00:00:02 Speaker 1: Bloomberg Audio Studios, podcasts, radio news. This is a breaking news update from Bloomberg, instant reaction and analysis from our three thousand journalists and analysts around the world. 00:00:20 Speaker 2: We really want to go all in on Alphabet and Tesla. We've got a great team effort to do just that. In the house with us is our Man Deep, saying he is, of course Bloomberg Intelligence excuse me, head of Global Technology. We've also got our Keith not in Bloomberg News Auto reporter. He joins us here in our Bloomberg Interactor Brokers studio along with Man Deep. And then we've got Edla Love, Bloomberg Tech host out. 00:00:44 Speaker 3: There in our San Francisco bureau. 00:00:45 Speaker 2: All right, where to start? I do want to start with you, Man Deep, only because I feel like the AI trade is so important. 00:00:51 Speaker 3: What do you make of it? 00:00:52 Speaker 4: I mean great print. I think overall the results were great Cloud eighty two percent growth. 00:00:59 Speaker 5: It's one hundred billion in our rundread. 00:01:00 Speaker 4: Business now, which is phenomenal, you know for a company like Alphabet, which was really consumer focused. But backlog is probably where I would say the visper number was higher simply because when I look at Microsoft's backlog number, it's higher than alphabets. 00:01:18 Speaker 5: And given it's growing eighty two percent and it. 00:01:21 Speaker 4: Has got Entthropic as one of its main customers of Google Cloud, I would have expected that to go up. I mean, Anthropic is signing deals left and right, so why is it not showing up in the Google backlog number. 00:01:34 Speaker 1: Gemini models now process twenty two billion API tokens per minute, and the Gemini app as nine hundred and fifty monthly active users. Contextualize that mandate for us compared to open Ai into Anthropic. 00:01:46 Speaker 4: I mean, it's great, but look, Gemini has an attached rate because of all the other properties that Google has the search, YouTube. So for me, until analysts they talk about usage of Gemini really taking off relative to the last quarter, it's hard for me to extrapolate that into you know, Gemini really taking share away from a chat CPT and all these companies are reporting very high MAU numbers, but it's really the usage that counts. And to my mind, the nine to fifty million is a reflection of the high attached rate that Google has because of the distribution through. 00:02:23 Speaker 5: Search and the operating system and browser. 00:02:26 Speaker 3: And we're going to come to you in just a moment on both. 00:02:28 Speaker 2: But I want to bring Keith not named Tesla, it is it is a lot of technology in that one. So as we talk about all of this, what do you make of kind of some of the numbers that we got from Telah. 00:02:37 Speaker 6: That's a big mess, you know, thirty three cents versus fifty one cents, and I get that Tesla is no longer really a car play. 00:02:44 Speaker 5: It's an it's an AI play. 00:02:45 Speaker 6: It Well, here's the problem with that, Tim. The thing is is that to fund that twenty five billion in capex they have planned for this year, they need to sell a lot of cars. So they did sell well in the second quarter, right, but yet we're coming in low. You saw the growth margins also, yeah, yeah, so they made money. They went negative cash flow. We expected that. They didn't go as negative as was expected, so that's good, but they did go negative cash flow. So you know, you got to generate revenue and profit from the car side of the house in order to pay for the robotics and the AI and the cyber caps. 00:03:21 Speaker 1: Yeah, ed Love, look, come on in on in this conversation. Is Tesla in your view and based on the folks you talked to is it's still a car company. 00:03:28 Speaker 7: Yeah, the street wanted to see Tesla spend a lot of money, more money than they are spending currently based on the trajectory of Capex, to make some progress on robotaxis and robotics. And you know, looking back at the court, it was the problem that they had is even though they had record vehicle deliveries, you know they are spending on R and D. That's an impact. Stock based compensation is a big impact. Remember like talent and stock based compensation in the valley. On the software and engineering side is like a really important factor. They had lower average selling prices, so you have record vehicle deliveries, but lower ASP has not gone well for them. It's such a simple story. Capital expenditures came in in line with expectations, but on the buy side, just put your money where your mouth is. Is Elon Musk spend more money on the AI story and it hasn't really translated. But again it's just an earnings deck. The real meat of it probably comes in the call. 00:04:24 Speaker 2: All right, So yeah, I want to bring well ed before we I want to bring Mandy back in your thoughts on also alphabet here. 00:04:31 Speaker 7: Yeah, I was listening very carefully to everything that man