From Mahard where Innovation of Money and Power Collie in Silicon Valley, NBN. This is Bloomberg Technology with Caroline Hyde and Ed Ludlove.
I'm Caroline Hyde at Bloomberg's Weld headquarters in New York and am Ed Ludlow in San Francisco.
This is Bloomberg Technology.
Coming up. We break down the market action as equities trade in the red and inn video drops for the first time since it's a blowout earnings report.
Plus, we'll talk artificial intelligence as the Biden administration weighs how aggressively to regulate new artificial intelligence tools like chat, GPT, and.
A mid recent termoil in the US banking system. One fintech startup saw revenue surge. We'll discuss why, plus so much more throughout the show, including where public markets are currently trading ed. We're seeing a little bit of a dismal feel to the NASDAC today. We're dropping about seven tenths of a percent nervousness. Nervousness out there that maybe some of the economic data in the US is just too strong. The Federal Serve is still going to have
to call this economy the Jolts data. Basically job openings looking much better than any single economist out there had anticipated. But what's interesting is you weave in therefore perhaps the sell off, well, the buying that we see in two year years actually just dipping a little bit. That's all around where we go in terms of the debt ceiling as well, whether we get some sort of agreement finally push forward. We're hearing from Kanie Lynes just at the
end of the previous show on that. But what's notable is also the fact that yes, economic data in the US perhaps signing too much on the positive side, and over in China and Europe that PMI and manufacturing data looking weak. So we sell off once again in terms of those US listed Chinese stocks. Let's move on and see what's happening elsewhere in the world of technology. I'm looking at what's happening in the world of crypto, because what a dire amount month. We end up the month
of May down more than eight percent. This is the worst since November when FTX collapsed. Of course, so there's significant perhaps pullback and some of the games we've seen going on in the crypto world. But dig into some of the microne names that we're seeing on a move today.
Yeah, some of that nervousness you're talking about certainly coming from earnings. Hewlett Packard Enterprise, the maker of networking or office infrastructure tools, down seven percent, pretty big decline. It's forecasts of the current period really coming in below expectations.
HP Inc.
Not to be confused with the maker of laptops missing sales in the court, had just gone the fiscal second quarter that giving some concern around demand for PC. But Bloomberg Intelligence our colleagues, they're saying, actually there was evidence that the market for PC's bottoming out. Based on what HP Inc. Had to say, tut the down two percent. Elon Musk over in China, giving back some of its recent games, coming off a two month high. Elon Musk
photographed having dinner at a restaurant in China. Heats at restaurants just like the rest of us.
Not sure me. I'm having much waiting on the stock.
But a lack of news for us mere mortals driving Tessa action in the chip space. Actually really interesting, Intel making significant gains now up four percent, had been up significantly higher after the CFO said revenue in the current period will come at the upper half of its previously guided projection or range, that giving investors some confidence, and then in Vidia down three point seven percent is actually
it's biggest drop since February. But remember we're coming off of record close Tuesday evening, and of course twenty four hours after Nvidia touched a one trillion dollar market cap for the first time.
Some profit taking.
Perhaps, yeah, maybe, just maybe. And we've got to think about the profit taking more broadly, because an awful lot of the rally that we've seen so far this year has been thanks to just a very few names, and that's highlighted by the recent report coming from Joanne Feenim with Police to welcome her to the show. Ed partner and portfolio and manager over at Advisor's Capital Management. And I read your note with real interest as we were, of course still trying to muddy through the macro what's
happening in terms of debt ceilings and the like. But talk to us are just how important a certain few names of ultimately tech stocs have been to the rally so far this year.
Yeah, Caroline, you know, we got a lot of inquiries from our clients about how much the market has gone up this year and is this a good time to get into equities. And when you break down, there are
really two stock markets. There's the super seven, the biggest tech and consumer companies like Amazon, Tesla, Microsoft, Apple, Google, obviously in Nvidia that have served on average as of Friday's closed, they are up sixty seven percent year to date, whereas the rest of the market is only a zero point three percent.
And that really does tie into two narratives.
One the growth opportunity related to AI on the one hand, and then the rest of the market, which is all about recession risks, intrast rate risk, inflation risk as we're seeing materialized today.
