From my Heart where innovation, money and power Collie in Silicon Valley, NBN. This is Bloomberg Technology with Caroline Hide and Ed Ludlove.
And Caroline Heinda Blomberg's weldad quarters in New York and Ludlow who's off. This is Bloomberg Technology coming up full earnings coverage ahead.
Netflix sept for its worst day.
Of the year, that says its outlook full short of estimates. Will break down the results. Plus, let's look at Tesla's results as Elon Musk warns and more blows to the company's profitability, and Chip Giant TSMC drops as it warns the AI frenzy may not last. We'll have that and so much more ahead. First, let's get you up to speed with a lackluster day in the markets.
We send it over to Blomberg's affaguilty a little.
It is a little lackluster, Caroline. In fact, it feels like the first worst down day in many days for the major indexes here in the US, because we've had this melt up. But some of the big tech reporters reports while putting a bit of a wrinkle in that right now we have Tesla down six point four to two percent of course on its report. That's where profitability
is an issue. Elon Musk even talking about the idea that if rates continue to rise, they will have to bring prices down that could weigh on profitability more so. And then the fears around some of the other big tech companies that have not yet reported, Meta, Alphabet and Apple all lower ahead of those reports. This idea that we were talking about yesterday. So much movement to the upside,
but will these reports deliver? Now if we flip up the boards, we are going to see that what this means for the s and P five hundred and the NASAC well, we are looking at a down day, of course, and the nicy Fang Index down even more, down three point two percent. The socks not immune from this. We have been video down quite a bit. We also have AMD down. And then finally, if we take a look at the other big earnings a lagger that is Netflix, So I think we have a chart of that here.
We're going to see, Wow, gosh, the worst day I believe since April twenty twenty two. And Caroline, there's this continued puzzle. I'm sure you're going to solve it on this show. In the next hour. But record blowout subscribers. Well, I don't know if a record, but blowout subscribers. But where's the revenue And they're talking about still keeping that double digit for the end of the year.
Putting a lot of pressure on the fourth quarter.
Quite frankly, Yeah, Bank of America looking optimistic towards that second half, but as you say, for now, the revenue not quite following that uptick in subscriptions. Let's go back to Tesla for a moment, because it is falling hard too, warning of more hits to its already shrinking profitability. CEO Elon Musk said the company will have to keep lowering the prices of its evs is interest rates continues to rise.
I think it makes it does make sense to sacrifice margins in favor of making more vehicles because we think in the not to do in the future they will have a chromatic valuation increase. I think the Tesla fleet value increase of the point which we can upload full of you know, full self driving, had it's approved by regulators, will be the single biggest step change in asset value maybe in history.
He's got more analysis than mention. Akaine as an Austin In Shawn, I don't really understand why the market's so surprised by this. These price cuts, This focus on volume over profitability has been.
A theme all year, But were they're hoping for some sort of change.
I think, you know, if you look at the cuts that they've made and the subsequent adjustments, there have been a couple of little price pumps over the last month or so, they have flattened out a bit, certainly not as volatile as they were in the first quarter. So I think maybe we have some investors who were expecting that to be completely closed, and especially given the fact that there was a chance maybe we were done with interest rates hikes, people were thinking that maybe this was
all over. But obviously Elon Musk says that if interest rates keep going up, they're willing to cut. I mean, he said last quarter, and he sort of reiterated it.
In that clip.
He's willing to cut even deeper to the bone because he believed so much in the ability to deliver full autonomy and what that would do to Tesla's margins. You know, he's so bought into that idea that he really kind of doesn't care about the margins, right now, so I think it's part that and also part Remember they've added like five hundred billion dollars to their market cap this year.
They've had a pretty big rise over the last couple of weeks and months, and so you know, maybe this is just some resetting of expectations.
I mean, good context there, Sean.
What's interesting about this focus on autonomous driving is the need for compute right now, and of course a supercomputer Dojo takes limelight a billion in.
Terms of investment in that area.
That seemed to surprise people, even though the CFO tried to calm some nerves.
