# Rocket Lab Targets SpaceX's Starlink Dominance in New Deal

An episode of [Bloomberg Tech](https://invisiblenetwork.ai/podcast/1611767434), 29 Jun 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/276bb033-f0c0-4251-b75d-b478011e3565/audio.mp3?utm_source=Podcast&in_playlist=9caafc41-289c-4115-995d-ae7c0030ab75

Bloomberg’s Ed Ludlow speaks with the CEO of Rocket Lab about the company's biggest bet yet: buying Iridium in an $8 billion deal to challenge SpaceX in the orbital economy. Plus, South Korean firms, including Samsung and SK Hynix will spend at least $880 billion on chips and data centers, as the country seeks to maintain its edge in the AI era; and Anthropic wins US approval to restore some access to its Mythos 5 AI model after resolving concerns about the technology's potential threats to national security. See omnystudio.com/listener for privacy information.

## Transcript

[0:02] Speaker 1: Bloomberg Audio Studios, podcasts, radio news. Bloomberg Tech is live
[0:15] Speaker 1: from the heart of Silicon Valley with ed La though
[0:18] Speaker 1: in San Francisco.
[0:22] Speaker 2: This is Bloomberg Tech coming up.
[0:24] Speaker 3: Rocket Lab makes its biggest bet yet, buying a Ridium
[0:27] Speaker 3: in an eight billion dollar deal to challenge SpaceX in
[0:30] Speaker 3: the orbital economy.
[0:31] Speaker 2: We'll speak with CEO Peter Beck.
[0:34] Speaker 3: Plus South Korean firms including Samsung and sk Heinix, will
[0:37] Speaker 3: spend at least eight hundred and eighty billion on chips
[0:41] Speaker 3: and data centers as the country seeks to maintain its
[0:44] Speaker 3: edge in the AI era and anthropic wins US approval
[0:48] Speaker 3: to restore some access to its Meethos five AI model
[0:52] Speaker 3: after resolving concerns about the technology's potential threats to national security.
[0:58] Speaker 3: M and A Monday is blast off into orbit here
[1:01] Speaker 3: on Bloomberg Tech. Rocket Lab making its biggest acquisition ever,
[1:04] Speaker 3: agreeing spy Iridium in an eight billion dollar deal. Is
[1:08] Speaker 3: it looks to challenge SpaceX in the rapidly growing low
[1:11] Speaker 3: Earth orbit economy.
[1:12] Speaker 2: This is how the shares are reacting.
[1:14] Speaker 3: The deal terms fifty four dollars a share in a
[1:16] Speaker 3: cash and stock transaction for the satellite telephone pioneer, valuing
[1:21] Speaker 3: a ridium at an enterprise value of about eight billion.
[1:24] Speaker 2: The offer fifty four dollars.
[1:26] Speaker 3: That's about a twenty seven percent premium. Friday's close, we
[1:29] Speaker 3: just showed a Ridium shares just a little shy of
[1:31] Speaker 3: fifty four.
[1:32] Speaker 2: But why joining us now?
[1:34] Speaker 3: Is Peter Beck, Rocket Labs, founder and CEO, live from Auckland,
[1:38] Speaker 3: New Zealand.
[1:38] Speaker 2: Let's start there. Peter, why buy iridium and why now?
[1:43] Speaker 4: Yeah, it's great to talk to you, but look, I
[1:46] Speaker 4: think it's become pretty obvious, and we've talked about it
[1:49] Speaker 4: for quite some time that the really large space companies
[1:52] Speaker 4: of the future are going to be a little bit
[1:54] Speaker 4: blurry about what a space company, even as in rocket
[1:58] Speaker 4: Lab today with a sick and most frequently lawn rocket
[2:01] Speaker 4: and we have a large spacecraft business and component's business,
[2:04] Speaker 4: but really the third letter leg of the store was
[2:06] Speaker 4: always an application. And by combining rocket Lab and Iridium
[2:10] Speaker 4: together we really make that complete picture. And what that
[2:14] Speaker 4: essentially means is we are a self launching company. And
[2:17] Speaker 4: I think if you look at other companies that have
[2:19] Speaker 4: their own rockets and can build their own satellites, it's
[2:22] Speaker 4: a pretty powerful combination.
[2:24] Speaker 3: I think, you know, very early this morning, people got
[2:26] Speaker 3: very quickly to this deeper level of vertical integration.
[2:30] Speaker 2: The basics of it. There's also the.
[2:32] Speaker 3: Idea of spectrum, right, so ridium has valuable but a
[2:36] Speaker 3: limited source of spectrum. Can you just quantify that for us,
[2:41] Speaker 3: how much spectrum we get access to and kind of
[2:44] Speaker 3: like how you leverage that to rocket labs advantage.
[2:49] Speaker 2: Yeah, look, you're exactly right.
[2:50] Speaker 4: I mean, you can have all the spacecraft and all
[2:52] Speaker 4: the rockets and the hangars you want, but if you
[2:54] Speaker 4: don't have the spectrum to actually utilize them, especially for communications.
[2:59] Speaker 2: Then it's all for nothing.
[3:00] Speaker 4: And the very unique thing about a ridium spectrum it
[3:03] Speaker 4: is it's our band spectrum. So you know, our band
[3:06] Speaker 4: is particularly effective in those very difficult and harsh conditions.
[3:11] Speaker 4: So not surprisingly, you know, Iridium has built a very
[3:14] Speaker 4: very strong business for you know, service servicing, safety critical
[3:18] Speaker 4: things and the defense uh you know, the defense network,
[3:24] Speaker 4: and so you know this this obviously enables us to
[3:27] Speaker 4: build on this. You know, it's a it's a very
[3:29] Speaker 4: established and profitable business to date, but you know, the
[3:33] Speaker 4: spectrum is an incredibly valuable asset for sure.
[3:38] Speaker 3: Peter, going forward, does rocket Lab just more generally become
[3:42] Speaker 3: a buyer of spectrum and an acquirer of assets that.
[3:45] Speaker 2: Give you access to more spectrum. No, no, not at all.
[3:48] Speaker 2: I think I mean this.
[3:49] Speaker 4: The spectrum gives us a fantastic you know, baseline and
[3:54] Speaker 4: headstart if you think about the opposite. You know, we've
[3:57] Speaker 4: always said we want to do an application that would
[3:59] Speaker 4: take years and years and years to first build the
[4:02] Speaker 4: constellations of satellites, deploy them, and then then of course
[4:06] Speaker 4: you know that's before you even see.
[4:07] Speaker 2: Your first dollar of revenue.
[4:09] Speaker 4: So this really supercharges, you know, our kind of interest
[4:13] Speaker 4: into applications, and you know, it's it's a wonderful business
[4:16] Speaker 4: to build upon. But our intention here is to to
[4:19] Speaker 4: you know, leverage the spectrum and what already are built
[4:21] Speaker 4: to provide a whole bunch of new services.
[4:25] Speaker 3: Okay, so we have launch, we have satellite manufacturing, and
[4:28] Speaker 3: now communications. One of the questions I got from the
[4:31] Speaker 3: audience is for you, Peter, is how do you take
[4:34] Speaker 3: Iridium's technology and improve it? You know, they themselves have
[4:38] Speaker 3: been very focused on chips right in their own custom electronics.
[4:43] Speaker 3: What can you do to kind of make best use
[4:45] Speaker 3: of it? But also make them better.
[4:47] Speaker 2: Yeah.
