Bloomberg Audio Studios, podcasts, radio news. Bloomberg Tech is a live from coast to coast with Caroline Hyde in New York and Edlow in San Francisco.
This is Bloomberg Tech coming up.
Micron wanns of heavy spending on production to meet demand for its memory chips. Will break down the company's results. Rasali Baba says it's aiming to generate one hundred billion dollars in cloud and AI revenue within the next five years, even as earnings come on pressure and Uber and Rivian. They team up with Uber set to invest up to one point twenty five billion dollars in Rivian's Robotaxi fleet. The details on the deal, but first we have the
details on the macro picture. Once again, stocks are under pressure today. This is because oil pushes high, the conflict intensifying in the Middle East, and energy infrastructure under fire. We're currently seeing up three percent on Brent, but boy headed spiral hire earlier huge moves in the bond market start to dial down. We've seen eighteen basis points on the two year, extraordinary moves in the UK on its guilts. Look, we are worried about inflationary pressure and the ability of
central banks to make a move. We saw the ECB, the BOE, and of course I've fed yesterday in the Japanese overnight. All keep rates on hold, but all of them eyeing whether or not policy moves when we needed in the near term, move on to individual stocks, because Micron down two point six percent when the chip maker posted a quadrupling in its revenue, it saw six hundred
percent increase in its earnings press share. However, it is spending big and we want to get the details with Dinabas because this stock has run up because well, the pricing pressure in memory chips in particular, that aids the revenue and the profitability of Micron, but it's going to have to spend to keep up with demand.
Yeah, I mean, you would think that the numbers they posted yesterday would cause the stock to go up. As you mentioned, there's a bit of an issue here that the reason the stock is going up, as you said, is because we are seeing an almost unprecedented increase in the price of memory. What's gone on is that there's a thing called high bandwidth memory, which is really critical for all of the AI chips that in Nvidia has
been talking about this week at its TC conference. And you know, while that's demand has accelerated, the three makers of that memory, that's Micron, Samsung and s k Heinex, have been focusing on that product, which is you know, short supply, and they have focused less on other memory products, things like d RAM. So the people that use DRAM, we're talking about PC makers, phonemakers, car makers, are paying dramatically higher prices for that. That's good for Micron, and
you saw that in those numbers. But Micron and you know, Samsung has also announced extra extra spending, so they're trying to expand you know, their capital expenditures, build more facilities, make more of this memory, you both kinds. And so you're seeing, you know, a forecast for higher capex this year, for continued increasing capex next year, and so investors are trying to wrangle with what all of that means.
Was there a risk when we go to HBM four in particular and just the desire for more these high boundwidth memory chips in AI accelerators, Is there any risk that Innvidia uses.
Skhinex or Samsung more than Micron.
So there have been some reports, you know, some analysis said something out a couple of weeks ago, raising concerns about whether Micron will be you know, as important to in Vidia's next Vera Ruben chips as some of its competitors. Micron has said this week that they are now shipping that HBM four in volume, and they have said that it is designed for Vera rubin. I think we were, you know, waiting to see what that means, what exactly in Vidia is going to use who and how much.
But that has been a concern and that has been weighing on the stock at various points, even though this dock has been over overall has been up.
Yeah, I mean, maybe just a bit of profit taking on the day. Dinabas, who has been incredibly busy this week at GtC and now on Micron, we so appreciate you joining this morning. Now let's turn our attention to the Fedshair Japwell, who says that look, while AI holds long term promise, is near term impact may actually be adding to inflation, such as with these chips as companies ramp up investment, just take a listen.
In the short term, what's happening is we're building data centers everywhere, and that's actually putting pressure on on all kinds of goods and services that go into building these things. So that's actually probably pushing inflation up at the margin.
That perspective comes as investors continue to pass the real world payoff.
Or not from AI for more.
Martin Norton Empowered Chief investment strategist joins us, look, Micron speaks to this inflationary pressure that we see coming from AI. Marta just dig in there as to what fedshair powers, saying how much are we likely to see just bottlenecks being the issue?
Well, it is the question or what are the questions of the hour when we think about AI in twenty twenty six, in twenty twenty seven, we know that timeline for the AI build is quite long, we know it's quite expensive, and we know that it is creating bottlenecks not just for memory chips but.