Yeah, yeah, there was a funny moment. It was probably one of the more maybe lively moments of the call, which was otherwise pretty straightforward. You know, Elon coming out and saying basically about exactly how much or at least in the ballpark of what they want to spend on a super coupeter over the next year. And you know, something important to remember, this is a specific use case type of supercomputer in the sense that he's not using it to try to be something that they could throw
at his AI effort. But he just announced last week and do large language models on this. He says, this is very tailored to processing video and images. It's something that is specifically made to help improve the driving software Antesla's cars. And so now we finally had a ballpark. And yeah, immediately after we got this sort of lawyered dep response from the CFO, who said, hey, you know,
that sounds like a lot of money. But we baked this into the guidance that we've been giving you guys, so you may have a little bit more clarity on exactly how we're spending some of that. But don't get spooked. This isn't some sort of new expense.
Nevertheless, expenses racking up, well, perhaps the profitability that margin under pressure at just eighteen point one percent, seawann O Caine. Great to get your analysis. We thank you, and let's got to invest to Tate now a peace to Welcome
to the show. Sylvia Jabronski defines ets CEO CIO overseeing nine hundred million dollars an assets onder management of course Tesla when your key holdings and your pure electric vehicle ETF and Sylvia, were you upended by this still focus on volume of a profit or had you anticipated so?
I anticipated that you know, the earnings announcement and call was sort of go like this. There was so much focus on what will profit margins look like when you take out kind of the tax credits and you know, different types of incentives coupled with the lowering of prices, and you know, the street didn't like it, but I actually thought it was a great call, and I would have said that there's actually calls for the stock to rally. You know, what Elona Musk is doing is continuing to
corner the EV market by making EV vehicles accessible. You know, there is a volume game here, so while you're sacrificing some margin in the near term, you're increasing you know, capacity to produce more vehicles, interest in purchasing vehicles because they're affordable and really you know, taking that stake while everybody else is trying to catch up.
I also think that you know, he has enough of kind of you know, shots to fire here.
He has the cyber talk coming out, and that's arguably going to be an expensive type of vehicle that people seem to be you know, really seeking and waiting for. And you know, so when you have days like this, when when a stock like Tesla kind of falls precipitously in one day from you know, what looks like bad news but is actually really good news.
And then you know the other EV makers fall in line with it.
I mean, this in my mind is a great kind of dollar cost averaging day on you know, the top EV stocks.
Okay, so buying in on the weakness even though it's still priced what eighty times forward earnings?
How do you vindicate that valuation?
Are you the person who's also in this idea of a ROBOTAXI the idea that ultimately there's a big bump in valuation.
I am, and you have to be right. You have to be if you're buying the stock at those valuations. It's a fair point.
But there is a whole lot of evidence that, you know, the R and D is moving there in the right direction. Of course, you know investors want everything to happen more quickly than it usually does.
But this is very much a growth stock. You know, they're they're working on things that do not yet exist, and this is what you want when you invest.
In innovation, the future of EV. So there's a few things. One TV market was fourteen percent of sales globally. That's projected to be thirty percent by twenty six percent. We know that Tesla has you know, sixty percent of the market. Now that shrinks, but they get more orders, more volume, so I think they continue to kind of benefit there. Number two is the new you know vehicle, the cyber truck.
Number three is you know that you mentioned the dojo before, the potential for AI their their self you know, kind of driving database, data center computer. The technology looks like it's kind of being set up to support them. And I think the robotaxi thing, if anyone is going to get it done, it will be Tesla, and I you know, we have no reasonably that he's not going to get it done.
Sylvia, what about the global nature of this company?
Of course they're talking about perhaps having to slow down on some of the factory output as they update them. Of Course they've got exposure in Berlin here in the US, but also China, and I noted that byd Lee Auto Ready managing to ramp up their own deliveries at the moment. What's competition like globally?
Yeah, and I think that you know, when you're in bad staying in the EV space. This sort of makes the argument for having a basket of all of these names right, because you want the test, you want Tesla, and then you also want who is the Tesla of China?
Right?
Is that Xkig, is that Li Auto? And it's good to kind of have that broad based diversification. I think the answer is that these are all going to be winners.
You know.
The piece of the entire pie of EV again is growing globally.
Right, we talked about that fourteen percent going to thirty percent, So massive upside for anyone who sort of gets it right.
In terms of Tesla.
The fact that they're talking about building a factory that's larger than Volkswagen and BMW in you know, Germany, the home of German autos, which is like you know, I mean, the Michigan of Europe is just mind boggling to me. So that tells you that the demand will continue there and the opportunity for car sales continues there.
