In short
NVIDIA expands its record share buyback by $150B (total $235B) and unveils “OpenShell” plus “NVIDIA Sentry” to rein in rogue AI agents in milliseconds; the episode also covers SpaceX’s Starship reaching true orbit and deploying 26 V3 Starlink satellites despite an engine cutout, plus Meta hiring MongoDB’s CJ Desai to lead a new enterprise AI platform.
Guests (backgrounds)
- Maggie Eastland (Bloomberg reporter covering AI safety/security).
- Erica Clower (Founder & CIO, Science and Technology Partners; tech/AI investor).
- Scott Morton (former SpaceX Starship engineering manager; CEO of Revel).
- Lauren Grush (Bloomberg space reporter).
- Robert Schiffman (Bloomberg Intelligence senior credit analyst).
- Jake Lusreri(an) (CEO, Gecko Robotics; physical AI/robotics safety).
- Alex Levine (Bloomberg reporter on Meta/MongoDB).
- Mike Shepherd (Bloomberg senior tech editor; AI policy coverage).
- Kunjan Sabani (Bloomberg Intelligence analyst).
Key claims + examples
- NVIDIA says Sentry could have stopped the Hugging Face breach involving OpenAI models.
- OpenShell runs on NVIDIA CPUs; Sentry isolates suspicious agents on NVIDIA DPUs.
- Starship Test Flight 14: one Raptor engine cut out early; reached 275 km orbit; deployed 26 V3 Starlink satellites; deorbited early (~3 hours vs ~10).
- Meta Enterprise Platform: CJ Desai (MongoDB president/CEO) leads enterprise agents/infrastructure; MongoDB shares drop sharply.
Written by AI. May contain mistakes. Listen to the episode to check what was said.
Chapters
Tap a time to open that second in VONVIDIA's Record Buyback Plan
1:54 to 2:56
Discussion on NVIDIA's massive share buyback plan and its implications.
“taking the total amount to$235 billion, plus a milestone for SpaceX's Starship rocket reaching true orbit for the first time after a brief scare due to an engine failure during Ascent.”
AI Safety Solutions by NVIDIA
2:56 to 4:51
Exploration of NVIDIA's new tools for AI safety and their potential impact.
“The company even going so far as to say it could have stopped that recent hugging face breach which involved OpenAI's models.”
AI Industry Insights and Challenges
4:51 to 7:20
Analysis of AI adoption trends and the implications for the industry.
“Just very quickly, what's the NVIDIA school of thought about what this means for the training runs that are happening in industry?”
Impact of Economic Factors on Tech Investments
7:20 to 11:58
Discussion on how fiscal policy and economic climate affect tech investments.
“So I think it's quite a powerful message coming from NVIDIA today.”
SpaceX's Starship Milestone
11:58 to 12:11
Overview of SpaceX's recent Starship test flight and its significance.
“Coming up, SpaceX was able to deploy the V3 Starlink satellites on board the first true orbital test flight for Starship, but the milestone did have some challenges.”
Details on Starship's First True Orbital Flight
12:22 to 14:00
In-depth look at the challenges and achievements of Starship's orbital flight.
“Elon Musk's company was able to deploy Starlink satellites that were on board, but the trip did come with some hiccups, which actually throughout the pre-market and into today's session caused volatility in the shares.”
SpaceX Starship's Orbital Milestone
14:00 to 14:58
Learn about the significance of SpaceX's successful orbital test flight of Starship.
“It's important that Starship could reach that orbital altitude of 275.”
Insights from a Former Starship Engineer
15:00 to 20:46
Hear firsthand insights from Scott Morton on the complexities of Starship's software and operations.
“Scott Morton spent nearly a decade at SpaceX as an engineering manager for Starship, focused on software before leaving to found Revel, which builds software to help companies design, test, operate complex machines.”
Upcoming AI Developments at Meta
20:47 to 21:02
Discover Meta's new AI initiatives and leadership changes.
“to lead its new AI pushes for businesses, for the enterprise.”
AI and Music Creation with Suno
22:14 to 23:18
Explore how AI is transforming music creation and the implications for the industry.
“AI is creating a new path for musical stardom.”
Show all 27 chapters
AI and Music Creation with Suno
23:32 to 24:29
Explore how AI is transforming music creation and the implications for the industry.
“Your logo, your message, your presentation, so your gift feels personal and on-brand.”
Meta's New Leadership Role in AI
24:31 to 25:26
Discussion on MongoDB's CEO transition to Meta and its implications in the AI landscape.
“It doesn't always work the way people expect it to.”
Meta's New Leadership Role in AI
25:30 to 28:14
Discussion on MongoDB's CEO transition to Meta and its implications in the AI landscape.
“Okay, take a look at shares of Meta and also MongoDB.”
Introduction to Gecko Robotics
28:14 to 28:34
Discussion about Gecko Robotics and their partnership with NVIDIA.
“Yeah, and we're showing some of those that kind of took a hit after those headlines hit this morning in the enterprise software space.”
Meta's Enterprise Platform
28:34 to 29:01
Analysis of Meta's strategy shift towards enterprise solutions.
“And Gecko Robotics is the physical AI case study.”
NVIDIA's Massive Buyback Announcement
29:01 to 30:06
Discussion on NVIDIA's stock buyback strategy and its market implications.
“Generally, the sell side is really positive about the idea that Meta has a mechanism now to do business with other companies rather than just focus on the consumer, leveraging its work in agents and AI.”
Robert Shiffman's Insights on NVIDIA
30:06 to 33:54
Expert insights on NVIDIA's financial strategy and AI ecosystem.
“You were so quick out the gate with the React.”
Gecko Robotics CEO on AI Safety
33:54 to 36:23
Jake Lusererian discusses AI safety measures in robotics.
“NVIDIA is pushing its AI safety efforts into the physical world.”
Partnership Implications for Military and Robotics
36:23 to 39:40
Exploration of the implications of the NVIDIA and Gecko Robotics partnership.
“but do so in a way that doesn't compromise safety and quality standards that humans set and prioritize.”
SpaceX's Starship Mission Update
39:40 to 40:09
Overview of SpaceX's Starship reentry mission and results.
“And that's what's leading to such incredible outcomes.”
