Meta CEO Weighs In On AI Safety Debate

21:18Ed Ludlow:we're talking about and we're showing space tugs, like for the completely uninitiated, those of us that haven't been to space, I appreciate that's most. What's a space tug? Like why is that a critical cog in the space ecosystem? Well, that image you have right now, that's Helios, which is that big rocket on a rocket that brought all this ride share up to geoseconders orbit. And then you can see the space tugs, the Mira spacecraft that are undocking from that and going to their orbit. So it's basically, it's a satellite like that right there that has power systems, carries payload. But mainly what makes it a space tug is it has a very powerful propulsion system on it.

22:08So a lot of propellant and a lot of thrusters. So it can move around very precisely in space. It can move around within an orbit or it can change orbits.

22:16Ed Ludlow:And so who then, Tom, are your customers? Who needs the space tug? For the Mira, mostly has been Space Force. We have three of those flying right now. In fact, we just did a flyby between our LEO-2 and our LEO-3 missions. They got within about 200 meters of each other. Those previous flights were mostly commercial and demo flights for us. And going forward, we have mostly government flights. although we do have another commercial one, Leo 4, going up later this year. And then we have, when Helios flies for the first time next year, we have our first GeoMira going to geosynchronous orbit. So it's a mix of both government and commercial.

22:59Ed Ludlow:Tom, when we say flies, which launch system are we talking about? SpaceX, right? And principally Falcon 9? Yeah, everything we book for the most part right now that's flying soon is on Falcon 9. we're also going to be flying on others in the future. So those others are interesting. There's been a lot of debate and reporting about how SpaceX, of which you're an alumni, transitions the Falcon 9 program to Starship. There are these reports that beyond 2029, SpaceX has stopped taking orders on Falcon 9. What's your experience of that, to be able to get reliable launch over that kind of longer time horizon?

23:43yeah um fortunately for us we're already booked on falconines out through that that date and then a lot of our customers are commercial i mean i'm sorry our government customers which are already booked through block buys so we're mostly covered for the next few years and then um i think this is a huge opportunity for all the you know the new uh other launch vehicles to step up we've got a few that have flown for the first time in the last couple years we got um several that are going to fly for the first time in the coming year. So there's a lot of demand and I think there's going to be a lot of supply.

24:17Ed Ludlow:You've also hired your first CFO. I don't know, were you doing the pencil pushing on the spreadsheets until this point, Tom? But how much, no, go ahead. Tell me what's the thinking about behind this, like precursor to IPO? I've been mostly functioning and the CTO, you know, with the share of my CEO. And Eric Romo, Mr. Romo, has been doing, you know, most of the helping with the financial stuff. And, you know, now that we've got so much customers and really so much capital to manage, we just, it was time for a C-level finance officer. So we were very lucky to find Adam, the guy's remarkable, a perfect fit for us.

24:59So yeah, it really worked out, a great time.

25:02Ed Ludlow:just from a pure founder perspective you know is impulse being a public company one day an ambition for you tom uh we'll see uh yeah it could happen it's uh not something we're really focused on right now right now we're just addressing the market and trying to build up our uh you know our capabilities can we just talk to end the conversation about building um you and i've gone over this in the past this idea that you are part of an ecosystem in southern California where there is a concentration of talent there is the ripple effect of SpaceX coming out of Hawthorne just talk about that environment right now and and let's end on on how it's going yeah there's you know a lot of founders of new companies like mine that uh that that were at SpaceX or have a lot of key employees uh that were at SpaceX that have joined and it's just I think really if you think about Silicon Valley is doing uh bits you know doing software I think about LA is doing hardware.

25:59So we're making parts here. We're making spacecraft. We're making drones. We're making nuclear reactors. There's all kinds of things going on within just a few miles of where I'm at here in Rudano Beach.

26:12Ed Ludlow:Tom Mula of Impulse Space, back on Bloomberg Tech, extending the Series D to$808 million. Thank you very much indeed.

26:26Ed Ludlow:Welcome back to Bloomberg Tech. The AI safety debate is heating up in Washington. Anthropic met with senior Trump officials yesterday as leading AI companies discuss working together on safeguards. Smaller rivals warn new rules could make it harder to compete. And gosh, spare us thought for the venture capitalists. They must all be there a little bit worried right now. Bloomberg's Natasha Mascheranis and Maggie Eastland join us now. So, Natasha, I'm going to go to you first. You have done what you always do, which is go out to industry and say, what do you guys think about what's going on here?