Also for energy.
So there's all these kind of hot pockets that are related to the build. I don't know if we're looking at a FED chair or a committee that's saying that this is going to excessively move core inflation, but I think it's coming at an inopportune time when, of course we also have an oil shock, and we have an
inflation picture that continues to linger. So I think it's more the sum total of all the different questions from AI, from energy, from lingering inflation that puts us in the position we're in today.
Go to the oil shock, Malta, go to the conflict with the Middle East, and what that means for invested demand right now.
Well, it's so interesting right now because Powell couched his comments yesterday around inflation uncertainty by also noting that typically wisdom would suggest that you look through an oil price
shock because it's not a permanent change in inflation. And in fact, when you look at the summary of Economic projections yesterday, yes, the committee mood inflation expectations higher for this year, but then the expectation and of course these have a wide range of outcomes around them, but the expectations for next year weren't really that much higher than they already were, and so that would reflect this idea that we're looking at a near term disruption but not
necessarily longer term. But of course the juries out we're seeing destruction of energy infrastructure in real time, and that's one of those tipping variables that could put us in a much more difficult position from an energy and oil price shock perspective.
I mean, boy, does it put a difficult position on Asia and Europe more so than the US.
In particular.
It's interesting that we have a meeting between the Japanese leader and the United States today and we are expecting actually nuclear deals to.
Be announced from that.
But how are you seeing in particular a long term change in geopolitics and what that means more ultimately for investor risk on attitude.
Well, I mean, the observation that this is a very different scenario for the US versus the rest of the world is spot on, because of course the US is the er is the world's largest producer of natural gas and of oil. That doesn't necessarily shelter it from all shocks on the oil front. It certainly provides protection on the natural gas front. In Asia and Europe are in very different positions, and we're seeing them kind of thread
the needle on the geopolitics front. They're not necessarily putting pressure on the US to stab as attacks, but they're certainly not joining in it. And of course we also have all the new dynamics that were released last year around tariffs, so there does seem to be a shifting of how countries are inner or are relating to one another.
I think one of the things we want to watch longer term is what kind of regionalization does this create from how economies operate versus how they've been operating on a much more global scale. But I don't think we want to jump to conclusions there. There's still a lot to understand about how things are unfolding.
Look, I'm going to thread a difficult needle here of how it impacts technology, because there have been stark warnings that if infrastructure is decimated in the Middle East, those pools of capital that have been very important to some of the big foundational modules here in the United States.
The AI growth story has been.
Dependent on Middle East of money and that might be diverted to far more immediate concerns.
March or Is that something you're hearing about.
That's a great observation. It isn't something that I've heard about, but I think it's a very astute topic to raise. I've also heard folks talk a little bit more about the supply chain disruption that could come from AI if we're looking at places like Taiwan and we're looking at
manufacturing capabilities and there's an energy import element there. What can we expect in terms of what that disruption looks like and how does that ultimately play through to AI cost Again, a lot of you know, the jury is still out, so I don't want to assume too much. But right now we're seeing tech mag seven software. We're seeing strength and a lot of those areas at least
relative to the sell offs that we're seeing elsewhere. But should we start to see some disruption both from a funding perspective and from the perspective of the AI supply chain, I think that's something that could you know, at least be a headwind for some of those areas up have been holding up well?
Does it?
And I might not be a out of expertise, but doesn't mean anything for the IPO pipeline and the idea that we were set to be having some of the biggest companies private companies coming.
To the market well IPO and M and A.
Those were areas that people were looking at as being a real source of excitement for twenty twenty six with a lot of momentum building. Again, I don't want to assume too much on what this geopolitical event means for the broad economy and whether it unhinges it all together. My base case at the moment, even though we're seeing energy infrastructure getting destroyed in real time, is that this does have a near term end point. We're not necessarily looking at the most severe scenario. So a lot of
the trends that were in play. My estimation is that they remain in play, but we're on a pause for the moment. Of course, we have to keep our options open on that front.
We do Martin norton a time of uncertainty, and we so appreciate you joining us on it.
From Empower coming.
Up, ali Baba sets an ambitious one hundred billion dollar target for cloud and AI revenue, even as earnings come under pressure.
That's next. This is brtting their tech.
