WEBVTT - Markets Brace for Microsoft and Meta Earnings 0:00:02.560 --> 0:00:15.000 Bloomberg Audio Studios, podcasts, radio news. Bloomberg Tech is live 0:00:15.120 --> 0:00:18.560 from the heart of Silicon Valley with ed La though 0:00:18.760 --> 0:00:20.400 in San Francisco. 0:00:22.600 --> 0:00:24.160 This is s Bloomberg Tech Coming up. 0:00:24.239 --> 0:00:27.920 Sk Heinix reports a sixfold jump in profit and raised 0:00:27.920 --> 0:00:30.520 capital spending to a record thirty one billion. 0:00:30.640 --> 0:00:32.960 But it's Korean shared slumps. We'll tell you why. 0:00:33.200 --> 0:00:36.440 Plus we're joined by SOFI CEO Anthony Nodo. Following the 0:00:36.440 --> 0:00:39.880 company's earnings topping Wall Street estimates but showing high expenses 0:00:39.960 --> 0:00:43.400 and big tech, big earnings, big AI spending. Next up 0:00:43.600 --> 0:00:48.000 Microsoft Meta and a FED decision will break it all down. 0:00:48.520 --> 0:00:51.600 There's an AI spending paradox careers. S k Heinix just 0:00:51.640 --> 0:00:54.800 reported a sixfold jump in profit and raised capital spending 0:00:54.840 --> 0:00:55.959 to a record. 0:00:55.760 --> 0:00:56.520 Thirty one billion. 0:00:56.520 --> 0:00:59.760 Now, under normal circumstances, that would be a positive for 0:00:59.760 --> 0:01:02.960 the memory chip maker. Instead, investors sent Korean shares of 0:01:03.000 --> 0:01:05.960 the market leader in high bandwidth memory for AI down 0:01:06.000 --> 0:01:08.960 nearly seven percent. It's US listed ADR is basically flat 0:01:09.040 --> 0:01:12.640 right now. The concerns over AI spending whether it'll ultimately 0:01:12.680 --> 0:01:17.520 pay off, overshadowed another blockbuster quarter that skepticism weighing on 0:01:17.600 --> 0:01:18.120 chip stocks. 0:01:18.160 --> 0:01:19.360 More broadly, look at the. 0:01:19.400 --> 0:01:21.200 NAZAK one hundred and the socks then NASA that one 0:01:21.319 --> 0:01:24.000 hundred is down for a six straight session, longest streak 0:01:24.000 --> 0:01:27.119 of decline since October twenty twenty two. The Socks down 0:01:27.160 --> 0:01:29.959 for a fifth day, matching the run of losses that 0:01:30.040 --> 0:01:34.880 it saw in December. Up next in earnings, Microsoft and Meta. 0:01:35.640 --> 0:01:36.920 But for now, let's. 0:01:36.760 --> 0:01:39.360 Kick it off with Memory Bloomberg Intelligence anast Jake Silverman 0:01:39.400 --> 0:01:40.880 here on sk heinix. 0:01:41.480 --> 0:01:42.480 What was the takeaway for you? 0:01:42.840 --> 0:01:45.840 I mean, as I said, in any other quarter, this 0:01:45.920 --> 0:01:49.200 would have been a big blockbuster blowout for sk Yeah. 0:01:49.240 --> 0:01:52.440 I think part of what investors are focusing on right 0:01:52.480 --> 0:01:58.120 now is just pricing. Pricing increases quarter over quarter, and 0:01:58.400 --> 0:02:02.840 overall Skhynicks kind of missed these lofty expectations, and that's 0:02:02.840 --> 0:02:05.639 been building just because AI demand has been so strong, 0:02:05.680 --> 0:02:09.800 it's really been crowding out demand for other products. So 0:02:10.120 --> 0:02:12.720 a lot of the capacity that sk heinez has is 0:02:12.760 --> 0:02:17.200 allocated towards AI, and so that's just created an environment 0:02:17.280 --> 0:02:20.120 of a lot of exuberance around that. And that's really 0:02:20.240 --> 0:02:24.960 ultimately the main takeaway, But another very important couple of 0:02:24.960 --> 0:02:27.120 things just to keep in mind, is one is just 0:02:27.800 --> 0:02:30.480 there seems to have been some idiosyncratic issues in a 0:02:30.560 --> 0:02:34.200 quarter around timing of product mix, and then also the 0:02:34.240 --> 0:02:37.320 long term contracts, which is really what investors are focusing on. 0:02:37.480 --> 0:02:40.120 They want a little bit more visibility in terms of 0:02:40.120 --> 0:02:41.120 demand long term. 0:02:42.080 --> 0:02:45.399 That's so interesting because you know, sk would say there's 0:02:45.440 --> 0:02:49.680 been