Bloomberg Audio Studios, podcasts, radio news. Bloomberg Tech is alive from coast to coast, with Caroline Hide in New York and Edlavelow in San Francisco.
This is Bloomberg Tech Coming up. Meta finally finalizes a fourteen billion dollar investment in Scale AI and has the CEO amping up its AGI efforts. Plus Adobe sales outlook. It tops estimates, but it's not enough to quell investors concern that it's losing the AI race, and reports of stable coin plans by the world's biggest retailers puts pressure on Visa a MasterCard. We discussed with the CEO of crypto firm Circle, but first we check in on what
the markets are doing today. There is a clear geopolitical risk that is fund and center for investors. We think about what the Israeli A tacks on Iran means for the search for safety, and we do. Then that's that one hundred off by eight tenths of a percent. There are idiosyncratic news within the benchmark, but Bitcoin not your savior of choice at the moment, not the area to go for safety. We're currently off by nine tenths percent on bitcoin move on and have a look at some
of the individual movers. Look, we've got to get into, in particular some tech news that is all about Meta, another aquahier, but this one costing more than fourteen billion dollars. We're off by five tens of percent, but generally the market likes the move, the spending of Meta on the future of AGI. Let's get to it all with Kurt Wagner,
who's been breaking this story. Set by step. You brought us the idea that, of course Alexander Wang was going to be part of this crucial team that Mark Zuckerberg is forming. This does seem to be yet another example of an aquahier by a big tech company.
It does because, as you mentioned, Caroline, there is that fourteen point three billion dollar investment. But Alexander Wang is joining Meta, so are some of the other employees from Scale Lai. So while Meta is obviously, you know, now a forty nine percent stakeholder, and Scale it's taking the CEO over to Meta, So a lot of people are sort of viewing this as a very expensive aquahier. Obviously, there's still Scale is still going to be operating independently.
It's going to have an interim CEO. But you know, given the talent that is moving from Scale to Meta, this is this is a very expensive talent grab. And this is something that I think we've seen from Mark Zuckerberg before and something that he is clearly very very focused on AI right now.
We've seen it from other key tech companies. This aquahar idea, the fact that maybe you skirt any regulatory oversight, particularly when perhaps metas in frontline. When it comes to the FTC investigation, You're thinking Microsoft with inflection, You're thinking Amazon with a debt to thinking of Google with character AI Kert, What does this show about the anxiety that Mark Zuckerberg has right now.
Well, part of the momentum here is because he felt like they are behind, right. We had a story a couple of days ago about him building this new super intelligence AI team and that was really motivated by the fact that their lamafour launch back in April was underwhelming and that he felt that Meta was not keeping pace in the way that they should. So I think there's
a real paranoia there. This is, you know, a sort of a classic Mark Stuck Colt Bring move for him to just get incredibly in the weeds on something that he feels his business is sort of lacking. But I also think it just speaks to the you know, the limit in terms of how much talent there is out there in the AI space, like really good talent. Like the fact that they're willing to spend this much on these on these folks sort of says that it's a
very competitive landscape when it comes to hiring people. And I think we've seen that across the board, and it's only going to get tighter now that everyone sort of picked their various startup that they're backing here.
It's interesting that we've already heard from Open Ai saying they're going to stick working with Scaleai, which is now going to be helmed by Jason Droges. He was previously the strategy officer, now him becomes a CEO. But talk to us about what Alexander wanglings in particular, He's got a very good way of networking, particularly on Capitol Hill.
Yeah, so, you know, I've been told by several folks over the last week that Mark Zuckerberg just thinks that Alexander Wang is the best guy. You know, he thinks he is so talented, He's he's super bullish on this guy. I believe he's just twenty eight years old, and you're right, Caroline, he brings a great network. Everyone I've talked to is that this guy knows everybody in the AI space. He also,
as you point out, knows folks on Capitol Hill. This is a time when I think Meta and Mark Zuckerberg in particular are trying to build relationships with the Trump administration. You know, we've seen them sort of announced plans to
partner with Andreil on potential military tech. So it's a bit of a new, you know, direction for this company trying to get into perhaps government contracts, and I think Alexander Wang brings some of that connectivity to Meta and Mark Zuckerberg, and so in addition to the AI experience, you know, this is also someone who might be able to help him out relationship.
