UK's CMA on the Microsoft Deal and Big Tech's AI Safeguards - podcast episode cover

UK's CMA on the Microsoft Deal and Big Tech's AI Safeguards

Jul 21, 202338 min
--:--
--:--
Download Metacast podcast app
Listen to this episode in Metacast mobile app
Don't just listen to podcasts. Learn from them with transcripts, summaries, and chapters for every episode. Skim, search, and bookmark insights. Learn more

Episode description

Bloomberg's Caroline Hyde breaks down the latest on Microsoft and Activision as UK regulators reject claims they're bowing to pressure to clear the deal. Plus, Big Tech commits to safeguards for AI at the White House's request.

See omnystudio.com/listener for privacy information.

Transcript

Speaker 1

From the heart where innovation, money and power Collie in Silicon Valley, NBN. This is Bloomberg Technology with Caroline Hyde.

Speaker 2

And Ed lud Love.

Speaker 3

And Caroline Heind and Bloomberg's well head quarters in New York and Ludlow He's off.

Speaker 4

This's in Bloomberg Technology.

Speaker 3

Coming up, we'll have the latest on Microsoft and Activision as UK regulators reject those claims they're.

Speaker 4

Bowing to pressure to clear the deal or ahead.

Speaker 3

Thus we'll talk ai regulation of big tech commits to safeguards for the technology that's at the White House's request. We'll bring you the details and bankrupt Crypto Exchange FTX sues San Bankman Freedom executives over one million dollars in bad deals, including a plan.

Speaker 4

To buy an island nation.

Speaker 3

We'll have more on the new laysuit and so much more throughout the hour. Meanwhile, let's go across the pond in the UK. Britain's top ad trust enforcerers pushing back against those claims that have been out there that it was kind of forced by outside pressure to reconsider its veto of the Microsoft Activision Blizzard deal. Here's the CMA CEO, Sarah Kuddle at the moment our decision on prohibition stands.

We investigated very thoroughly the deal. We found that there were serious concerns in relation to cloud gaming and.

Speaker 4

The proposals that were put forward.

Speaker 3

By Microsoft at the time did not resolve our competition concerns. So Microsoft have indicated that they are considering how they might restructure the deal in a way that resolves our concerns. And the regulator had found itself increasingly isolated, pretty much the loan watchdog standing in the way of the deal.

Speaker 4

Now, after a week of drama.

Speaker 3

The acquisition has some pretty fresh momentum in the US. For example, the FDC has just paused its in house trial against Microsoft, leaving room for potential settlement talks.

Speaker 4

Let's break it all down.

Speaker 3

Bloomberg's Catherine Gammel is with us and what else did the Competition and Market Authority have to say? The CEO about this deal and how it's progressing in the UK, thanks.

Speaker 4

So much for having me.

Speaker 5

Yes, so in this morning we had said a Cardell in the London studio to talk us through where we're at with the Microsoft activision probe. And you know, one thing that she made really clear is that the UKCME has still vetoed this deal, and that veto will still stand until Microsoft offers up officially fresh restructuring ideas. So that's a big one that we need to keep in

mind that those are still to be officially submitted. Another thing that was an interesting point that she made is that you know, after if they consider Microsoft's new proposals, if they still don't think that those actually address the competition issues at the heart, then this deal could be vetoed again and we might go back to the phase that we're at before, where we're at the competition appeal tribune and we're getting a full of peel. I think.

Also the most important thing that you touched on a moment ago was that Cardile really put to bed any claims that they UKCME had bowed to UK political pressure or that had in some way its decision to reconsider this had in some way been a fit impacted by the FDC's loss at court on the eleventh.

Speaker 3

Yeah, trying to still stand by their independence. What then does a deal, What recommendations, what remedies are likely to actually cut the mustard when it comes to the CME.

Speaker 5

Yes, So Cardile touched on this early in her interview, and I think the only sort of remedies are going to cut the mustard at this point as divestitures. Bloomberg News reported, according to sources earlier that that this could possibly be a divestiture of their UK cloud unit. It's yet to be made public or clear what the actual remedies as that Microsoft is going to offer up the CME, so that still remains to be seen what exactly will cut the mustard.