Let's dig deeper into the AI equation because yes, in Nvidia now back below that one trillion, but still basically what a surge. Are we putting too many eggs in just a very few baskets? Are you on the sort of Kafy Woods side of the equation that there are other companies that are going to win here and you should start looking at better valuations out.
Yeah, Caroly, it's hard to tell how much we're going to get out of AI over the long term, but it's pretty clear that the companies that are enabling AI, whether it's through the chips like Nvidea provides, or whether it's the software and deployment like Amazon Web Services, Google, Microsoft are enabling. What I think investors recognize is that even though we don't know all the details, we know that AI and it's spread across the economy is going
to provide years of growth for these companies. And that's a step change in what we expected these companies to be able to do, you know, just some months ago, and so I think that's warranted. It isn't to say that there aren't opportunities out there, and you know, investors can look for those, but we think these are solidly in the middle of this and it's a less risky way I think to play the AI move over time, and then you can look beyond that into companies adopting
AI in industrials, in healthcare. There are plenty of opportunities out there. We'll learn more about them as time goes on.
Johan Kaffee Wood our programming our colleagues in Asia. Overnight she's moved her focus on from Nvidia Tesla to the next big bet.
Have listened to what she had said.
We are really happy that investors who own benchmarks like the Nasdaq and QQQ still own Nvidia.
They own it in some of our portfolios. But we're onto the next thing. Nvidia is a hardware stock, and sure it has some software, but the history of hardware and software is the bigger beneficiary.
Over time is going to be software, is.
Going to be software. I think you note that investors are looking to move money into software, which can mean a broad range of things, right cloud, SaaS enterprise. What are the corners of that sub sector that interests you?
Oh yeah, there are going to be many. I mean the adoption of AI.
Into software tools across multiple and markets I think is significant. You know, we own a company, for example, called Slunk in our growth strategy. They are a big data analytics company, and AI is going to play an increasingly important role there. From Nvidia's perspective, you know, the tying of hardware to software is actually pretty critical to what they're enable enabling
their customers to do. It's also why their gross margins are going up, and so you know, we see in hardware and software to be more and more tied over time, which I.
Think is why you know, in Vidia is such an interesting opportunity here.
Not to say one should potentially take a little bit of profits at this time, which we did for clients actually after they recorded. But going forward, yeah, there are going to be plenty of opportunities in software as well.
Yeah, Karry.
The point that I make as well is like this obsession with a small group of US domicile names, right the megacap tech, But if you look at the commentary of the broader market, particularly the bomb market, when we think about this economy and global inflation, everyone's got an eye on China.
And well, Elon Musk is there at the moment, isn't he as well that we were just highlighting a bit earlier? I mean, joan to that point, how much when you're looking at the macro, when you're looking at what's happening here in the US and everyone seeing the seven winners the rest leaving behind. When we think about where inflation's at, how much are you looking for your investor base at oh, China, Europe being opportunities or being just there as you need to stay well away from at the moment.
You know, there's always a role for international exposure in portfolios. We have actually separate international strategies in addition to some strategies owning some international names.
China clearly has become more of a concern.
I think investors were overly optimistic about how much the reopening we'll be able to drive demand in the near term, Which isn't to say that China is not going to be a very important source of conservat demand. They have massively moved a big chunk of their population into the middle class and that's going to be a foundation for buying over many, many years.
So you know, overall, there are some attractive valuations.
Our approach is to look for companies in countries as opposed to investing in countries as a whole, and that way we can hand select, as we do in all of our strategies. You know, back to the hardware side, the advantage we're seeing that Ed mentioned the focus on US I think comes from the deep modes right that
our companies have created around this technology. Now as a semiconductor analyst for ten years, and one of the key things there was digging into the technology advantage that companies like Nvidia and Google teaming up with Broadcom to develop an alternative big chip to be able to do these AI training computations shows that mode is very deep.
It supports their margins. So I think it's not wrong for investors.
To recognize that the source, right of a lot of the gains are going to be in these US companies.