Ali Baba says it's aiming to generate one hundred billion dollars in cloud and AI revenue within the next five years.
This is CEO dy Wu really laid out the.
Ambitious target out of the company reported a sixty seven percent drop in quarterly earnings alongside some pretty sluggish revenue growth. For more Bloomberg's Asia stuctural Portos here in New York for US.
Hyukang Lee Yang. Why the pressure on the profitability in particular.
Yeah, I think that's a really good question that all the investors are asking, and I think the sixty seven percent drop in that income pretty much shows how much profit or how much investment in the quick commerce or the AI are eating into the profitability of the company. I mean, the sixty percent drop, it was a huge miss and it was the smallest that income since early twenty twenty four, So it was the biggest disappointment to investors, that being.
A big disappointment even with these big Bowld ambitions to drive up the revenue side the profitability side by charging more for cloud, charging more for what is their AI agentic offering. How are we seeing sort of investors pay attention to this in the stop because at the moment, it's having its worst day since October of last year, at one point of April of last year.
Yeah, the stock was down almost like ten percent earlier this morning, and investors are apparently very unhappy. And I think, especially under this macro environment, investor tends to punish these statement unless they're accompanied by some clear profitability paths or the monetization plan and especially the competition the AI is
getting heated in China. Alibaba is facing steper competition from Tencent, and if you look at the share price in March, ten Cent has been performing a lot better than Ali Baba in the month of March, so you can see there how investors are more hopeful and more optimistic about
one company's AI plan than the others. So how it is going to profit rather than how much reavenue is going to be making in the coming years might be the bigger focus for investors going forward, if you want to, if you want to.
Guess about his future share price, you're currently I'm very pleased to say with us when it matters on all things in Asia.
Thank you very much.
Indeed, look, let's stick with the Alababa story and Jacob Cookies with us. Wpic CEO joining us and your expertise really has been in the e commerce in marketing areas within Asian businesses and companies looking to penetrate Asia, but you're also looking much more on AI adoption as well. First, with Alibaba, just general numbers, how worrying is it that their profitability is being so hit by competition.
It's not that worrying. I think this was very well announced.
They were telling us about the investments in twenty twenty five that were going into AI, and they've also announced that that's going to double in twenty twenty six, so that probably holds. In terms of pressure on income. Some of the numbers were pretty optimistic, and certainly in terms of the AI announcements and what they're doing with their token hub, you know, we see pretty good potential.
And again, I think this has.
Been part of a well announced plan that's in a couple of years of execution right now and seems to be relatively on schedule.
Okay, let's talk about what's on schedule when it comes to AI adoption and competing against say like a ten cent We were just hearing from yuk Lang about why ten cents high and many people feeling that the we chat optionality that it gives you, particularly with openclare and agentic AI penetration is just far clearer. How is Ali Barbar going to make sure that it's AI tech stat works.
Well, I think there's two things. One, quens open source model is incredible. I've got it running on my new M five laptop right now. And it actually is one of the best models that you can get running locally. So they've got incredible adoption there. And you talk about open claw, and then again, what are these agents and the agentic ai going to be doing. They're going to be buying token So opening up that token hub is great.
And let's not forget.
I mean, this was their tenth quarter of triple digit AI revenue, so when they're talking about these massive ambitious goals of one hundred billion in revenue, you know, I think there's a real clear path for them to get there. And you know, another thing that we notice in China as well is you know, they may not be.
Dealing with the top models or winning the.
Contest there, but the adoption rates among users even open Claw, you know, a little later adoption there, but now it's having more uses installations inside China than the US. So I think all of those adoption rates in China in general with the a the way the AI market looks like in the willingness of employers now to spend on tokens to enhance the productivity of their offices, I think is very positive and should lend to more triple growth.
You know, throughout twenty twenty six.
I mean, Jacob, you've always been this great read on the consumer over in Asia and China in particular. What's it like on the streets, the sort of fanfare, the hype around open tool.
It's not just on the streets, it's all over social media. It's pretty amazing. And the levels of innovation that are coming out, you know, they're using different models, it's running so much more locally than the installations did. But it's basically allowed on premise to come back to being an option. They don't need to, you know, have a very tech background to open up you know, a WUS or alien equivalence.
