, even.
[18:21] Speaker 7: back at the very beginning of the program, we had,
[18:23] Speaker 7: you know, we had, uh, kind of CAD drawings of
[18:27] Speaker 7: what is there today. Uh, I mean, this is exactly
[18:30] Speaker 7: what we were hoping for and, uh, what ultimately happened. And,
[18:33] Speaker 7: and obviously much further beyond that, that Starship would be, uh,
[18:37] Speaker 7: you know, it takes, takes people to Mars. That is
[18:40] Speaker 7: a hundred percent what we were thinking about day to
[18:41] Speaker 7: day as we're putting in the extremely long hours. Um,
[18:46] Speaker 7: so yeah, I think, um, You know, this is exactly
[18:50] Speaker 7: this is like for me, this is like absolutely incredible
[18:52] Speaker 7: that we, you know, Starship has gone to orbit. It
[18:54] Speaker 7: is actually now successfully done the first year of the burn,
[18:57] Speaker 7: which is kind of one of the most concerning, I'd say,
[18:59] Speaker 7: parts of the whole mission. But, yeah, I mean, extremely exciting,
[19:02] Speaker 7: exciting for everyone at SpaceX and everyone that's contributed to
[19:05] Speaker 7: it to date.
[19:08] Speaker 4: Scott Morton, now the CEO of Rebel, formerly at SpaceX,
[19:12] Speaker 4: managing the software of the Starship program.
[19:14] Speaker 1: It's great to have you back on the show. Thank
[19:16] Speaker 1: you very much indeed.
[19:17] Speaker 4: Coming up, Meta has tapped MongoDB's chief to lead its
[19:21] Speaker 4: new AI pushes for businesses, for the enterprise.
[19:24] Speaker 1: It's been a big story this morning, and it's next.
[19:27] Speaker 1: This is Bloomberg Tech.
[19:40] Speaker 4: Okay, take a look at shares of Meta and also MongoDB.
[19:43] Speaker 4: MongoDB quite clearly down much more severely. It announced that
[19:48] Speaker 4: its CEO was leaving to go to Meta. CJ Desai
[19:51] Speaker 4: is stepping down effective immediately. He's then set to lead
[19:55] Speaker 4: Meta's new enterprise platform, bringing its AI models, agents, infrastructure
[20:00] Speaker 4: to business customers. Bloomberg's Alex Levine joins us with the reporting. Okay,
[20:06] Speaker 4: there's a lot going on here. Let's start with the
[20:09] Speaker 4: enterprise platform.
[20:10] Speaker 1: What is that?
[20:12] Speaker 8: So for well over a year at this point, we
[20:14] Speaker 8: have been writing all about the ways that Meta has
[20:17] Speaker 8: been continuing to invest enormously in AI. And I think
[20:23] Speaker 8: if you're reading the tea leaves, especially over the better
[20:25] Speaker 8: part of the last year, this may not be all
[20:28] Speaker 8: that surprising that Meta is now moving into this enterprise
[20:31] Speaker 8: platform to offer its services like agents, like AI infrastructure,
[20:35] Speaker 8: and other services for businesses.
[20:39] Speaker 4: The sell side's really positive about Meta Enterprise Platform. The
[20:43] Speaker 4: stock is down 4%. Let's go to the CJ Desai
[20:47] Speaker 4: piece of this. MongoDB down 18%. At one point, it
[20:50] Speaker 4: was down 27% and on track for its worst day
[20:54] Speaker 4: since March of 2025. This is a talent that Mark
[20:58] Speaker 4: Zuckerberg wanted for this.
[21:00] Speaker 1: What's the story?
[21:01] Speaker 8: Yeah, so Desai is an industry veteran for MongoDB, this
[21:05] Speaker 8: document database. I think as we've been covering with other
[21:10] Speaker 8: major recent hires, including Alexander Wang last year, this is
[21:14] Speaker 8: sort of the next iteration of what Zuckerberg sees as
[21:18] Speaker 8: is necessary to really extend superintelligence, not only to all people,
[21:23] Speaker 8: as he has said, but also to all businesses and
[21:25] Speaker 8: the developers behind them. And so he sees Desai as
[21:28] Speaker 8: really fit for that.
[21:30] Speaker 1: Okay.
[21:31] Speaker 4: People will be watching going, okay, meta, the parent of Facebook,
[21:35] Speaker 4: social media, more recently AI, but consumers, I get that.