You know, we learn that one morning a market does not make But it is interesting that there's more emphasis right now on current risks rates recession. How much is that going to impact the technology sector going forward? Do you think from this moment right now?
Yeah?
Great, great thought there. As I said, we have two stock markets right now. We have the investor focus on which companies out there can actually provide years of growth with or without a recession. Right, that's secular growth and now it's been revealed in AI that that's going to be significant. And then you have the rest of the market ed and I think that's the market where they're going to be really focused on those recession risks, and
rightly so. Right the jolt state of this morning suggests that the Fed's going to have to raise rates again instead of pausing, and it really throws into doubt the FED being able to cut rates towards the end of the year, and that makes investors worry that those higher rates are going to constrain lending that's actually going to trigger a recession.
The FED obviously would like to avoid that.
So there's just a massive amount of uncertainty right now, and we're seeing it in the prices of stocks in the rest of the market, where secular growth isn't as big a driver, where the sickle could concerns are far more important.
But the patient investors should really look beyond this.
Equities do a terrific job over the long term of delivering the ability to build wealth. So it's an opportunity to get some of those companies in the rest of the in the rest of the market and build those positions now while.
Risks are high, and if you're a long term investor, you.
Can look beyond that and build really solid, diversified portfolios.
And take profit in the seven names that have driven us thus far.
We've done some of that absolutely.
You know, we're actually slightly now underweight and video we've been overweight for the last year. And I think it's appropriate to make sure that the risk exposure you have in a portfolio is balanced against you know, how the potential return from here is now confronting the level of risk that we see, you know, across the global economy.
Jo Am Feenie Partner, Profolio Manager Advisor's Capital Management one morning. A market does not make, but it's good to get a snapshot of view.
Thank you very much. Credit Card starts up.
Brex has seen a surge and usage of its products following this year's regional banking term or the think tech company on track to reach roughly five hundred million in recurring revenue over the next twelve months of it's if its current pace of growth continues. Joining us now to discuss is Brex co founder Enriq dubigraph in Rik what's driving this. That's a pretty big build up in forward looking revenue.
Right, Yeah, the first thank you so much for having me here. So we're actually announcing that there's two products that crossed.
Over one hundred million in revenue.
Which is our empower product, which is replacing concur for travel and expense for most businesses. And also our business Accounts right which can do operational payments, deposit checks, all the kind of like traditional banking activities for our business in one place.
And both of them crossed over one hundred million.
So now we have three products that crossed one hundred million revenue.
That's something that we're like really really excited about.
So the one hundred million in annual recurring revenue for the business account, it's unit and Enpower Bloomberg sources saying that company wide and you're recurring revenue on track for about five hundred million.
What is it that makes bres so.
Attractive to startups that you've done business with since the fallout of SVB.
So, I think there are two kind of main growth drivers.
The first one is actually our enterprise segment of the business. So Empower, which is our spend management solution, has been serving customers such as door Dash, Coinbase, indeed that has glassed or Elimonade, et cetera truck and I think both startups, mid sized companies and a large enterprises have been attracted for a product because it's super easy to use for employees. You know, you can just slip your car and your
expenses are done. It's extremely global that works all over the world for all your subsidiaries all over the world, and that's been like growing a lot as remote work grows and people are hiring internet a lot of focus on experience. The second one is our banking product, and I would say that, yes, the regional banking crisis definitely had We didn't know if it was going to be a net negative for Brax.
We got that question a lot. Now we know it wasn't a positive.
Customers did come and stay with Brax and that also drove a lot of growth for the company.
You know, Caroline, I think you and I will never forget that Friday where the SVB story unfolded.
We did that special program.
But we learned so much about how startups manage their finances, the type of funding that they seek beyond just raising capital from vcs.
And what was interesting was Silicon Valley Bank was so intertwined with the startup sector because not only was it offering lending, but offering menu ways, venture, debt and enriquey. That's an area that you've been working in as well. How much are you seeing startups getting into cash burn needing to get further money to deploy but don't want to raise equity? Is venture that's still a big They're part of your business.
I would say that vengjur Donna is a very small part of you know, it's definitely an experiment that we try it and continue to do it in small amounts, but it's not mainly why customers come to us. The main way our business account customers come to us is because of their operational banking, so running payroll wires, they ch checks, making their kind of day to day banking.