There's some amazing stats out there.
If you look at EV sales last year in China, fifty percent of evs sold or fifty percent of car sold were EV's and the Nordics, though it was eighty nine percent of all vehicles sold and Europe that numbers closer For sixty percent, So you know, he's following the of interest.
There at a time where macro headwinds also there.
Just that's ran out of this conversation a little bit with where you know, Musk started to focus on interest rates still rising and the impact on the consumer.
How are you feeling about the macro picture right now?
Yeah, I'm actually a little more bullish on the macro picture.
And you know, we've talked Caroline before about some of my stock picks that were getting crush last year in twenty twenty two with interest rates rising.
On the tech space, and I think that that's playing out now.
You know, there was a bear market last year and tech interest rates were rising, and now we're coming to the end of that. So I am in the camp of if we get another hiker two, okay, it's it's sort of manage what we're already there. I think that, you know, the economy has withstood a major recession. Perhaps you get a soft landing, but in terms of rates specifically towards ev or other goods and services, you know, I don't see prices necessarily kind of coming down, but
inflation is coming down. Prices will probably stabilize and it will become you know, kind of easier to borrow money to produce more vehicles, to you know, to take on debt for innovation for growth companies.
So I think we'd have you know, some arrange bound volatility.
It will be kind of a slower year into the year end, but if earnings hold up, then I think the rally continues.
A little bit of dissinflation visa v deflation right now, Sylvia Jablonsky, really great to have your take across Macro, and of course so Micro and Tesla CEO and CIO of Defiance ETF.
Only a small percentage of our members are on the ads to here, even with the moves we just mentioned. Nice growth in the ads to here, but still off a small base, and we're really early in terms of paid sharing impacts including extra member for the reasons that Greg mentioned, that's going to build up over multiple quarters.
That's a newman there.
Netflix is CFO and trying to articulate why they're all so early in this shift towards an advertising model. That's continuing the conversation with Julie Alexander, director of strategy over at Parent Analysts, and it was a really interesting set of numbers. The fact that Netflix is down the most it's Eightril twenty twenty two. Clearly people had some anxiety around the fact that the revenue just doesn't seem to be following the uptick in subscriptions.
Yeah, I think it speaks too.
I mean, let's be clear, Netflix had a really great quarter when we're looking at Netflix's previous quarters.
This is the kind of uptic that we are hoping to see from Netflix.
But that softer revenue is concerning.
It's only concerning if we think about this in terms of that ARM number, that average revenue permember. Other companies refer to this as typical RPO, that average revenue per user.
And so we see that three percent decline year over year.
Now that we know that as Netflix tries to onboard these new subscribers, right, so we see that nice subscriber growth in the most recent quarter. The question is what are revenue are they generating for these customers that now have different pricing options. They can come in at a cheaper add tier, which actually is better for Netflix overall. The ARM front because of that advertising revenue alongside the
subscription review is really strong. But a lot of these countries where they're experimenting with the password sharing crack, and where they're looking at how to kind of increase their subscribers and leading into this immense demand that they see
across their portfolio. The questions are these members coming in at a cheaper plan that maybe does not include subscribers We know that Netflix includes options, for example in Latin America, for some of these subscribers who are getting kicked off their parents' plans, are getting kicked off their friends'.
Plans to go in and say, hey, I want to come in a cheaper tier for the service.
And so how much revenue is Netflix really generating on average across these different subscribers is going to be the long term question. I think that's why you see Spencer
and you see Greg and you see tests Rando's. The co CEOs, Greg Peters and teds Rando's speak to this really important aspect of pushing people towards the advertising tier, taking away certain plans like these basic ad free plans right and kind of moving into this idea of having a better position pricing power to really generate the revenue that they need, especially.
As they kind of level out content spending.
So JUDI they did do that of course in the UK and the US countries where people, of course canford they uptake, can decide Okay, I'll go for the AD supported cheap A model, or I'll have to pay more to get AD free.
What do they do about emerging markets?
How do they ensure that they can have the revenue they need from areas that they're still able to grow in.