AI-Generated Music Revolution
42:00 to 43:01
Discover how Suno is changing music creation through AI technology.
“As Bloomberg's Lucas Shaw and Ashley Carman report, Suno, an AI-generated music company, lets users generate a song in any style based on a text prompt.”
AI-Generated Music Revolution
43:13 to 44:12
Discover how Suno is changing music creation through AI technology.
“It's a thank you, a milestone, a moment of appreciation.”
Optum's Healthcare Innovations
44:12 to 45:09
Explore how Optum is integrating patient care to improve healthcare efficiency.
“Let's talk about healthcare for a second.”
AI Safety Debate: Trump and Amodei
45:19 to 48:01
Analyze the implications of the AI safety debate between Trump and Amodei.
“A White House official has confirmed that it took place.”
Nvidia's Stock Movements and Insights
48:01 to 49:25
Get insights into Nvidia's buyback plans and its market impact.
“An IPO would add another layer to the broader SK group's already complex ownership structure, raising concerns for shareholders.”
Lockheed Martin and Nvidia Partnership
49:25 to 53:24
Learn about Lockheed Martin's long-standing partnership with Nvidia in AI.
“There's also been some disclosures on their backlog, how much NVIDIA gear Anthropic is using.”
SpaceX's Starship Test Flight
53:24 to 54:18
Explore the details of SpaceX's recent Starship test flight and its significance.
“to what's happening in defense technology.”
Transcript
Automatic transcript. May contain errors.0:00The most effective people at work aren't working harder than everyone else. They're working smarter, inside better systems. Superhuman Go, from the makers of Grammarly, is the AI assistant that works inside every tool you already use, always ready, already aware of what you're working on. It's a teammate whose only job is to help you be better at yours. With Go working with you, you can show off what you do best. See what Superhuman Go can do at superhuman.com. The thing about AI for business, it may not automatically fit the way your business works. At IBM, we've seen this firsthand. But by embedding AI across HR, IT, and procurement processes, we've reduced costs by millions, slash repetitive tasks, and freed thousands of hours for strategic work.
0:49Now we're helping companies get smarter by putting AI where it actually pays off, deep in the work that moves the business. Let's create smarter business. IBM. Never bet against American grit or American energy. Through innovation, Venture Global is not only building some of the largest energy facilities in the world right here in the United States, but delivering American energy at a fraction of the cost and a fraction of the time. So while others are busy talking, we're busy building. That's Venture Global. That's unstoppable energy.
1:31Bloomberg Audio Studios. Podcasts, radio, news.
1:42Bloomberg Tech is live from the heart of Silicon Valley with Ed Ludlow in San Francisco. This is Bloomberg Tech. Coming up, NVIDIA boosts 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 true orbit for the first time after a brief scare due to an engine failure during Ascent. And Meta taps MongoDB's president and CEO to lead a new AI platform for enterprise customers it deems as the next major pillar of its business. Let's get right to today's big number,$150 billion. That's how much NVIDIA has added to its share buyback plan, taking the total remaining authorized amount to$235 billion, a massive number, reflecting CEO Jensen Wong's confidence in its continued growth.
2:39The company expects to complete the program through fiscal 2028, according to a statement. They're currently in the third fiscal quarter of their fiscal year 27. At the same time, NVIDIA says it has a new way to rein in rogue AI agents, introducing new security tools that can shut them down in milliseconds. The company even going so far as to say it could have stopped that recent hugging face breach which involved OpenAI's models. Bloomberg's Maggie Eastland has the details, and it's kind of in two parts. Explain them. Absolutely. So this new product release has two layers. So first, there's OpenShell, that is an open source software program that can run on NVIDIA's central processing units.
3:19And that's designed so that a user can impose certain guardrails and enforce those guardrails. Now, this was first announced in March back at their GTC conference, being rolled out now, in combination with a completely new product called NVIDIA Sentry, which is another separate system that runs on NVIDIA's data processing units. And essentially, it's an added layer that can, like you said, isolate an agent that's acting suspiciously, the company says, in milliseconds. The analogy that I've been going with this morning is like it's putting an AI agent in a locked room where part one controls what that room looks like and the doors in and out of it.
3:55Part two is a security guard that's always outside. But it's kind of in line with Jensen Wong's school of thought that this is an engineering issue at the heart of the AI safety debate. Well, he's given us an AI engineering solution. Absolutely, Ed. That's the line I've been using. You know, Jensen Wong, one of the biggest beneficiaries of the AI trade, has, you know, really continually voiced that AI progress needs to continue and that the executives who want to slow down just need to find a solution to their problem. And now he's offering them the solution. Importantly, you know, we don't know whether OpenAI and Anthropic are going to adopt this for their training run specifically, which is where they've had the most problems with these rogue agents.
4:37But if this really can solve the issue, it could really quiet some of these calls that we've been hearing for a slowdown. There is an element where each of those solutions runs on other parts of NVIDIA's ecosystem hardware-wise. Just very quickly, what's the NVIDIA school of thought about what this means for the training runs that are happening in industry? Yeah, I think the idea here is that, you know, NVIDIA has said this potentially could have stopped the hugging face incident, which was an incident where agents got out during a training run. So if the technology works in the way that they're saying it works, it could enable OpenAI and Anthropic to continue with some of these training runs that they've been more hesitant about.
5:21We saw just last Friday OpenAI said they were pausing training after new disclosures about some government-related incidents and a breach of an Australian government website. So, again, this could be really a breakthrough if the technology does work the way NVIDIA is claiming. NVIDIA shares up 2 % on multiple pieces of news. Bloomberg's Maggie Eastland on the AI safety component of it. For more, let's get to Erica Clower, founder and CIO of Science and Technology Partners. And this is so interesting because this is NVIDIA acting on their warning that this is an engineering problem. Here's a solution.
5:58What do you make of that? NVIDIA just offering that to industry. So good morning, Ed. Nice to see you. It's such an interesting time once again for NVIDIA showing and flexing its muscle of innovation. I think one of the things that's sometimes hard to remember is that over half of NVIDIA's engineer are software engineers. These are engineers who are experts, not only in terms of software expertise, but also in cybersecurity. So it makes total sense that the company would be stepping in to alleviate some of the concerns around the cyber area. Yeah. What's the link between the share buyback story, this AI agent safety story and the other disclosures that that NVIDIA made this morning?