It's a lot of startups.
They keep up a lot of their funds in some of the big banks, but they keep you know, one, two, three, four months of money with us to be their day to day partner that really understands them.
And where are these companies coming from. You're obviously focused on allowing these companies to be global, to travel, to grow, but are they largely US based companies? Are you seeing this becoming much more of an international play?
These are a lot of them are global companies.
So over fifty percent of the companies that we serve have global employees.
All of them have US employees.
We only serve you if you have some sort of US operation for a bunch of regulatory reasons. But a lot of them are global companies and we serve as I said, from startups just were born to you know, customers that were from SEB also to kind of like larger enterprise is like indeed, door adacting point base.
I have a sninky suspicion as a few more AI players looking to bank or being born at the moment and Ed I mean every day with discussing the exuberance around that.
Yeah, it reflects the conversation we had twenty four hours ago, right, Rex Salisbury. It's about what AI can do for existing fintech players And I guess, Enrica, that's my question for you. How is AI going to boost your offering? Where are you investing to jump on the bandwagon, so to speak.
I think AI is going to completely transform our industry. And I would say there's a couple different places.
Right.
You can think about it something the employee experience. How can you make it even easier for employee to do their expenses or not have to do them at all, you know, make all of that automated. We also announced earlier this year kind of like you know, a CFO for everyone where if anyone has any questions, you know, you can just ask this chatbot and they will respond to you, so you don't have to waste the finance team time of questions. They already know the answer to
all the way to like accounting and operations. I think it's going to completely revolutionize and transform our industry.
And we're investing very, very heavily into the.
Development of new capabilities, and you know, I think we're over the next few months you'll see a lot of very innovative things coming out.
Of Rex Henrique. When does Brex go public? We don't have a timeline yet. We're not against being a public company, you know, in any shape or form.
That being said, I would love to be a low volatility public company. We're not yet profitable, so we'll probably wait for that to happen before we go public. Other things to remember is we're a six year old company. We were started in twenty seventeen, so I would say, you know, don't we don't have any pressure to go public. Yes, we're still quite young for a startup.
And we understand you don't have any pressure to raise new bench capital yourself. I'm interested though, where you are in terms of the focus on being profitable, being CAPEX positive and all this sort of good stuff. You still trying to hire because you seem to be talk about developing leaning into AI. How are you managing those costs?
Yeah, I would say that, you know, we're still investing a lot in one our go to market organization because the product is having so much traction that you know, I think it is higher or y for us to invest in sales and marketing to acquire you know, more customers faster. So we are investing there and our investors are really excited about it. And then in the second area, as are. I do believe this is going to be
transformation on a step function change. So I think if you know, if you're a company and you're not willing to invest in AI, you know, you're probably missing the.
Next big shift.
That being said, we do have a good business of a good business model. I think we can we have a plan to being you know, cashul positive without raising any more money.
Ry kay, great test in time with you again, Thank you. Enrique Ubrogress is the bex Coach CEO. Meanwhile, coming up, we're going to talk about a guest the email master tests the CEO's been making the rounds Beijing. More on who he met with and any projects that he's mulling. That's next. Meanwhile, we've got to remember we've still got earnings coming dripping through the system salesforce. Investor is going to be watching that one closely today after the Bell
Company gonna update us on its own cost cutting campaign. Look, how is it going to provide us with the long run revenue goals that it's set itself. This is Bloomberg. It's time now for talking tech. First up iPhone maker fox Con Technology says it expects to more than double its revenue in the second half of the year thanks in large part two sales and guess what, artificial intelligence
and the servers thereof. It's part of a move to generate revenue from other fields such as electric cars and AI on high that's the listed vehicle for fox Con Technology also working with Nvidia on autonomous driving applications, and Baidu is looking to beef up China's AIC and the internet company has set aside roughly one hundred and forty
million dollars to fund Chinese startups. Has specialized in generative AI by Doo and its VC partners will accept pitches from prospective founders who will use its ernie bot to build their own large language model before Baydu determines well who gets the seed funding. Plus day two of Youla Musks visits China and he's already met with more government officials. That's as he looks to bolster opportunities in the country. Musk met earlier with China's Minister for Industry and Information
Technology and Scott by the Ministry of Commerce. He also met with the head of the Battery Giant's ATL, sparking rumors of potential collaboration, and Musk arrived in Beijing Tuesday, where he met with Foreign Minister Quinn Gang. And I mean, overall, this is a focus another CEO. We've heard from Mercedes, We've heard from GM, We've heard from well Jamie Diamond, really having to talk up that. Look, these are still focused on Chinese despite some of the trade tensions, the tech tensions.