Well. So far, investors have trusted the steerage of Mark Zuckerberg. Just a significant out performance of eighteen percent so far on the year Blue Most Kurt Wagner, thanks so much on all things Meta. Let's bring it out. Michael Reynolds is with US vice president and investment strategy at Glen Mead, manages forty five point six billion dollars in assets and just focus for a moment on the AI trade and this rush to spend by the hyperscalis of which Meta is now one. What do you make of this?
Aquaha, thanks for having me.
So we think without a doubt AI is transformational, and anyone who disagrees with that just probably hasn't been using it properly. But I think we're cautioning investors on our end to think through the connection between a transformational technology and what it actually leads to in terms of investment opportunities. History is littered with examples of great technology that just was not a good investment. Gutenberg invented the printing press,
but he died in poverty. Rare Road bonds were notoriously poor investments. We're just not seeing a lot of pure plays in AI right now in the public markets. But perhaps if investors are turning to the private markets, that may be a little bit more plentiful with opportunities for AI as there's certainly going to be winners and losers in this space, but we're certainly in early stages.
Okay, So how are investors getting access to the private markets on this particular area. Are they looking more to access well VC liquidity in some way.
It would think it's a combination of trying to get that direct exposure with those that are actually building the software and delivering the AI tools that the end user is actually using, but also some of sort of the ancillary providers that can help on the margins, whether it's delivering the hardware on the tech or it's maybe the data centers that are delivering the energy that can help
facilitate all of this. That the best solution is probably a combination of the two, as you don't want to sort of put all your bets on one side of a trade here and end up missing some of the bigger beneficiaries that we're perhaps a little long intuitive as you were thinking through it in real time, Mike.
So's sticking with public markets, you're seeing people trying to bet on whether it be the energy stack, whether it be the AI hardware or the software stack. From an equity perspective, or they're going into the bob market.
Too, a little bit of both.
I think the equity markets are probably a little sexier from that perspective. Data centers have been really great performers over the last couple of years in expectation that there's just going to be a big energy sink to be able to use these models at scale. We've seen a little bit of pushback on that, given some of the opportunities we've seen coming out of Deep Seek where they were pay able to perhaps do things a little more
efficiently and pull back on some of those expectations. But it's probably a little bit of both on both sides of the capital markets, stocks and bonds, where there's going to be opportunities here driven by technological innovation.
I like that you bring that up, Mike, because it actually was quite recently that we were all panicking around actually whether you needed the scale of investment, you needed a scale of infrastructure because of Deep Seat. Well that's almost been put to bed. You look at Oracles numbers this week. You think about the anxiety around supply. Really, that's what the nature of the story has been, people can't build it fast enough. Is that where we're back to in terms of the market psyche.
To some extent, I think over the last year or so, perhaps pre Deep Seek, there was a lot of concerns that just wasn't enough energy to go around, that some of these data centers were going to have to build out alternative energy sources, maybe flirting with nuclear, but it's really I think it falls into the historical precedence where when you have a breakthrough technology, it often seems like there's a lot of startup costs to really break through
and get the technology working at scale. But innovation is a beautiful thing and finding things, finding ways to do things more effectively and efficiently always just seems to occur. I think that's happening in real time this year, and we could continue to expect that to unfold over the years ahead, just finding better ways to build it and building a better mousetrap.
Okay, what about the mousetop? Is it all about the United States? Are you looking at other opportunities? Are your investors, the people that you're speaking to, wanting to get out into Europe and Asia?
Yeah?
There was certainly this theory that US exceptionalism in Ai was, you know, an uncrackable theme, and I think there are some sort of fissures into that theme this year, where again you had the deep seek news out of China. You have several other countries that are really doing a lot to incentivize domestic development of these types of technologies. It's happening in the UAE and European Union, many parts
of Asia. They're trying to build out this technology. So it's you know, is the US going to have a monopoly on this technology for the foreseeable future remains unseen, But you know there's quite a few players that are emerging in this market that could end up sharing market share on this at.