[4:47] Speaker 4: Sure, So, I mean I think, you know, just the
[4:50] Speaker 4: economics alone and that the synergies alone, when you have
[4:54] Speaker 4: you know, control over the most expensive and the most
[4:57] Speaker 4: longest lead time items being launched and spacecraft manufacture, I
[5:01] Speaker 4: think that just puts you know, that that level of evento
[5:03] Speaker 4: vertical integration just puts everybody into a totally different league. So,
[5:08] Speaker 4: you know, instead of scratching out business plans on a
[5:10] Speaker 4: whiteboard where you have to spend billions of dollars on
[5:12] Speaker 4: launch in spacecraft manufacturing, when all of a sudden that
[5:15] Speaker 4: those costs evaporate, you know, your ability to innovate and
[5:20] Speaker 4: execute different business plans becomes, you know, vastly superior.
[5:25] Speaker 3: Let's talk about the economics real quick. Is debt going
[5:28] Speaker 3: to play a role in financing this?
[5:31] Speaker 2: Yeah? Initially for sure.
[5:32] Speaker 4: You know, rocket Lab is is a very strong balance sheet,
[5:36] Speaker 4: and you know, Iridium historically has had had some debt,
[5:40] Speaker 4: but of course you know it's throwing off a lot
[5:41] Speaker 4: of cash to service it. But you know, you know,
[5:45] Speaker 4: we we have a you know, a debt bridge that
[5:47] Speaker 4: we intend to take out.
[5:50] Speaker 3: I get a lot of questions for you about Neutron
[5:53] Speaker 3: next gen launch system that you're working on. I think
[5:56] Speaker 3: one reason we get a lot of questions from the
[5:59] Speaker 3: audience now about Neutron is is that people want you
[6:02] Speaker 3: to kind of explain how dependent this this future business
[6:06] Speaker 3: line is on Neutron constellation based connectivity.
[6:11] Speaker 4: Look, I mean the Achilles Hill in space applications and
[6:15] Speaker 4: space infrastructure.
[6:16] Speaker 2: Right now is launch.
[6:18] Speaker 4: You know, launch is an incredibly valuable and rare asset.
[6:21] Speaker 4: And you know, as I mentioned before, where we launched
[6:23] Speaker 4: the Electron rocket, which is the second most frequently launched
[6:26] Speaker 4: rocket in the world, and Neutrons coming online by the
[6:29] Speaker 4: end of the year, and you know, luckily also, you
[6:33] Speaker 4: know they're ready in constellation. It's a relatively fresh constellation,
[6:36] Speaker 4: so it doesn't require immediate launch. There's a number of
[6:39] Speaker 4: years left in that constellationally over a decade, so we've
[6:43] Speaker 4: got a little bit of time to refresh the constellation.
[6:45] Speaker 4: But I mean, you know, we can throw stuff up
[6:48] Speaker 4: with either Electron or Neutron to experiment and develop new
[6:52] Speaker 4: technologies almost immediately.
[6:55] Speaker 3: Okay, the big question is rocket Lab developing the sort
[6:59] Speaker 3: of make a scale constellation specifically to take on SpaceX
[7:05] Speaker 3: Amazon as T not just direct to sell satellite based
[7:10] Speaker 3: Internet connectivity, et cetera.
[7:12] Speaker 4: Well, certainly you know Iridium has a director device system
[7:18] Speaker 4: and planning that they're rolling out later on this year.
[7:21] Speaker 4: And look, this is the very beginning for us. This
[7:24] Speaker 4: is our first foray into applications. We've talked about it
[7:27] Speaker 4: for a long time, and I think this is a
[7:29] Speaker 4: very smart deal where you know, we're not We're not
[7:33] Speaker 4: you know, acquiring a field of dreams and starting from scratch.
[7:37] Speaker 4: We are acquiring a very valuable spectrum, a profitable business.
[7:42] Speaker 4: And you know it's a very typical smart rocket Lab deal.
[7:44] Speaker 4: But you know, can our clear intention here is to
[7:48] Speaker 4: not stop we're at this acquisition and continue to grow.
[7:51] Speaker 2: For sure, a lot of folks.
[7:53] Speaker 3: This morning kind of pointed the idea that you know,
[7:55] Speaker 3: Starlink has has a lead in terms of bandwidth speed
[7:59] Speaker 3: late in see. So I think people would love to
[8:02] Speaker 3: just hear you be very granular and specific about the
[8:05] Speaker 3: niches that you can go after. What you see, it's
[8:08] Speaker 3: the total addressable market, but also like definition of that
[8:11] Speaker 3: total adjustable market that rocket Lab is like zeroed in
[8:15] Speaker 3: on as opposed to like going after all of it.
[8:18] Speaker 4: Yeah, look, I mean to your point before you know,
[8:21] Speaker 4: spectrum is key, and this spectrum, the alband spectrum, is
[8:24] Speaker 4: very very unique in the fact that it's ineffected by
[8:27] Speaker 4: weather and harsh conditions, and as a result, you know,
[8:29] Speaker 4: this alband spectrum is widely used in all the safety
[8:33] Speaker 4: critical stuff you would imagine, whether it be mariners, pilot's
[8:37] Speaker 4: defense forces. I mean, I'm my perhaps self personally, you know,
[8:41] Speaker 4: a user of the Iridium network. I'm a helicopter pilot.
[8:44] Speaker 4: When I get like ten minutes of my own time,
[8:47] Speaker 4: and you know, there's a little little thing on the
[8:49] Speaker 4: top of my desh with an Iridium logo that I
[8:51] Speaker 4: push the button that I know if I have a
[8:52] Speaker 4: super bad day and end up in the bushes, someone's
[8:55] Speaker 4: going to come and get me for sure.
[8:57] Speaker 2: So so you know.
[8:59] Speaker 4: Our focus and then you know, initially is on maintaining
[9:04] Speaker 4: you know, that super important constellation. I mean it is
[9:07] Speaker 4: completely global, so no matter where you are on the
[9:10] Speaker 4: planet ac land, where you are connected, and that's something
[9:16] Speaker 4: that really only Iridium has.
[9:20] Speaker 3: Two final quick questions. I guess that given the landscape,
[9:24] Speaker 3: the regulatory risk on this, you guys feel confident that
[9:27] Speaker 3: this deal will close and that all get done.
[9:30] Speaker 4: Yeah, No, I think I think from a regultary standpoint.
[9:32] Speaker 4: This is pretty straightforward. I mean it's it's you know,
[9:36] Speaker 4: a very very strong deal and you know, not controversial
[9:40] Speaker 4: at all.
[9:42] Speaker 3: And then just for twenty twenty six, the balance of
[9:44] Speaker 3: this year next year, has rocket Lab got a refreshed
[9:48] Speaker 3: launch target for what we could see on Electron and
[9:51] Speaker 3: then Neutron in aggregate.
[9:54] Speaker 4: Yeah, look, you know, certainly Electron is launching a crazy
[9:58] Speaker 4: flat out. You know, we just broke the world rec
[10:00] Speaker 4: called for the fastest time that someone calls up a
[10:02] Speaker 4: rocket and launchers, so we did dead and sixteen hours
[10:05] Speaker 4: forty two minutes for the Space Force, and Neutron comes
[10:08] Speaker 4: online by the end of the year.
[10:11] Speaker 3: Rocket Lub CEO Peter Beck live from Auckland, New Zealand,
[10:14] Speaker 3: where it's the middle of the night, back on Bloomberg Tech.