Speaker 4

I started that phrase and we're going to keep on running with it. A little bit of.

Speaker 3

British sotuly in the morning and it's great to catch up with you. Go have a great weekend, Kafum Gammel. She has been one busy woman throughout this week. Meanwhile, well, the whole market's been busy, hasn't it In terms of fun flow data, for example, come from Bank of America showing that despite the distancing themselves from broader equity markets,

investors have continued to pile into tech stocks. Tech funds, for example, attracted one point eight billion dollars in the past week, pulling in money for the fourth consecutive week and while outperforming the rest of the sectors. How long can this last Mahony Asset Management. See you, Ken Mahoney, who is invested in Microsoft, Apple, and Vidia, Tesla to name but a few, and peace to say, welcomes us

and Ken. It is interesting that tech has just continued to fight off all that demongering, the fear of the economy.

Speaker 4

And plow forward.

Speaker 3

Will that come to an end perhaps with the rebalancing and then as that one hundred for example, Look.

Speaker 6

I don't think it's gonna happen because the rebalance of the NAZAG one hundred. I think it's got legs and the AI we keep talking about, and a lot of people feel like they missed it. I think when the second or third inning of this AI, I mean, this is gonna be a multi year move here. And you know the picks and shovels of the last decade or nineteen ninety nine, where was you know, Oracle and Cisco pixing shovels of this gold rush, Microsoft and the videos

and the names you mentioned before. So I think in the early innings, yes, it's gonna be violatile. I don't think the nats QUE one hundred be shuffling is going to change anything because I think the fundamentals are very strong for this group.

Speaker 3

Well, the fundamentals be shown in these earnings. Will we get a guide of how much AI is going to power revenue?

Speaker 6

I think so On Microsoft, by the way, which come out next week. They're very good for beating estimates and raising guidance, and then the analysts have to scramble to catch up with them and raise targets and so forth. So I suspect we're going to see that again for some time until the levels off, and this AI is helping companies longer their costs and helping their top line.

Speaker 2

So it's a creative.

Speaker 6

So it's not just a line item on an expense sheet. It's actually something that can actually help these companies.

Speaker 3

Ken a bit tactical, perhaps, but how do you play the run up and the aftermath of an earnings?

Speaker 4

Do you try and trade of it in any way?

Speaker 3

Do you just wait until the print comes and well move within that volatility?

Speaker 4

Do you trade or do you just push through?

Speaker 6

I think you kind of pushed through again. It's so difficult sometimes, like it's like rushroa roulette. We never know which ones are going to sell off, which one's going to go last week a couple of days ago Tesla Netflix we saw kind of sell the news again a little soft guidance, So we don't know exactly how they're going to go. But and I just see past this quarter,

and that's probably with the investors. They kind of get in, they trade wrong, they get out, and again these are multi year I hope to think multi year growth because of what's happening AI and some other thinking about Microsoft by the way, all the different verticals you get when you own Microsoft, from gaming to network security, to Windows to you know, cloud to now AI integrating all that, So this to me is not just a good quarter

to jump in and jump out. I think there's there's a lot more tailwind here.

Speaker 3

But valuations ken thirty six times future earnings currently for Microsoft, it was eighty or something for no One.

Speaker 4

Do we see a bit of softness?

Speaker 3

Do you think these valuations can be vindicated in the short term, in the near.

Speaker 6

Term, Well, that's what we were going through this digestion process because these multiples are really kind of whacko. We get it, but the old side is too. You see some valuations with the PE six and PE seven, you look aback a year and now later and you still find some of those same companies still trading with multiples

that are stuck at six or seven. So what we've seen with top line growth, with the analysts trying to play catch up, we've seen some of these multiples expand and justifiably sol because there's only a handful of companies that can really roll the way they're growing. And that's why again those multiples.

Speaker 3

Oh I'm interested in that only a handful? Are there only a handful? Are we not looking hard enough? There's sort of Kathy Wood perspective that in videos just so obvious, try and go for some slightly different AI plays.