Ultimately, I think what's really smart about Netflix's plan, And let's be clear, I think there would actually be more of a concern if the ARM didn't change, in part
because Netflix was not experimenting with its pricing power. I think the fact that Netflix is saying we're going to experiment with pricing power, We're going to experiment in markets that we know local content is really important, and we know that customers don't necessarily have the disposable income that customers in the US, the UK, Canada tend to have.
We really want to figure out how we can be the local, dominant entertainment source and all these different countries right where Netflix is trying to monopolize attention is not just as the global distributor of one type.
Of content, but being a local powerhouse and then sitting under.
This global umbrella that is Netflix.
And so when we think about that, what Netflix is really I think building is the s of having a prestige type of platform, a premium platform and then a casual platform, and the casual platform and the premium platform, unlike in other companies where that might be limited access to certain titles, whether that might be.
The inability to actually access certain things on the platform.
And think about how Peacock has Premium Plus, how HBO Max, how they differentiate between or now max rather how they differentiate between the ad free platform and the ad supported platform.
Netflix is taking this into account of well, if we get if we get rid of these basic plants that don't have advertising, and we can make up that kind of average revenue per member by bringing people into that ad free plan, sorry, that ad supported plan, and then continuing to build upon that and actually bringing an additional revenue.
Then in these.
Markets where we can keep our pricing really low and see increased household penetration, especially in markets like in VR or parts of Europe where the linear and paid TV system is still pretty fundamental.
These are countries where they don't necessarily need streaming.
Internet access is still being developed in many ways, or they're still kind of moving on to those types of plans. Netflix is saying, we know that we need to penetrate, and we know that that is our first goal. You'll got a country like India where they're seeing strong penetration, but they're taking it at a stronger loss on the revenue front.
And unlike Disney, who's who's suggesting with Bob Buyer's recent.
Comments, we might move out of this market a little bit, you have Netflix saying we really want to double down. How can we use two different pricing power plans globally to ensure that our revenue does not kind of see these consistent slowdowns as we bring on these new customers.
Juday, what about content, Because what's interesting is is perhaps Netflix less exposed to some of the strikes here in the United States because they can make much more content locally in these local markets.
Yeah.
I mean, you know, just some quick data points from our firm, parent Analytics. You know, Netflix leads the pack when it comes to on platform demand share, and what that really suggests to us is this ability to retain customers. This idea that when people are looking at an entire platform across original licensed programming, rarely spending the majority of their time.
Netflix has actually beaten out Hulu in Q one twenty twenty three.
In Q two it beat out HBO Max again now Max and with those combined Discovery plus Max assets, and so Netflix still is kind of the home in many ways, especially in the United States, to a lot of these customers. Where Netflix is most insulated is that Netflix has two or three main advantages. It has this global pipeline, and we've seen demand for global content to increase. Now, I want to be very clear here, the story of squid games should not be that Netflix can create global hits
on the fly. It should be that they can create really strong local hits that might be able to find audiences outside of those territories and therefore increase the revenue that they that they originally had planned for that.
And so if we think about we'll have to wrap it up.
We've loved there's so much meat to the bones of everything that you're bringing in terms of your proprietary data and analytics, and we'll get into it a lot more later. But thank you so much that that was a very wide ranging conversation when it comes to all things Netflix parent analytics Judio Alexander. Meanwhile, coming up a conversation with our investment CEO and founder Kath Wood.
This is a bit back technology with tech earnings upon us.
How much will we actually see AI driver of revenue or more just a driver of conversation? And what are the risks still posed by the technology? Talked about all of this and much more with the Arch investment CEO and founder Kathy would and that Christie's art and tech summuch.
Take a listen.
One of the questions we get we're asked a lard is about the displacement of jobs. I think you're worried about the nefarious uses of artificial intelligent of humanity.
The people seem to be talking.
Yeah, so that on the second, all technologies can be used for nefarious purposes. I think what's scaring people is that this one is evolving so quickly. And one of the reasons it's evolving so quickly is the rate at which costs are declining. So AI training costs are dropping seventy percent per year. And why is that it's actually twice as fast as Morse law. It's the pace is the learning curve associated with this technology, So just to get put help people understand what this means. So chat
GPT came out of GPT three. If that model had been developed in twenty fifteen, it would have cost eight hundred million dollars. Instead it was twenty twenty cost four and a half million dollars inceday, energy usees, intent, the cost yes, the cost of yes, developing the model to hardware software today, So eight hundred million eight years ago, two and a half years ago, four and a half
million today less than four hundred thousand dollars. Now we're building bigger models, but the costs are collapsing and the opportunities therefore are exploding.