[10:17] Speaker 2: Thank you very much.
[10:18] Speaker 3: Sticking with M and A, I want to take a
[10:20] Speaker 3: look at shares of Comcast now up seven percent, kind
[10:23] Speaker 3: of off session highs for the company unveiled plans to
[10:26] Speaker 3: split itself in two, spinning off NBC Universal and Sky
[10:31] Speaker 3: into a standalone public company, while leaving Comcast focused on
[10:36] Speaker 3: its broadband, wireless and cable businesses. The company says the
[10:39] Speaker 3: separations designed to sharpen each business is strategic focus and
[10:44] Speaker 3: create more value for shareholders who remain shareholders in both.
[10:47] Speaker 3: It's about a year away from that vision being realized.
[10:51] Speaker 2: Okay, still ahead.
[10:52] Speaker 3: AI spending is booming, but will those investments deliver? We
[10:57] Speaker 3: are seven am Shanty Kellerman that question. This is Glomberg Tech.
[11:15] Speaker 5: South Korea has laid out an ambitious plan to cement
[11:19] Speaker 5: its status as a global tech powerhouse. Among the mega
[11:23] Speaker 5: investment pledges, Samsung Electronics and eske Heinix, the world's two
[11:27] Speaker 5: biggest memory makers, are set to build four chip fabs
[11:31] Speaker 5: in a roughly five hundred and eighteen billion dollar project
[11:34] Speaker 5: in the country's southwest. Now, Korea aims to double its
[11:37] Speaker 5: DRAM production capability in the next five years. They're also
[11:41] Speaker 5: built chip packaging facilities in the south central region, and
[11:46] Speaker 5: Korea plans to spend about six hundred and fifty billion
[11:49] Speaker 5: dollars in AI data centers by twenty thirty five. Now,
[11:54] Speaker 5: what stands out from the investment pledsure so far is
[11:57] Speaker 5: how different regions will focus on specific technologies and industries,
[12:03] Speaker 5: which underscores the need for more balanced regional development. At
[12:07] Speaker 5: the time where there are concerns that the benefits from
[12:10] Speaker 5: the AI boom are not being spread evenly across the country.
[12:15] Speaker 5: President Niche Jong said Korea needs to accelerate building its
[12:18] Speaker 5: chip base to meet demand and that Career has to
[12:21] Speaker 5: move faster than its rivals. You can almost hear the
[12:24] Speaker 5: sense of urgency for Korea to maintain its competitive edge,
[12:28] Speaker 5: and of course this comes a mad questions of sustainability
[12:32] Speaker 5: of profits as competition heats up. All in the numbers
[12:36] Speaker 5: are more or less in line with what Eskehiinix and
[12:38] Speaker 5: Samsongi Electronics have already committed to spend for twenty twenty seven,
[12:44] Speaker 5: and underscores how CAREA is looking to capitalize on the
[12:47] Speaker 5: AI boom with chips and AI front and center as
[12:51] Speaker 5: part of Korea's long term industrial strategy. Abral Hon Bloomberg
[12:56] Speaker 5: News Singapore.
[12:59] Speaker 3: Bergs Abrahan reporting that brings us to the bigger question
[13:02] Speaker 3: for global investors, can the current AI trade last joining
[13:05] Speaker 3: us now? As Shanty Kellerman, co chief investment officer at
[13:08] Speaker 3: seven i AM, who says markets have rewarded the beneficiaries
[13:11] Speaker 3: of AI spending while penalizing the company's funding. It that
[13:15] Speaker 3: dynamic isn't sustainable if those investments fail to generate real value.
[13:20] Speaker 3: Senty welcome back to the program. I think in recent
[13:23] Speaker 3: days and weeks, we've been talking about this idea that
[13:25] Speaker 3: there is a distinction between capital expenditure deployers and capital
[13:30] Speaker 3: expenditure recipients, and that seems to be very top of
[13:33] Speaker 3: mind for you.
[13:35] Speaker 2: South Korea is a case study.
[13:36] Speaker 3: But why do you think that that psychological approach to
[13:39] Speaker 3: the trade is not sustainable?
[13:42] Speaker 6: Well, I think in the long run, the investments will
[13:44] Speaker 6: stop at some point if they aren't delivering value. A
[13:47] Speaker 6: lot of those companies that are investing in AI, they
[13:49] Speaker 6: do have very long term time prizes. I don't think,
[13:52] Speaker 6: you know, they're not going to get out of the
[13:54] Speaker 6: game anytime soon, but if you start looking ahead, you know, three, five,
[13:58] Speaker 6: ten years, there will be pint where it has to
[14:00] Speaker 6: deliver returns. I think we're at the point where you're
[14:03] Speaker 6: starting to see that across the corporate world. You're starting
[14:06] Speaker 6: to see more companies focus on what's the ROI from
[14:09] Speaker 6: an adoption. So I think there's you know, there's certainly risks,
[14:12] Speaker 6: but I think the early signs that it can deliver
[14:14] Speaker 6: value were there.
[14:17] Speaker 2: What is A seven.
[14:17] Speaker 3: IM's attitude towards those South Korean names. I mean principally
[14:20] Speaker 3: it's Samsung and Skhinex, but the cost b had had
[14:24] Speaker 3: some real outperformance so far this year on the idea
[14:27] Speaker 3: that a lot in the AI trade is happening in
[14:29] Speaker 3: South Korea.
[14:31] Speaker 6: Yeah, I think a really interesting demic has been how
[14:33] Speaker 6: emerging markets has kind of been a beneficiary of the
[14:37] Speaker 6: AI trade, where historically the US and emerging market equities
[14:40] Speaker 6: were a little bit uncorrelated. So that's been to change
[14:43] Speaker 6: for this year. In our strategy, we run strategies they're
[14:46] Speaker 6: linked to different global sectors as well as doing general
[14:49] Speaker 6: acid allocation, and so we've had more in technology companies
[14:53] Speaker 6: for some time, and also communications services, so things like
[14:56] Speaker 6: Meta and Alphabet have done really well. I think what
[14:59] Speaker 6: we see we look at it is though even though
[15:01] Speaker 6: we've had a lot of growth, the valuations are still
[15:04] Speaker 6: fairly attractive because earnings growth has been so strong. So
[15:07] Speaker 6: you throw together the momentum the earnings growth, there's certainly
[15:11] Speaker 6: our risk and we've had so much momentum so quickly,
[15:14] Speaker 6: but when you look at those quantitative factors, it's still
[15:17] Speaker 6: a pretty strong case to be in technology stocks.
[15:21] Speaker 3: That's the other story that we've been really focused in.
[15:23] Speaker 3: Like Micron as a case study, has had really interesting
[15:26] Speaker 3: performance but actually trades at a pretty reasonable multiple. Sk
[15:31] Speaker 3: heinez looking at a US listing in part because they
[15:33] Speaker 3: want to bridge the valuation gap with Micron.
[15:37] Speaker 2: Could you talk a little bit more about that.