Speaker 6

You know, I kind of think about Picte and Shovel are kind of the gatekeepers of this, and now would be Microsoft in the video and some of those names. There are gonna be some secondary names. It's gonna be like the dot com and the nineteen nineties. We're gonna have go from three to twenty. Well, there are one stra ponies. What I like about is that these companies

have a lot of verticals. Again, you think about the video with the timeless striving, the video with a gaming and everything that they're doing and a creative to earnings is now this AI place, So I'd rather play this way with the kind of proven winners that already have a lot of verticals, have a lot of businesses, have a loud opportunities within those businesses. Oh and then let's at AI on top of that, and then then you get this kind of great growth rate.

Speaker 4

Ca Mahoney.

Speaker 3

Great to have some time with you, Thank you, Mahony, Asset Management CEO. It's going to be busy week ahead, so you have a fun weeknd ahead of it. Meanwhile, coming up, we're gonna be talking more about investing in ourltivision Intelligence. But at the earlier stage in private stage one received bending six hundred million dollars with a new fund backing startups focus solely on artificial intelligence at all, with Vege capitalists behind the investment. Next, I mean, well,

look we're just talking about earnings. Let's talk about one company that has been under strain from a share price perspective today following its earnings, American Express ames actually lower after investors concerned by some slower payment volume growth. But I had some time with the CEO C Square a little bit earlier on the phone.

Speaker 4

He's less concerned.

Speaker 3

He was trying to reiterate, look look at our record revenue, look at our record at earnings for share, look at our record billings. And he's also showing really about some of the risks surrounding this economy. He feels, actually those recession risks are fading, that hard landing fading as we go forward. So really we see a company that he says is standing by its growth plans firing on all cylinders.

And it was just some tough comparisons if you're looking from the previous year in extraordinary numbers that we had in twenty twenty two post COVID interesting take on AI, though, he's trying to push back on some of the dangers of generator AI, particularly when you're looking at credit scoring, worries about bias in the system. He says, look, our models, like everyone else in the industry, heavily regulated, heavily reviewed.

Speaker 4

Our algorithms are reviewed.

Speaker 3

This doesn't come cheap, and the FED and not the oversight bodies have them in their sites, so he's.

Speaker 4

Not too fearful of that bias.

Speaker 3

From New York as a Bloomberg Technology.

Speaker 4

Wing Bench Capital is betting big on artificial intelligence.

Speaker 3

Now, the firm closing a six hundred million dollar fun that's forth to back startups focused on just really AI.

Speaker 4

Let's get right to it. The wing founding partner goa gard.

Speaker 3

I mean, is there anything out there that isn't in some way AI powered?

Speaker 1

Now? Well, our view is that.

Speaker 7

In today's day and age, as we move from digital transformation to AI transformation, if you're investing in business technology, it's not worth doing unless it's AI focused. And an example might be do you think people are even going to buy a tractor that is not A powered and cannot drive itself? So we feel that way about the entire landscape.

Speaker 3

Okay, so do you put money to work on AI powered tractor companies or is it really more about the foundational models, about the underpinnings about these companies that are going to be almost plugged into other corpus out there.

Speaker 7

Well, first, thank you for having me on the show. About half of our investments are in business applications and the other half are in software infrastructure that powers AI, data security, developer led tools, things like that. Let's mostly software and why.

Speaker 3

Are those entrepreneurs where are those companies being built? There's been this narrative that Silicon Valley is back where it's at.

Speaker 4

When it comes to AI.

Speaker 3

Is that true or is this more democratized across the US and internationally?

Speaker 7

Yeah, it is fairly significantly in California where we have companies all over the country, and we also have some in Europe and Israel.

Speaker 3

We've just been looking at some of your current portfolio already Bedrocks System inserting Unicorn, and you've had successful exits in the past. I think of fire Eye, I mean long ago, Jasper required by Cisco Mobile Iron These a lot of these not only were acquisitions or IPOs. How does that market for exits even look now at the moment for.

Speaker 7

You, well, the IPO windows kind of locked up right now, and in fact, i'd ask you when you think it's going to open.

Speaker 1

You've talked to everybody.