26:57Ed Ludlow:Right. If you think about Anthropic and OpenAI in private markets, let's say Meta in the public markets, there are lots of investors that are bracing for what this net results in, right? Absolutely. I mean, I spoke to over a dozen founders in the first 48 hours since Dario posted his call for AI pacing. And I would say the headline is we are worried about a regulatory wall emerging here. We are worried that the bigger companies are going to set rules that us smaller rivals are going to have to follow. So there is a lot of fear, I would say, on the investor side, a little bit more optimistic of, as per usual, of, you know, let's keep moving.

27:34Everything is fine. No chilling effect here. You know, maybe keep your fundraising dollars a little closer to heart in case you need to work on your own models instead of relying on the other one.

27:43Ed Ludlow:OK, the U.S. government's really important here. Maggie Eastland out in Washington, D.C., bring us the reporting. So we know there's been some contact between Anthropic and U.S. officials at least. yes and we know that going into that meeting Howard Lutnick, Commerce Secretary as well as Emil Michael were both eager to push back on Anthropic about this idea all of this doom that has been spreading and has kind of burst onto the public scene you saw President Trump post really just a barrage of comments on his true social platform talking about how he did not want to slow down AI and we can definitely bet that that's the message that Anthropic's Tom Brown received on that call that we reported on yesterday.

28:27Ed Ludlow:Natasha, you know, timing in the world of Silicon Valley and venture capital is everything. There's the report out about OpenAI considering a private market round, pre-IPO. You know, is there anything in the reporting that suggests a slowdown or a pacing at the frontier results in reduced activity across the ecosystem, in both investment and sort of the activity of founders? I think that will be answered by how successful OpenAI is in getting this valuation and in getting an actual round done. I mean, our sources tell us that this is something that investors have approached the company with. And so we don't know exactly where it will land, but I do think it will definitely be a sign of where energy and activity is.

29:12Ed Ludlow:Maggie, yesterday you and I participated in a Bloomberg Live Q &A. on what's happening broadly. And there is this idea, let's go back to the David Sachs interview 36 hours ago, that actually there are enough pieces of regulation and laws in place that no more government interventions needed. Kind of summarize that camp in this debate. Yeah, so there definitely is one side of the story here that says, look, these AI companies are liable for the products they make. And if they make dangerous products, they're going to be held responsible for that. That one interesting twist in this story is that I also reported yesterday that OpenAI said that it's working with Anthropic and with Google DeepMind already to set safety standards.

29:58And they said they actually don't see any need for the government to get involved or to issue any kind of antitrust waiver for the companies to be able to do that.

30:06Ed Ludlow:Maggie, let's actually get some more specifics on that. So this was Chris Lehane doing a briefing, essentially. Who is Chris Lehane? and like literally in concrete terms, what has OpenAI agreed to do? Yeah, Chris Lehane leads global policy for OpenAI and he spoke to reporters yesterday in Washington. And specifically what OpenAI is saying is they've agreed to look into setting safety standards. Now, this is along the terms of what, you know, Demis Hasevis, Google's DeepMinds leader, laid out in a proposal that compared this to something like FINRA for the financial industry. So this would be the company is getting together and deciding for themselves what the standards should be and potentially even grading each other's models or having independent evaluators do that.

30:54But right now, it seems that it's still early discussions about this among those three companies.

30:59Ed Ludlow:Natasha, there's another party in your reporting, more from Maggie's World, which is the U.S. Federal Trade Commission chair coming in. What did that part of this debate have to say? I mean, that is the thing that a lot of startups are still trying to figure out, if I'm being honest. Like, how seriously do we take this perspective when trying to, you know, decide what the next even week, month, year looks like in terms of startup spend and how much to kind of change the way we're pacing our own models? They are talking about the antitrust exemption, which, you know, Maggie, come back in here real quick.

31:37Ed Ludlow:on the antitrust exemption. So this was something that Dario Amadei had mentioned in his Saturday essay, right? The antitrust exemption. And that is a particular point that David Sachs in his response was like, absolutely not. Do not ask for that. But could you try and explain what an antitrust exemption would look like? Why it would be that Anthropic or others would ask for that? I think the details of how exactly that would look aren't clear. But obviously, anytime you know, companies are coming together and trading information, there is a risk there. OpenAI and Anthropics seem to differ on this issue, and OpenAI thinks, you know, there's precedent for this, there's things like the airline industry.

32:15There's some times when sharing safety information is not a competition risk. But in terms of a waiver, we actually saw yesterday as well, FTC chair saying, you know, he'd be deeply suspicious of that. So it seems like the U.S. government is unlikely to issue that, and now OpenAI has put forward this plan that they say they can move ahead without any kind of antitrust exemption.