G forty two making inroads into Europe and the UK will discuss that later in the show. You've had Jensen Wang over there in the United Kingdom and Europe making deals. Mike iland it here. Therefore, because we've gone global, and day is a global day in nature, how people are feeling about potential increased tensions in the Middle East, maybe even more, How are your investors responding to that at this moment.
So we're cautioning our investors not to react too quickly to unfolding geopolitical conflict here so far, I think we're following a pretty standard geopolitical blueprint where markets are reacting to the uncertainty that just didn't exist a few hours ago. But in general, markets tend to see through this type of thing, especially for those countries that are not directly involved this time around. I think inflation is going to
be really important to watch here. If energy infrastructure in the region starts to come in the crosshairs figuratively or perhaps even literally speaking, that could have potential inflationary impacts for energy, which is just not just an input cost for a direct input costs for consumers, but also for a lot of the goods they consume also consume energy, So it's we're looking at that pretty close. As well as the shipping lanes throughout that region, which can be
pretty important for global trade. Those are going to be really important to watch over the weekend.
Michael Raynolds, we hope you get some rest over the weekend, Vice President of the Investment Strategy at Glen Mead. We appreciate your time. Now let's discuss that key story, Israel launching multiple strikes across Iran, targeting nuclear facilities, killing top military officials. President Trump urging Iran to accepting nuclear deal in an effort to avoid further attacks. Iran vowing to continue its nuclear program despite the Israeli strikes. For more,
we turn to Bloombogs. Jomana Vassetchi joins us. Now actually based in London just for this week and Jomana you're usually in the Middle East, And it's interesting we're getting a US official telling Bloomberg News that indeed we are still seeing talks lightly to be conducted between the United States in Iran.
Yes, perhaps there is a potential for a de escalation here, especially as those sixth round of talks between US and Iran we're slated to take place in Oman this weekends. Giving the events over the last twenty four hours, at the the eventual possibility of those happening was thrown into doubt. But of course what we have seen, just as a recap is the most significant attack on Iran going back to the eighties.
What Israel did in the.
Last twenty four hours supersedes all of the events that happened last year, both back in April and in October, and really constitutes a major blow to Iran, not just in terms of the fact that Israel went after specific nuclear facilities and even though no radiation has actually been recorded from those nuclear facilities, what they did succeed at doing is that taking out some senior IRGC. So these are Revolutionary Guard commanders, the chief of staff of the
Armed forces. Some of these people are key masterminds behind the attacks that had been conducted on Israel and the last year, but also on the IRANQO facilities in Saudi Arabia back in twenty nineteen. So some of the key players have been taken out with these strikes. Now where we go from here. Iran has vowed a harsh retaliation already this morning they fired over one hundred drones directed towards Israel.
Those were mainly intercepted.
But the question from here onwards is how many more ballistic missiles they're looking to send towards Israel, whether or not this will encapsulate other proxies in the region as well, whether the likes of Hezbalahthi's Hamas may get involved in the retaliation, and then also whether or not the international community look to de escalate from this point onwards. And to your point, US plays a significant role here because of the possibility of a furthering of those nuclear discussions.
And let's see whether Iran actually does come to the table in good faith.
Jermanavasseatchi, we really appreciate your time today. Thank you. Now coming up, we go back to tech. Apple sets a new target for revamp Siri or on that next this is pretty bad Tech said to be targeting a new internal date of spring twenty twenty six where it's revamped SyRI upgrade of the multiple delays. Now that's according to sources. Let's get to our own Mark German for more. It was but this week that WWDC left us hanging on Siri and you bring us some clarity.
That's exactly right. So they did not talk about serie features pretty much at all at WWDC other than to say they're still delayed and they'll be arriving in the coming year. So I looked into it, and the new launch timeframe is likely going to be as part of what will be known as iOS twenty six point four, and that's currently scheduled for the springtime next year, so around March April, and this will deliver the delayed serie features.
These are the features e to an AI overhaul for the voice assistant that we're announced back in June of twenty twenty four. So the three main features are the ability to tap into your personal data to fulfill serie queries. So if I asked for a question such as what was the big I sent Caroline last week. It would be able to pull up that video or a podcast or a song. If I wanted to get analysis of something on my screen using Siri, it could do that
as well. And it has precise control of applications. So for instance, I could say, pull up this photo of Caroline and edit it and send it to her as well as a few other people in our group chat. Right. Yeah, those are impressive and big features, all delayed now coming about nine months from now.