Yeah, and in the statement issued by the Chinese government, Elon Musk is quoted as saying that he is against the decoupling from China. They'll continue to invest in China. It's a really interesting two way situation. Fifty percent of Tesla's output was in out of Shanghai globally last year. And you know, the Chinese government have given Tesla the freedom to operate there as an independent US company, which is rare.
But again it's a chorus of names. As you point out, and.
Interesting that potential rumoring around the big battery maker, because that's of course a big part of trying to have supply chain in the US rather than depending on the Chinese made goods.
Right well, when Ford announced that it was going to make batteries in Michigan licensing Coatl Technology, there was a big fallout, particularly from the Republican Party. That relationship really closely watched. Welcome back to Bloomberg Technology. I med love Vo in San Francisco.
I'm Caroline hid in New York. Let's check in on these markets. Halfway through the show, halfway through the training day therein and we're looking at slightly from lower and now's that one hundred At the moment, we're off by eight ten seven percent. Bit of a concern profit taking call it after the massive run up that we've seen in this particular benchmark, but also the warriors that basically we're running too hot in the US. The Federal Reserve's
still going to have to tame the growth. The jolt stata showing that people are still really hiring. We see a little bit of movement in the two year yield on the back of that data that came roaring through but remember tepid growth coming from China in terms of PMIS and manufacturing data there. Europe to interestingly taking the window of the sales of bitcoin once again. We're actually off by more than eight percent over the course of
this month, the worst month since when FTX collapsed. Let's moving on and have a little look at what's happening in terms of individual movers, because we do still have earnings we used to have real news, but we also have perhaps the first fall in Nvidia since it's waring earnings that we saw last week. We're off by about five percent, called it profit taking after it hit that
one trillion dollar mark. We also, though, see Intel on the back seeing an uptick a more point three point eight percent higher as we see the CFO speaking at a td count event really discussing how maybe they're going to manage to be at the top end of the range when it comes to their forward looking guidance Packet enterprises, though that disappoints on the back of its numbers ed down by some seven percent. Are we worried about compute spending at the moment. We are still in a very
difficult macroeconomic environment. We're going to remember that.
But AI is still a big factor in those markets.
Now.
The US and EU are meeting up this week at the USU Trade and Technology Council gathering in Sweden, where the EU is discussing plans to subject generative AI to additional rules.
This is bigger than Europe. US is important, but it is bigger than the US. But if the two of us take the leads with close friends, I think we can push something that will make us all much more comfortable with the fact that generality of AI is now in the world and is developing at amazing speed.
Meanwhile, Biden administration officials have been divided over how aggressively new AI tools should be regulated, a dissonance that has left the US without a coherent response to the EUS proposals that it's bringing Bloombergs and eguiton to break it down. And I thought we were in a place where the US and EU were on the same page. You regulate the use case the tool rather than the underlying technology itself.
Where have we gone wrong?
Well, we were kind of on the same page, you know.
The last TTC meeting came out with this joint roadmap that kind of agreed to pursue this risk based approach, like you said, focusing on the use of the technology rather than the development of the technology. So AI used for things like critical infrastructure would be subject to higher
levels of compliance. Now Chat GPT changed everything, and that really showed us how a general purpose AI product, you know, something that's not inherently high risk, can become very risky on a societal level when it's used by one hundreds of millions people every month for everything from you know,
helping with homework to know more nefarious purposes. So that's what policymakers are grappling right with right now, is how to treat these generative AI products and how to class sify the risk of those.