Is it exploding just for private companies for closed source foundational models? Are you thinking that open source which in many ways the art world is trying to struggle with this sort of ownership versus sharing model as well. What do you think will end up being the AI models or can they all live together?
So we're studying this very carefully. So foundation models, the open AI anthropic metas Lama, Google's palm. Now we have AJX, the foundation models. We thought we're going to commoditize, and we actually still think they will. You do open AI and Microsoft together now announcing charges for all of these AI assistants and so forth. But you have challengers. You have Meta out there saying no, we're not, We're not.
This is open source. You have academic researchers, You've got people in the technology community who who are passionate about this open source movement.
Kathy with there, we'll talk much more about open source in our VC spotlight. Wellcome back to bling back technology. I'm Caroline Hyde in New York. Let's get you a check on the market's halfway through this trading day, and we are seeing negativity warries, particularly in the tech space. We're looking over all, and then as that pulls down one point three percent, those earnings looking lackluster for some
of the key tech names. How much can we justify a record first half for the NASA one hundred in particular. Now remember this is just falling that little bit lower, but it is the worst day that it was see since Dune the seventh for this index two year yield. Also a movement away from the bond market is we're seeing basis points move nine almost ten basis points on the two years. So once again perhaps we see that there's going to need to be a tackle of inflation.
Maybe this is more what's just ultimately happening in some of the sell off that we've seen in terms of yields pushing lower over the last few weeks, that we've seen the CPI print come in, and we look forward to all other inflationary prints we've had. Over in the UK, it was looking good, and indeed in Europe we're seeing the Bloomberg Commodity Index though up on the six tenths percent.
Maybe this is what's feeding into anxiety as to why the Fed might still have to continue to increase interest rates. It's because inflation is still evident, particularly in grains. We're thinking of wheat pushing higher. Therefore, the Bloomberg Commodity Index just on the upside. Moving on, let's look at the
individual names when it comes to our sector. Right here, when I'm looking at tech IBM, look I focused on one of the very few big tech names are on the higher side today, IBM up more than three percent as they managed to actually releave some anxiety about their forward guidance in terms of yes, their second quarter didn't perhaps match expectations on current but people really feel that this is still a turnaround story that they're managing to bring in the free cash flow at the moment.
So some analysts liking that number.
SAP, I'm showing the American Depository receipts of this German company. It's the biggest software company over in Europe with down more than five percent, even upgrade their outlooks for profitability, but not enough and maybe some of the second quarter looked a little bit shy, were worried about spending on it over in Europe. And Tesla, I mean, down by
more than seven percent. Pretty crushing day. But remember how far we have rallied on this particular stock as we look at a company that's still focuses.
On volume over profitability.
Interesting that also Tesla was talking about that supercomputer that's going to be driving its autonomous driving. We understand the billion being spent there, dojo. Let's move across to another supercomputer, Cerebras, known for its AI compute processes and cloud services. It's operating and managing one of the largest AI supercomputers in the world.
It's called Condoor Galaxy one or CG one for short hows in Santa Clara, California.
It can be built to help train generative AI models at a faster rate than ever before. They claim, CG one has already been purchased by one of its strategic partners based over an Abidabi. Let's talk now to the CEO, Andrew Feldman of Cerebras. Andrew, it's great to have some time with you. And look, it's a lot of money, what one hundred million for one of these supercomputers, and I'm interested as to what Abu Dabi is going to be using it for.
Well.
First, you know, today we announced a strategic partnership with G forty two and the first deliverable from that strategic partnership was one of the largest AI supercomputers in the world. And as you and your audience knows, there's a yawning demand right now for AI compute. And this demand is
not just domestic in the US. We're seeing it globally, and by working together with G forty two, we think we can build a constellation not just one, but nine AI supercomputers and fundamentally change the global inventory of compute.
Okay, Chris, many might rightly or wrongly have really put all their eggs in the in video basket, feeling that is where we're going to be satiated in terms of our compute power and necessity. And you're saying that's not true, or at least we don't need to just focus on some of these large, big tech American players.