I'm waiting for that photo Mark put it on our group chat. But Mark, I'm also just remind us what the problem has been here? Why has it been so difficult to integrate with Siri?
Apple has been facing some serious engineering and management problems. But just to break it down and not to get too technical, essentially, Apple had to split Siri into two brains, right. The first brain, the more basic one, has to do with features like setting alarms, phone calls, all the legacy stuff.
Then they built a second brain, more based on large language models at the core of modern AI, for those more advanced features I talked about, right, But making one SERI that has two brains combined to do two different set of actions had compatibility problems in real world use, and I'm told it didn't work properly up to a third of the time, which is a lot based on
how much usage these things are going to get. And so what Apple is doing now is rebuilding the underlying serie architecture to have one brain based on that newer generation technology to handle everything, and that would eliminate the compatibility issues, but obviously is going to take more time to put.
Together brilliant storytelling. Marcum and thank you so much for the breaking news as well. But let's turn our attention to Adobe. Now shares under pressure after the company actually gave a sales outlook that was pretty good, but it failed to come invest as skeptical about its ability to really compete with the AI focused startups. Whimberg's Brady Ford joins us now and look shares are off by five percent worst days it's March twenty twenty five. Even though they've beaten they.
Raised, right. I mean, so Adobe is this kind of poster child of a company which was the kind of leader of their industry for twenty years. Nobody could touch them, And the question is okay AI's coming. A lot of these apps are much.
More lightweight, cheaper, easier to use.
Does Adobe stay relevant That's a big question, and investors are not convinced. I mean, I think they're down ten percent over the last year Adobe's stock and down further today. And it's all because of those AI disruption anxieties.
You're thinking mid journey for example, we're thinking even Soora making roads when it comes to making the video side of the equation. But Firefly has it been getting some use? What was it like two point four billion content stayed over the course of it's being used out there in the wild. It's been getting used.
But when I was growing up, if you wanted to be like a cool creative person, you had to go find a copy of photoshops somewhere Adobe Premier's video editing. It's not that way anymore. If you talk to people who are young, who are really kind of making cutting edge, aren't there using the like these mid journeys Google's vo I mean, when it comes to the enterprise level where you have a CIO who's worried about cybersecurity, they're still
buying Adobe. But the question is that next generation, where is the mind share going.
What was also interesting is they tried to win the IP battle. I'm trying and say, look, we've got you covered. You don't need to be worried. We'll even pay out if we find that any of the images you're creating are based on some other IP. But people aren't worried about that so much anymore.
It's an interesting question. I mean, we saw Disney suing mid Journey a couple of days ago. I'm sure the CIO of our company or others is not gonna go and pay for a company getting sued by Disney. But at this current moment, it's still unsettled. It appears that the anxieties around hey, scraping data to build models, is that gonna be okay? It seems like a lot of buyers at this point are swayed.
Against it, like I'm more convincing to do Brinberg's berdie Ford brilliant as always, Thank you. AMD CEO Lisa Soon said her company's latest day I processes can challenge in video chips in a market she expects to saw past five hundred billion dollars in the next three years now. This is as the company aims to play catch up with in video in AI accelerators with its new products. Let's talk about them. Bring meg Intelligence analyst Contrins Savani's
hey with us. You're at the event, you're assessing what's been discussed the Mi three fifty. Will it be fast enough to stop winning more market share?
I mean, if you look at their event and announcement in isolation, we think there were many positives. One the customer adoption is broadening some of the new names like Xai, open Ai, Oracle. The deal with Oracle, which we think will be the largest three to fifty five deals so far yet for and will show results in the second half. These were really surprisingly positive. Remember these have been traditionally and VIDIA only houses, so them getting to these customers
is a pretty big deal. They showed continued progress on the hardware with the announcement of the roadmap with the four hundred Elio server rack, and it's more important the software, which they have been lagging until as of last year. So a what a lot of impressive progress on all aspects. But the challenge is the competition also continues to speed ahead, and they have had a headstart.