[21:39] Speaker 4: This new evolution, this pillar, it's about doing business with
[21:43] Speaker 4: other companies.
[21:45] Speaker 1: Exactly.
[21:46] Speaker 8: And I think it's really important to also look at
[21:47] Speaker 8: where this plants meta in the larger space with its
[21:50] Speaker 8: competitors in the AI race. So you've got other giants
[21:53] Speaker 8: in the industry like Microsoft, like Amazon Web Services, like Google,
[21:57] Speaker 8: who are already offering meta. many of these same services
[22:00] Speaker 8: to their customers. So I think not only is this
[22:03] Speaker 8: one way that Zuckerberg can diversify the company's revenue streams
[22:09] Speaker 8: as it pursues these enormous AI ambitions, but it is
[22:12] Speaker 8: also very much a strategic and competitive move as Meta
[22:15] Speaker 8: seeks to keep pace with its rivals.
[22:17] Speaker 4: Yeah, and we're showing some of those that kind of
[22:19] Speaker 4: took a hit after those headlines hit this morning in
[22:22] Speaker 4: the enterprise software space. Welcome back to Alex Levine. Thank
[22:25] Speaker 4: you very much. It's great to have you back on
[22:27] Speaker 4: the program. Coming up, NVIDIA and Gecko Robotics are building
[22:31] Speaker 4: new guardrails.
[22:32] Speaker 1: For autonomous machines.
[22:33] Speaker 4: We're going to speak to Gecko Robotics CEO Jake Lucerarian next.
[22:36] Speaker 4: This is the AI agent safety story that we did
[22:39] Speaker 4: at the top. And Gecko Robotics is the physical AI
[22:43] Speaker 4: case study. Super interesting. Halfway through the program here in
[22:46] Speaker 4: the city of San Francisco. Thanks for being with us.
[22:49] Speaker 4: This is Bloomberg Tech. Welcome back to Bloomberg Tech. Meta
[23:03] Speaker 4: Enterprise Platform, it's a big story in markets this morning. Generally,
[23:07] Speaker 4: the sell side's really positive about the idea that Meta
[23:09] Speaker 4: has a mechanism now to do business with other companies
[23:12] Speaker 4: rather than just focus on the consumer, leveraging its work
[23:15] Speaker 4: in agents and AI. It has tapped CJ Desai, who
[23:19] Speaker 4: was the CEO and president of MongoDB, and that stock
[23:23] Speaker 4: is down accordingly. Off session lows, but down 17%, a
[23:26] Speaker 4: pretty severe reaction as CJ Desai leaves MongoDB and goes
[23:30] Speaker 4: to Meta to lead. meta enterprise platforms. And they brought
[23:34] Speaker 4: back a former CEO at Mongo to bridge the gap
[23:37] Speaker 4: where they searched for a long-term replacement. It's had a
[23:39] Speaker 4: lot of effect on the market in different ways to
[23:42] Speaker 4: different enterprise stocks. The other big kind of financial market
[23:44] Speaker 4: story is NVIDIA, putting another $ 150 billion behind its own stock,
[23:49] Speaker 4: taking its remaining buyback authorization to $ 235 billion. The stock
[23:52] Speaker 4: higher 2% off its session highs Bloomberg Intelligence says the
[23:59] Speaker 4: size of that commitment underscores NVIDIA's financial firepower, but with
[24:04] Speaker 4: plenty of room to keep funding the broader AI build out.
[24:08] Speaker 4: Robert Shiffman, Bloomberg Intelligence Senior Credit Analyst joins us now.
[24:11] Speaker 4: You were so quick out the gate with the React.
[24:14] Speaker 4: Just explain your thesis on this buyback.
[24:17] Speaker 9: Well, I think NVIDIA has been reading Bloomberg Intelligence research
[24:20] Speaker 9: and they realize that there's no AI bubble. And because
[24:23] Speaker 9: of that, they're going to be so cash rich They
[24:26] Speaker 9: can not only fund friends and family and keep this
[24:29] Speaker 9: AI ecosystem going, but they can also spend a significant
[24:33] Speaker 9: amount on shareholder returns, benefiting, I think, both creditors and
[24:39] Speaker 9: shareholders alike. And this sort of stuff, I think, over
[24:41] Speaker 9: time is only going to get bigger because we see
[24:44] Speaker 9: their free cash flow potentially going to a trillion dollars
[24:47] Speaker 9: over the next couple of years.