Speaker 7

But in the long term, we feel that capital markets go up and down, but the role of technology in the economy is growing, and particularly this time around, we feel that anything that can tolerate ninety five percent accuracy will be impacted or significantly by AI. So we just feel the opportunity is so enormous and will pervade.

Speaker 3

All of GDP and the LPs saw that opportunity to how quickly were you able to raise the six hundred million.

Speaker 1

Yeah, it went pretty quickly.

Speaker 8

It was a slightly more difficult environment than last time, but in some ways our last one was raised right at the start of COVID, so this one felt a little easier and it took about the same amount of time in a few months.

Speaker 3

And you really are focused on the very early stage. How broad are these companies already when you're starting to look at them and you're looking at.

Speaker 4

Entrepreneurs with an idea.

Speaker 3

Are you having to see revenue already being generated? How are you deciding which companies to back in what is a very noisy field.

Speaker 7

Half the companies are seed stage, the other half our series age. We write checks in the seed up to three or four million, and in the series age up to twenty million as our first check, So it's a pretty wide remit. I would say that we are quite distinct in that ninety five percent of our companies have less than a milllion dollars in revenue when we invest.

So we invested Inception pre product in very early revenue as well, and we're willing to take on that uncertainty because it results in the best long term returns.

Speaker 3

Well, come back when you start to write those checks, when we start to hear about or maybe even an Ibo window opening for your larger, more well older mature companies. Wing Value Partner, thank you so much for your time. Time now for talking tech. First up TikTok Shop. It's teaming up with a fintech platform at Home to expand its online retail platform in Malaysia. And platform will integrate a tomes by now pay later service check out rivaling

the likes of c and Ali. Babba's BMPL options. Parent company fit Dance says it's looking to invest billions of dollars in Southeast Asia markets in the coming years. Meanwhile, Open Ai seems to support certain requirements here government regulation. This is according to an internal policy memo drafted Forralling a meeting with the White House officials and other tech executives back in May.

Speaker 4

Now.

Speaker 3

The document laid out a series of policy commitments and a government licensing system. Executives from leading AI firms like Well, OpenAI, Meta Alphabet will join President Biden today to discuss transparency and security of AI.

Speaker 4

Will have that plus FTX.

Speaker 3

It's banking on a new lawsuit against its centorious founder and former Topic associates.

Speaker 4

In order to recover cash for its creditors.

Speaker 3

Now the suit Sam Mackwin Freed, co founder Gary Wang, and others, it's all as from making fraudulant transfers that benefited them personally but in nothing for FTX more broadly, and the complaint is filed Thursday, and it seeks to undo more than a billion dollars from those transactions. And while those details just keep on coming. And what's so interesting is some of the intricacies within this recent lawsuit and the co founder Sam Mamuin Freed and former top executives.

Let's get straight to it with Hannah Miller to really discuss well, some of the more questionable behavior. And the thing that caught my eye and we'll catch everyone's eye is a private island.

Speaker 4

Really what was that for?

Speaker 9

Yes, So, the complaint alleges that there was this memo exchanged between an executive at the FTX Foundation, which was the charity arm of FTX, and Gabriel Bankman Freed Sam Bankman, Freed's brother, and it basically laid out this plan that you know, hey, maybe we should buy the country of Nauru uh and build a bunker there that could be used to protect effective altruists in the event of an apocalyptic you know, crisis that would take over the world.

And we know that Sam Bankman Freed publicly said that he embraced effective altruism, so it makes sense that that's.

Speaker 4

Who they'd want to protect.

Speaker 3

A wonder who would have been in that crew, uh if any la. Some of the other allegations are well just about the sheer amount of money that was already realized to have been lost prior to the ultimate demise, particularly by well heads of alimedia, and like can you just detail some of the other more shocking revelations and just the knowledge of how badly this was doing.

Speaker 9

So the complaint alleges that in March twenty twenty two, Caroline Ellison, the former CEO of Hour Meter Research, estimated that there was a ten billion dollar deficit around the same time she gave herself.

Speaker 4

A twenty two million dollar bonus.