They have and boy does everyone just relentlessly compare them to Jensen's powerhouse. You say on your note that really that MI four hundred ramp up in twenty twenty six is gonna be the key inflection point for potential share gain in a way, though, do they need it, Kenjym because in many ways the issue for Nvidia is a supply side issue. They can't make them fast enough. So is there still so much market for Lisa's who to.
Take there is?
And why that's a inflection point is when you think about the current competition or the playground, right, it is all about server level rack solution, not about chips anymore. AMD is still not there yet. So the Helios RAC will be their first iteration using the ZT acquisition talent that they acquire and expertise where they can really compete head to head with Nvidia, assuming and Media doesn't move forward another year from there. But that is what they
really need. What they're really missing is a big cluster level server RACKCHI view solution.
What's interesting is are they all still fighting for the training side of the equation. We've all just been dominated about the inference and there are some really powerful scartups coming into that space trying to take marketshap too.
I think they're leading by competing in inference because that is where they have the edge given their higher or I would say apples to apples, higher memory bandwidth and efficiency there and a lower TCO. So they're getting into the customer doors through inference, and once they're proven and starting to gain a little bit more share, then they're showing up like, hey you can still try us for training, and you know that's how they're trying to increase a
little bit of a share in training. We still like the inference play. Remember, going forward, inference we expect to grow faster than training. Yes, training will be a big chunk still, but there will be a lot more broader customer based for influence.
Bloomberg Intelligence analyst Savani So great to get your analysis. Thank you, Welcome back to Bloomberg Tech. I'm Caroline Hide in New York. Let's get a quick check on these markets. But we are consumed by geopolitics today, anxiety around the Middle East after Israel's track on Iran. We see the NASA that one hundred off by six ten percent, as we see that risker version play in, but not for
all names. Oracle up five percent. It is just managing to ramp up higher a new record high after they posted earnings earlier in the week saying that they are going to be getting seventy percent gaining cloud infrastructure sales in this fiscal year. Adobe is down on its earnings. We discussed it with Brodie Ford of by five percent, not enough to quell anxiety that it won't be an AI winner. Move on, have a look at what's happening
in broader assets, though. I want to shine a light on what's happening in the world of crypto because Bitcoin, look isn't actually playing out as a haven today so much. It's coming off of its lows. Look, we're still one hundred and five pound, but it is off by three tens percent as we sell risk. But let's just talk more broadly about the movement within crypto and stable coins in particular. This comes as a Senate nears the passage of the Genius Act, which would regulate integrating stable coins
more into mainstream finance. I want to shine out, like what's happening with these names Visa and MasterCard currently off by five percent four percent because reports coming out of the Wall Street Journal that look, big players Amazon Walmart are studying their own stable coins. Does this cut out the middlemen, the card providers thus far, and of course we're going to speak to one beneficiary of all of this legislation, all of the move towards stable coins and
digital payments. It's Circle Circle. CEO Jeremy Elaire joins me now on set this report that potentially we see Amazon Walmart already really galvanizing themselves to have their own stable coins. How quickly could that up end Visa and MasterCard.
Well, look, I think if you're a major technology company, or your major commerce firm, or you're a major financial institution, when you get legal certainty that stable coin money is a new form of electronic money that is available to the global financial system, you're going to pay attention. And obviously you're going to look at the technology, the progress
of the technology as well. And I think the conclusion that a lot of major firms are reaching is that this is the new money layer of the Internet, and it provides tremendous technology to innovate beyond.
What we've been able to innovate with money.
Before, and it offers tremendous benefits ultimately to these firms bottom lines. And so it's a it's a technology that's ready and it's you know, a technology with the Genius Act which you refer to, I think, which gets the legal certainty to really utilize this at scale. And so from my perspective, first of all, it's not surprising at all, and I think it's a it's a tremendous opportunity as the world connects to UH to this new form of currency on the Internet.
Is it not surprising a tool because they're coming to you to help build the infrastructure there, how much would USDC or indeed what circle does be what Amazon and Walmart eventually adopt.
We see tremendous opportunities to collaborate with major technology firms, major payments companies, major financial institutions, and we already do. We work with some of the biggest household names. I mean, even just yesterday we saw Shopify announce that they're making USDC payments a default option automatically for every merchant on their platform in the US and then in Europe.