[15:40] Speaker 6: Yeah, I think, you know, if you look at the
[15:42] Speaker 6: Russell one thousand value index, that's now like tech is
[15:45] Speaker 6: the second highest weight and it's got you know, some
[15:47] Speaker 6: of those companies you're like, I think Microns in it, Intel,
[15:51] Speaker 6: you know, So I think kind of those distinctions between
[15:53] Speaker 6: value and growth don't work so well anymore because you've
[15:56] Speaker 6: had huge growth, but also that earnings growth. I think
[15:59] Speaker 6: the discount on valuations from the US versus other places
[16:02] Speaker 6: is you know, is a perpetual thing, and some of
[16:04] Speaker 6: that will be linked to corporate governance that you have
[16:07] Speaker 6: where there probably is still you know, more that can
[16:09] Speaker 6: be done in some of those nations like Korea and
[16:12] Speaker 6: Japan who have been improving dramatically, But some of it's
[16:15] Speaker 6: also about where the capital is and where people want
[16:18] Speaker 6: to allocate their money to.
[16:19] Speaker 3: So, yeah, a lot of new equities fooding the market,
[16:23] Speaker 3: but it's not just IPOs, right, And later in the program,
[16:26] Speaker 3: we're going to look at the data to kind of
[16:28] Speaker 3: where we are so far in twenty twenty six. SpaceX
[16:31] Speaker 3: the easy example. But what do you make of that dynamic,
[16:34] Speaker 3: the relationship between wall streets fundraising activity in public markets
[16:39] Speaker 3: and like what is happening in private markets still being
[16:42] Speaker 3: pretty insane.
[16:44] Speaker 6: Yeah, I mean there is still a lot of cash
[16:46] Speaker 6: on the sidelines in private markets, and that is cash
[16:49] Speaker 6: you know, that could go into private markets or public markets.
[16:52] Speaker 6: The amount of money that's kind of waiting in the
[16:54] Speaker 6: pipeline for the IPOs is you know, if it does
[16:57] Speaker 6: all come through, will be one of the biggest issue,
[16:59] Speaker 6: you know, years ever. And I think we need to
[17:01] Speaker 6: indedicake account that people will put cash into those but
[17:05] Speaker 6: they probably will also reduce a bit of what they're
[17:07] Speaker 6: already holding to make space for those things. So I
[17:09] Speaker 6: think that's a little bit of a downward pressure on markets.
[17:12] Speaker 6: And I think it's also you know, keeps in mind
[17:14] Speaker 6: that I think investors will scrutinize some of these deals.
[17:17] Speaker 6: I think the interesting thing is less some of the
[17:19] Speaker 6: equity issuance and more the bond issuance with.
[17:22] Speaker 2: The equity issuance, you.
[17:23] Speaker 6: Do have a bit of you know, you can have
[17:24] Speaker 6: retail fanfare around it. With the bond issuance, there's no
[17:28] Speaker 6: you know, index inclusion. It's just kind of you know,
[17:31] Speaker 6: looking at the credit risk. So I think that's been interesting,
[17:33] Speaker 6: the divergence between performs we've had some of the bond
[17:36] Speaker 6: issuance and also the equity.
[17:37] Speaker 3: Issuance, SpaceX bond market, Nvidia bond market. Are you saying
[17:43] Speaker 3: that that's something that worries you or you think that's
[17:46] Speaker 3: a good mechanism for funding what's happening in AI.
[17:50] Speaker 6: I think right now the amount of funding that's going
[17:52] Speaker 6: to bond markets is very healthy. It's by no means overstretched,
[17:56] Speaker 6: and a lot of those companies have very solid long
[17:59] Speaker 6: term cash flows. I think more what I'm saying is
[18:01] Speaker 6: if you look at I think some of the equity
[18:03] Speaker 6: markets can get caught up in a bit of the hype,
[18:05] Speaker 6: whereas the bond markets you have a bit less of that,
[18:08] Speaker 6: and I think you get sometimes a truer picture of
[18:10] Speaker 6: the outlook.
[18:11] Speaker 3: Chanty Kelleman of seven Am back on Bloomberg Tech, Thank
[18:14] Speaker 3: you very much. Indeed, coming up, Bob Eiger and Joshua
[18:18] Speaker 3: Kushner a considering a bid for an NBA expansion team
[18:22] Speaker 3: in Las Vegas, got the Bloomberg exclusive.
[18:24] Speaker 2: Next, This is Bloomberg Tech.
[18:39] Speaker 7: It's time now for talking tech. I'm your Hi, Rahna
[18:41] Speaker 7: and first up. Millennium Management is stepping up its AI drive,
[18:45] Speaker 7: launching a dedicated lab to accelerate early access to cutting
[18:49] Speaker 7: edge tech tools. A new lab will go live over
[18:51] Speaker 7: the next few weeks and coincides with weeping changes in
[18:55] Speaker 7: the hedge fund industry to adopt AI capabilities. Plus where
[19:00] Speaker 7: is taking a heavy hit to grow a global footprint.
[19:03] Speaker 7: The Telecom Giants projects his second quarter losses up to
[19:06] Speaker 7: eight hundred million dollars after striking a deal with the
[19:10] Speaker 7: UK's BT Group to merge their international businesses. The new
[19:14] Speaker 7: four billion dollar joint venture includes a six hundred and
[19:17] Speaker 7: twenty five million dollar equalization payment from Verizon to BT
[19:23] Speaker 7: and Hollywood animation industry is facing an AI reckoning. Some
[19:27] Speaker 7: predict the tech will elevate storytelling, but others expressed it
[19:31] Speaker 7: will slash production costs by thirty to ninety percent and
[19:35] Speaker 7: eliminate many jobs. Leading major players like Disney, DreamWorks and
[19:39] Speaker 7: Netflix proceed with caution.
[19:42] Speaker 8: Ed.
[19:43] Speaker 2: So we're really focused right now.
[19:44] Speaker 3: Media entertainment and sports former Disney CEO Bob Ayga and
[19:48] Speaker 3: Thrive Capital founder Joshua Kushner the joining forces the two
[19:52] Speaker 3: discussing making a bid for an MBA expansion team in
[19:56] Speaker 3: Las Vegas. According to sources, Clombergs and Tashi and Mascarnis
[20:00] Speaker 3: break the story. Okay, give me the details here. This
[20:03] Speaker 3: is interesting. I don't know the historic relationship between mister
[20:07] Speaker 3: Kushner and bobaiga.
[20:09] Speaker 2: An expansion team. What have you learned?
[20:12] Speaker 9: Yeah, I mean absolutely not on my bingo board this year.
[20:14] Speaker 9: But this is a really interesting extension of a relationship
[20:18] Speaker 9: that the two have actually had for quite some time.
[20:20] Speaker 9: So Bob Iger has previously been a venture partner at
[20:24] Speaker 9: Thrive in between both of his stints as CEO of
[20:28] Speaker 9: Walt Disney Company. He stepped down from that CEO role
[20:32] Speaker 9: earlier this year and rejoined Thrive as an advisor. Now
[20:35] Speaker 9: we're seeing both Bob and Josh team up together as
[20:40] Speaker 9: they discuss whether they want to make a bid for
[20:42] Speaker 9: the NBA expansion team in Las Vegas through Thrive Eternal,
[20:45] Speaker 9: another new effort which Bob is advising on, and the
[20:48] Speaker 9: whole idea between Thrive Eternal is that they want to
[20:51] Speaker 9: invest in iconic franchises, cultural assets, everything from collectibles. Two
[20:57] Speaker 9: sports teams in this case.
[20:59] Speaker 3: Okay, look, I am a fair weather Warriors fan at best.
[21:03] Speaker 3: The Bleed Tech team in New York is still pretty
[21:07] Speaker 3: high on the NIXT success. What I do know is
[21:10] Speaker 3: that the NBA has looked at the idea of a
[21:13] Speaker 3: potential franchise expansion in Las Vegas.