[24:49] Speaker 4: Chip, sincerely, and just like in very basic terms, what
[24:52] Speaker 4: is the credit analyst's interest when you get a headline
[24:55] Speaker 4: about what is now, I think, the biggest buyback in U.S.
[24:59] Speaker 1: Company history?
[25:00] Speaker 9: Yeah, it's bigger than Apple's $ 110 billion. It certainly shows
[25:04] Speaker 9: that they are king of the hill when it comes
[25:06] Speaker 9: to AI these days. Unfortunately, there's not so many of
[25:10] Speaker 9: their hyperscaler friends that are in this position anymore. I
[25:13] Speaker 9: think from a creditor standpoint, it doesn't really make any
[25:17] Speaker 9: difference to the way NVIDIA trades. What I do think, though,
[25:20] Speaker 9: it suggests is that they're going to have enough cash
[25:23] Speaker 9: to give out in order to fund everybody else. If
[25:27] Speaker 9: you look at what we think CapEx is going to
[25:30] Speaker 9: be spent through 2031 from the hyperscalers, it's now going
[25:33] Speaker 9: to reach $ 9 trillion.
[25:34] Speaker 1: That's with a T.
[25:36] Speaker 9: And I think next year, we're going to see IG
[25:39] Speaker 9: borrowing probably in and around $ 500 billion. And to know
[25:44] Speaker 9: that there's capital around to help build that system, I
[25:47] Speaker 9: just think brings confidence to People have been very, very
[25:50] Speaker 9: concerned that there's both an AI bubble, there's a credit bubble.
[25:54] Speaker 9: We don't see it at all. We've never seen demand stronger,
[25:58] Speaker 9: and we haven't heard anything different from either investors or
[26:01] Speaker 9: the companies themselves.
[26:03] Speaker 4: Okay, so as of the end of the fiscal second
[26:04] Speaker 4: quarter of fiscal year 27, the most recent quarter, NVIDIA
[26:08] Speaker 4: is sitting on just north of $ 99 billion, cash, cash quibs. You, A,
[26:14] Speaker 4: don't see a circular financing concern, And you see their
[26:18] Speaker 4: ability to make different kinds of investments, grease the wheels
[26:22] Speaker 4: for the build out.
[26:23] Speaker 9: Yeah, listen, circular financing is a great media buzz line
[26:27] Speaker 9: to try to bash this space. And I just don't
[26:29] Speaker 9: think it's working. You know, what they're doing is they're
[26:33] Speaker 9: making equity investments. They're making credit style investments where they're
[26:38] Speaker 9: both leasing facilities. They're providing residual value guarantees. They're providing
[26:42] Speaker 9: support behind the value of their chips beyond their depreciable life.
[26:48] Speaker 9: I just think what they're doing is they're just acting
[26:50] Speaker 9: as a bank for everybody, which is not a bad thing.
[26:52] Speaker 9: I mean, this is the whole story of Robinhood. You
[26:55] Speaker 9: take a little bit from the rich and you give.
[26:56] Speaker 1: To the poor. That's all they're doing.
[26:57] Speaker 9: It's a timing delay when it comes to making money
[27:00] Speaker 9: in this business. You know, those that are collecting right now,
[27:03] Speaker 9: the semis, are sharing that money with those that are building,
[27:06] Speaker 9: the hyperscalers.
[27:07] Speaker 1: And the neoclouds.
[27:08] Speaker 9: That is going to change in the next couple of
[27:10] Speaker 9: years when those companies have facilities that are up and running,
[27:14] Speaker 9: and you see AI monetization start to boom. They're all
[27:17] Speaker 9: going to turn free cash flow positive, and they're all
[27:20] Speaker 9: going to start buying back masses and masses of stock again.
[27:23] Speaker 9: And I think this sort of revolution is just going
[27:25] Speaker 9: to end up continuing, and I don't see NVIDIA slowing
[27:29] Speaker 9: that down, and I don't think they care about circular referencing.
[27:32] Speaker 9: other than they're also now raising third-party money. Remember, they
[27:36] Speaker 9: just announced a deal with the biggest alternative asset managers
[27:39] Speaker 9: out there to help raise $ 500 billion. And I don't
[27:43] Speaker 9: think that that money is getting any smaller. I'm surprised
[27:45] Speaker 9: it wasn't a trillion or it's not going to grow
[27:47] Speaker 9: to a trillion or two trillion in the next couple
[27:49] Speaker 9: of years.