That's like a million merchants, so.
Like a PayPal button, and then the USDC people will be there by default, and in fact, not only that they're offering, they'll be offering merchants half a percent cash back to the merchant.
So they're going to actually compensate the merchant, which is incredible. And so I think you're seeing platforms like this adopted USDC and so you know, as the operator of the world's largest regulated stable coin network in the world, we want to you know, we're really a market neutral infrastructure. We don't compete for consumers or merchants or businesses. We want to work with everybody to take advantage of this breakthrough technology innovation.
The breakthrough comes with the regulatory support. As you say, Genius Act. When do you expect it to come into play?
Well, yesterday the Genius Act passed a critical cloture vote with a super majority bipartisan, and so that sets it on a glide path to pass the Senate on Tuesday.
Of next week.
We actually previously did hear from Tim Scott. Of course he's been leading the charge in many ways on crypto more broadly, but he's been speaking to Bloomberg about some of the opportunities around stable coin and leading the charge here in the United States. As of course a key representative just take a listen.
With a bipartisan genius ad we can do more than just pass a bill. We can deliver results for the American people. We can bring clarity for a sector that's been clouded by uncertainty, and we can make it known the United States will lead, not follow, in the digital asset revolution.
Chairman of the Senate Banking Committee, an important voice. What's interesting with Tim Scott is now he's also pushing us towards where we go with broader crypto adoption. Yes, we get the Genius Act, but then what of the Clarity Act as it's called. How much further would that take your industry?
Well, look, there's this incredible innovation of blockchains, there's this incredible innovation of digital assets, digital tokens, and we need legal clarity on what these are, how you can offer them, how they can be used, the platforms that where they need to be registered or regulated, and where you know, there's just very clear rules for issuers. So I think the Clarity Act is essential. I think it is a
very strong piece of legend. I know there's more work to do on that ultimately in the Senate, but I think the combination of both the Genius Act and the Clarity Act ultimately creates an incredible environment for a rules based system in the United States a competitive environment, and I believe does accomplish the objective of making the United States sort of the capital of this new form of technology innovation.
It's a great opportunity for American companies as well.
Briefly, it's been good timing to go public. Your shares absolutely soored since they first started training last week. How does it feel.
I mean, look, we're very excited to be a public company. It's critical from our perspective to just enhance the trust, transparency, compliance, good governance that we're known for. And I think in particular at this moment when we're getting this legal clarity around the world on stable coin money, and in particular, as you said, these major companies are looking at this. We think that being a public company will serve us well in.
Working with the leading institutions of the world.
So from that perspective, you know, we're really pleased and excited to build tremendous partnerships around the world.
Oh and you sink some of those partnerships do come back on to talk all about it circle CEO Jeremy Elaire, we appreciate it coming up. Felisa's raises nine hundred million dollars in its tenth round. This is a fund of course, and Felice's managing partner, Sunday p Choo is going to be joining us next as the Bloomberg Tech. It is
time now for talking tech first up. Chinese authorities have released draft guidelines on the transmission of car data from overseas vehicles look as a move that could benefit Tesla and it's autonomous driving features. In marks the first time Beijing has really clarified its stance on data generated and exported from within China. Plus Abu Dhabi's G forty two when it's planning to launch a European unit G forty two Europe and UK, but it's set to deploy a
solutions for the private sector, partnering with governments too. Now. The announcement comes as in video CEO does and one forecasts a tenfold increase in Europe's AI computing capacity over the next two years. And GameStop Welz says the company's future is in training cards, saying the likes of pokemonon sports cards aligned with the company's heritage, maybe not gaming
game stops. Collectible's business made up twenty nine percent revenue in the first quarter of the company planning to add two hundred and eighty more stores that will offer training card evaluation services. Right, it's time for today's VC spotlight. Now and Felisa's has announced a nine hundred million dollar
tenth fund, the fund's largest ever to date. Let's bring in Sunday Peache is Felie's managing partner joining us now and Sunday just tell us who the LPs are, what they're excited about you allocating.
Towards HI carline. It's wonderful to be here. It's it's a very exciting week for us. It's we've raised nine hundred million, our largest fund yet from mostly US institutions that we're super proud to take money from. And these are hospital systems, museums, university endowments that are all doing like great things in the world, and we're proud to be in business with them, and we're proud to like multiply their money and you know, have it, you know, go back to some great causes.