They can actually just go buy older Mac hardware that's readily available on the market and get some of these systems running up in their apartments or houses or offices. So that's really allowed it to go from being a maybe having a very technical background, to general acceptance among the masses.
Why Mac minnie is are flying off the shelves, why Raspberry Pie and the UK share prices spiraled higher, Jacob, what though of the pushback from a regulatory perspective. If you own Ali Baba, if you own Tenson or any of the Chinese giants, you know about regulatory risk. While we're starting to see regulators and state governments slightly pushed back against open claw just running rife andin state backed businesses.
Well, I would say that that would be the case. Put the government in terms of where we operate in Junk Sun Ju John just outside of Shanghai, there's been very supportive. There's even been local government classes, you know, for getting people to set these up, which so we've
never seen adoption from the governments like that. So I don't think that there's there might be some security with the open club, but let's not also forget Windshot launched Kimney claud too, So there's going to be alternatives that do roughly the same thing that might be just a little bit more suited to the Chinese market too as well, a little bit more firewall compatible as well.
Let's just therefore put this all in context, Jacob.
As people are as the brands that you work with are looking to access an Asian virus, people are looking to take on the various ways of getting to consume AS via chap. How does that landscape look for you right now?
Yeah, Well, I think in terms of AI in general,
It's made accessing Asia way more easy. The chatbots are probably not the best way to interact with your customers at that point, but being able to use you know, generative so you know, you're maybe taking model shots for China and being able to adjust them for Japan and Korea, not having to do three shoots for example, being able to change your languages, being able to interact with your customers in multiple languages throughout Asia has really kind of
laid economies of scale to being able to get access to the entire region and not looking at this like you know, maybe twelve different markets and more looking at it like just a couple AI is going to make that easier and easier. And inventory projections, you know, we're looking at social media now to see you know what you know might be selling really well in the summer.
All these things are really helping, you know, brands get access to the market at cheaper you know, or a less barrier, a smaller barrier to entry than they were doing free AI.
Who else should a potential market here be wanting to access Asia be looking at We talked Alibaba, Tencent, but there are plenty of other options.
Well, Kimmi ka is one of the best on premise models that exists out there too, and that's by Moonshot. So there's just a ton I mean, we have an office in Honjo too, and there's just AI companies all around us. I mean, Deep Seek's not too far from Ali Babadian's sandwiched in between them Red Notes in that neighborhood. So there's just a whole, you know, environment of innovation, very similar to what's been going on in Silicon Valley.
So you know, we don't know all of.
The total winners, but the barrier to entry again with AI, it's smaller and smaller. Deep Seak was launching you know, their models with only one hundred and twenty five people, so we would expect a lot more innovation and you know it's going to be an exciting year, that's for sure.
Play can cook wpic CEO. It's always great to catch up with you.
Thank you, Thank you.
Now coming up, the pressure on social media ramps up here in the US, California joining the chorus of localities calling for bands.
We're on that. Next. This is a blue megtech open AI.
When it's expanding, it's pusher to coding, the plans to acquire Astral, and that's a startup that makes Python tools for developers. The deal hasn't get closed, but the companies say that it will bring Astral's team into open AI's codex effort and broaden its suite of developer services.
Open Ai, of course, has.
Been racing with competitors like Anthonopic, Microsoft and startups like Cursor to appeal to corporate customers using AI as coding assistants.
Meanwhile, social media companies were they're facing increasing pressure.
From mounting lawsuits and laws dictating age limits on their platforms, in a particular blow to the industry. California home to Meta. Of course, TikTok operates there too, but it's home to Google Snap. It's joined the list of US states in countries globally proposing a ban on social media for children. Let's get the latest from Lumos Kirk Wagner, who's covering the industry, who's put out a tech in depth about this all today and really home Turf.
Could this be a reality?
I think it could be a reality. We're definitely seeing a snowball effect here. And Caroline, to the point you just made the fact that California is getting in on the mixture. I think is pretty significant. It's certainly symbolic of how big this issue has become and how strong this conversation has happened. Now, I want to be clear, a lot still has to happen for these sort of proposals,
these bills to become laws. But we are talking about a thing that has really been a little bit on the periphery for the last several years, and now we're seeing California, several countries in Europe, of course Australia start to float and push for these social media bands. So I think that's pretty significant.