[21:15] Speaker 2: Where is that process at.
[21:18] Speaker 9: Yeah, I mean this is something that they have approved,
[21:22] Speaker 9: you know, the actual expiration of doing this. And so
[21:24] Speaker 9: this would mean that they would be expanding the NBA
[21:27] Speaker 9: to include teams in both Las Vegas and Seattle. In
[21:31] Speaker 9: this case, the bid is for a majority investment. It's
[21:35] Speaker 9: yet to be submitted according to my sources, but this
[21:38] Speaker 9: is an attempt at taking a controlling stake in that
[21:42] Speaker 9: potential team. And what's also interesting here is it's not
[21:45] Speaker 9: Thrive's first investment in sports. Earlier this year, we reported
[21:49] Speaker 9: that Thrive was making a smaller investment in the San
[21:53] Speaker 9: Francisco Giants, a little bit closer to home for both
[21:55] Speaker 9: of us, and so this is, you.
[21:57] Speaker 2: Know, a bigger deal.
[21:58] Speaker 9: Literally, but really it kind of show is interesting given
[22:01] Speaker 9: that they have made a few bats in sports.
[22:05] Speaker 2: Just to point out that.
[22:06] Speaker 3: For the story, representatives for Thrive Capital and Bobaiger declined
[22:09] Speaker 3: to comment, but I we'd check out and Attasha's story
[22:12] Speaker 3: one of the most read things today on bloombergs on
[22:14] Speaker 3: attashm mascarinas, thank you very much. Coming up, SpaceX helps
[22:18] Speaker 3: hower a record year for Wall Street fundraising. We look
[22:21] Speaker 3: at how AI is driving the deal boom. We're about
[22:25] Speaker 3: halfway through the program and right now we're in a
[22:29] Speaker 3: place where market's filling pretty good. Now's that one hundred
[22:32] Speaker 3: upper percentage point. Semiconductors up almost a percentage point. Just
[22:36] Speaker 3: to quick check on bitcoin, fifty nine dollars per token
[22:41] Speaker 3: a little softer, but there's nothing in the news cycle
[22:44] Speaker 3: really driving that. As I said, it's half time. We'll
[22:47] Speaker 3: be right back from San Francisco. This is Bloomberg Tech.
[22:57] Speaker 3: Welcome back to Bloomberg Tech. I want to get back
[22:59] Speaker 3: to that Comcast deal. So Comcast is going to spin
[23:03] Speaker 3: off NBC Universal and Sky, basically separating into two separate
[23:09] Speaker 3: public company. This is something that's about a year out
[23:11] Speaker 3: and existing investors will have exposure to both. But it's
[23:14] Speaker 3: really interesting to see how the streets.
[23:17] Speaker 2: Kind of reacting to this.
[23:18] Speaker 3: In fact, Bloomberg Intelligence and our own analysts kind of
[23:21] Speaker 3: see this as something that opens up and sets the
[23:24] Speaker 3: stage for more M and A in the cable media space. Right,
[23:28] Speaker 3: so you have a separate entity that is focused on
[23:31] Speaker 3: the media business NBC Universal, while Comcast doubles down on
[23:35] Speaker 3: more of that kind of legacy infrastructure business and as
[23:37] Speaker 3: Hetherranger Nation rights, the transaction also enhances strategic flexibility and
[23:42] Speaker 3: can pave the way for M and A. Interesting, it's
[23:45] Speaker 3: kind of like what happens next. Let's take a look
[23:48] Speaker 3: at today's big number as well, five hundred and seventy
[23:51] Speaker 3: billion dollars. That's the amount of market value Microsoft is
[23:54] Speaker 3: set to shed in the month of June, with the
[23:57] Speaker 3: stock at one point down as much as twenty percent
[24:00] Speaker 3: during the month. So right now it's on track for
[24:03] Speaker 3: its worst monthly drop since December of two thousand, or
[24:07] Speaker 3: its worst month of performance in almost twenty six years.
[24:11] Speaker 2: One.
[24:12] Speaker 3: We're tracking a wave of fundraising tied to the AI
[24:15] Speaker 3: buildout has pushed us IPOs and share sales to record
[24:19] Speaker 3: two hundred and fifty one billion dollars this year so far,
[24:23] Speaker 3: blockbuster offerings from SpaceX and Alphabet have led the way,
[24:26] Speaker 3: and bankers say there's plenty more in the pipeline for
[24:30] Speaker 3: more most Baily Lipshaltzer here on the IPO inequity beat, like,
[24:34] Speaker 3: the big picture story is one we've talked about, right,
[24:36] Speaker 3: lots of equity is going to flood the market. But
[24:40] Speaker 3: the line of what I just read that it's interesting
[24:42] Speaker 3: is that you know, the bankers hope and the bankers
[24:45] Speaker 3: see that there's more to come.
[24:47] Speaker 10: What do we know bankers get paid to be optimistices,
[24:50] Speaker 10: so we got to take everything they say with a
[24:51] Speaker 10: grain of salt. But when you look at the record
[24:54] Speaker 10: year so far, the record half we've already seen again
[24:57] Speaker 10: north of two hundred and fifty one billion dollars raised,
[24:59] Speaker 10: that doesn't even fully encapsulate that Google raise because it's
[25:03] Speaker 10: an ATM, which means at the market offering, which it
[25:05] Speaker 10: technically hasn't been fully penciled in, so that number should
[25:08] Speaker 10: be even larger. Again, we're looking at numbers we have
[25:11] Speaker 10: never seen before, beating twenty twenty one. The expectation for
[25:14] Speaker 10: the weeks and months ahead, s k Heiinez expected in
[25:17] Speaker 10: the coming weeks that's going to be potentially thirty billion
[25:20] Speaker 10: dollar fundraise here in the US, So that'll take the
[25:22] Speaker 10: baton and get.
[25:23] Speaker 2: Us off to a quick start in July.
[25:25] Speaker 10: Expectations do broaden out quite a bit as we look
[25:27] Speaker 10: for the likes of private equity, maybe in the retail
[25:31] Speaker 10: consumer space. Jersey Mics is a name that continues to
[25:34] Speaker 10: come to mind, but certainly continuing to keep an eye
[25:36] Speaker 10: on anything and everything around artificial intelligence. I also want
[25:40] Speaker 10: to call out that we're expecting c Square, which is
[25:42] Speaker 10: a data center company, to kickstart its process on the
[25:44] Speaker 10: other side.
[25:45] Speaker 2: Of the fourth of July holiday.
[25:46] Speaker 3: It's gonna say thank you for bringing it back to tech,
[25:49] Speaker 3: bringing Jersey.
[25:50] Speaker 2: Mics to the show. Unbelievable.
[25:53] Speaker 3: Part of it is like we're showing this amazing chart.
[25:55] Speaker 3: Let's go back to the number so far in twenty
[25:58] Speaker 3: twenty six. This is a big number. This is something
[26:02] Speaker 3: that we have not seen for quite a few number
[26:05] Speaker 3: of years now.