Deep said. I mean the way that I look at it. You know, we looked at the backlog, Mandeep explained Gemini, and the trajectory of the cloud business search is where the slight miss is. And so I guess the other way of looking at it is that there's a concern out there that the core search business gets more impacted by the behavior of using a chat bot in lieu of the search engine, and so it's a slight miss, right, that's not evidence of that. Again, a very high ibar quarter for alphabet I just can't get over the cloud growth, like Mandy, like, just save me a bit, like cloud growth eighty two percent, pretty good like relatives what you and I thought it's about earlier in the week. 00:05:13 Speaker 4: Yeah, look, I think overall it's hard to find any fault in the print. It's just, you know, because everyone is expecting Capex to go up to three hundred billion. I mean, look, believe it or not, this company will have negative free cash flow next year. So from that perspective, you have. 00:05:31 Speaker 5: Wild why because of Capex investment. 00:05:35 Speaker 4: Right now, there are probably you know, ten to fifteen billion dollars free cash flow for this year. Next year, if it goes to three hundred billion, there's no way they're going to be positive free cash flow. So from that perspective, despite that twenty four percent top line growth, we're talking about a company that will have negative free cash flow at their scale. And that's where you know, you want to see all these businesses really doing well. Right now, it's cloud that's carrying all the weight, but you want to see that, to Edg's point, and search and you know, YouTube and other businesses, and you're not seeing that kind of lift. 00:06:11 Speaker 3: And you're laughing why at the amount? 00:06:12 Speaker 7: No, no, are Actually I take it super seriously. But like remember when Oracle flipped a negative free cash flow for the first time since the nineties, the markets melted down. When Amazon goes to negative free cash flow, So what to our audio listeners, I shrugged my shoulders and made a funny face. But like you know, the interpretation of mandi It's analysis and the research that Bi's done on this. The market's very sanguine about that. They want to see capital expenditure high. They also want to see top line growth directly evidenced as a result of the capex. But on the cash flow thing, Like everyone seems pretty calm about that. 00:06:50 Speaker 1: Mandy, when's the payoff on this spand is it already happening. 00:06:54 Speaker 5: It's happening in a big way with the cloud business. 00:06:56 Speaker 4: I mean, then have you seen you know, a company get to one hundred billion dollar new business line in a matter. 00:07:04 Speaker 5: Of you know, three four years. 00:07:05 Speaker 4: So they are seeing that in the cloud business, and it's a great investment. It's just I think with Alphabet, search is always the cash cow that funds everything. And even though the top line growth, to my mind, seventeen percent isn't bad in terms of top line growth, it's just I think the backlog number combined with where search would be two three years from now, that's where you start to get a little worried. But you know, maybe they come out in the call and say Gemini nine hundred and fifty million users saw engagement growth of x percent, and then suddenly everyone will be okay, But you really want to see Gemini delays not carry forward. Remember they have seen a delay in their Gemini Pro three point five release. Now, all that is adding to the anxiety. Is Alphabet really falling behind when it comes to the frontier mind yes, they're doing very well on the cloud side, but what is it that will prevent search from really going down or you know, the company falling behind in the front of your model race. 00:08:10 Speaker 2: It's interesting you have to spend right to build out some of these businesses in a big time and that is certainly a metric in terms of how you're judged. I want to go back to Tesla, same thing, though they want Elon to spend to kind of do what he needs to do. 00:08:22 Speaker 6: Yeah, I mean he's made some very large promises, hasn't he about AI and about the cybercabs who are supposed to all be writing in them by now, right, and that hasn't happened. The launch is slower. Optimist is not being built yet out in Fremont, although he says that will happen by the end of the year. But we've heard that before, right, Yeah, So until he actually delivers some you know, deliverable, some tangible results, you know, we're still relying on the car business to deliver the mail and you know, this report isn't showing that that's meeting expectations. 00:08:53 Speaker 1: And is the case being made by investors or by at least Elon's investments Tesla right now that yeah, base X could absorb this company at some point in the near future. 