I'm just really interested sending more broadly in where you therefore put the money, because at the moment it sounds as though you're going to be in general to AI more for Throttle. It's going to be the infrastructure air but that feels like a very busy, very crowded space. Who are the companies are depending on?
Yeah, I mean just to give you know, take a step back here. Felicius was actually born in twenty tens. Our very first institutional fund was just at the beginning
of the cloud and the mobile tech supercycle. And you know, we've been very fortunate that we found some like amazing, amazing businesses as part of that household names today like a shop of I add in, a Canva and a notion and so we're now at another exciting text supercycle in AI, you know, just in the first or second innings, if you will, And so there's a lot of excitemente In the past two years, two thirds of our investments
have been AI native companies. Most of our investments are like usually first or second money in so we're playing very very much into the AI supercycle. Household names hopefully like the canvas of the world ten years from now that go public us both the infrastructure stack as well as the application stack.
It's interesting. We had Meno Ventures partner on yesterday and Sean was telling us that in certain areas the funding just seems extraordinary. The valuations once again almost fail a kin to twenty twenty one. Is that what's happening right now? How are you getting in at a decent price point?
Yeah?
Look, I think valuations, you know, continue to be market driven in many ways. And we're not the only ones excited, BODYI But if you take a step back, the reason why everybody's excited, aboudyie is that even like you look at like the broad arc of tech history, what tech is really good at is providing us products and services that are previously expensive or like inaccessible except to like a very few, and bringing it all the way down tomarketizing that you know, to the masses if you will.
And you know a great examples of these and the last tech cycle where companies like Canba that allowed everybody to design. Now over two hundred million people use Canba to design. You know, you have something like a Shopify, anybody can become an entrepreneur and millions have become entrepreneurs because of Shopify. An Uber gave us all private driver and Airbnb gave us all a vacation rental. And so you know, you take that forward and you think about
like what AI can do for us. It's you know, every household I mean, think about how many times we keep talking about, you know, I wish I had like a private tutor for my kids. I wish I had, you know, a therapist like and they're so expensive. Every
session is expensive. They're hard to find. Think about all of the labor that kind of goes into household services or if you move that to businesses, you have you know, things that are available to senior management, like executive assistant order, an executive coach, that are not available to everybody else. And EI is expanding the surface of software so drastically. And if you think think about this, Caroline, just one to two percent of world GDP currently spent on software.
And EI is so radically restructuring the surface area of software. And you can be an AI pessimist or optimist, and those numbers in ten years are going to be different for each one of us, but none of us are going to disagree that that number is going to move up drastically.
Something very briefly like you have had some superb exits and think a shop of fire Addy, and you are in some of the hottest names. When I think about Runway and in dcanva. How much pressure have you seen from LPs being like, we need the liquidity now. How quickly do you think these doors can remain open? Briefly?
I think I think you know, for the very very best companies you think about a canva, like their shares can trade every single day. They don't have to even raise any primary money. They've been so profitable for so long that in the secondary markets you always have the ability to exit if you want. And we've selectively chosen that for our LPs and giving them back like some distribution, and so we've done that like an honor. Many other funds have done that, so I'm not worried about that.
I think one of the things you know, think about is that given this new AI tech cycle, if you assume that the software as a percentage of GDP doubles, let's say conservatively, that's hundreds of billions of dollars of new spend that is going to get allocated to new products and services that are AI. Our job is to basically go and find those outlier businesses for our LPs and find the next Shopify and canby and add and you know before they're obvious.
Sname Petru thanks for bringing it. Foreleases on the new tenth fund Media LGBTQ plus dating app Grinder. As popularity is growing globally, averaging fourteen and a half million monthly active users. Shares with the company are up thirty three percent year to date. We spoke with the Grinder CEO, George Arrison about the role AI is playing in that growth and how they handle privacy in the age of AI.
Grinder privacies by farther paramount thing because we have users in countries where it's illegal to be gay, or in areas where while it might be legal, you still can't really be yourself, and so protecting them is really really crucial. We don't ask for a lot of information from users when they join the app, we don't require a photo, We don't try to be as minimal as we can in that regard, again, because privacy is so important to users.