[27:51] Speaker 4: Robert Schiffman of Bloomberg Intelligence with a credit perspective on
[27:54] Speaker 4: this massive buyback program.
[27:56] Speaker 1: Thank you very much.
[27:57] Speaker 4: NVIDIA is pushing its AI safety efforts into the physical world.
[28:01] Speaker 4: It's working with Gecko Robotics to test new safeguards that
[28:04] Speaker 4: put hard limits on what AI-powered robots can do and
[28:08] Speaker 4: stop them when they cross boundaries. Gecko CEO Jake Lucerarian
[28:12] Speaker 4: is with us and This was our top story today, right?
[28:14] Speaker 4: Part one, make an agent fit within parameters that you
[28:18] Speaker 4: set in a box, right? Part two, a security guard
[28:22] Speaker 4: for when that agent misbehaves or acts suspiciously. How does
[28:26] Speaker 4: that work in your physical AI world?
[28:30] Speaker 10: Well, thanks for being able to have me on to
[28:32] Speaker 10: talk about this. What's so important, and we're all talking
[28:34] Speaker 10: about today, is the criticality of these agentic tools being
[28:37] Speaker 10: able to be deployed in the real world, especially being
[28:40] Speaker 10: deployed in such important environments, but doing so in a
[28:42] Speaker 10: way that's safe and ensures that we have the utmost
[28:45] Speaker 10: standards as it relates to safety, quality, safety. especially with
[28:49] Speaker 10: some of the most critical sectors, whether it be on
[28:51] Speaker 10: the military side or on the commercial side with energy
[28:53] Speaker 10: mining and manufacturing. And that's exactly what this partnership is
[28:56] Speaker 10: all about, is ensuring that we keep humans in control,
[28:59] Speaker 10: but also supercharge the benefits of these agentic tools coming
[29:03] Speaker 10: out of the Foundation Lab. So really exciting announcement, and
[29:07] Speaker 10: it puts safety and humans first.
[29:09] Speaker 4: This is a partnership NVIDIA has with about 100 different
[29:13] Speaker 4: entities across software, hardware, robotics companies. Give me a case
[29:18] Speaker 4: study that can make it real for the audience. It
[29:20] Speaker 4: does all sound a little bit abstract. So an AI
[29:23] Speaker 4: agent in the context of something like a robot in
[29:26] Speaker 4: the industrial and defense settings that.
[29:28] Speaker 1: You work in.
[29:31] Speaker 10: Well, what's becoming very clear is that, I think, impressive
[29:35] Speaker 10: and critical these models are to such an incredible, ecosystem
[29:41] Speaker 10: of tech. What's becoming really clear is that deployment into
[29:44] Speaker 10: the real world on the physical intelligence side, that's where
[29:46] Speaker 10: real large moats are coming from. By gathering information, data
[29:49] Speaker 10: sets never existed before to power these abilities to model
[29:53] Speaker 10: and simulate what the real world looks like and how
[29:55] Speaker 10: to make sure it optimizes for better. A way to
[29:58] Speaker 10: understand one of the use cases is actually with the
[30:01] Speaker 10: military One of our robots, Komodo, is actually using this
[30:06] Speaker 10: partnership in the system that the open agent safety platform
[30:09] Speaker 10: that NVIDIA and Gecko are pioneering to actually ensure that
[30:12] Speaker 10: we're gathering information and data sets by letting agents run
[30:16] Speaker 10: the robot. but ensuring that parameters are set and humans
[30:19] Speaker 10: are able to control certain sorts of actions that the
[30:21] Speaker 10: robots are taking in the real world. And so the
[30:24] Speaker 10: agents need to be able to path plan and gather
[30:25] Speaker 10: information and supercharge where they're taking actions and getting data,
[30:29] Speaker 10: but do so in a way that doesn't compromise safety
[30:31] Speaker 10: and quality standards that humans set and prioritize. And this
[30:34] Speaker 10: is exactly the key to ensuring that we can get
[30:37] Speaker 10: as much information and data as possible at low cost
[30:40] Speaker 10: and do so in a way that keeps humans in
[30:43] Speaker 10: control and safety paramount.
[30:46] Speaker 1: Take me into the backstory here.