As you say, a lot has to happen for the bands to come into force and hook, We've been waiting decades for any sort of talk to become a walk in terms of law on social media, and it hasn't occurred in the United States. But what's interesting is we are seeing more courts ruling or potentially about to rule on the addiction of these sorts of offerings as well. Can you bring us the latest and what's happening in LA and New Mexico.
Yeah, So, for the last several weeks in Los Angeles, meta and YouTube, So Instagram and YouTube have been facing these allegations that they are intentionally creating their products to be addictive, and right now the jury is deliberating whether that is in fact in the case. The jury has been deliberating for three and a half days at this point, so we're sort of on pins and needles waiting for
an answer. But this is really the first in thousands of similar cases accusing these platforms of creating addictive technology, and so it's meaningful and that it's going to sort of set the narrative around the rest of these cases that are coming later this year. And as you can imagine, Caroline, if these companies are deemed to be, you know, intentionally making their products addictive, that only adds more fuel to this fire for lawmakers who are already trying to rain them in.
Of course, companies such as Meta would dispute that sort of typecasting, and in many ways have put in place a lot of new safety tools for parents and.
Children to use.
But want to keep a track of Bloomberg's Kirk Wagner, we so appreciate you joining us now. Coming up, electric boat maker ARC looks to expand beyond the consumer market into work and actually defense vessels too. We'll talk to the CEO about its latest fundraise to do just that. This is bluebg Tech. Welcome back to bloom Beg Tech.
Let's check in on markets because they're intent to hooks, to put it mildly, as we see conflict seemingly intensifying in the Middle East, particularly involving the infrastructure of energy. That's why brencrud is up four point four percent.
We've seen it.
Spiral ever higher NASDAC and stock markets. Risk assets more generally have been on the downside. Particularly keep an eye on the bond market as we see yields and spiked on the two year here in the US, significantly in the UK. In Europe there is a concern about inflatiary pressure. Certainly, that's a lot of the questioning for all of the central bankers that we heard from overnight and into today. That's Japan, Europe, UK and the United States course have
been yesterday. Let's move into what's happening in terms of individual stocks.
That there is AI news as always.
And we're up by two point eight percent and Shao me this is as the company announces an update to its SU seven electric vehicle, but also released its latest ultificial intelligence models. Really, this refresh seems to be boring investor sentiment when the rest of the market have been down over there, So it's a key out performer.
Let's talk elsewhere in the EV.
Space, though, because another big development happening closer to home, Uber Rivian teaming up with the Uber set to invest to one point two five billion in Rivian's robotaxi fleet for more those consumer apps and gig economy report.
Natalie Lang joins US now and Wow, Uber has been busy.
This is what it's upteenth partnership that it's been announcing.
Yes, there's been more than twenty partnerships, and most of them struck in the past year, and today it's about Rivian. They're already committed three hundred million of that investment and they're going to buy ten thousand robotaxis to launch in
San Francisco and Miami starf in twenty twenty eight. So if it will take some time to ramp up, but they're starting to, you know, put together these arrangements to piece together this narrative that they can be a commercialization partner for these developers.
They're the platform, they're not on the competitive threat from these robotaxi made because they can just harness them. And then it's all about the infrastructure that's needed. When we look at perhaps what Rivian needs to get these on the road's charging facilities and cleaning, what happens.
Yeah, So there's a lot of sort of the unglamorous work that goes behind running fleets, the charging, the daily cleaning. What if people lost an item they have to retrieve if someone has to pick it up. And so Uber wants to be that sort of partner for these developers.
And what do we know about Rivian and the rto robotaxis.
So they're trying to add a new in house autonomy chip and a new lighter sensor into this new art vehicle starting in twenty twenty seven. So that's when you see production happening, and that's when Uber will probably make the purchases and ramp it up. This is really part of their broader ambitions to want to produce these consumer vehicles for individuals.
Hi Lang, it's great to have you, Thank you very much.
Indeed, on all things Uber and Rivian, we can stick with autonomous or at least electric and go from road to waters. Electric boat maker ARC has raised fifty million dollars and it aims to expound me on leisure vehicles, with tubboats, ferries, defence vehicles. Let's discuss this all with our CEO, Mitchley Mitch for the cynics among us, maybe someone would say, of course they've become a defense company. How much was this a pull in terms of demand, not a push from your vcs.