And from there, the other really critical thing for us is to be able to keep all the data on Grinder, So we work with Amazon's Bedrock service, which is a virtual private space that allows us to actually keep everything in our end. Our data does not kind of stay with anybody else, and our data does not train anybody
else's models, and that's been really helpful obviously. And the other thing we're now starting to do is actually take open source models and put them into our system and be able to train them to make them be a little more gay. And that's also quite interesting because that's kind of pushing us closer to the frontier of what's happening in THEI. But no one else is doing anything like that, and we think we can provide some really useful insights for our users.
Through that grinder. CEO George Arrison. There, let's just talk about New York State as well. It is begun asking companies to disclose when artificial intelligence contributes to mass layoffs. It's a symbolic move as officials look to possibly regulating AI. The technology's impact on the workforce could have an uneven effect between men and women should be roused with us. She is the founderacy of up level. It's a platform AI platform designed to accelerate women in the workplace. Shuby
is also an engineer, former tech executive. Yourself, Shuby, is there going to be an unequal impact of generative AI's use on the workforce?
Do you think, well, when you look at what is happening in the world of AI and where the applications, at least today are. It's about automating jobs that are largely held by women. It just so happens that those jobs tend to be more administrative in nature and are perfect and right for automation, And so it will be really interesting to see what the data shows, especially if
this does become legislation in New York. It is how it impacts women and how disproportionately that impact then creates further gender inequities.
How is your technology perhaps trying to play a counterforce here.
Yeah, you know, if you take a step back and think about today how the models are being trained from a generative AA perspective, Those large models, whether it's the Opening models or is Gemini or Cloud, have been trained largely from data of the Internet.
And when you look at what is the data on the Internet, large spats of data of course are Wikipedia, Reddit, GitHub, you know, popular pop gender data sets, which tend to be highly male dominant. Then you look at the other data sets, whether that's related to academic papers. You know, publications are on business books and use cases healthcare where again, you know, up until recently, women were not even included in those clinical data sets. Really there's not a lot
around caregiving unpaid labor. Right, so when you look at what that data set highly represents or over represents men than it does women, so you can so so that is the basis of the training data sets for majority of the models today.
So the counter of course for Uplift, what we're trying to do at up level is create the world's largest gender data set so that this gender data set actually represents women both contextually from an experienced perspective, understands women's challenges around career, health, wealth, and more importantly, can be that counter force whereby you know, in theory, someday we could blend both those data sets, then then we'll have that gender equality across the broader training sets.
Is really interesting. We you're hearing from the Grinder CEO saying they're training their large language models to be more gay from the perspective of maybe you start to counteract the bias within the applications that come from the large language models. But what about just unsaid biases about women and actually adopting this technology. Are we being first movers, are people being early adopters in the workforce, do you think at the same level as men.
Yeah, you know, this is a disturbing number that you know, I keep seeing where it feels like the gap was maybe twenty percentage points between men adopting. And I'm talking about generative AI, not just AI broadly, but just even generative AI tools versus women. Part of that has to do with, you know, if you look at broader adoption of women in terms of big tech, yes, women might be on those platforms, but are not merely as active in terms of posting, sharing comments, reacting than men are.
And so that'd be and why is that?
Because you know, when you look at a lot of these platforms and there is has been a mass exodus of women, is because of the misogyny, the nursism, you know, the sexualization, etc.
So there is this inherent issue that we have around trusting tech and that's bleeding itself into also using these generative AI tools. So a big part of that is just can I trust this or if I use this tool, then can I actually.
Say this is my work?
You know?
So there is that bigger issue. So there have been impediments to adoption for women.
In thirty seconds should be is it getting better.
No, it's not getting better, it's actually widening. So or is it differently, Maybe women are still at the same rate of adoption and men are adopting it faster.
More work to do, more reason to see up Level going in there and building these connections with big companies should be raw. So good to have you, the up Level founder and see how we appreciate it and that does it. From this edition in Bloomberg Tech, do not forget to check out our podcast. You can find it on the terminal as well as online on Apple, Spotify, and iHeart. From New York. This is Bloomberg Tech.