Speaker 9

That's what the complaint alleges. So it's very interesting to see that contrast there, but it highlights how the executives really you know, were very generous to themselves that they took you know, huge bonuses equity even in your common shares.

Speaker 4

I mean some of the details.

Speaker 3

What was it potentially alleged fraudulent transfers of four hundred and seventy seven million dollars worth of FTX common shares to the former director of engineering, payments made by Sam Macmunfried allegedly granted himself, right some more than six million dollars in equity. I mean, all of this starts to

add up. Can you just remind us of the timeline of all of this to when these sorts of accusations are going to be laid, whether we can really understand whether there's truth to them or not, and ultimately whether the creditors are going to see much So.

Speaker 9

The allegations spread out over a period of several years leading up to the collapse of FTX. You know, there's stuff happening in twenty twenty one, twenty twenty two, according to the complaint, but that is prior to the collapse of FTX, which was in November twenty twenty two. Prior to that, the company had given this perception that it was extremely strong and engaged in a series of bailouts helping out struggling crypto companies. During the summer of twenty twenty two.

Speaker 3

I'm just kind of also interested, I suppose in robin Hood. Within all of this, can you just remind us as to how much money ended up being purchased of a public company. We're all so aware of that purchase and then felt the one point FTX wanted to actually buy out robin Hood.

Speaker 4

They used money from Alamedia.

Speaker 9

Though, yes, the complaint alleges that and Bankment Freed and Gary Wang, his co founder, actually took over five hundred and million, five hundred million dollars from Alameda and used that to buy robin Hood shares. So they just took that money. They didn't really have to give anything in return, according to the complaints allegations. So it's pretty crazy to see that they were really, you know, taking huge steps in terms of you know, looking at robin Hood.

Speaker 3

I feel I've become numb to pretty crazy allegations and you keep them coming. Thank you so much, Hannah Miller on All Things FTX. Remember you've got to go and listen to the podcast. So they have about Sammi Munfried as well. Welcome back to Bloomberg Technology. Now let's talk a little bit what's happening over in Washington, because leading USAI companies publicly admitted to safeguards to the booming technology

at the White House's request. Today, President Biden joined by executive from Amazon and Alphabet Meta, Microsoft, Open Ai, and others for a key event. We understand now one advisor to the administration has been Neverinasing Creato AI.

Speaker 4

CEO.

Speaker 3

Creato is a responsible AI governance platform and you empower organizations basically to deliver to embed AI responsibly by trying to measure, monitor and manage some of the AI risks. And what do you make of these safeguards that are really being driven by the administration right now?

Speaker 10

Well, Carolyn, so good to see you again and thanks for having me on the show.

Speaker 2

You know, just a reminder Credo AI.

Speaker 10

We are on this mission to ensure that AI is always in the service of humanity, and we do so by providing an AI governance software that provides that continues oversight and accountability of artificial intelligence by managing, mitigating, and monitoring AI introduced risks. So today's announcement, we are very encouraged to see the Biden Harris administration, you know, continuing to depict their commitment to responsible innovation as well as to AI governance.

Speaker 2

I think a lot will be seen in the.

Speaker 10

Coming months because right now these are still voluntary commitments and what is going to be.

Speaker 2

Important is the executive.

Speaker 10

Order that is going to be coupled with these commitments. That's where the eyes are on right now.

Speaker 3

They claig over a meta saying these voluntary commitments are important first step, and let's just talk about some of the commitments there because big names also private names like Anthropic and Inflection AI are in on this and they're saying they're going to share information most importantly trying to mitigate some of the risks for governments and civil society. But what do you make of the sort of water marking virtual water marking and being understood basically when you're looking at AI.

Speaker 10

Yeah, Carolyn, I think watermarking is a really important aspect of really figuring out, you know, the provenance these systems.

Speaker 2

But before we go there, I want to step back and.

Speaker 10

Just focus on what these voluntary commitments are. As you indicated that these voluntary commitments are being committed to by seven leading generative Aian Foundation model providers, and they are focused on security, safety, and trust, and as you can imagine these are not comprehensive, but a great forst step to ensure that, you know, the US citizens and the federal government deserve the transparency from these large language model providers.