[26:06] Speaker 10: No, we've never seen it, and I think when you
[26:09] Speaker 10: look at the breadth So we're talking about obviously the
[26:12] Speaker 10: big fundraise from SpaceX again, record setting number kind of
[26:16] Speaker 10: bolsters this as Kahinich would potentially be the second largest
[26:20] Speaker 10: IPO ever. So these would be two deals in consecutive months,
[26:23] Speaker 10: something we've never seen before. Again, going back to the
[26:26] Speaker 10: broader equity issuance, whether that's regular way equity or convertible debt,
[26:29] Speaker 10: that big fundraising from Alphabet potentially puts the pressure on
[26:34] Speaker 10: other hyperscalers to sell equity, so diverting away from debt,
[26:37] Speaker 10: which has kind of been the bread and butter for
[26:39] Speaker 10: the build out of this AI boom, this data center boom,
[26:42] Speaker 10: and the expectations that there will be more lying on
[26:45] Speaker 10: the other side of the midway point again on the
[26:47] Speaker 10: other side of the fourth of July holiday. Again, just
[26:49] Speaker 10: want to call out things have been working across the
[26:52] Speaker 10: AI ecosystem Sarah Bras again not really trading as well
[26:55] Speaker 10: as many had expected, or it certainly was day one,
[26:57] Speaker 10: But that's a north of six billion dollar IPO, and
[26:59] Speaker 10: then you just go down the list, there's about a
[27:01] Speaker 10: dozen raising a billion dollars or more. Bankers expect that
[27:05] Speaker 10: number to continue and be matched in the second half,
[27:07] Speaker 10: and even some of these conversations with these big bankers
[27:10] Speaker 10: at the likes of a Morgan Stanley or Goldman Sachs,
[27:12] Speaker 10: they expect the third quarter to be critical. As we
[27:15] Speaker 10: kind of get towards potentially Anthropic coming in the fall
[27:18] Speaker 10: towards year.
[27:18] Speaker 2: End the most faded Upsheltz, thank you very much.
[27:22] Speaker 3: Now coming up, and Thropic wins US approval to restore
[27:26] Speaker 3: some access to its powerful Mythos five AI model.
[27:30] Speaker 2: We'll discuss what.
[27:31] Speaker 3: That means for model development and AI regulation with Hugging
[27:35] Speaker 3: Face CEO clemd Along.
[27:36] Speaker 2: That's next. This is Bloomberg Tech.
[27:47] Speaker 3: Late on Friday, and Thropic won US approval to restore
[27:51] Speaker 3: some access to its powerful Mythos five AI.
[27:55] Speaker 2: Now.
[27:55] Speaker 3: You may remember two weeks ago the US government abruptly
[27:58] Speaker 3: barred Anthropic from letting foreign nationals have access to its
[28:02] Speaker 3: most powerful models over cybersecurity concerns. In response, and Thropic
[28:07] Speaker 3: disabled all access to those models. And right now, after
[28:12] Speaker 3: furious negotiations, Mythos five is back, sort of. In a
[28:17] Speaker 3: letter to the company seen by Bloomberg News, Commerce Secretary
[28:20] Speaker 3: Howard Lutnick wrote the model could be released to quote
[28:23] Speaker 3: certain trusted partners. The Fable five model or variant wasn't mentioned.
[28:29] Speaker 3: Let's get into it, starting with the business impact. Teresa Payton.
[28:33] Speaker 3: She is the CEO of Cybersecurity Advisor for Delice Solutions
[28:36] Speaker 3: and a former White House CEO during.
[28:38] Speaker 2: The George W. Bush administration.
[28:40] Speaker 3: I rely on both your expertise in industry and experience
[28:44] Speaker 3: of being on the government side of the table. The
[28:47] Speaker 3: June twelfth Export Control the parameters of it are very interesting,
[28:53] Speaker 3: and maybe you can unpack them a little. But what
[28:55] Speaker 3: they said was there's been significant progress in risk mitigation,
[28:59] Speaker 3: and I'm finding out to understand what that progress looks like.
[29:04] Speaker 11: I mean, where do we begin. I mean, the greatest
[29:07] Speaker 11: AI breakthrough is one that people can actually use. And
[29:10] Speaker 11: what we're saying in this instance is it's the greatest breakthrough.
[29:13] Speaker 11: It scares us, but we're not gonna let people use it.
[29:16] Speaker 11: But my understanding of what's been reported is that there
[29:20] Speaker 11: will now be roughly more than one hundred but not
[29:23] Speaker 11: much more than that, federal departments and agencies, private sector
[29:27] Speaker 11: companies that have already been vetted, which will have access
[29:30] Speaker 11: to this model, and they will not have sort of
[29:33] Speaker 11: the constraints around foreign employees to be able to use
[29:37] Speaker 11: this model. And then I guess more to come. It's
[29:39] Speaker 11: sort of a stay tuned and let's see.
[29:43] Speaker 3: The thing the tension here is that because of the
[29:46] Speaker 3: capabilities of the model industry wants access to it right,
[29:51] Speaker 3: they want to use it for its benefits.
[29:55] Speaker 2: From the government.
[29:55] Speaker 3: Side of the table, how do they hear that? How
[29:58] Speaker 3: do they get that part right?
[30:00] Speaker 11: I think this is a tough one, and you know,
[30:04] Speaker 11: this should be a wake up call for everybody, not
[30:07] Speaker 11: just the AI companies who are producing you know, agentic AI,
[30:11] Speaker 11: generative AI and frontier AI models, as well as businesses
[30:15] Speaker 11: implementing them that you know, we saw open AI actually
[30:19] Speaker 11: just got told recently that they needed to limit model
[30:22] Speaker 11: access for new capability. So what we need to do
[30:26] Speaker 11: here is, you know, we know that the rest of
[30:28] Speaker 11: the world, including China, is investing aggressively increasingly. There's capable
[30:33] Speaker 11: open source, in some cases free models that are narrowing
[30:37] Speaker 11: the gap. So leadership in this moment really depends on
[30:40] Speaker 11: responsibly getting these breakthroughs as fast as possible into the users.
[30:46] Speaker 2: So what's going to.
[30:46] Speaker 11: Have to happen here is a much faster vetting process
[30:50] Speaker 11: for allowing these incredible capabilities to be in the hands
[30:54] Speaker 11: of people in business who need to have this to
[30:56] Speaker 11: protect and defend their enterprises.
[31:00] Speaker 3: So you teed us up really nicely for our next
[31:01] Speaker 3: conversation with hugging Face CEO clemmed Along in a moment
[31:04] Speaker 3: particular focus on open source. You said, what needs to happen.
[31:08] Speaker 3: I guess there's a distinction with what is likely to happen,
[31:11] Speaker 3: and so you know, again reflects you on your experience.
[31:14] Speaker 3: Do we need to expect tighter US restrictions, particularly on
[31:18] Speaker 3: the frontier labs each generation of powerful model they put out.
[31:24] Speaker 11: I think we run a huge risk if we don't
[31:26] Speaker 11: create a framework that provides certainty in the process and
[31:31] Speaker 11: fairness in the process. And you know, so if you
[31:34] Speaker 11: don't create a framework and we don't know who's next.
[31:37] Speaker 2: You know, the markets don't like.
[31:38] Speaker 11: Uncertainty, and certainly our you know, global partners don't like uncertainty.
[31:43] Speaker 11: You're already starting to see global partners say, well, maybe
[31:45] Speaker 11: we need to develop our own models, Maybe we need
[31:48] Speaker 11: to do things our own way, maybe we need to
[31:50] Speaker 11: install open source models here within our country. So we
[31:54] Speaker 11: really need to see leadership come out from Washington, d C.
[31:58] Speaker 11: To say this is the framework, this is the certainty,
[32:01] Speaker 11: here's the process. It may change, but at least publish
[32:03] Speaker 11: a process so we all know what the test looks
[32:07] Speaker 11: like and we can all perform according to the test.