I think that's exactly right. I think first, all your eggs in one basket has historically been a poor strategy. I think dependence is anywhere in your supply chain is probably not the right strategy. But there are visionary companies like G forty two around the globe. We're seeing models both foundation in closed source and open source in Europe, coming from the Middle East, coming from Abudhabi, coming from Singapore, from Japan, from South Korea. And this is a phenomena
that extends well beyond the six or seven US hyperscalers. Right, this is a need for compute that we thought a partnership with G forty two could help me to a global demand.
Can you push us into the future a little bit.
We were talking with Kathy Wood about closed source open source and you over at Cerebris are basically helping train GPT models and then putting them into the open source community.
Why are you doing that?
What do you think ultimately the AI landscape looks like in five years or so.
Ok. I think one of the things she spoke about in her session was just how fast the AI community is moving. And it's been moving that quickly in part because of the open source and not just open source, but open publishing mentality. People have put ideas into the space again and again, and within weeks or months, those ideas are built on, they're improved, they're furthered and this is produced a sort of a Cambrian explosion of ideas. And we have been leaders in putting large models into
the open source community. Our partner G forty two through its companies, has been pushing open source models into the community, and many others have. We're not alone at all, other hardware makers, other software makers. I think the community, the ecosystem is more healthy if it's not just a small number of compute providers and a small number of model makers.
What about the China US divide, though.
It's an important divide, and you know there we have to think carefully that there are no easy answers about powerful technology and the right way to manage their distribution around the world. I think it's it's an extraordinarily challenging problem and one that the politicians have to have to work through.
Meanwhile, though, the politicians have to balance the risks with the real wards. Many would say key rewards being the application of AI in science and healthcare and the like. Which industries have you been most impressed with the way in which they've adopted and used AI, particularly using your supercomputer.
That's exactly right.
I think the regulation has to balance the drive for innovation and how startups like ourselves and large partners of startups like G forty two can drive innovation. I mean, just recently we've seen through our partnership with G forty two use of AI models and healthcare. We have partners who do drug design with AI models. We see digital assistance helping the elderly. We see in nearly every realm of life and how we live, or work or play,
there's a role for AI. And you know you're showing some of our customers, ask Zeneca, Total Bear, Jasper all using AI in profoundly different ways, I think, and G forty two will use it in still different ways too. I think they are underrepresented languages. Arabic is one of them where there's an opportunity for tremendous expansion in.
The role of AI.
But I think we have to work together. I think your point about export is real and a challenge. But the community has come together.
In the past.
We've built standards together, competitors come together and said this is the standard. We can overcome geopolitical challenges as well.
Andrew Fellman, thanks for the time and the insights. The CEO A cerebris No, it's a pleasure.
Thank you for having me.
Meanwhile, let's talk about the power of artificial intelligence and what is US lawmakers doing about it? Scrambling to impose limits on AI disinformation in particular ahead of twenty twenty fours elections. So far, Democrats have unveiled some pretty modest proposals calling for more transparency and campaign ads. One bill seeks labeling on political ads the use of deep fakes and other forms of AI. But Republicans it's been kind
of slow to sign on. Nevertheless, both actual sides of the aisle have been pretty slow to grasp for really comprehensive understanding, Like many of us are the rebby evolving technology, and in fact we understand there at least months away from introducing comprehensive legislation to mitigate AI's most serious disinformation threats. Meanwhile, let's look at us social media giants or ready taking
action in India complying with the government. There, Facebook, Google and Twitter are already removing a violent video There was actually from May, but that's gone viral overnight. According to people familiar, some social media companies began removing photos and videos of an incident as it violated their rules even
before New Delhi issued emergency blocking orders. Video triggered the first public comments from Prime Minister Modi on the violence in Manipur state, where ethnic groups have clashed for nearly two months. Though some journalists have voiced concerned about blocking orders are obstructing news publishers' ability.
To report on the event.
Meanwhile, coming up, why all companies will need to incorporate II into that operations to survive. Calling to a Gray Lot partner Jerry Chen, we'll discuss that this is blue meg technology.
Time Now for VC roundup.