So coming back to water marking, I think one of the core things is really understanding the entire provenance of.

Speaker 2

How these systems are built, where.

Speaker 10

The data is coming from, and water marking is a mechanism to really distinguish what is human generated versus what is AI generated. So though an important commitment, there needs to be more work done in that space.

Speaker 3

We understand actually the system needs to be developed also, I mean we understand that you required to report some of the risks, but they're not actually required to eliminate the risks.

Speaker 4

How do you go about doing that?

Speaker 2

Yeah, I think that's a really important question.

Speaker 10

You know, as we look at some of the poor things missing in today's White House announcement. One is, as you mentioned, not only reporting, but how are you going to hold these companies accountable to what they're reporting against.

There is no indication in today's announcement around those commitments, those benchmarks, and I think those are really critical thing that we should be you know, watching out for the second thing is these are completely voluntary, So I think going back to we've seen in the history that voluntary commitments don't really have the impact that we are you know, looking to see from these important but.

Speaker 2

Limited set of model providers.

Speaker 10

So the next steps, as I mentioned, is the executive of order and more importantly, what would that executive order mandate?

Speaker 2

Because I think.

Speaker 10

What these large POWERFOLII systems need right now is set informed regulations and mandates to make sure that there is delivery against what is expected of them.

Speaker 3

Okay, so in executive order coming from the administration again because Congress is well, unsurprisingly perhaps a little bit slow to truly try and get through any changes in law and overseeing AI there.

Speaker 4

But what about how it compares to what's Europe's upting.

Speaker 10

I think right now one of the great things that I've seen with today's announcement is that there has been discussions with the allies, you know, notably we've seen engagement

with France, Germany, India, UK and others. But I think what's going to be really important in the coming weeks and months is how there is a global EI governance framework aligned to you know, what we care about in the United States, but also the great work that European Commission has done with respect to the Risk Based EUAI Act, which is you know, in works right now and most likely going to be past it this year.

Speaker 3

From your perspective, what is self regulation like VISA be perhaps and is there some sort of way in which we can fundamentally oversee these companies. Is there some sort of new part of a regulator that needs to oversee it?

Speaker 10

Yeah, Carolyn, So you know, I do want to note that I Advice President Biden, and I sit on the National AIA Advisory Council, but.

Speaker 2

Today I'm speaking on personal capacity and as a CEO of Predo AI.

Speaker 10

So when we look at these self regulations though, an important step to really showing commitment that you know, these powerful AI technologies need to be transparent and in service of US citizens and federal government. You know, we need to make sure that as we progress, there is a bigger focus on what does that continuous oversight look like?

Speaker 2

And right now there's two pathways.

Speaker 10

In the short term, we should be looking at our current agencies and building capacity to ensure that they provide that oversight of.

Speaker 2

These powerful systems.

Speaker 10

But we are also you know, looking ahead in long term, and do believe that there's room for a global regulatory body that is really tasked with keeping pace with this very fast changing, very you know, massive scale technology and providing oversight to it.

Speaker 3

And can that break through the geopolitical tension between US and China?

Speaker 2

You know, we hope so.

Speaker 10

But I think right now what is really important is making sure that United States and its allies come together on this global regulatory framework and start putting the foundation for what this regulation regulatory body globally could look like.

Speaker 3

Avaruna, great to get your expertise when you are indeed helping advise the administration, but here is a capacity, as credo A CEO, Averrena is saying me, thank you.

Speaker 4

Meanwhile, sticking with Biden, the president seeing a.

Speaker 3

Major setback to one of his key legislative achievements after TSMC has been forced to delay production and its project in Arizona, key battleground state next year's election. The company won't start production of chips at the Arizona plant until after the US presidential elections in twenty twenty five, citing look a lack of skilled labor as.

Speaker 4

Well as costs. In America, there's reasons for the postponement.

Speaker 3

Let's talk about VC money a little bit more now next, because coming up, safire Adventures is joining us. Of course, it's going all in on AI with one billion dollar commitment.

Speaker 4

But we'll discuss the firm's president and co founder I do a little bit more. Next. This is Bloomberg Global Software VC firms. Safire Venture has.