32:34Ed Ludlow:Bloomberg's Natasha Mascarenas here in SF and Bloomberg's Maggie Eastland in Washington, D.C. Thank you very much. Let's go to robotics. Agility Robotics has a new humanoid robot, and this one is designed to work right alongside people. It's called Digit5 and is meant to take on jobs like loading materials or inspecting items. And with more than 300 million dollars in multi-year orders, Agility is now preparing to go public through a SPAC. CEO Peggy Johnson explains why. We looked at all of our options available to us and, you know, whether that was staying private and continuing to fundraise or going versus a standard IPO or choosing the route.

33:17We're going to through the SPAC, which is the most efficient, quickest way to market. And we have orders pending. We have over$300 million of orders for multi-year commitments for Digit 5. And we wanted to ensure we were able to deliver on those as well as continue to work with the number of customers in our pipeline. So it was really about acceleration, speed to market. But it also does give us the ability to define the category. Being first out the door, the first pure play humanoid to go public, we can show other companies what will be needed. The high safety bar, reliability, the ability to plug into existing IT systems in these facilities.

34:02The enterprise hardening of the robots is where we'll get to set some of the standards going forward.

34:08Ed Ludlow:Well, the first pure play humanoid robotics public company may be in the United States. Elon Musk might have something to say about that with Tesla and Optimus. But how closely did you watch what happened in China and in Asian capital markets with Unitary, for example? You know, it's on a regular basis. We're kind of comparing and contrasting the progress on humanoids here in the States versus China in terms of like the scale of output, where they're deployed. What do you take from that? Well, what we take is we want to be able to deliver what our customers need. We haven't seen this, to tell you the truth, across the industry.

34:48We have a robot that can lift 50 pounds over and over again. We have hands that can be replaced with different kinds of tools, swappable end effectors, and a battery that allows for three shifts of work with one robot. It's got a 10 to 1 run to charge ratio, which means it can run for 90 minutes and very quickly in nine minutes have the battery recharged. So it was listening to our customers, understanding the needs and deploying that functionality in digit five, which allows us to get paid to work. We really haven't seen that deployed anywhere else yet.

35:28Ed Ludlow:What's agility going to focus on for the next one to two years? You'll go public via the SPAC, but in terms of the technology scaling, what's the big priority for you, Peggy? The big priority is with AI, continuing to teach Digit 5 new skills and listening to our customers, what they need, how we can fill in any gaps in their workflow in these facilities. So with AI now, teaching Digit new skills can go much more quickly than in the past when an engineer had to program every movement of the robot. So you'll see a lot more functionality, capability from Digit 5 in the next few years to come. That was Agility CEO Peggy Johnson.

36:16Ed Ludlow:Coming up, we're going to stick with physical AI. Among other early-stage investing opportunities, Slater Stitch of Bain Capital joins us to talk about the firm's latest fund target, fund targeting early-stage AI, Slater Stitch of Bain Capital Ventures. That's next. This is Bloomberg Tech.

36:39The 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.

37:19Go 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. Some people treat ChatGPT like some kind of smart search engine, and some use it to get work done. ChatGPT Work is a new way of working in ChatGPT that can take action across your apps and files, stay with a project for hours if needed, and turn a goal into finished work.

38:02It's designed to help you move from a chaotic starting point to a reviewable first version. So all the source materials, briefs, and scattered information that you have to grind through to turn into something useful can just become something useful. Put ChatGPT to work on your most ambitious ideas and projects. Get started at ChatGPT.com by selecting Work Mode, available on Plus and Pro plans. Let's talk about health care 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.

38:38The system should just work better for everyone. That's exactly what the people at Optum are trying to do every day. They're a health care 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. Technology designed to help doctors spend less time on busy work and more time with their patients.

39:12And 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.

39:33Ed Ludlow:Bain Capital Ventures has raised a$1.6 billion fund to back early-stage AI companies for building on a playbook it established with its previous Fund 10. BCV partners Slater Stitch is with us, and this fund is for life after AGI. Explain. That's right. Well, first of all, Ed, great to see you. So good to be here. Thanks for having me on the show. You're right. So one of the big ideas that we have for this fund is that we have stored up this incredible ability with artificial intelligence, and we've seen some applications for it. In some cases, there may be the more obvious applications. But as the technology gets even better, we should expect an explosion of new applications.