6:46So I think that perhaps the takeaway is that despite all of the concerns about the amount of spending that's going into AI infrastructure and whether or not there's an AI bubble, NVIDIA is saying not only are we stepping in to buy more of our own shares and a vote of confidence about the future of the industry, but we also have the cash to do it. We are generating sufficient cash flow to not only make these investments, but also to be very shareholder friendly. So I think it's quite a powerful message coming from NVIDIA today. The company had also kind of disclosed that their commits or their backlog continues to grow.
7:29Anthropic is a case study to date,$180 billion of NVIDIA hardware commitments across its different deployments. You're a long biased fund. You're focused on like the absolute leading edge of what's happening. We are through the third quarter of earnings season, right? What did you conclude about the AI story from that earnings season? I think that the most important takeaway from the third quarter was just how broad-based AI adoption has been, whether it's a financial or a drug development company or a consumer-facing company that's looking to improve customer service and response. the overwhelming adoption of AI tools has been phenomenal.
8:15And I think that that really speaks to the idea that instead of prior problematic areas where capacity was built ahead of demand, we have the flip side, which is that we simply don't have enough compute and capacity and power and all of this to meet the demand, which continues to grow across many, many different industries. The story that happened in parallel with Earnings Season was the AI safety debate, pacing at the frontier, what kind of third-party evaluation is needed to make AI development as safe as possible. Is there any direct line to how that impacts that supply-demand on the compute side that you've just outlined?
8:57I think that the, I think, Ed, you bring up an excellent point, which is that there has to be more innovation around the security and especially around agentic security in order to give continued confidence in adoption of these key technologies. So I do think that this provides an excellent opportunity for many of the cybersecurity software names. But I also think that this now will become table stakes for other chip vendors, whether it's AMD or Marvell or Broadcom. Having a security aspect to the platform is going to be table stakes. Erica, how much of a warning signal was the developments around AI safety to you?
9:41I do think that the AI safety is a profound issue. And I think that the speed at which we are seeing response is very reassuring. I think the other thing that we heard a lot from many of the cyber software companies was that they saw a flip in terms of trying to go out to sell their wares to customers, to actually customers coming in to talk to them about how they could thwart any kind of cyber threats. So clearly this is a big issue and what I think is very interesting is just how fast the response time has been. We're showing Nvidia is higher but generally speaking technology shares are under pressure today.
10:22Yields are at multi-year highs and the headline that hit just before we came on air is that the US 10-year treasury yield has moved up about 10 basis points, 5.27%. It's back down 5.26%. Just how closely you're paying attention to this U.S. economy and the actions of both fiscal policy and the Fed right now in the context of your investments in the technology sector? Well, I mean, we've seen a tremendous correction in technology shares since June. In many cases, there are companies, premier companies, that are down 30, 40, 50 percent. So I think we're in the midst of a bottoming process against a backdrop of fiscal tightening.
11:11And also we're heading into midterm elections where many concerns are being surfaced about the pace of data center builds in the United States and the pressure on the grid prices and also environmental concerns. But we've been able to navigate through most of those. My suspicion is that at the end of the day, earnings will matter most. And we've seen extraordinary financial results across the board for the best in class AI players. And I think that this is probably an interesting moment in time to really think about seeing these companies and buying these companies at extraordinary values relative to their growth prospects on a longer term basis.
11:53Erica Clower of Science and Technology Partners. Great to have you back on the show. Thank you very much indeed. Coming up, SpaceX was able to deploy the V3 Starlink satellites on board the first true orbital test flight for Starship, but the milestone did have some challenges. We're going to have the details next. This is Bloomberg Tech.
12:21SpaceX's Starship rocket lifted off just a few hours ago for its 14th test, its first true orbital flight. Elon Musk's company was able to deploy Starlink satellites that were on board, but the trip did come with some hiccups, which actually throughout the pre-market and into today's session caused volatility in the shares. starship now returning to earth earlier than planned let's get the absolute latest with bloomberg space reporter lauren grush um overall successful but where are we at right now in this 14th test right it wouldn't be a starship flight if it didn't have a little drama associated with it so yes it was the first orbital mission they were able to achieve orbit after some confusion or consideration of whether they were able to do so one of starship's raptor engines did cut out early during that initial climb to space.
13:11So there was a part where they weren't sure if they were able to do that orbit insertion, but they were able to push ahead as planned and successfully deploy all 26 V3 upgraded Starlink satellites on board. But they ultimately opted to bring Starship home early. I think right about now it's deorbiting and coming back home. We'll probably hit the atmosphere soon. It was supposed to be a nearly 10-hour test flight, but now it's going to be about three hours instead. Exactly. So at the bottom of the spacecraft part of Starship, there are six raptors. One of them misbehaved in the ascent. They did reach orbital altitude, 275 kilometers.
13:52And there'll be people watching, Lauren, that are like, hold on, I've seen Starship in space. Like, what's the difference? So I think get into that. It's important that Starship could reach that orbital altitude of 275. Yeah, up until now, all of Starship's test flights have been suborbital, meaning it's been on a trajectory where no matter what, the gravity of Earth would take it back. Just so that, you know, if there were any issues with Starship on board, that it would come back and fall in a place that wouldn't have any kind of issues for people or uninvolved property. And so going to orbit is a big decision because of, you know, if there was an issue with the engines, they want to make sure that they can be able to take Starship out of orbit safely, safely, which is hopefully what they're doing right now.
14:38But yeah, it's a big moment. And, you know, we'll see what the FAA says and what SpaceX says following the data review after this, but it could potentially open the door for more orbital flights to come, more Starlink deployments to come. And as we know, that's a big moment for SpaceX. Bloomberg's Lauren Grush, thank you very much. Let's talk to someone who's worked on Starship. Scott Morton spent nearly a decade at SpaceX as an engineering manager for Starship, focused on software before leaving to found Revel, which builds software to help companies design, test, operate complex machines. I think we'd agree, Scott, that Starship is a complex, multifaceted machine.