[32:10] Speaker 3: Let's recap where we stand from Friday. There are one
[32:12] Speaker 3: hundred US companies and federal agencies that will have been
[32:15] Speaker 3: approved for access. They don't require an export license for
[32:19] Speaker 3: themselves or for their non US employees. But how much
[32:23] Speaker 3: scrutiny will that group of one hundred be under how
[32:26] Speaker 3: they use the model, what they were able to get
[32:28] Speaker 3: from it, rather than just anthropic itself sort of being
[32:32] Speaker 3: under the microscope of this administration.
[32:35] Speaker 2: Yeah, you bring up.
[32:36] Speaker 11: A great point and sort of be careful what you
[32:37] Speaker 11: wish for. So now that you're in this group, if
[32:40] Speaker 11: you're in a heavily regulated private sector industry, your regulators
[32:43] Speaker 11: are going to ask you when did you run it,
[32:45] Speaker 11: what did you find, and did you fix what you found?
[32:48] Speaker 11: So there is going to be another layer of undue
[32:51] Speaker 11: scrutiny on kind of these businesses. But they also have
[32:54] Speaker 11: an advantage. They get to have access to this tool first.
[32:58] Speaker 11: I think the other reminder for people is is that
[33:00] Speaker 11: running this tool isn't free. It does require token, so
[33:04] Speaker 11: it does cost money. It costs computer resources, it costs
[33:07] Speaker 11: staffing resources to look at the results. I know, by
[33:10] Speaker 11: the way, if you find something, you definitely need to
[33:12] Speaker 11: get it fixed. Right away, and so having the capability
[33:15] Speaker 11: to actually mitigate and remediate anything that's found and doing
[33:19] Speaker 11: it at machine speed is not something that all businesses have.
[33:23] Speaker 3: Three Sepeton, CEO of Faughterly Solutions and former White House coio,
[33:28] Speaker 3: thank you very much. This isn't just about one company
[33:30] Speaker 3: or ONEAI model anymore. It's about what happens when government
[33:33] Speaker 3: scrutiny becomes a market signal. Our next guest has argued
[33:37] Speaker 3: that being labeled too dangerous may actually be good marketing
[33:42] Speaker 3: for frontier AI firms Joining us now is clemmed Along,
[33:45] Speaker 3: CEO of Hugging Face and if you're runfamiliar, Hugging Face
[33:49] Speaker 3: is where that more than seven million developers discover tests
[33:52] Speaker 3: deploy AI models from leading labs to open source, giving
[33:56] Speaker 3: Clem a pretty unique perspective on how AI is actually
[33:59] Speaker 3: being adopted across industry and clim. I want to start
[34:02] Speaker 3: with that post you put on X the idea that
[34:05] Speaker 3: that designation is good marketing.
[34:07] Speaker 2: Could you explain a little bit more why you think that?
[34:11] Speaker 12: Well, First, I want to point out that a lot
[34:13] Speaker 12: of people are saying that, you know, it's not completely
[34:17] Speaker 12: unfair for them to start being regulated, given that They've
[34:21] Speaker 12: been doing what we call doom marketing for quite many
[34:25] Speaker 12: years now. If you remember GPT two that was released
[34:28] Speaker 12: I think five six years ago, was already deemed like
[34:31] Speaker 12: two dangerous to release. So I feel like it's fair
[34:35] Speaker 12: for the US government to at least try to get
[34:38] Speaker 12: more transparency about what these models can and come to.
[34:43] Speaker 12: And you know, I think it's important to remember that
[34:46] Speaker 12: these are gigantic companies, you know, fastest growing in the world,
[34:51] Speaker 12: on their way to maybe be the most valuable companies
[34:54] Speaker 12: by the end of this year. So I feel like
[34:57] Speaker 12: they can take these kinds of regulation and an interaction
[35:01] Speaker 12: with the US government. I think one thing that we
[35:04] Speaker 12: need to be careful of is to kind of like
[35:07] Speaker 12: spread these constraints and these regulations to the rest of
[35:11] Speaker 12: the ecosystem, like startups, little tech university academia, who certainly
[35:17] Speaker 12: don't have the same kind of like legal and policy
[35:20] Speaker 12: capabilities than these companies have.
[35:23] Speaker 3: I want to get to that distinction of regulating sort
[35:26] Speaker 3: of the frontier labs and APIs and open source In
[35:29] Speaker 3: just a minute, but to end the conversation on that post,
[35:34] Speaker 3: do you think that the US government has handled this
[35:36] Speaker 3: correctly with anthropic and methos and fable.
[35:41] Speaker 12: Well, I think it's important to get more transparency for everyone, right,
[35:45] Speaker 12: the government needs to have more transparency about what these
[35:49] Speaker 12: models are capable or not capable of doing, which is
[35:52] Speaker 12: hard to do through APIs right because there are guad rails,
[35:55] Speaker 12: there are limitations that makes it hard to actually know
[35:58] Speaker 12: what they're capable and not capable of doing.
[36:01] Speaker 2: So I think it's it's.
[36:02] Speaker 12: It's quite quite fair for the US government to try
[36:05] Speaker 12: to get more transparency on this very complex black box system.
[36:11] Speaker 12: You know, obviously, like it's hard to do that perfectly,
[36:15] Speaker 12: given how fast the field is moving and how uncertain
[36:19] Speaker 12: some of these risks are. But I think it's it's
[36:22] Speaker 12: not it's not unfair overall.
[36:25] Speaker 3: You argue that there should be a distinction how governments
[36:28] Speaker 3: regulate frontier AI and open source. Why why should open
[36:34] Speaker 3: source be treated differently? What is it that the ecosystem
[36:38] Speaker 3: needs differently on the open source side?
[36:42] Speaker 12: Well, first, in terms of capabilities, I think it is
[36:44] Speaker 12: badly accepted that the most dangerous capabilities are concentrated behind
[36:51] Speaker 12: like in these kind of like frontier AI labs, where
[36:56] Speaker 12: the majority of kind of like open source models are
[36:58] Speaker 12: more like specialized aller more kind of like broadly beneficial
[37:03] Speaker 12: models that aren't really creating more risks. So there's a
[37:09] Speaker 12: different in terms of capabilities. There's a different in terms
[37:11] Speaker 12: of transparency.
[37:12] Speaker 2: You know, Like open.
[37:13] Speaker 12: Source models, it's really easy to evaluate and to look
[37:16] Speaker 12: at what they can and can't do from the get
[37:19] Speaker 12: go because you know, you get you get the models,
[37:22] Speaker 12: you can test it really easily. Versus an API, it's
[37:25] Speaker 12: really really hard to do that. So you need kind
[37:27] Speaker 12: of like some some more kind of like access, I
[37:31] Speaker 12: would say, to really assess them. And also, of course,
[37:36] Speaker 12: like the provenance is different, right Like open source is
[37:40] Speaker 12: like an ecosystem of smaller companies, smaller organizations, you know,
[37:45] Speaker 12: versus you know, the Frontier Labs is a couple of
[37:48] Speaker 12: companies that are concentrating a lot of power behind closed doors.
[37:53] Speaker 3: There's been a lot of press coverage of late and
[37:56] Speaker 3: other data points like your own revenue run rate, right sources,
[38:00] Speaker 3: your lifeblood. Open source models seem to be competitive in
[38:05] Speaker 3: a market against closed models. What else would you point
[38:08] Speaker 3: to to evidence.