[30:50] Speaker 4: Part of this platform, NVIDIA talked about it at GTC
[30:54] Speaker 4: earlier this year, right? There has been so much in
[30:57] Speaker 4: the headlines about AI safety, models misbehaving, agents going rogue. Like,
[31:03] Speaker 4: was there a frantic phone call from Jensen Huang last
[31:05] Speaker 4: week saying, okay, this is what we're doing, everyone get
[31:08] Speaker 4: on board?
[31:11] Speaker 10: Yeah, I'll keep the phone calls between the two confidential,
[31:14] Speaker 10: but you're 100% right, which is that you need to
[31:17] Speaker 10: prioritize safety, but you also need to be moving at
[31:21] Speaker 10: a pace that's really bringing to bear the best technology
[31:24] Speaker 10: in the world for the people keeping our borders safe
[31:27] Speaker 10: and keeping freedom for the world. It's so exciting that
[31:32] Speaker 10: Jensen at NVIDIA and myself at Gecko are able to
[31:35] Speaker 10: be doing the partnership like this for the military as
[31:38] Speaker 10: the first example of many to come as it relates
[31:41] Speaker 10: to the technology that's being deployed to ensure that we
[31:44] Speaker 10: keep ships out of dry dock and in the fight
[31:47] Speaker 10: and ensure that we can deter conflict around the world.
[31:50] Speaker 10: And this is putting such a positive spin on the
[31:54] Speaker 10: benefits and impacts that this technology.
[31:57] Speaker 1: Stack is able to have.
[31:59] Speaker 10: And also the game plan of how robotics has to
[32:03] Speaker 10: be used to give us.
[32:04] Speaker 1: The advantage in the West.
[32:06] Speaker 10: And so it's an exciting partnership because it pushes the
[32:08] Speaker 10: frontier and ensures that technologies and technology companies like Gecko
[32:13] Speaker 10: and NVIDIA are leading the way and creating the rules.
[32:17] Speaker 4: Jensen Huang's school of thought is that the safety of
[32:23] Speaker 4: a model and an issue of alignment, does the AI
[32:27] Speaker 4: do what it's intended to? That's an engineering problem. Is
[32:31] Speaker 4: that what you see in your physical AI realm? It's
[32:34] Speaker 4: just an engineering problem that you can now solve for
[32:37] Speaker 4: because of what NVIDIA has made open source.
[32:41] Speaker 10: Yeah, it also is ensuring that we put into perspective
[32:45] Speaker 10: the power of these tools. I mean, we hear so
[32:48] Speaker 10: much conversation from certain CEOs about the doom of AI,
[32:52] Speaker 10: but I think what's important is this partnership highlights how
[32:56] Speaker 10: these technologies, robotics and artificial intelligence models, are actually being
[33:01] Speaker 10: used for good, for keeping humans out of harm's way,
[33:04] Speaker 10: to ensuring that we get critical assets and components back
[33:07] Speaker 10: up and running, ensuring that they're deterring conflict or keeping
[33:10] Speaker 10: lights on, empowering homes for heating and cooling. This is
[33:14] Speaker 10: the kind of way that we want to imagine the
[33:16] Speaker 10: future looking like for robotics and AI, and not some
[33:20] Speaker 10: doomerism that can be sold to assist humans. whatever is
[33:24] Speaker 10: trying to be pushed by these companies. And so similar
[33:28] Speaker 10: to Jensen, what we're here talking about is such a
[33:31] Speaker 10: positive impact that these technologies are having for real people
[33:34] Speaker 10: and ensuring as well that we keep those real people
[33:37] Speaker 10: with those sort of better expertise in control.
[33:40] Speaker 1: And we're building technology stacks right alongside them. And that's
[33:43] Speaker 1: what's leading to such incredible outcomes.
[33:46] Speaker 4: Jake Lucerarian, Gecko Robotics CEO on that technology partnership with NVIDIA.
[33:50] Speaker 4: Thank you very much indeed. Let's talk about SpaceX again.
[33:54] Speaker 4: We're seeing some live pictures now. Live pictures. This is
[33:57] Speaker 4: Starship re-entering Earth's atmosphere. It did reach an orbital altitude
[34:03] Speaker 4: of 275 kilometers, but SpaceX made the decision to bring
[34:08] Speaker 4: Starship back to Earth and back home early. It was
[34:12] Speaker 4: originally the mission goal to do six orbits of the planet.
[34:16] Speaker 4: And then after about nine hours and 50 minutes, bring
[34:19] Speaker 4: it back. But they came back early after successfully deploying
[34:22] Speaker 4: 26 V3 Starlink satellites. The feed breaking up there as
[34:27] Speaker 4: it passes through Earth's atmosphere.