15:19But just your reaction, you've been inside the beast, right? You understand how important this milestone was for TestFlight 14. Yeah, I mean, I think, I mean, overall, this was a monumental step forward, like actually going to orbit. I mean, that really meant that SpaceX and the FAA thought that Starship was ready. You know, they did have a hiccup with one of the engines, but, you know, they deployed Starlink satellites in orbit. And then they've also, it sounds like they've successfully completed the first deorbit burn. So I think overall, this is like, I mean, if I were internal at SpaceX still, I would be seeing this as a massive success.
16:02When you, the time you'd left SpaceX, you were the software manager for the Starship program, right? We talk a lot about the iterations on the hardware, the heat shield in particular, but generation to generation, those hardware changes they make. But software is absolutely critical because the launch system and the spacecraft itself is essentially autonomous, right? Explain that part. Yeah, I mean, it's actually autonomous, I'd say, even further out on the launch pad getting ready for flight. um but yeah uh you know uh if you can plan for everything that you want to do with the vehicle and encode all that in software uh it's a much easier uh kind of thing to reason through than having a lot of manual commanding as much as you can automate as you possibly can it's definitely the best approach um and even like the scenario that happened today uh this was uh definitely fully planned for there was software already ready to go for all these types of contingencies I am familiar with over-the-air software updates for my Tesla vehicle, right?
17:08And that gives, with existing hardware, new capabilities to the car, improves performance. How does that work in the context of the most powerful rocket and spacecraft launch system that humankind has developed? Does it work the same way? at least at my time at spacex we could not kind of do over the air updates i'm not sure where they are today on that um there are a lot of things that you can tweak uh you know parameters and things that you can update but generally you want to have you know the software be quite locked down and not be kind of making changes especially not on flight um i assume long term they'll have some type of capability for this when you have you know vehicles on orbit for very long durations but at least in my time on the program, you know, you want to have the software very locked down.
17:59There's a lot that you need to make sure is kind of that you're operating within the bounds of what is expected and kind of not kind of going too far off of those plans in real time. When people kind of question whether or not a mission is successful against its original goals, one of the things that SpaceX always talks about is the data that is generated from whatever the outcome of the mission. How do the software teams take the data and then use it? Like if there was a sort of iterative process on software, how meaningful is the data from the mission that's gathered? I mean, it's massive. They're going to spend weeks and weeks, probably actually months on this one, going through every single system, all the software that was run, literally point by point in some cases, like telemetry point by point, to make sure everything performed as expected.
18:54So yeah, the data is actually, I'd say, the most valuable thing coming out of this flight. They actually went to, I just like, I was on the program from the very beginning. I think I was probably the first software engineer down in Boca Chica, Texas, down in a Starbase. And the fact that it has now gone to orbit is just incredibly exciting for me. But yeah, the most valuable thing coming out of this, all the data that they're gonna be going through, I'm sure they've learned a tremendous amount from this flight. take me back again to your time there you know in the when prior to spacex's ipo like the prospectus made it crystal clear that all of those kind of out of this world but future business lines they're all predicated on starship working being fully reusable rapidly reusable but when the program started what was the kind of vision for starship like what were you all talking to each about hoping to achieve with it at that time?
19:48Yeah, I mean, even back at the very beginning of the program, we had kind of CAD drawings of what is there today. I mean, this is exactly what we were hoping for and what ultimately happened. And obviously much further beyond that, that Starship would be, you know, it takes people to Mars. That is 100 % what we were thinking about day to day as we're putting in the extremely long hours. So, yeah, I think, you know, this is exactly, this is like, for me, this is like absolutely incredible that we, you know, Starship has gone on to orbit. It has actually now successfully done the first sea orbit burn, which is kind of one of the most concerning, I'd say, parts of the whole mission.
20:30But yeah, I mean, extremely exciting for everyone at SpaceX and everyone that's contributed to it to date. Scott Morton, now the CEO of Rebel, formerly at SpaceX. managing the software of the Starship program. It's great to have you back on the show. Thank you very much indeed. Coming up, Meta has tapped MongoDB's chief to lead its new AI pushes for businesses, for the enterprise. It's been a big story this morning, and it's next. This is Bloomberg Tech.
21:02The people who seem to get more done than everyone else. They're not working longer hours or running on more caffeine. they've just stopped wasting time on the stuff that doesn't move work forward. Switching apps, re-explaining context, hunting for files, those aren't small inefficiencies. They're hours wasted every week. Superhuman Go gives you those hours back. From the makers of Grammarly, Go is an AI assistant that sits inside every tab and tool you already use, always available and already aware of what you're working on. Ask it to draft something, summarize a long thread, pull up a file, or prep you for a meeting.
21:43Go handles it without you ever leaving the page you're on. This is what it looks like when AI actually fits into your work, instead of adding to it. It's like having a teammate whose only job is to help you be better at yours. Go keeps up so you can move forward. With Go working with you, you can show off what you do best. Superhuman Go. Find out more at superhuman.com. That's superhuman.com. This is the Bloomberg Tech Minute brought to you by ChatGPT. Now with ChatGPT work. I'm Carol Masser. AI is creating a new path for musical stardom. As Bloomberg's Lucas Shaw and Ashley Carman report, Suno, an AI-generated music company, lets users generate a song in any style based on a text prompt.
22:29People can upload their own lyrics or start from scratch. record their own voices, or rely on ones provided by Suno. While most of the songs generated on the service go unheard of by the masses, a few have taken off, including a recent TikTok trend that involves users uploading text messages to Suno and turning them into musical performances, like taking the text from a crazy night out and making it a gospel song. Suno's pitch to investors is that it'll democratize music creation, leading to a more than$5 billion valuation for the company. The music industry, it's watching, as its biggest concern about AI is whether record labels, rights holders, and artists will be compensated for the work they claim built the training models for companies like Suno.
23:13Another question? Will these AI bangers have any staying power or just be one-hit wonders? That's the Bloomberg Tech Minute brought to you by ChatGPT. Put ChatGPT to work on your most ambitious ideas and projects. Get started at ChachiPT.com today by selecting Work Mode, available on Plus and Pro plans.