At the moment, though, Anna doesn't even really matter if the US, for examples, and parts of the administration are divided, because to me, it feels like the EU is going to be the first mover here with the AI Act.
Well, and that's why the the EU's AI Act is most important, is because US companies are going.
To have to comply with this if they want to operate in Europe.
So much like the GDPR did for privacy, the AI Act is going to set a defacto floor for compliance. So that's what US companies are worried about is as they develop these foundation models that underpin this generative AI technology, if they want to operate in Europe, they're going to have to comply with those rules, whether or not the
US has passed its own policy. And these companies are really looking for a champion in the Biden administration to go to the EU and say, listen, we don't think this policy is going to work the way it's imagined when it's implemented eventually. So the reason why it matters if the Biden administration is divided is that these companies don't have that champion. They don't have anyone going to these multi latteral discussions and really expressing their point of view.
There was a time where we thought that Vice President Kamala Harris was the champion, right she hosted that meeting with Sam Altman's statues at Nadella. The administration has made noises about they want to do something. But I'll go back to one point you made policy. We haven't really heard any arguments for I side of where to focus regulation, particularly from this White House.
That's right, Well, I will give the administration credit for putting forward some very well received frameworks. We have the Risk Management Framework from the National Institute of Standards and Technology that had a lot of input from industry, from civil society as a very well thought out document. You had the White House putting forth their Bill of Rights
for people who use AI products. But we don't have binding policy in the United States, and that's because the US Congress is just now starting to hold their initial hearing to educate while makers and what AI even is. So while the Biden administration has done a lot of work to kind of set non binding guidelines, actually binding regulation is going to come from Europe.
Once again, Anna Juton, I mean thank you giving us well the global perception of AI risks, let's dive into some of the perceptions here in the United States, because there's been another open letter signed by a group of industry leaders.
Warning of this time the risks of extinction due to artificial intelligence. So it was published by the Center for AI Safety.
Which has mission to reduce societal scale risks from artificial intelligence, joining us now to deeper dive on all of this and explain the center zone thinking.
This is Executive director Dan Hendrix it's great to have you. Thank you, Dan. I'm just interested in how this first started, this whole mission statement that you put out there about summer mitigating the risks in line with potentially a nuclear war or indeed a global pandemic. How did you have first off conversations with Sam Altman and demis of deep Mind. Do they come to you, you go to them.
So we created the letter largely because I knew that many people had concerns that AI could lead to extinction and that we should not just be treating it like every other tech issue, but instead as a global priority. So we created this letter to succinctly convey the shared concern, and then we disseminated it among.
Some profits and from there it spread organically.
And so there are some surprises in that open Ai signed it, Google, deep Mind signed it, Microsoft signed it, as well as many of the scientists that built the current wave of artificial intelligence.
What's interesting is I'm trying to understand also how you were born as a center for AI safety, you funded by any of these businesses. What is the ultimate way in which you continued to deliver and do your research.
So we started, we started a while ago and We're largely funded by a philansphy. We're not funded by Elon Musk or things like that. So we're just trying to reduce the risks from AI. I've been when I was as a graduate student at Berkeley, I'd been researching safety for a very long time, So this has been a concern that I've had for many years. And now finally the AI technologies are getting to the point where it's becoming a lot more obvious to the scientific community that
this is a large concern. A letter like this probably could not have existed six months ago, but given the rapid developments, a lot of people are changing their minds, a lot of experts are changing their minds of just about how severe the risks can be done.
I actually want to go back to Caroline's previous question quickly. Have you taken money from any AI executives that were signatories to that letter? The previous letter basically key names in the field of AI.
There, I'd have to check if we're having any funding from any employees that are also concerned about safety.
The AI community is particularly large.
Ninety plus percent of our funding is from Open Philanthropy, which is an independent.
Philanthropy One question that arose quite quickly when you put out the statement twenty four hours ago, is the rationale or motivations behind the signatories?
Why are they doing this?
I get it there's a long term, broad concern about the risks from AI. But why do you think these leading names who want the field to advance signed the letter that you organized.