It was very much a pull for us from day one of the company. We've always had the vision that we wanted to go electrify and power everything on the water, and we are just doing that even faster than we expected because of the incredible demand we've seen across both the commercial and the defense sectors.
How will you use your technology that you've proved out in the consumers realm into the world of commercial I mean, already you've been.
Developing a tugboat.
How does that look like some sort of well case study that you bring into the defense realm.
Yeah, I'll start with electric propulsion is just a fundamentally better technology for the water than combustion. This is very similar to the transition that rail went through many decades ago, and now the marine industry is transitioning over to better
technology we are. We've developed that technology, we've hardened it in the consumer space with our consumer boats, we've scaled up production of that, we've built a team around that and the credibility that we know what we're doing, and now we're taking that same technology and deploying it into the commercial and defense sectors to electrify those fleets and make them more reliable, give them greater uptime, better performance, and better operating costs.
When we think about how much you will be making, I mean we see the boat, the consumer folks one you're building that, how much you're having to work in partnership with with you know, the clean key primes. Will you be building out an entire for them or is it more about just the propulsion and the engines.
What we specialize in is in that electric powertrain and basically the technology platform powering these vessels, so the motors and the battery packs, the wiring, the thermal controls, the firmware, software, telemetry, all of those systems. We do design the vessels around those powertrains as well, but we do that in partnership
with commercial operators with defense primes. So we are approaching this through the lens of partnerships where we bring what we do best, which is that electric powertrain, and partner with vessel manufacturers or shipyards boatyards.
We've just seen the vcs that are backing you on this latest round fifty million dollars Series C. A lot of them very focused on defense tech, very focused on American dynamism, to use a phrase. Has that been helpful in terms of the partnerships.
Absolutely, These are a fantastic group of investors. We are thrilled to have their support behind us as we continue to scale as a company. I mean Eclipse and Memlo, lower carbon, necessary interes and horror with These are amazing investors and we're lucky to have their support and their enthusiasm for what it is that we're doing.
What is is that you now need the fifty million? Is it about going to market? Is it about sales? Is it about still getting the technologists in house.
There's so much demand for the technology that we're bringing to market. This is very much a supply side problem for us. We need to ramp up our supply of the technology that we're building. That's increasing production of the consumer boats we're making. It's increasing production and the speed at which We're bringing the power trains to the commercial and defense markets as well. So this is really an accelerant on our bild to supply these sectors with this this electric technology.
I mean, they look really cool and our team has been on them before. How many of you made in the consumer realm? How many are already being contracted in the commercial and the defense realm?
Yeah, we are now making multiple boats a week out of our Los Angeles based factory on the consumer sides. On the smaller vessel side, we are under contract for one hundred and sixty million dollars worth of tugboats. And I can't disclose much more about this, but we are working with a major defense prime on autonomous surface vessels so basically drone ships out on the water Mitch.
Unfortunately, defense is front and center right now, and we think about a Middle Eastern conflict that does impact supply chains. How much are you able to keep the supply chain us and local? How much are we now self sufficient in the terms of electric vehicle batteries that you need.
Yeah, we are a Los Angeles based company. We have two hundred people working out of our headquarters here in Los Angeles. We pride ourselves on making things here in America. It certainly helps our supply chain that we do this locally. That doesn't mean that we are entirely insulated from disruption in supply chains across the world. But we've got a big jump start on this and have since we first started this company five years ago.
Oks Mitchley, it's great to catch up with you. Congratulations on the fundraise and the scaling of the business.
Now thank you for having me.
Coming up, we'll discuss the race for leadership between AI labs, what it means for the labor market to Ethan Choy of Coaslaventure's stick around as a BLUEBG Tech time now for Talking Tech and first up K Pop demon Hunters.
They're heading out on a world tour.