23:55Every detail. Your logo, your message, your presentation, so your gift feels personal and on-brand. And with expert support, dependable service, and thousands without a setup fee, creating something thoughtful doesn't mean making it complicated. Every order is backed by 4imprint's 360-degree guarantee, so you can be 4imprint certain it'll arrive exactly as expected, on time, and with the care your brand deserves. Because when the moment matters, the right gift speaks volumes. And the right partner makes it easy. Explore gifting with purpose and certainty at 4imprint.com. 4imprint. 4certain. Let's talk about healthcare for a second.
24:32It doesn't always work the way people expect it to. If you've ever waited on a prescription refill or had a hard time getting the care you needed, you know the feeling. The system should just work better for everyone. That's exactly what the people at Optum are trying to do every day. They're a healthcare company linking patient care and pharmacy services and using data and technology to drive the whole system so care is connected, not complicated, for patients and providers. Things like making it easier to get care that looks at the whole person, from primary care doctors to mental health support and even in-home care, and then using technology to make sure they all work together.
25:09Technology designed to help doctors spend less time on busy work and more time with their patients. And those prescriptions, Optum is working to bring costs down, save patients money, and make it easier to get refills. Little by little, Optum is helping make healthcare work as one for everyone. Head to business.optum.com to see how.
25:36Okay, take a look at shares of Meta and also MongoDB. MongoDB quite clearly down much more severely. It announced that its CEO was leaving to go to Meta. CJ Desai is stepping down effective immediately. He's then set to lead Meta's new enterprise platform, bringing its AI models, agents, infrastructure to business customers. Bloomberg's Alex Levine joins us with the reporting. Okay, there's a lot going on here. Let's start with the enterprise platform. What is that? So for well over a year at this point, we have been writing all about the ways that Meta has been continuing to invest enormously in AI.
26:19And I think if you're reading the tea leaves, especially over the better part of the last year, this may not be all that surprising that Meta is now moving into this enterprise platform to offer its services like agents, like AI infrastructure, and other services for businesses. The sell side's really positive about Meta Enterprise Platform. The stock is down 4%. Let's go to the CJ Desai piece of this. MongoDB down 18%. At one point, it was down 27%, and on track for its worst day since March of 2025. This is somebody, a talent that Mark Zuckerberg wanted for this. What's the story? Yeah, so Desai is an industry veteran from MongoDB, this document database.
27:03I think as we've been covering with other major recent hires, including Alexander Wang last year, this is sort of the next iteration of what Zuckerberg sees as is necessary to really extend superintelligence, not only to all people, as he has said, but also to all businesses and the developers behind them. And so he sees Desai as really fit for that. Okay. People will be watching going, okay, meta, the parent of Facebook, social media, more recently AI, but consumers, I get that. This new evolution, this pillar, it's about doing business with other companies. Exactly. And I think it's really important to also look at where this plants meta in the larger space with its competitors in the AI race.
27:49So you've got other giants in the industry like Microsoft, like Amazon Web Services, like Google, who are already offering many of these same services to their customers. So I think not only is this one way that Zuckerberg can diversify the company's revenue streams as it pursues these enormous AI ambitions, but it is also very much a strategic and competitive move as Meta seeks to keep pace with its rivals. Yeah, and we're showing some of those that kind of took a hit after those headlines hit this morning in the enterprise software space. Bloomberg's Alex Levine, thank you very much. It's great to have you back on the program.
Read the full transcript
28:24Coming up, NVIDIA and Gecko Robotics are building new guardrails for autonomous machines. We're going to speak to Gecko Robotics CEO Jake Lusrerian next. This is the AI agent safety story that we did at the top. And Gecko Robotics is the physical AI case study. Super interesting. Halfway through the program here in the city of San Francisco. Thanks for being with us. This is Bloomberg Tech.
28:58Welcome back to Bloomberg Tech. Meta Enterprise Platform, it's a big story in markets this morning. Generally, the sell side is really positive about the idea that Meta has a mechanism now to do business with other companies rather than just focus on the consumer, leveraging its work in agents and AI. It has tapped CJ Desai, who was the CEO and president of MongoDB, and that stock is down accordingly. off-session lows but down 17%, a pretty severe reaction as CJ Desai leaves MongoDB and goes to Meta to lead Meta Enterprise platforms. And they brought back a former CEO at Mongo to bridge the gap while they searched for a long-term replacement.
29:35It's had a lot of effect on the market in different ways to different enterprise stocks. The other big kind of financial market story is NVIDIA, putting another$150 billion behind its own stock, taking its remaining buyback authorization to$235 billion, the stock higher 2 % off its session highs. Bloomberg Intelligence says the size of that commitment underscores NVIDIA's financial firepower, but with plenty of room to keep funding the broader AI build out. Robert Shiffman, Bloomberg Intelligence, senior credit analyst joins us now. You were so quick out the gate with the React. Just explain your thesis on this buyback.
30:13Well, I think NVIDIA has been reading Bloomberg intelligence research, and they realize that there's no AI bubble. And because of that, they're going to be so cash rich, they can not only fund friends and family and keep this AI ecosystem going, but they can also spend a significant amount on shareholder returns, benefiting, I think, both creditors and shareholders alike. And this sort of stuff, I think, over time is only going to get bigger because we see their free cash flow potentially going to a trillion dollars over the next couple of years. Shif, sincerely, and just like in very basic terms, what is the credit analyst's interest when you get a headline about what is now, I think, the biggest buyback in U.S.
30:56company history? Yeah, it's bigger than Apple's$110 billion. It certainly shows that they are king of the hill when it comes to AI these days. Unfortunately, there's not so many other hyperscaler friends that are in this position anymore. I think from a creditor standpoint, it doesn't really make any difference to the way NVIDIA trades. What I do think, though, it suggests is that they're going to have enough cash to give out in order to fund everybody else. If you look at what we think CapEx is going to be spent through 2031 from the hyperscalers, it's now going to reach$9 trillion. That's with a T.
31:33And I think next year we're going to see IG borrowing probably in and around$500 billion. And to know that there's capital around to help build that system, I just think brings confidence. People have been very, very concerned that there's both an AI bubble, there's a credit bubble. We don't see it at all. We've never seen demand stronger, and we haven't heard anything different from either investors or the companies themselves. Okay, so as of the end of the fiscal second quarter, fiscal year 27, the most recent quarter, NVIDIA is sitting on just north of$99 billion, cash, cash, quibs. you, A, don't see a circular financing concern, and you, B, see their ability to make different kinds of investments grease the wheels for the build-out.