[34:30] Speaker 1: Mission successful? Mostly.
[34:32] Speaker 4: Coming up, Anthropic CEO Dario Amodei met with President Trump
[34:36] Speaker 4: as the debate around AI safety continues.
[34:39] Speaker 1: We'll have those details next. This is Bloomberg Tech.
[34:42] Speaker 9: Thank you.
[34:54] Speaker 1: President Trump and.
[34:56] Speaker 4: Anthropic CEO Dario Amodei taking their AI safety debate face
[35:00] Speaker 4: to face. The two met for dinner after Amodei called
[35:03] Speaker 4: for slowing the development of the most advanced models as
[35:06] Speaker 4: a growing number of AI security breaches add urgency to
[35:09] Speaker 4: the fight over guardrails. Bloomberg Senior Tech Editor Mike Shepard
[35:12] Speaker 4: is with us. We know the dinner happened. What else
[35:15] Speaker 4: do we know?
[35:17] Speaker 6: We do know that dinner happened. A White House official
[35:19] Speaker 6: has confirmed that it took place. It was late last
[35:22] Speaker 6: night here in Washington. But we don't know anything more
[35:25] Speaker 6: beyond that right now. There's been no readout from either side.
[35:28] Speaker 6: And Ed, really, the key takeaway is that this meal
[35:31] Speaker 6: would have brought together two men on opposite ends of
[35:34] Speaker 6: the AI safety debate. As we all know, Dario Amadei
[35:37] Speaker 6: has for years been warned of the potential existential risks
[35:41] Speaker 6: from the technology. And on September 12th, he published that
[35:44] Speaker 6: high-profile essay promising to slow development of the most advanced models.
[35:49] Speaker 6: President Donald Trump, on the other hand, has pushed back
[35:52] Speaker 6: against those doomsday fears, calling them a hoax and also
[35:56] Speaker 6: rejecting calls embraced by Amadei for more regulation.
[35:59] Speaker 1: As a globalist conspiracy.
[36:01] Speaker 6: So we'll be eager to see whether the two men
[36:04] Speaker 6: were able to meet somewhere in the middle. And we
[36:06] Speaker 6: may get some details on that from President Donald Trump
[36:09] Speaker 6: during a policy announcement later this afternoon. And we're also,
[36:13] Speaker 6: of course, out to Anthropic for any readout on their side.
[36:18] Speaker 4: Over the weekend, I think ahead of this meeting, the
[36:20] Speaker 4: president did speak to reporters and he basically said, here's
[36:23] Speaker 4: my point of view on AI. Could you summarize what
[36:28] Speaker 4: the president's standing is on it?
[36:31] Speaker 6: Well, for weeks, he has been in response to a
[36:35] Speaker 6: lot of the safety fears that have been surfaced by
[36:38] Speaker 6: leaders in Silicon Valley, including Amadei, but as well as
[36:41] Speaker 6: his arch rival, Sam Altman, that something must be done
[36:45] Speaker 6: on a global basis to create some standards to address
[36:48] Speaker 6: these risks that AI agents could get out of control.
[36:52] Speaker 6: But the president has pushed back against that. He is
[36:55] Speaker 6: interested in two things. He wants development of this technology
[36:58] Speaker 6: to proceed faster. largely unfettered because of its economic importance
[37:02] Speaker 6: to the United States and, of course, also to the
[37:05] Speaker 6: importance to the U.S. of staying ahead of China, the
[37:08] Speaker 6: top geopolitical rival in this technology. And that was really
[37:12] Speaker 6: one of the centerpieces of the conversation between the president
[37:15] Speaker 6: and his Chinese counterpart, Xi Jinping, here in Washington last week.
[37:20] Speaker 6: A lot of other tech CEOs were at that state
[37:22] Speaker 6: dinner for Xi Jinping on Thursday evening, but one absence notably.
[37:26] Speaker 1: Was Dario Amadei.
[37:28] Speaker 6: So perhaps we will see a little bit of movement
[37:31] Speaker 6: in the debate here in Washington, including with the president himself.
[37:36] Speaker 4: And as Shep mentioned, the president is due to make
[37:39] Speaker 4: a statement in a couple of hours' time. Bloomberg's Mike Shepard,
[37:41] Speaker 4: thank you very much indeed. Let's get over to New York.