Well, so many of them were on the voter. We get out a move as quickly as possible. Developing I will be a generally good thing, But a lot of the scientists who signed this have recently changed their tune. Jeff Hinton, of course helped create artificial intelligence modern artificial intelligence, so did Yashchu A Benjio. But in recent months now they're substantially more concerned about there being rogue behavior, that these AIS could potentially lead to human extinction as they.
Get more advanced.
So for that reason, I think that it's just because people are starting to see that this is moving extremely quickly. We don't understand how these technologies work. It's difficult to steer them, and so that could potentially.
They could be misused.
Those things could potentially lead to catastrophic risks or potentially, in the longer term extinction.
Very quickly.
In response to your letter, there were those that know present and much more near term basic risks have listened to SASHALUCCIONI of hugging face you joined us yesterday.
I personally see it a bit as a magic trick, as misdirection right. I care very very strongly about, for example, data consent, for example, disclosing data sources, disclosing transparency and model documentation. And instead of focusing on that we're being directed towards these unsolvable risks.
Are we paying enough attention to more near term and immediate risks from AI.
I should certainly hope that the risk of extinction is not unsolvable, or else we're in big trouble. So I think it's important as a society to manage multiple different risks.
I think we can certainly do it.
I also think that many of the risks from that currently affect us can take on more extreme forms later on. For instance, if the people developing these AI technologies have decisive control over them, that gives them extremely high power and power inequality with respect AI is a concern that could potentially get out of control where a few people are calling the shots in society. So I think that it's important any competent form of risk management will address current ongoing arms as well.
As tail risks. So I think we need to have a complementary approach.
Yeah, and Hendricks, the outcome from this, You've certainly stirred a lot of interest people, perhaps either aligning them to thinking that this is too much or indeed that we need to shine a light on this. How do you think we ultimately take on these risks? You're very good at highlighting what the risks are. Is it regulation, is it self regulation? What do you think the outcome will be?
Well, I think that self regulation would be a useful start, But we wouldn't trust something like a pandemic or nuclear technology just to the scientists. We wouldn't say solve the AI or the nuclear arms race scientists.
It's not necessarily completely their purview.
There are technical aspects to this problem, but there are also social aspects to this problem. So we wanted to just make sure that the public is aware that many of the AI scientists, we have AI scientists from all the top universities, and many of the people who created it.
Are concerned that it could even lead to extinction.
So then policy tends to be more of a negotiation process, and so hopefully we can get those conversations started recognizing that there are severe effects on the horizon, such as potentially extinction. So I think we need to treat as a global priority, and we need to work toward cold operating domestically but also internationally so that we're not caught in an AI arms.
Race between different countries.
The nuclear arms rate brought us to the brink of catastrophe, and we don't want another arms race where we build extremely powerful technologies that could potentially.
Destroy us all and we just keep stockpiling them.
That was not a good outcome for humanity that could have gone substantially worse. I don't want the same to happen with AI. So hopefully we can cooperate starting today.
Have you seen decent self regulation thus far on some of the more near term, immediate societal damage we're seeing, as many have referenced the bias already baked into some of the data that it's being built upon, the worries that the wrong people are being potentially stoped by police, the worries that the wrong misinformation is already circulating. Have you seen that tackle before we even get onto the much larger and more significant concern that you have about extinction.
Yes, so I think we ought to be.
If we can't address many of those current risks, then I don't have much help us addressing some of these larger risks as well.
I should say.
That there's a bit of a mix in whether companies have been successful at making their EI technologies safe. There's some notable failures, like with Being's initial rollout where it was threatening users, but some other technologies have been useful in being relatively reliable and mitigating their amount of bias. So there are fortunate signs that there's some progress being made.
I'm concerned though, that since the AI companies are competing with each other and are locked in an arms race with each other, that there won't be enough time for safety, that they're going to prioritize development over safety, because if they don't prioritize development and making it as powerful as quickly as possible, then we won't be able to make it safe and bring the risks down to a negligible level.
Is it a risk that by regulating, self regulating or otherwise we start to stifle out the new competitors, the smaller players, the academics that you're talking to within labs, rather than the large players such as a being or a bard or an open AI.