First up, Sequoia Capital shaking up its team after a turbulentlyar of market upheaval, a breakup, and at least five investor departures, including two crypto focused investors Michelle Frodin Daniel Chen. In the breakup, the firm also spun off Sequoia Heritage. It's a wealth management business, which Michael Morrits will be
focusing on following his departure. Meanwhile, m and A and investment activity at the largest tech in life science firms finished the first half of twenty twenty three on a disappointing streak, with pending and completed deals down forty five percent compared to the start of twenty twenty two. If it continues on twenty twenty three, total deal count could
be one of the lowest for Silicon Valley decades. Plus, a US Congressional committee is investigating for Benure Capital firms for their investments in Chinese tech companies GGV Capital, GSR Ventures, Walden International, and qual Con Ventures. We're all being probed at the moment. This is both the White House and members of Congress acrafting policies to track and potentially block
US investments in certain fields in China. Well, let's get more into the micro of what checks are being written right now in the world of Benure Capital, and on today's VC Spotlight, I'm pleased to bring in Jerry Chen, partner at gray Lot Partners. Who's focus I'm sure, along with a lot of your capadres at the moment, has been artificial intelligence. How much are your current portfolio companies managing to embrace and a D maybe pivot towards it.
Hey, Caroline, thanks for having me today. Everything's AI today. So I think we said in a recent blog the new new modes, it's AI or die. So it's not like am I AIVC or is this AI VC fund? Pretty much AIS touching healthcare, fin tech, consumer tech, door, storage, security, So pretty much every partner at graylock folks on AI, my self included.
Can you help us understand where the immediate opportunities are?
What's sort of been interesting.
There's been various reporting about perhaps how Jasper's been having to cut back on its headcount, how actually experimental CEOs and their engineers within big companies are willing to adopt maybe open source models rather than purchasing the sort of easier just plugins that many anticipated they would go with. First, how are you seeing actual adoption of some of the companies you've been investing in.
So I would say two things.
One, your interest from customers consumers is spot On. So it'd say, like fifteen years ago, every at large customer header had a cloud strategy, right, they had to talk about how we're going to move to Amazon or as or a Google cloud. Now every company out there, big and small, has to have an AI strategy. So first and foremost every portfolio company, Graylock is kind of benefiting
from the instant AI. But then the second question asked is the difference between plugins, between models, between open source and that's a question we had Graylock have been asking, like where are the motes right? Where are the defential business models in AI going forward? These some world where we have chat GBT or Lama tea for Meta or Palm from Google, Like what can you do to build a mote around the AI models?
And so we've been debating that.
In one hand, you can talk about system engagement like chat, like chat Gibt or bard. Chat has also become the way to interact to AI, and it's debatable what that persists or not. We're investors in a company called Pie Inflection starre by Mustafa and my partner Ried Hoffmann. That's all about using chat to talk to AI. There's infrastructure layers.
Companies like Lama Index, a new investment we just did that talks about taking your enterprise data, interacting with these open source models, and then comes like adept or inflection and building these fundamental models themselves.
Carolyn.
So it's pretty cool because we're investing up and down the stack. I think it's early innings to try to figure out where all the money's going to accrue to.
Do you think at the moment some of the money is being spent wrongly?
Have we got a sort of mini bubble on our hands in terms of valuations and where it's being allocated.
I don't think it's it's right or wrong.
I think we're going to look back in time in five or ten years to say that was a mistake. I think there's a lot of experimentation going on right. I think we're experimenting at the foundation models. We're experimenting as a chat layer, we're experimented infrastructural layer. So I think there's a lot of experimentation on how AI is
going to change a consumer tech, enterprise, tech security. I think in the next two or three years we're going to figure out where one which models work too, which verticals make sense, be I, legal, be a healthcare be a CRM. I think we're going to figure that out over time, and then I think as investors we're going to learn also, you know what business models make sense, And you know my bias or my spoiler alert is I think you know the old modes of our network
effects go to market strength, brand marketing. All those weapons you employ as a startup founder in the past apply even more going forward in AI. Right, So AI is great, but I think Carolyn, like, the things that made money in the past will make money in the future.
Where are those founders, sorry, where are the founders?
Well, I mean I think founders today AI are largely coming out of two areas. One, they're coming from research around AI. So you see a cluster of AI researchers from open AI, from meta, from Google, folks that actually know the deep technology. Number Two, I think you see a cluster of founders coming out of domains.
So people are thinking like, Okay, I want to.
Apply AI, but I really know healthcare, legal, CRM customer support very very well, and trying to marry the two together.