[38:09] Speaker 12: That, well, it's not just us, right, Like, the whole
[38:13] Speaker 12: ecosystem right now is booming. We see tons of companies
[38:19] Speaker 12: doing super well, like the influence providers nail cloud or posting,
[38:23] Speaker 12: kind of like fantastic fantastic growth. A million new models
[38:30] Speaker 12: in data sets have been sued on hugging face just
[38:32] Speaker 12: the past past quarter. And you see that in usage too, right,
[38:37] Speaker 12: the usage of open source by American companies, American startups,
[38:43] Speaker 12: American small companies is also booming. So this is really
[38:47] Speaker 12: exciting because it basically enable empower more more people to
[38:53] Speaker 12: really own AI themselves, whether than renting it with APIs.
[38:59] Speaker 12: So this this is quite exciting Clem.
[39:02] Speaker 3: Another area of real focus of the show of late
[39:05] Speaker 3: it's been robotics. We had deep Utala from a video
[39:07] Speaker 3: on last week talking about Halo's hugging face is also
[39:12] Speaker 3: looking a lot at robotics. What I would be really
[39:14] Speaker 3: grateful for is to understand the role that open source
[39:17] Speaker 3: is playing in robotics and physical AI, but also like
[39:21] Speaker 3: the benefit of leaning on open source in that field.
[39:25] Speaker 12: I think it's even more important than in general AI,
[39:27] Speaker 12: just because you know, if you think of having a
[39:29] Speaker 12: robot at home that is interacting with your environment, with
[39:34] Speaker 12: your kids. I just got two daughters a few months ago,
[39:37] Speaker 12: so when I think about a robot interacting with my daughters,
[39:42] Speaker 12: like I don't want to have to trust a black
[39:45] Speaker 12: box controls by one megacorp being able to do anything.
[39:51] Speaker 12: I want to have some transparency about what's going on
[39:54] Speaker 12: on this robot, how is it built, how does it
[39:58] Speaker 12: decide to interact one way or another? And for that
[40:03] Speaker 12: open source is the only way, right Like, it gives
[40:06] Speaker 12: you transparency, it gives you control.
[40:08] Speaker 2: It creates an eco.
[40:09] Speaker 12: System of different companies that are going to be able
[40:11] Speaker 12: to build robots. So that's that's why we excited about it.
[40:16] Speaker 12: We created this loyal robot called Richie Mini, and we've
[40:19] Speaker 12: been really surprised by how into it people have been.
[40:23] Speaker 12: We've we shipped over ten thousand of them all over
[40:25] Speaker 12: the world in the past past few months. And that's
[40:29] Speaker 12: a testament to how excited people are by you more open,
[40:34] Speaker 12: more transparent, more open source approaches to robotics.
[40:38] Speaker 3: Claimed along hugging face CEO back on Bloomberg Tech, thank
[40:41] Speaker 3: you very much.
[40:42] Speaker 2: Indeed, now coming.
[40:43] Speaker 3: Out, AI bots are taking over remote meetings, creating new
[40:48] Speaker 3: challenges for the corporate's fare. We have a really interesting
[40:52] Speaker 3: one next this is Bloomberg Tech. AI note takers are
[41:07] Speaker 3: infiltrating the corporate world meant to increase productivity, AI bots
[41:11] Speaker 3: are instead posing privacy concerns and security risks. As AI
[41:16] Speaker 3: is increasingly incorporated into the workplace, the debate is growing
[41:19] Speaker 3: on what conversations constitute robots and which ones are strictly human.
[41:24] Speaker 2: Bloomberg BusinessWeek contributor to.
[41:25] Speaker 3: Issi Labowski has more brilliant BusinessWeek story than new etiquette
[41:30] Speaker 3: for navigating AI. Note takers start with the overview of the.
[41:33] Speaker 8: Story, right, So, I think we've all seen this happen
[41:37] Speaker 8: in our sort of virtual remote work world. We're joining
[41:41] Speaker 8: these Zoom calls and Google meets, and when you get
[41:44] Speaker 8: on the call, it's you and whoever you're supposed to
[41:46] Speaker 8: have a call with and their note taker. Sometimes that
[41:50] Speaker 8: says you know otter AI or read AI, or sometimes
[41:53] Speaker 8: it doesn't even announce itself at all. And what these
[41:56] Speaker 8: note takers do is kind of take a live transcript
[41:58] Speaker 8: of what you've talked about. Afterwards, they hand you back
[42:01] Speaker 8: a really handy summary, and a lot of people feel
[42:04] Speaker 8: like it's great, this is saving me time on doing
[42:06] Speaker 8: the whole recap, and I can go back and look
[42:08] Speaker 8: at my notes afterward. The question is, are people really
[42:11] Speaker 8: doing the work up front? To make sure the other
[42:13] Speaker 8: person on the other side of the screen is comfortable with.
[42:16] Speaker 2: That as well.
[42:17] Speaker 3: You know, that is also a question of human responsibility.
[42:20] Speaker 3: You make the piece in the point in the Business
[42:22] Speaker 3: Week piece that you know, the AI note take is
[42:25] Speaker 3: a quite commonplace. Now some of these tools have already
[42:28] Speaker 3: got millions of views as a piece, but it's up
[42:30] Speaker 3: to the shutdently to say how is this working?
[42:34] Speaker 8: Absolutely, I mean that's what the story is about. I
[42:36] Speaker 8: think we all know that there are legal invocations here
[42:39] Speaker 8: of having every single word recorded and shared and saved,
[42:43] Speaker 8: but there are also interpersonal consequences. Right when you are
[42:47] Speaker 8: the person who has your note taker, you know, plugged
[42:50] Speaker 8: into your calendar and it goes to every meeting and
[42:53] Speaker 8: sometimes you don't even show up because you forgot about
[42:55] Speaker 8: the meeting but your note taker didn't. That's a really
[42:58] Speaker 8: bad look. So this story is really about the norms
[43:01] Speaker 8: and common courtesy that we have to develop around these
[43:04] Speaker 8: AI note takers because we are not all at the
[43:06] Speaker 8: same place of comfort with regard to AI. We are
[43:09] Speaker 8: in some cases quite polarized, and these meetings, these AI
[43:14] Speaker 8: note takers are sort of the one place where my
[43:17] Speaker 8: drive for you know, AI enabled efficiency is running up
[43:21] Speaker 8: against your discomfort with having AI in your life, and
[43:24] Speaker 8: so we need to navigate those things on the human
[43:27] Speaker 8: to human level, and often that deals with issues of
[43:29] Speaker 8: consent and all of that and just personal responsibility.
[43:35] Speaker 3: That was Bloomberg BusinessWeek in tributy, Issie Lapowski, Thank you
[43:38] Speaker 3: very much. Indeed, let's go back one last time to
[43:41] Speaker 3: Comcast shares markedly high. There is a plan to spin
[43:44] Speaker 3: out NBC Universal and Sky in Europe into a standalone
[43:49] Speaker 3: public company, so there will be two and in a
[43:51] Speaker 3: year's time, if it all goes to planned, shareholders will
[43:54] Speaker 3: have exposure to both to stop up seven percent, on
[43:57] Speaker 3: track for its best day since mid April. That does
[44:00] Speaker 3: it for this edition of Bloomberg Tech. So much in
[44:03] Speaker 3: the program. Check out the podcast to recap all of it.
[44:05] Speaker 3: You know exactly where to find it. This is Bloomberg

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.