32:21Yeah, listen, circular financing is a great media buzz line to try to bash this space. And I just don't think it's working. What they're doing is they're making equity investments. They're making credit-style investments where They're both leasing facilities. They're providing residual value guarantees. They're providing support behind the value of their chips beyond their depreciable life. I just think what they're doing is they're just acting as a bank for everybody, which is not a bad thing. I mean, this is the whole story of Robinhood. You take a little bit from the rich and you give to the poor.
32:54That's all they're doing. It's a timing delay when it comes to making money in this business. You know, those that are collecting right now, the semis, are sharing that money with those that are building, the hyperscalers and the neoclouds. That is going to change in the next couple of years when those companies have facilities that are up and running and you see AI monetization start to boom. They're all going to turn free cash flow positive and they're all going to start buying back masses and amounts of stock again. And I think this sort of revolution is just going to end up continuing. and I don't see NVIDIA slowing that down and I don't think they care about circular referencing other than they're also now raising third-party money.
33:32Remember, they just announced a deal with the biggest alternative asset managers out there to help raise$500 billion and I don't think that money is getting any smaller. I'm surprised it wasn't a trillion or it's not going to grow to a trillion or two trillion in the next couple of years. Robert Schiffman of Bloomberg Intelligence with a credit perspective on this massive buyback program. Thank you very much. NVIDIA is pushing its AI safety efforts into the physical world. It's working with Gecko Robotics to test new safeguards that put hard limits on what AI-powered robots can do and stop them when they cross boundaries.
34:07Gecko CEO Jake Lusererians with us. And this was our top story today, right? Part one, make an agent fit within parameters that you set in a box, right? Part two, a security guard for when that agent misbehaves or acts suspiciously. How does that work in your physical AI world? Well, thanks for being able to have me on and talk about this. What's so important and we're all talking about today is the criticality of these agentic tools being able to be deployed in the real world, especially being deployed in such important environments, but doing so in a way that's safe and ensures that we have the utmost standards as it relates to safety, quality, especially with some of the most critical sectors, whether it be on the military side or on the commercial side with energy, mining, and manufacturing.
34:51And that's exactly what this partnership is all about, is ensuring that we keep humans in control, but also supercharge the benefits of these authentic tools coming out of the Foundation Lab. So really exciting announcement, and it puts safety in humans first. This is a partnership NVIDIA has with about 100 different entities across software, hardware, robotics companies. Give me a case study that can make it real for the audience, right? It does all sound a little bit abstract. So an AI agent in the context of something like a robot in the industrial and defense settings that you work in. Well, what's becoming very clear is that I think impressive and critical these models are to such an incredible ecosystem of tech.
35:38What's becoming really clear is that deployment into the real world on the physical intelligence side, that's where real large moats are coming from. by gathering information, data sets never existed before, to power these abilities to model and simulate what the real world looks like and how to make sure it optimizes for better. A way to understand one of the use cases is actually with the military contracts that we have. One of our robots, Komodo, is actually using this partnership and the system that the open agent safety platform that NVIDIA and Gecko are pioneering to actually ensure that we're gathering information and data sets by letting agents run the robots, but ensuring that parameters are set and humans are able to control certain sorts of actions that the robots are taking in the real world.
36:20And so the agents need to be able to path plan and gather information and supercharge where they're taking actions and getting data, but do so in a way that doesn't compromise safety and quality standards that humans set and prioritize. And this is exactly the key to ensuring that we can get as much information and data as possible at low cost and do so in a way that keeps humans in control and safety paramount. Take me into the backstory here. Part of this platform, NVIDIA talked about at GTC earlier this year, right? There has been so much in the headlines about AI safety, models misbehaving, agents going rogue.
37:00Like, was there a frantic phone call from Jensen Huang last week saying, okay, this is what we're doing, everyone get on board? Yeah, I'll keep the phone calls between the two confidential, but you're 100 % right. which is that you need to prioritize safety, but you also need to be moving at a pace that's really bringing to bear the best technology in the world for the people keeping our borders safe and keeping freedom for the world. You know, it's so exciting that Jensen at NVIDIA and myself at Gecko are able to be doing a partnership like this for the military as the first example of many to come as it relates to the technology that's being deployed to ensure that we keep ships out of dry dock and in the fight and ensure that we can deter conflict around the world.
37:47And this is putting such a positive spin on the benefits and impacts that this technology stack is able to have. And also the game plan of how robotics has to be used to give us the advantage in the West. And so it's an exciting partnership because it pushes the frontier and ensures that technologies and technology companies like Echo and NVIDIA are leading the way and creating the rules. You know, Jensen Huang's school of thought is that the safety of a model and an issue of alignment, does the AI do what it's intended to? That's an engineering problem. Is that what you see in your physical AI realm?
38:31It's just an engineering problem that you can now solve for because of what NVIDIA has made open source. Yeah, it also is ensuring that we put into perspective the power of these tools. I mean, we hear so much conversation from certain CEOs about the doom of AI. But I think what's important is this partnership highlights how these technologies, robotics and artificial intelligence models, are actually being used for good. For keeping humans out of harm's way, to ensuring that we get critical assets and components back up and running. Ensuring that they're deterring conflict or keeping lights on, empowering homes for heating and cooling.
39:11This is the kind of way that we want to imagine the future looking like for robotics and AI and not some doomerism that can be sold to assist whatever is trying to be pushed by these companies. And so similar to Jensen, what we're here talking about is such a positive impact that these technologies are having for real people and ensuring as well that we keep those real people with those sort of expertise in control. And we're building technology stacks right alongside them. And that's what's leading to such incredible outcomes. Jake Lusararian, Gecko Robotics CEO, on that technology partnership with NVIDIA.