Speaker 3

Just announced well a commitment of some undeployed capital specifically towards AI investing, particularly technology setups that are specializing, perhaps even in generative AI. The firm will focus on all areas of emerging AI techtact though we understand foundational models and neighbors, middleware, next gen AI applications, let's stick into

all of it. Safire Ventures president and co founder Jadas joins us more for today's VC Spotlight and ju are the second VC on today talking about a commitment in AI.

Speaker 4

The previous was a new round of funding.

Speaker 3

Yours is money you'd raise, but really want to signal basically to entrepreneurs that this is your sweet spot that you want to remain and show commitment to AI. You already have loads though, right, how much have you already got deployed across ais.

Speaker 1

First of all, Caroline, thank you for having me on the show. We have about ten billion under management. Ron Off that about two and a half billion has been

invested already in sixty AI companies. This new commitment and in AI companies is really a signal to entrepreneurs that you know, not only are we open for business for investing in AI oriented companies, but more importantly, we are creating a full suite of services that will help these entrepreneurs build and you know, the next generation of AI companies.

Speaker 3

Oh that almost takes me back to the heady days of what twenty one, where all vcs are out there trying to say, why choose me? Isn't that fierce the competition for the latest and great stay I entrepreneurs.

Speaker 1

No, that's not the case. You know. I think it's a little bit of a signal within our firm to help kind of solify the goals of what all the

investors should be looking for. But it's also a set of services that we are building up around AI where we kind of go in and help our portfolio companies and the company that we invest in in how they're going to use AI and how they're actually going to take it to market and figure out how they can use AI to solve certain solutions and certain problems that's out there.

Speaker 3

You tend to invest in companies that already have some sort of product market fit. What pain points are you really focused on that you want the AI to solve.

Speaker 1

Well, first and foremost, you know, I think this new generation of AI, the generative AI, will really really help increase the way that people are actually change the way people interact. The user interface is going to change completely, Right. We typically interact with each other through voice and text and in the natural way, but when we interact with an application, we're always using a keyboard and announce, right, and that will completely change with generative AI that's coming out.

But I think fundamentally, you know, this technology that has come up recently with large language models is going to change the way we actually do and solve all our tasks. Right. The way we develop software, the way we you know, go in and create and do certain tasks will all be changed because of this new technology.

Speaker 3

What are valuations like to some of the companies that you're currently looking at.

Speaker 1

Yes, the valuations are back to the heady days, right where if you have AI companies you know, you can raise that practically any valuation you want, but we actually believe in the fundamental still. So we have been pretty looking at companies that have efficient business models but are using AI to kind of change the way they are going to market, the way they are changing the problem that they're solving.

Speaker 3

I feel like I'm asking this day in day out, but I'm not yet sure. I didn't know if you are as to how the overall open source versus closed foundational models ends up being used.

Speaker 4

How do you see that evolving?

Speaker 3

What are the key areas that you feel startups should be building in.

Speaker 1

Yeah, Look, I think that the fundamental thing that has changed with this new generation of technology is the data mode. As we talk about, right, models that are out there, you know, whether they're open source or closed source is as good as they can be. Actually, the application that are being built on the models are far behind and the models themselves. So we fundamentally think it doesn't matter if it's an open source or a closed source model.

It's how you use the data to train those models and provide the intelligence and solve the problem that you need to solve. Right, And I think the startups that will be successful are the ones who are able to use this data that they're able to collect and preserve

and solve the problem. It's not going to be about the foundational models, right I think those things have been done, and there might be some closed source models that might become popular, but overall, I think open source models are going to win out, and it's all about the data and how you use the data train those models.

Speaker 4

So data is value proposition here Jabe.

Speaker 1

Exactly.

Speaker 4

That's sort of changing as well.

Speaker 3

I mean, individuals suddenly become aware of how much their own data is valuable. We have certain institutions trying to stop the amount of scraping that's going on.

Speaker 4

How are you seeing startups.

Speaker 3

Navigate this and ensure that they do build these modes and they have access to the right data to create.