40:17and we wanted to approach this fund with the philosophy of really thinking about that hard from first principles. What does it mean like when we go through a technological revolution in intelligence just as profound as electrification or mechanization earlier? What would it mean to sort of have a post-AGI economy?

40:36Ed Ludlow:There are three different sort of, well, four probably flavors of this. Infrastructure, physical AI, which can mean robotics, security, which given in the events of the last seven days is massively important. And then also services of those four to your mind, which is most present in the post AGI life that you're envisaging? I think that all of those will have important roles to play. And it's just a question of sequencing, you know, sort of in what order will we see these different things? Do you see an order? I think that, you know, things that seem sort of like knowledge work on a computer flavored are going to be the form factor where it's easiest to make progress fastest.

41:15We sort of know how we would ascend up the scaling curves there. The models, the base models get more intelligent each generation. The harnesses get better. The sort of task horizon for agents gets better. The tool usage. It's just easier to do things in a purely digital domain. So that's where we see it happening first. But of course, we're incredibly optimistic about applications and robotics and other physical world applications over time.

41:40Ed Ludlow:You guys talk about life after AGI. The difficulty is AGI is not universally defined. Some would say we're at AGI. I'm thinking more recently about OpenAI's release of Astra and Greg Brockman is an example saying, this is the first step to true AGI. How are you defining it and measuring it? Well, you know, one thing I think about is if I just gave my past self information, would I have called this AGI? If I went back to how I thought in 2016 or something like that. And by that measure, absolutely. I mean, I think if you had told me that we would be solving millennium problems or being able to run these incredibly useful long running agents, that would have surprised me at the time.

42:28So I would say that from an economic and also even from a scientific perspective, a lot of the promise we're already seeing there. I do think AGI is one of these terms that can never be fully defined. No matter how good the systems are, there will always be sort of some next thing. But I personally am most interested in the ways it will affect the economy, and I feel like the technology is already there to do so much.

42:54Ed Ludlow:This is an early-stage fund, right? So it's$1.6 billion to invest at that early stage for companies where you feel they have an opportunity in that future economy. So if we take physical AI, robotics, for example, where would you apply that to the economy today? Where are the gaps in this U.S. economy where you think, okay, this is ripe for robotics to be the solution? Yeah, I think that the nature of, you know, again, I think lots of things will happen in the long run. In the short run, application areas with predictable environments, with sort of relatively set tasks are the ones that will, of course, be easiest.

43:31But in some ways, in the long run, that might be the least important. So, for example, we're investors in a company called Sunday Robotics that's working on home robots, sort of fully integrated to that use case. And I do think that when you take an application like that really seriously, you can make progress very quickly. You know, I do think that there's also this question of, you know, the industrial applications will probably have some of the biggest impacts. We have, you know, sort of good industrial robots for relatively predictable manufacturing tasks. For manufacturing tasks that are less predictable, that's where intelligence can actually play a role.

44:14Can we talk a little bit about investing?

44:15Ed Ludlow:So this is a specific fund. It's Fund 11, technically. it comes after fund 10. What's realistic in a calendar year? Do you go after high volume or do you just say, okay, like this is high conviction in this environment. We have a thesis for this fund. So we'll do a handful of investments and kind of cap it. How does that work? We don't think about it in that way. Like the way that we think about it is that we know what it looks like when we see a incredible founder pursuing an outlier opportunity that we believe in. You know, we might get the conviction on that opportunity in different ways. In some cases, it could be thesis-driven, like you're describing.

44:52I would say that our investment in Fleet, for example, the Arl Environments Company, happened that way. In other cases, it's really through conversations with the founder and learning the opportunity that we get to conviction. But I don't think about it in terms of capital deployment over some period of time or some sort of strict rules about how we would deploy it.

45:13Ed Ludlow:Take us inside Bain Capital Ventures, how it's working as a firm at the moment. So if you have an opportunity or an idea, you were the sponsor of it, how did the partners kind of get around it and decide, like, this is the bet we want to make, particularly when you're saying, okay, we want to invest in an opportunity for life after AGI? Yeah, I think it's, so one way that I think about this is opportunities that we think could be tremendously valuable and that we understand the technological path to achieving them. So for example, in companies like Decagon or Cognition, we understand the promise of what those companies are in automating software engineering or customer support.

45:59And it's sort of clear that the plan, if the founders are able to do the things that they want to do, that we should be able to achieve that. There's other cases, a company like Periodic Labs, for example, which is building an AI scientist initially focused on material science applications where we spend sort of more time on the technology and sort of why we think the founders are going to be uniquely able to achieve that.