39:48Thank you very much indeed. Let's talk about SpaceX again. We're seeing some live pictures now. Live pictures. This is Starship reentering Earth's atmosphere. It did reach an orbital altitude of 275 kilometers. But SpaceX made the decision to bring Starship back to Earth and back home early. It was originally the mission goal to do six orbits of the planet and then after about nine hours and 50 minutes, bring it back. But they came back early after successfully deploying 26 V3 Starlink satellites, the feed breaking up there as it passes through Earth's atmosphere. Mission successful, mostly. Coming up, Anthropics CEO Dario Amodei met with President Trump as the debate around AI safety continues.
40:36We'll have those details next. This is Bloomberg Tech.
40:45The people who seem to get more done than everyone else. They're not working longer hours or running on more caffeine. They've just stopped wasting time on the stuff that doesn't move work forward. Switching apps, re-explaining context, hunting for files. Those aren't small inefficiencies. They're hours wasted every week. Superhuman Go gives you those hours back. From the makers of Grammarly, Go is an AI assistant that sits inside every tab and tool you already use, always available and already aware of what you're working on. Ask it to draft something, summarize a long thread, pull up a file, or prep you for a meeting.
41:25Go handles it without you ever leaving the page you're on. This is what it looks like when AI actually fits into your work, instead of adding to it. It's like having a teammate whose only job is to help you be better at yours. Go keeps up so you can move forward. With Go working with you, you can show off what you do best. Superhuman Go. Find out more at superhuman.com. That's superhuman.com. This is the Bloomberg Tech Minute brought to you by ChatGPT. Now with ChatGPT work. I'm Carol Masser. AI is creating a new path for musical stardom. As Bloomberg's Lucas Shaw and Ashley Carman report, Suno, an AI-generated music company, lets users generate a song in any style based on a text prompt.
42:12People can upload their own lyrics or start from scratch, record their own voices, or rely on ones provided by Suno. While most of the songs generated on the service go unheard of by the masses, a few have taken off, including a recent TikTok trend that involves users uploading text messages to Suno and turning them into musical performances. like taking the text from a crazy night out and making it a gospel song. Suno's pitch to investors is that it'll democratize music creation, leading to a more than$5 billion valuation for the company. The music industry? It's watching, as its biggest concern about AI is whether record labels, rights holders, and artists will be compensated for the work they claim built the training models for companies like Suno.
42:56Another question? Will these AI bangers have any staying power or just be one-hit wonders? That's the Bloomberg Tech Minute brought to you by ChatGPT. Put ChatGPT to work on your most ambitious ideas and projects. Get started at ChatGPT.com today by selecting Work Mode. Available on Plus and Pro plans. A business gift is never just a gift. It's a thank you, a milestone, a moment of appreciation. It's a message about how much someone matters and what your brand stands for. At 4imprint, you'll find thousands of customizable options, like premium apparel, branded drinkware, tech, totes, and more, each chosen not just for function, but for meaning.
43:37You can tailor every detail, your logo, your message, your presentation, so your gift feels personal and on-brand. And with expert support, dependable service, and thousands without a setup fee, creating something thoughtful doesn't mean making it complicated. Every order is backed by 4imprint's 360-degree guarantee, so you can be 4imprint certain it'll arrive exactly as expected, on time and with the care your brand deserves. Because when the moment matters, the right gift speaks volumes, and the right partner makes it easy. Explore gifting with purpose and certainty at 4imprint.com. 4imprint.
44:124certain. Let's talk about healthcare for a second. It doesn't always work the way people expect it to. If you've ever waited on a prescription refill or had a hard time getting the care you needed, you know the feeling. The system should just work better for everyone. That's exactly what the people at Optum are trying to do every day. They're a healthcare company linking patient care and pharmacy services and using data and technology to drive the whole system so care is connected, not complicated, for patients and providers. Things like making it easier to get care that looks at the whole person, from primary care doctors to mental health support and even in-home care, and then using technology to make sure they all work together.
44:51technology designed to help doctors spend less time on busy work and more time with their patients. And those prescriptions? Optum is working to bring costs down, save patients money, and make it easier to get refills. Little by little, Optum is helping make healthcare work as one for everyone. Head to business.optum.com to see how.
45:19president trump and anthropic ceo dario amode taking their ai safety debate face to face the two met for dinner after amode called for slowing the development of the most advanced models as a growing number of ai security breaches add urgency to the fight over guardrails bloomberg senior tech editor mike shepherd is with us um we know the dinner happened what else do we know we do know the dinner happened. A White House official has confirmed that it took place. It was late last night here in Washington. But we don't know anything more beyond that right now. There's been no readout from either side.
45:53And Ed, really, the key takeaway is that this meal would have brought together two men on opposite ends of the AI safety debate. As we all know, Dario Amadei has for years warned of the potential existential risks from the technology. And on September 12th, he published that high-profile essay promising to slow development of the most advanced models. President Donald Trump, on the other hand, has pushed back against those doomsday fears, calling them a hoax and also rejecting calls embraced by Amadei for more regulation as a globalist conspiracy. So we'll be eager to see whether the two men were able to meet somewhere in the middle.
46:30And we may get some details on that from President Donald Trump during a policy announcement later this afternoon. And we're also, of course, out to Anthropic for any readout on their side. Over the weekend, I think ahead of this meeting, the president did speak to reporters and he basically said, here's my point of view on AI. Could you summarize what the president's standing is on it? Well, for weeks, he has been in response to a lot of the safety fears that have been surfaced by leaders in Silicon Valley, including Amadei, but as well as his arch rival, Sam Altman, that something must be done on a global basis to create some standards to address these risks that AI agents could get out of control.
47:17But the president has pushed back against that. He is interested in two things. He wants development of this technology to proceed largely unfettered because of its economic importance to the United States and, of course, also to the importance to the U.S. of staying ahead of China, the top geopolitical rival in this technology. And that was really one of the centerpieces of the conversation between the president and his Chinese counterpart, Xi Jinping, here in Washington last week. A lot of other tech CEOs were at that state dinner for Xi Jinping on Thursday evening, but one absence notably was Dariu Amadei.
47:52So perhaps we will see a little bit of movement in the debate here in Washington, including with the president himself. and as shep mentioned the president is due to make a statement in a couple of hours time bloomberg's mike shepherd thank you very much indeed let's get over to new york bloomberg's yara yara anand has the news hi yara hi ed it's time now for talking tech first up shares of sk hynix fell five percent in south korea and its adrs are falling here as well on reports its u.s subsidiary solid dime could go public as soon as next year at a valuation of up to 100 billion. An IPO would add another layer to the broader SK group's already complex ownership structure, raising concerns for shareholders.