I think that, well, if we're concerned about academics, I think that that would largely be the government to be taken care of that. I think that we can definitely target some of these catastrophic risks for models that are at a certain level of capability. That seems like a possible proposal. So we'll basically have to see with these negotiations.
I'm largely just pushing for us and initiating this process and trying to have broader societal cooperation, including international cooperation on these issues.
Dan, how does the public reconcile that the signatories to your petition, your initiative were the leaders in the field of AI, the AI evangelists for want of a better expression, who are now apparently the AI doomers. How have we moved so quickly from leadership to cann in that field.
I think that this is a natural part in any industry where initially it's move fast and break things, be risk seeking, tinker, throw stuff at the wall, see what sticks. But then we start shifting over to a more risk averse regime when the technology starts affecting society on a larger scale. Now people are actually using these AI technologies, and so we ought to be substantially more cautious with their potential outcomes as well. It's getting a lot closer
to human level intelligence. So earlier two years ago, they really weren't intelligent at all. But it's quite conceivable that in the next few years that in many cognitive domains we have AIS that are as good as us or better.
So I think that's one.
Of the main things shifting this underlying change in tone, Because this air arms race is making it move so quickly, people are starting to think we need to rethink our priorities.
Center for AI Safety Executive and research director Dan Hendrix, thank you for your time now. Elizabeth Holmes turned herself in at a minimum security prison in Texas, yes today to begin her eleven year, three month sentence. The disgraced tech founder had been convicted of defrauding investors at her failed blood testing startup Pharrhos, and she and her former business partner Ramesh Bahwani must together pay four hundred and fifty two million dollars in restitution to investors.
Cara, Well, let's get back to it, ed, because we've got more to discuss on artificial intelligence, if you can believe it. But we're going to talk about it on Wall Street. How hedge funds are actually dipping their toes into the pool of AI tools to handle all the boring stuff the grant work or on that next there's a Bloomberg Wall Street. It's already filled with quants with
computer wizards. But because not stopping hedge funds from utilizing the powers of CHATCHBT for the basics, be better to discuss all of this reg news Will Street repolitician atibaseconds. So what is it, junior hedge funds analysts?
Really?
Allelujah? You don't have to do the basics the grumble work here.
Yeah, there's a reason that it only really works for the basics. I don't normally like to use personal examples, but I had gotten my degree in quant finance and down over at NYU for the NBA program. What they do is they take all the students, they have them run through all this data. They put it through code under different market scenarios, and what they show you is that the code the models don't always work.
The point here being is two things.
Once, sometimes the data doesn't work, sometimes it's just wrong, and the other times the market is just unpredictable. It's why there's limitations to how chat GBT.
Could be used. But Caroline, there are things about.
Chat GBT that make this kind of rendition of quantitative finance different than before, and that is natural language processing generated AI that I'll make it easier to read company.
Reports, research reports.
News filings, and even company transcripts to hopefully make it easier to purse through investing signs down to a third minutes to get signals on how devish or hark is something is.
So it feels for now it's about productivity and workflow rather than perhaps getting an edge when actually comes to your investing. Shenani bassa short sweep but always to the point. We love it. And what is going viral? It's the finale z ed of Succession and Ted Lasso Warner Brothers Discovery, which calls owns HBO. It's so close to three million people watch the Succession finale on TV and it's streaming
app HBO Max. In the US sixty eight percent improvement over the last episode of season three, according to the company, And meanwhile, it's the top trending topic on Google Trends. Ted Lasso's own finale Tealasa has been born a boon, of course, to Apple TV, breaking into the Nielsen weekly streaming ranking in sixth place during the last week at April with this third and last season. Meanwhile, though our Brits aren't watching it much.
Right, it's unbelievable.
I mean, Apple doesn't disclose numbers, but it's all everyone's talking about on social succession is on fin Twitch. Three million doesn't seem like a lot in a country of three thirty three hundred and thirty million people, but as you know, we're both big fans. That does it for
this edition. Kara of Bloomberg Technology, don't forget so much to recap and what's been a short week packed with news on our podcast, Apple, Spotify, iHeart, wherever you get your podcast from New York and San Francisco.
This is Bloomberg