Netflix is seeking to capitalize on this most popular movie ever by planning live shows next year featuring songs from the film, head of a planned sequel. Plus, Samsung is devoting amount of capital seventy three billion dollars to seize the lead in AI semiconductors from Korean rival s k Heinex. Now Look, the tech giant is hiking investments twenty two
percent this year. Is it sees a surge in orders for high bandwidth memory and server grade storage, and Google is moving ahead with plans for a major data center in Michigan, and the project requires a company to cover
the full cost of adding new clean power. The deal is unusual, featuring a twenty year electricity contract for as much as one gig or what of power, but it may serve as a playbook for other firms desperate for electricity to power the AI boom, and discussions on the AI boom and productivity promised The impact dominated conversations by tech leaders and policymakers of late. But how far does the AI boom? I'd love to go? Our next guest
says infinitely. Ethan Joy, partner at Coast Adventures, joins us. Now, so where it's the first innings, how much more can actually the economy be able to support an infinite demand as it seems for compute and for the latest AI agent.
It's great to be here. Thanks for having me. It's great to be here, Thanks for having me. Look, we're just at.
The beginning stages. I say, we're not even in the first innings yet.
The biggest constraint today is data centers and gear watts of compute, but in terms of demand I viewed as infinite. We haven't even scratched the surface of robotics starting to infiltrate our workplaces, our homes, and we're just beginning to have agents in the enterprise that can do things across the enterprise at the intelligence level and at the speed of humans.
And so we're just at the very beginning, just at the very beginning of the pushback as well in many ways. And I'm interested actually as the first one that we push back against some of the valuations that we're seeing at the moment. Ethan I got to Sweetly Gradient yesterday, who'd just been raising a new fun talking about some of the numbers involved in the latest fun rounds. Darren Schure's just hey.
What he had to say.
We actually think that there is a bubble in certain parts of AI. There are bubbles. There is a bubble in foundational model companies, contenders to open Ai, contenders to Google, contenders to Xai, and Anthropic. When it comes to a lot of these other companies that are raising a billion dollars here or there, it seems as though those are a lot of venture capitalists or investors that want to chase the next open AI. But realistically the moat at
the model layer is just capital. How much money can you raise?
Cos was backed open Ai.
All people in VC still trying to chase investments right now? Are the rounds too high? Some of the foundational models not even the applications.
Look, it's it's clear that we have clear leaders at the LM level for the foundational models. And you know today here in the US it's really full companies Open Ai, Anthropic, XAI,
and Google, Gemini. But there are there's a lot of research being done novel model architectures, things like state space models, neurosymbolic models, math based models, and there is a lot of optimism that there'll be some kind of leap ahead of the transformer architecture that was so fundamental to chat, GPT and GPT three that calls.
This AI boom.
And so you know, to some degree it's justified that you know, vcs like us and others are going to back teams that are going to take a crack at building the next open AI and Anthropic And if there's a company that has that potential with a research team that has a caliber to produce a.
Novel model architecture. It's the kind of bets that we should be making as an industry.
Now.
Are the valuations too high? You could argue that, but we have to work backwards.
In some ways. These teams need to.
Hire very expensive researchers, they need they need to have giggle wats of computer to train models, and that's very, very expensive. And as you think about the cap table, we have to make sure that the founders and the team are incentivized and to some degree, if they're going to raise hundreds of millions or a billion dollars out of the gate, it means working backwards. If we're going to buy twenty percent as vcs, then the math kind.
Of takes care of itself in terms of the valuation.
So that's where we are today, somewhat justified, a little overvalued even.
Can Silicon Valley Adventure, American venture can develop world venture support the sort of numbers that are involved. If Middle Eastern money becomes harder to access at this moment.
Yes, Look, today there's a lot of debate about whether the model company should be profitable and self sustaining. I don't believe that's the right approach. We're just at the very beginning and there's a lot of R and D left to do.
There's a lot of business.
Model transformation that's happening today. Obviously with the SaaS pocalypse, fundamental existing business models are being disrupted heavily, and so it's a time where we are as an industry still figuring out what to do in terms of how to
fund all these investments. Today, at least globally, the foundational model companies have been able to raise capital, but at some point I do believe it becomes somewhat sovereign and at coastal ventures we have been investing behind themes and companies like Salvam and Sicanta in Japan where they're raising money and being funded by governments, and there is a question and we're seeing some of these play out with
a dow between open AI andthropic. Does the government step in and through either big contracts continue to fund some of the developments and the investments that need to be made into compute and to the R and D for the research models