00:09:07 Speaker 7: You know, it's still the prevailing sentiment in the industry and of the existing investigation on the SpaceX side, that this will happen with time. Right. The financial mechanics of that are a bit of a mystery, you know, one public company backing into another. But to lots of people, you just go back to why they believe that. They believe that the joint scale makes sense too, very deeply vertically integrated companies that have a shared initiative on the compute side and semiconductor's side, where there is already a lot of cooperation on the engineering side. You know, to lots of people, it's just logical. We just don't have an answer for that. I found it, you know, I put it in the blog, right, I found it amazing that, you know, But but is it to be expected? There's no mention really is SpaceX at all in the Tessa roundings deck. And the only thing is that they have a one billion dollar unrealized game from their prior equity investment in Xai which rolled into SpaceX. 00:10:06 Speaker 2: Interesting, Yeah, I do wonder too, and I want to bring this question to both man Deep and to Keith. I mean, Mandeep, do you you have to think about Tesla and you have to think about SpaceX and their role in AI and kind of where this company is going. Right, We've talked with you about this. I mean you've got to kind of think about where this goes next. 00:10:24 Speaker 4: I mean, to my mind, why did Google have to rent compute from SpaceX at such a high price? You know, when they have their own cloud business that well maybe they can monetize it better than SpaceX data centers can on their own. So from that perspective, Google cloud business is more established, they can get a lot more out of that compute than SpaceX can. But really that's the case to be made that Google should go big in terms of Capex increase because right now they're renting from SpaceX. 00:10:56 Speaker 2: Well the same thing though, to you, Keith, I mean, do you increasingly think about Okay, you know, Tesla is not just going to be this car company anymore. 00:11:04 Speaker 3: You really have to think about the whole elon universe. 00:11:06 Speaker 6: Right, and he and he has to start showing that and just talking about that, and he also has to make promises that he can achieve and so far he's over promised and under delivered. 00:11:17 Speaker 2: Yeah, it's just the date always moves. 00:11:20 Speaker 4: That takes his two gigabotts of capacity, and suddenly that will be another you know, fifty billion dollars in revenue that SpaceX can add. So it's not that hard right now, at least in this environment, more gigabot capacity is equal to twenty five billion dollars in revenue per gigawatt. 00:11:37 Speaker 3: Man, I do you think. 00:11:37 Speaker 2: It's odd that there's no mention, as Ed mentioned, really of SpaceX in the Tesla results. 00:11:44 Speaker 4: I mean, right now, these are independent companies. Why would they be mixing those up? 00:11:49 Speaker 1: Same boss? You know, ish Ish Keith. I want to play this out with you a little bit and what we were talking to Ed about in the idea of Tesla being absorbed by SpaceX at some point. I know there's no historical precedent for this, but you do have in some cases auto manufacturers that are part of big conglomerates. But you know, Tata comes to mind for me. But is there is the history mixed? 00:12:16 Speaker 5: Well? 00:12:16 Speaker 1: I think with this sort of like a huge company that also turns out cars. 00:12:22 Speaker 6: Tesla as a car company, is not really a growth story anymore. They are contracting or they had a good quarter, but the previous two years they were down in car sales. And they're not even a luxury car maker anymore. They stopped making models, they stopped making Model X. Their most common trade in now is a Toyota, a Toyota hybrid, so meaning people come with people trade in their Tesla Toyota. Now that's what's happening, Edmonds tells us today. Wow, people are getting out of Tesla's and going into Toyota hybrids. So they're a mass market car maker. Now that's why the margin shrink. Right, You're no longer run the luxury business. The most common trade out of a Tesla previously had been a luxury car maker, a German car maker, are going into Toyotas. So Tesla's growth engine is slowing on the car side of the business. And that and so linking up with SpaceX could give them new ways to find revenues and to monetize the AI side of the business, which so far it doesn't. 00:13:17 Speaker 1: But at ed to Keith's point, could the cyber cab or the you know, the robotaxi come to the rescue here. 00:13:24 Speaker 7: You know, there is a they stopped the S and X in the period, right, Those will higher margin, higher price point vehicles. So like what keeps outlining is completely correct. People forget the business model, and actually, what I would say the biggest tone shift in the deck it's just a document, is that this is Tesla the sort of complicated compute industrial robotics company. You know, that's the kind of inaggregate takeaway from it. The Robotaxi business plan is multifaceted. Cyber Cab is a vehicle that Tesla purpose produces and operates within a proprietary ride heading fleet itself, but also plans to sell the consumer. Makes not a lot of sense to a lot of people. But there's also like the Airbnb model, where you, as an existing Tesla owner, submit your vehicle to the fleet, so when you're not using it, it goes out and operates in the ride hailing fleet like an Uber, but it's your vehicle that you own. And what's not clear is the economics of that mixed fleet. People just don't understand it, and Tesla hasn't really explained it since they first made the proposals, So that doesn't answer your question. Tim but that's the problem with this print. The cord of gone was bad, and there's not a lot of evidence that the future is underway. 