46:23Ed Ludlow:I'd like to end the conversation talking a little bit about cognition. You know, they've just closed around. But the thing that's changed for me in the last calendar year or so is the ads in San Francisco. I'm paraphrasing, but it says Devin is actually good now. Yeah. Which is sort of an admission that it's taken time. But that's amazing. I mean, you see billboards over SF all the time, but just in Cognition's case, what you saw in the development arc. Totally. So I really love that campaign. And what I like about it is, to me, it's Cognition calling their shot really early on. So, you know, at the time that they sort of put a flag out and said, we're going to be the autonomous software agent that works, you know, I would say that the base models were not quite there yet.

47:06At that time. Well, that was an easy thing to do. But they understood exactly where things were going and sort of understood that that was a valuable place to be in the long run. And so I think it was like a courageous thing to do. And I think they called it perfectly.

47:20Ed Ludlow:Is there a next cognition then with this fund that you think you can find? We do. I think that there will be other applications that look like this. And then I also think that as you just sort of go up levels of intelligence automation, there will be all sorts of things that we can't even anticipate yet. Slay the Stitch, partner at Bain Capital Ventures. Really great to have you here on Bloomberg Tech.

47:46Ed Ludlow:There's a lot going on in the world. Let's get out to New York where Bloomberg's Yahaira Anand is standing by. Hi, Yahaira. Hey, Ed. It's time now for Talking Tech. First up, as you said earlier, Europe is joining the AI debt boom. Between$5 and$10 billion of data center bonds could reach European markets by year end. That's according to Goldman Sachs, with a bigger wave expected in 2027. This comes as the continent is pushing to reduce reliance to overseas technology. Plus, over in South Korea, its deputy prime minister says the country must continue advancing AI without pause or delay, saying the nation can't afford a slowdown like the one proposed by anthropic leader Dario Amude.

48:29The country has, of course, been one of the biggest beneficiaries of the AI buildout. And ByteDance co-founder Zhang Yiming is now Asia's richest person, overtaking Indian billionaire Gautam Adani. The Bloomberg Billionaires Index puts Zhang's fortune at more than$105 billion as ByteDance pushes deeper into AI and moves past that regulatory fight that nearly forced TikTok out of the U.S. Ed.

48:57Ed Ludlow:Thank you, Yahaira. Real quick, go back to Intel and the U.S. listed shares of SK Hynix, both higher. There was a report that the two are looking at a tie up in the United States over manufacturing. Intel now markedly higher up more than 5%. SK's ADR's 2.5%. There is another chip story out there, Apple. In the moment, there was some share reaction to a report from the information that Apple is working on an enterprise server that would use its own chips and has held discussions with NVIDIA about using its networking equipment. We've done a lot in the program of late about MVLink Fusion, etc. The stock is modestly high now, 4 tenths of a percent.

49:38Ed Ludlow:But in the moment, there was a move on that story, which Bloomberg has not verified. That does it for this edition of Bloomberg Tech. What an edition it was as well. Like a lot of you come to the show through the podcast. So thank you very much for that. You know where to find it. But recap on the terminal as well as online on Apple, Spotify and iHeart. Halfway through the week, this is Bloomberg Tech.

50:26Thank you. make healthcare work as one for everyone. Learn more at business.optum.com. Before you sign off, you tuned in for ways to help teams move faster, make sharper decisions, and turn scattered contexts into work they can use. ChatGPT for Business can help.

50:48Ed Ludlow:ChatGPT for Business gives teams a shared workspace with admin controls, permissions, and access to work and codecs in ChatGPT. This means your business can move from question to answer and code to rollout quicker. Join over 10 million business and enterprise users worldwide already using ChatGPT for work. Download the ChatGPT desktop app or contact sales to learn more. You already know how AI is changing how everyday work gets done, how much ground you can cover, and how fast a team can scale. To stay ahead, you need the tools that give you a competitive advantage built for this new era. Welcome to Agentic Revenue.

51:24Adio is the CRM for this world. It meets you where you work, compounds every customer signal into context, then acts on it across your pipeline to let you move it on match speed and scale. With agents and automations for every job in revenue, Adio orchestrates your work around the clock. Built to handle the scale of your workloads, extensible with API and MCP, and with the infrastructure to keep up with your most ambitious agents. Loved by high-growth startups like Granola, Modal, and Etched, Adio runs the work behind every win. That's Adio, the agentic CRM, the intelligent system that never sleeps.

51:58Picks up leads at 2 a.m., catches renewals before they slip, hands you the answer before you ask. Try Adio free at adio.com slash iHeart. That's adio.com slash iHeart.