[37:44] Speaker 4: Bloomberg's Yahira Anand has the news.
[37:47] Speaker 1: Hi, Yahira. Hi, Ed.
[37:48] Speaker 11: It's time now for Talking Tech. First up, shares of
[37:51] Speaker 11: SK Hynix fell 5% in South Korea, and its ADRs
[37:54] Speaker 11: are falling here as well on reports.
[37:56] Speaker 1: Its U.S.
[37:57] Speaker 11: Subsidiary, Solid Dime, could go public as soon as next
[38:00] Speaker 11: year at a valuation of up to $ 100 billion. An
[38:03] Speaker 11: IPO would add another layer to the broader SK Group's
[38:07] Speaker 11: already complex ownership structure, raising concerns for shareholders. Plus, China
[38:14] Speaker 11: expanding overseas travel restrictions for families of top AI professionals,
[38:19] Speaker 11: relatives of executives whose work is considered critical to state security,
[38:24] Speaker 11: now require Beijing's approval before traveling abroad. That's according to sources.
[38:29] Speaker 11: That's to prevent sensitive AI expertise from leaving for the U.S.
[38:34] Speaker 11: And OpenAI CEO Sam Altman and Anthropic CEO Dario Amadei
[38:38] Speaker 11: will not appear before an Australian Senate.
[38:41] Speaker 9: Inquiry into AI.
[38:42] Speaker 11: The hearing comes after an OpenAI model hacked an Australian
[38:46] Speaker 11: government website. Both companies will send other executives instead.
[38:50] Speaker 6: Ed.
[38:51] Speaker 1: Thank you, Yohaira.
[38:52] Speaker 4: Now, coming up, we're going to go back to NVIDIA's
[38:54] Speaker 4: buyback plans, one of several NVIDIA stories that are impacting
[38:59] Speaker 4: the shares today.
[38:59] Speaker 1: Stock up 2%.
[39:01] Speaker 4: There's also been some disclosures on their backlog, how much
[39:04] Speaker 4: NVIDIA Gear Anthropic is using, and then there's the AI
[39:08] Speaker 4: agent safety offering that they opened source this morning as well.
[39:11] Speaker 4: All to come, this is Bloomberg Tech. Let's get back
[39:28] Speaker 4: to our top story. NVIDIA's record buyback shares 2% off
[39:32] Speaker 4: session highs.
[39:33] Speaker 1: Actually, there were several pieces of news.
[39:36] Speaker 4: Kunjan Sabani from Bloomberg Intelligence out with his react saying
[39:39] Speaker 4: it's positive, but not thesis changing. Specifically when it comes
[39:44] Speaker 4: to that $ 150 billion buyback increase. Kunjan Sabani of Bloomberg
[39:49] Speaker 4: Intelligence joins us now. You know, it's such a big number.
[39:53] Speaker 4: And then in aggregate, a bigger number still. Why is
[39:56] Speaker 4: it not thesis changing for you?
[39:59] Speaker 12: Well, it does have some positive reads. Buyback is always good.
[40:02] Speaker 12: Shareholder returns is always a positive. Also signals that the
[40:05] Speaker 12: company believes that the stock is cheap and also helps
[40:09] Speaker 12: sort of those investors who are concerned that the company
[40:12] Speaker 12: is deploying too much cash. into those financing deals instead
[40:15] Speaker 12: of returning back. So definitely don't want to overshadow the positives. However,
[40:20] Speaker 12: doesn't do anything to address the fundamental risks that can
[40:23] Speaker 12: the CapEx from hyperscalers sustain, the rising competition from the
[40:27] Speaker 12: ASIC areas, and again, people who are concerned about sort
[40:30] Speaker 12: of the circular financing. So on one hand, does do
[40:34] Speaker 12: positive read-through on capital allocation, but on the other hand,
[40:37] Speaker 12: doesn't sort of sway away the risks that have been
[40:40] Speaker 12: hampering over the stock multiple.
[40:42] Speaker 4: There has been a lot of discussion about NVIDIA open
[40:45] Speaker 4: agent safety platform. You talk about security, but the idea
[40:49] Speaker 4: that NVIDIA is offering more than just GPUs, like in
[40:53] Speaker 4: a very real sense now, it's saying, you guys have
[40:55] Speaker 4: a concern.
[40:56] Speaker 1: Here's an engineering.
[40:57] Speaker 4: Solution to it for a piece of open source software.
[41:00] Speaker 1: What did you make of that?