00:14:33 Speaker 3: All right, I want to go back to Alphabet if I may. 00:14:35 Speaker 2: Tesla's still under pressure, selling off a bit after the market trade here, Alphabet's been bouncing around. So I want to go back to you, Mande, because I'm looking at our live blog, our market's live blog, and they're pointing out. 00:14:48 Speaker 3: Most of its biggest businesses. 00:14:49 Speaker 2: We're head of alst expectations, with the exception of search, which was a touch below. What is it that you think we need to hear the call? Is it about what's happening, what's going on with spending like the outlook here more. 00:15:03 Speaker 4: I mean, the world of AI is measured in terms of token consumption, and even though they gave a token consumption metric around API usage going to twenty two billion from sixteen billion last quarter, so that's a nice uptick, but really on the whole, you want to see them continue to grow that token consumption across the family of apps, and I think that's where you will see the usage of the model, how much Gemini is getting used. So token consumption along with that Kapex guide to me, those are the two key metrics. 00:15:36 Speaker 1: Do we have metrics on AI overview versus traditional Google search? 00:15:40 Speaker 5: They do talk about how much that does that matter to you? 00:15:43 Speaker 4: I mean, I care more about the aggregate, even if there are some offsets that they are moving some of the traffic to AI overviews and AI mode. At the end of the day, it's the time spent on Google Family of Apps. 00:15:56 Speaker 5: That's what. 00:15:56 Speaker 1: Okay, So that's okay, ed, come on back in here. I don't think we've talked to you about this, which is sort of the little incremental updates that alphabet is making. 00:16:06 Speaker 7: Two. 00:16:07 Speaker 1: I don't want to call family of apps because I don't want to confuse meta. But it's like, you know, Google's Gmail having this sort of like AI inbox or you being able to ask Google Maps questions that are more conversational and AI. Does that move the needle in your view? Do analysts talk about that making this stuff more engaging? Because AI inbox is great for me? 00:16:28 Speaker 7: You know, I'm not deflecting. I really got I would go to Mandeep on this, but it's not the consumer that moves the needle, right, you know, look at what they did say, ninety percent of the Fortune one hundred are using the Gemini Enterprise. According to the statement, Gemini models process twenty two billion API tokens per minute. You know, this is the token economy. That's how we're judging a success of the utilization of different AIS that the frontier labs and the hyperscalas are developing. That's in the enterprise. You know, it has very little to me to see the needle move on the existing suite of software that Google ads, just like Microsoft's having a very hard time telling me that three point sixty five has anything to do with the AI story and then boosting cloud sales. It just doesn't. 00:17:15 Speaker 1: Yeah, that's a good point, I mean, Mandeep, But if a lot of consumers are sort of interacting with Google's AI through these tools that they've used for years, then certainly that makes them more engaging. It does it, But like Ed said, maybe it doesn't move the needle. 00:17:29 Speaker 4: I mean, to my mind right now, because the LLM companies don't have a premium models with ADS, the consumer side is somewhat shielded and the battleground is really the enterprise side to ADS point, because that's where the consumption is measured around tokens, and you are seeing that backlog really come to fruition. But the consumer side is important in the sense all the Internet platforms leverage the data to make the platforms better. That's why Google has been so successful over the years. So if you lose the engagement on the consumer side over time, it's going to affect how good your product is. And that's where people moving their queries to chatchpt or cloud will have an impact. Because right now Google has that monopoly ninety percent it used to have. I don't think that's the case anymore, but that's how the platform got so much better. The search box got so much better is because of the usage. So I won't underestimate the usage on the consumer side.