So I think we're investing in deep.
Tech founders, you're really academic researchers. And then we're investing in domain experts that say, hey, I understand what the problem here is. Oh holy shoot, Like I can solve this problem with AI that I could in the past. And AI is unlocking these domain ex first to do things they can do.
Three four years ago, jey Chen with the applications of AI, still trying to come to some really where all the value is going to come from, as we will our partner, a great up partner, is really interesting to have some time with you.
Thank you. Meanwhile, let's stick with.
The world of bench Capital because mag so how many changs sat down with Benchmark general partner Bill Gurley to discuss the keys to his success over the decades long career as a tech investor.
Is his take on what it takes to be a great VC.
To be great at it requires a level of hustle that's really hard to explain. And the reason is you're trying to maximize the optionality that you get in front of a home run pitch, and to do that, you have to look under every rock you possibly can. And so if you're not studying some new business model or new platform, you know, AI, if you're not just in every meeting, then you're lowering the chance that you're going to be success. So it's an exhausting games.
Catch the whole conversation circle with them and chang tonight ten pm A New York on Bloomberg Television, or stream it at eight pm on Bloomberg Originals.
Going to have a profound impact on virtually every element of the global economies. Obviously some areas more than others. There are businesses that will be more vulnerable. If you think about lower value add service businesses, maybe a call center business, You've got to be cautious. There'll be businesses like data centers where we've seen a step function increase in demand.
Backstam president John Gray on the impact, of course of AI, and let's stick with it again, artificial intelligence and well, the infrastructure that sort of powers it TSMC out with earnings. The chip maker is cutting its annual outlook for revenue more. Let's bring in Kunjin Sivanni from Bloomberg Intelligence. Well, it seemed to me that we had first and foremost a high profile executive saying we don't know if AI is here to stay, if this frenzy in demand is going to be long term.
Yeah, So I think from TSMC perspective. What the guidance suggests that the macroeconomic headwinds, especially demand weakness in China, is expected to result in a continued slow down in consumer markets like smartphones and PCs, where we are seeing a sluggish inventory digestion playing out. On the positive side, the AI frenzy driven growth was a positive tailwind, but was not enough to offset the global weakness in other markets.
And remind us, I mean, there was the double whammy of perhaps slow down, certainly for consumer based chips, but there's also the fact that they're trying to boost supply, particularly here in US and Arizona, but they can't get the right labor.
They're seeing spiraling costs. This supply chain issue is clearly a headache.
It definitely is. But you know, the Arizona project is again of course come out in the future, so it does not impact their supply currently. However, speaking of supply and going back to your question about the AI frenzy, whether it's short term or not, I think we are in the very early, early leaning subgenerative AI cycle. I don't think this is a short term frenzy. In fact, we are seeing very strong demand and a higher percentage of wallets spent in data center and enterprise shift towards AI.
But what we are limited right now is the supply interesting.
So would what do you think would give t SMC the confidence to see and gain your sort of optimism that generator AI is just at its early innings that they should be focused on supply that this is ultimately demand that steady for the long term.
I think one would be orders from the chip makers, which again translates into the backlocks of the chip maker.
And we are seeing.
Significant demand for the AI chip makers, like in media, they're having backlog and they're just not able to ship enough. So I think that's a good proof point that gives TSMC to backup supply and start shipping towards it. But again, it doesn't happen over a quarter, so I don't expect chip makers in AI to see a significant revenue step up quarter over quarter as we saw with in Vida last quarter. It will more be a steady growth until supply really catch yourself.
And to that end, do you think that more supply will be being based here in the United States? Will we see more openings such as Arizona despite the hold.
Up I'll be shuled. I mean, there might be some delay in the timing of the milestones, but we definitely don't think it will be postponed or I mean or canceled. We definitely will see. And beyond TSMC, we have players like Intel right which have announced their foundry strategy and are going all in on that, so we'll definitely see steps taken in that direction the near term. Gyrations might just slow it down and.
Jen great perspective context. Thank you so much. Conjen Savanni. There bloombag Intelligence all Things and the chip and AI sphere. Meanwhile, that does it for this edition of Bloombag Technology. Do not forget, though, to check out our podcast. You can find it on the terminal have one as well as online on Apple, Spotify, and iHeart from New York.
This is pretty bad technology.