[41:02] Speaker 12: That was a more fundamental positive breakthrough. I mean, look,
[41:06] Speaker 12: with all the talks about security, safety, with all the
[41:09] Speaker 12: attacks happening and the talks about slowing down training progression, remember,
[41:14] Speaker 12: all of that will require more compute. Even if you
[41:17] Speaker 12: slow down training, model training development to add more compliance, safety, security,
[41:22] Speaker 12: you still need more compute.
[41:24] Speaker 1: And guess what?
[41:24] Speaker 12: Where are people going to turn for that compute? To
[41:26] Speaker 12: the likes of NVIDIA. So that is definitely a fundamental
[41:30] Speaker 12: long-term tailwind here for the company.
[41:33] Speaker 4: Kunjan Sivani of Bloomberg Intelligence with a react to a
[41:35] Speaker 4: lot of what happened with NVIDIA this morning.
[41:37] Speaker 1: Thank you.
[41:38] Speaker 4: Sticking with NVIDIA, Lockheed Martin CEO Jim Ticlare says the
[41:43] Speaker 4: chip company has been a partner of theirs for two
[41:45] Speaker 4: decades now. As AI keeps playing a growing role in
[41:49] Speaker 4: Lockheed's roadmap, listen to this.
[41:52] Speaker 13: Our chief engineer, when I started the company in management,
[41:55] Speaker 13: six years ago, said AI is really critical. And of course,
[41:58] Speaker 13: to ask for budget to invest in our own AI
[42:02] Speaker 13: containerized instance where we would not be going off using
[42:07] Speaker 13: data sets off the internet, we would not be using
[42:10] Speaker 13: models and then feeding our information back into the system,
[42:13] Speaker 13: where we would have a basically a controlled environment for
[42:18] Speaker 13: our AI workloads.
[42:20] Speaker 1: And one of our partners, again, in.
[42:22] Speaker 13: The technology space has been Jensen Wang. Back in 2000, 2001,
[42:26] Speaker 13: where hardly anyone maybe had heard of NVIDIA, if you
[42:29] Speaker 13: weren't a gamer, we were teamed with NVIDIA to get GPUs.
[42:34] Speaker 13: And that relationship continues today.
[42:38] Speaker 4: So that was Lockheed Martin's CEO, Jim Ticlare. Tune in
[42:41] Speaker 4: on Friday for more of that conversation during Bloomberg's Defense
[42:45] Speaker 4: Tech Special, 11 a.m.
[42:47] Speaker 1: Eastern, 8 a.m. Pacific.
[42:48] Speaker 4: We're going to be live from Lockheed Martin's Skunk Works.
[42:51] Speaker 4: That's Lockheed Martin's Advanced Aircraft Development Facility.
[42:55] Speaker 1: In Southern California.
[42:56] Speaker 4: A really rare, incredible access to what's happening in defense technology.
[43:02] Speaker 4: From defense tech to space, let's get back to those
[43:03] Speaker 4: live pictures of Starship. Re-entering Earth's atmosphere, it did its
[43:10] Speaker 4: de-orbit burn, reached orbital altitude 275 kilometers.
[43:16] Speaker 1: That mission goal achieved.
[43:17] Speaker 4: It deployed 26 Starlink V3 satellites, commercial, operational, revenue generating
[43:24] Speaker 4: for Starship.
[43:25] Speaker 1: That was a first.
[43:26] Speaker 4: But SpaceX decided to return Starship in its 14th test
[43:30] Speaker 4: flight home early. The original mission design was to do
[43:34] Speaker 4: six orbits of Earth. over about a 10-hour period, but
[43:37] Speaker 4: they decided to come back. We still don't know why.
[43:39] Speaker 4: We will find out. The plan is for the Starship spacecraft,
[43:44] Speaker 4: which you're seeing live on your screen, to splash down
[43:48] Speaker 4: west of the coast of Chile.
[43:50] Speaker 1: South America. out of this world.
[43:53] Speaker 4: That does it for this edition of Bloomberg Tech. Lots
[43:56] Speaker 4: of conversations in the show, many market moves on the headlines.
[43:59] Speaker 1: Recap on the podcast.
[44:00] Speaker 4: You know where to find it, on the Bloomberg Terminal,
[44:02] Speaker 4: as well as online on Apple, Spotify, and iHeart.
[44:06] Speaker 1: What a way to start the week. This is Bloomberg Tech.

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.