Speaker 1

Real value exactly. Well, these are all opportunities for startups right where large enterprises do not want their data to go out to the open source community, and other people can use them use their data to train their models. So this is an opportunity. And also we are looking for startups who are able to get to the data and use that data you know in some way that

they solve the problem. So I think, you know, there's going to be going to see number of use cases where large enterprises will build applications because they have the data within you know, in their in their in house programs or software. And then there will be startups that enable these prices to build applications using the data that they have and provide you know, some kind of solution

for their customers. And then of course there's going to be companies that enable you to protect your data right so that it doesn't get corrupt or used by somebody else to train their models and make a profit out of it.

Speaker 3

What do you think these B to B companies are being grown right now? And how diverse are those founders?

Speaker 1

You know, the founders that we are finding again are coming from areas where they actually know the domain. Right in B to B software, having the domain is very very critical. So if you're coming from legal tech or if you're coming from healthcare, you know the domain and you know the problem, and then you're going out and using this new technology to actually solve this problem better

and easier. Right, So, we are still looking for entrepreneurs who have domain expertise because just because you are you know, in depth in AI or generative AI doesn't mean that you are understand how you can solve the problem for a lawyer or for a healthcare specialist.

Speaker 4

Jade As, thank you for joining us.

Speaker 3

SAFA Ventures president co founder on the reallocation of funds really specifically towards artificial intelligence.

Speaker 4

It's been going viral.

Speaker 3

New Jersey suing the federal government this in the hopes to block New York's congestion pricing proposal, which would charge drivers entering Midtown Manhattan as much as twenty three dollars.

Speaker 4

Bloomberg's Michelle Caski joins us now for.

Speaker 3

More and they call it a brazen money grab?

Speaker 4

Is it, Michelle?

Speaker 11

From New Jersey's perspective, it is. They're saying, Look, our residents New Jersey commuters heading into Manhattan already paid tools and crossings such as the Hudson Tunnel and Kinternnel, George Washington Bridge. And what they're saying is if you put this additional potential twenty three dollar charge on top of that, that's going to be it's going to be tough for New Jersey families.

Speaker 3

They're also trying to fight this from an environmental impact perspective, Can you explain that they're saying that, particularly to the north of Manhattan and the George Washington Bridge, that you're going to have a build up.

Speaker 2

That's that's the risk here.

Speaker 11

That's that is that is what could happen is you could have more traffic in certain areas of New Jersey, maybe even the South Bronx, increased pollution in those areas, while at the same time congestion pricing. The goal of congestion pricing is to ease traffic in midtown Manhattan and.

Speaker 2

Produce and reduce pollution in midtown Manhattan.

Speaker 3

How much of a risk is this? I mean, this has been done in other cities. I think of London's had a congestion charge for years now, but obviously doesn't have different states. They perhaps just have different counties to be marked to be navigating with. But ultimate there's always some nimbiasm to all of these things. Do you think that they can be navigated?

Speaker 11

There is a chance, I mean, maybe that is part of Governor Phil Murphy's New Jersey Governor Phil Murphy's plan is to go forward with this suit against the United States Department of Transportation in.

Speaker 2

The hopes that New York will compromise.

Speaker 11

In some way maybe give some sort of an exemption to New Jersey commuters.

Speaker 2

That's very possible. But the problem with trying to figure out a.

Speaker 11

Tolling structure is that the more exemptions you give that means the higher the actual toll needs to be because other motorists will be making up for those exemptions.

Speaker 3

Yeah, the point is to either have more evs so it's less dabveraging into the environment if you don't count the battery and element, or indeed more people on public transports so you cut down on those fumes. We'll see how the battle commences. Michelle Caskey, thank you sin great on is there? Meanwhile, look, that.

Speaker 4

Does it for this day, this week of Bloomberg Technology. Do not forget to check out our podcast though you can find it on the terminal if you've got one.

Speaker 3

Also go online on Apple or Spotify, on iHeart from New York and well from New Jersey.

Speaker 4

Just then as well, this is Bloomberg Technology. Have a wonderful weekend.

Transcript source: Provided by creator in RSS feed: download file
For the best experience, listen in Metacast app for iOS or Android