48:37Plus, China expanding overseas travel restrictions for families of top AI professionals, relatives of executives whose work is considered critical to state security, now require Beijing's approval before traveling abroad. That's according to sources. That's to prevent sensitive AI expertise from leaving for the U.S. And OpenAI CEO Sam Altman and Anthropics CEO Dario Amadei will not appear before an Australian Senate inquiry into AI. The hearing comes after an OpenAI model hacked an Australian government website. Both companies will send other executives instead. Ed. Thank you, Yohaira. Now coming up, we're going to go back to NVIDIA's buyback plans, one of several NVIDIA stories that are impacting the shares today.
49:24Stock up 2%. There's also been some disclosures on their backlog, how much NVIDIA gear Anthropic is using. And then there's the AI agent safety offering that they open source this morning as well. All to come, this is Bloomberg Tech.
49:52Let's get back to our top story. NVIDIA's record buyback. Shares 2 % off session highs. Actually, there were several pieces of news. Kunjan Sabani from Bloomberg Intelligence out with his react saying it's positive, but not thesis changing. Specifically when it comes to that$150 billion buyback increase. Kunjan Sabani from Bloomberg Intelligence joins us now. You know, it's such a big number. and then in aggregate and a bigger number still, why is it not thesis changing for you? Well, look, it does have some positive reads. Buyback is always good. Shareholder returns is always a positive. Also signals that the company believes that the stock is cheap and also helps sort of those investors who are concerned that the company is deploying too much cash into those financing deals instead of returning back.
50:41So definitely don't want to overshare all the positives. However, it doesn't do anything to address the fundamental risks that can the capex from hyperscalers sustain, the rising competition from the ASIC areas, and again, people who are concerned about sort of the circular financing. So, on one hand, it does do positive read-through on capital allocation, but on the other hand, it doesn't sort of sway away the risks that have been hamping over the stock multiple. There has been a lot of discussion about NVIDIA open agent safety platform. You talk about security, but the idea that NVIDIA is offering more than just GPUs, like in a very real sense now, it's saying you guys have a concern.
51:21Here's an engineering solution to it for a piece of open source software. What did you make of that? That was a more fundamental positive breakthrough. I mean, look, with all the talks about security, safety, with all the attacks happening and the talks about slowing down training progression. Remember, all of that will require more compute. Even if you slow down training, model training development to add more compliance, safety, security, you still need more compute. And guess what? Where are people going to turn for that compute? To the likes of NVIDIA. So that is definitely a fundamental long-term tailwind here for the company.
51:57Kunjan Sabani of Bloomberg Intelligence with a react to a lot of what happened with NVIDIA this morning. Thank you. sticking with NVIDIA, Lockheed Martin CEO Jim Teichlet says the chip company has been a partner of theirs for two decades now as AI keeps playing a growing role in Lockheed's roadmap. Listen to this. Our chief engineer, when I started the company in management six years ago, said AI is really critical. And of course, to ask for budget to invest in our own AI containerized instance where we would not be using data sets off the internet, we would not be using models and then feeding our information back into the system, where we would have basically a controlled environment for our AI workloads.
52:45And one of our partners, again, in the technology space has been Jensen Wong. Back in 2000, 2001, where hardly anyone maybe had heard of NVIDIA, if you weren't a gamer, we were teamed with NVIDIA to get GPUs, and that relationship continues today. So that was Lockheed Martin CEO, Jim Teichlet. Tune in on Friday for more of that conversation during Bloomberg's Defense Tech special, 11 a.m. Eastern, 8 a.m. Pacific. We're going to be live from Lockheed Martin Skunk Works. That's Lockheed Martin's advanced aircraft development facility in Southern California. A really rare, incredible access to what's happening in defense technology.
53:26From Defence Tech to space, let's get back to those live pictures of Starship reentering Earth's atmosphere. It did its deorbit burn, reached orbital altitude 275 kilometers. That mission goal achieved. It deployed 26 Starlink V3 satellites, commercial operational revenue generating for Starship. That was a first. but SpaceX decided to return Starship in its 14th test flight home early. The original mission design was to do six orbits of Earth over about a 10-hour period but they decided to come back. We still don't know why. We will find out. The plan is for the Starship spacecraft which you're seeing live on your screen to splash down west of the coast of Chile, South America, out of this world.
54:17That does it for this edition of Bloomberg Tech. Lots of conversations in the show, many market moves on the headlines, recap on the podcast. You know where to find it on the Bloomberg terminal as well as online on Apple, Spotify, and iHeart. What a way to start the week. This is Bloomberg Tech.
54:38At Edward Jones, we believe Rich is more than caring about the latest and greatest. It's also taking care of what gives your life meaning. That's why your dedicated financial advisor meets you where you are with personalized financial strategies that help protect what matters. So you can preserve your progress while creating a path forward. The key to being rich is knowing what counts. Let's find your rich together. Edward Jones, member SIPC. Healthcare doesn't always work great. If you've ever waited on a refill or couldn't schedule an appointment, you get it. That's the kind of stuff Optum is changing.
55:17They're using data and technology to integrate patient care, pharmacy, and everything else. So healthcare is connected, not complicated. What's that look like? Cheaper prescriptions that are easier to get and care that looks at the whole person. How you need it. Optum is helping make healthcare work as one for everyone. Learn more at business.optum.com. This dog salon? Operational excellence. Thanks to Genius from Global Payments. Scheduling? Personalized. Checkouts? Instant. Absolutely genius. Big League reliability for any business. That's genius. This is Robert Smith from Business History. If you're listening to this, there's a good chance you're a small business owner.
55:58And like every small business owner, you started with a dream to do what you love and watch it grow. What you probably didn't dream about? Keeping up with cyber threats. That's where MasterCard can help with access to tools that help identify cyber threats to better protect your business. Building your dream business? Priceless. For cybersecurity in a changing world, there's MasterCard. Learn more at MasterCard.com slash small business.
From the publisher
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.