a change in customer behavior. You know, historically memory 0:02:49.720 --> 0:02:53.360 is highly sit, boom and bust, but customers have said, 0:02:53.560 --> 0:02:56.120 we're willing to commit. I'm just talking a little bit 0:02:56.160 --> 0:02:58.600 more about that, Jake, the change in how the market works, 0:02:58.639 --> 0:03:01.000 but also why it's not a big posative signal right now. 0:03:01.440 --> 0:03:03.760 Yeah, So typically in the past, what we've seen is 0:03:04.360 --> 0:03:08.040 customers signing monthly but typically more like quarterly contracts. And 0:03:08.080 --> 0:03:10.400 a lot of that was because the demand was mostly 0:03:10.400 --> 0:03:14.960 coming from smartphones, PCs, consumer devices and increasingly over the 0:03:15.000 --> 0:03:18.600 last several years, cloud and now more AI demand. And 0:03:18.680 --> 0:03:23.200 so because AI demand is really it's growing at a 0:03:23.240 --> 0:03:26.080 pace that's really unprecedented. And a lot of this is 0:03:26.120 --> 0:03:28.959 just because memory is far more sorry, AI is far 0:03:28.960 --> 0:03:33.000 more memory intensive relative to other applications that we've seen 0:03:33.040 --> 0:03:35.760 in the past, and so because of this, as you're 0:03:35.760 --> 0:03:39.040 building out these massive data centers, and there's so many 0:03:39.080 --> 0:03:43.480 of them. Just the customers need assurances that they can 0:03:43.520 --> 0:03:46.000 get the supply because there's been an under investment in 0:03:46.000 --> 0:03:49.640 the past in terms of memory capacity. So now that's start. 0:03:50.240 --> 0:03:52.880 There's kind of a you know, a give and take 0:03:52.960 --> 0:03:55.800 in terms of supply and demand. The memory makers need 0:03:55.840 --> 0:03:57.960 to add more capacity, but they can't add it fast 0:03:58.040 --> 0:03:59.040 enough to meet the demand. 0:04:00.120 --> 0:04:03.680 Jake Silverman of Bloomberg Intelligence covering memory, thank you very much. 0:04:03.720 --> 0:04:06.040 The AI sell off is taking a toll on Kyosha. 0:04:06.160 --> 0:04:09.840 The Japanese memory maker, has received its first analyst price 0:04:09.880 --> 0:04:13.560 target cut in nearly a year, with Philip Securities slashing 0:04:13.560 --> 0:04:16.760 its target by roughly forty percent. The firm sided lower 0:04:16.839 --> 0:04:21.200 valuations across memory chip peers, but maintained its by rating, 0:04:21.279 --> 0:04:25.080 saying Kyosha's long term growth story remains intact as demand 0:04:25.160 --> 0:04:29.360 for AI storage and nan flash memory continues to build. 0:04:30.160 --> 0:04:33.160 The money keeps pouring into China's AI race. Moonshot Ai 0:04:33.360 --> 0:04:36.120 has raised three point five billion dollars at a thirty 0:04:36.120 --> 0:04:40.480 five billion dollar valuation, according to sources, far surpassing its 0:04:40.480 --> 0:04:44.320 original fundraising target. After the breakout success of its Kimmi 0:04:44.400 --> 0:04:47.000 K three model they're not stopping there. The most Bailey 0:04:47.000 --> 0:04:51.240 Lipschaltz on the IPO and Capital markets beat. We've been 0:04:51.279 --> 0:04:54.520 reporting about this for some time now, but it's confusing 0:04:54.600 --> 0:04:59.560 because Moonshot is basically doing this round seems like it's done, 0:05:00.000 --> 0:05:02.240 and then doing a quick follow on give us all 0:05:02.240 --> 0:05:03.120 the details we need. 0:05:03.400 --> 0:05:05.720 And it's crazy how fast some of these deals are 0:05:05.720 --> 0:05:07.400 coming together. So, as you mentioned, three and a half 0:05:07.400 --> 0:05:10.240 billion out a thirty five billion dollar valuation above what 0:05:10.279 --> 0:05:13.440 Moonshot was targeting, but it has been out to investors 0:05:13.440 --> 0:05:16.599 looking to raise another private round at about a fifty 0:05:16.640 --> 0:05:19.159 billion dollar valuation, So you're already talking going from thirty 0:05:19.160 --> 0:05:22.240 five to fifty billion pre money, and that's all ahead 0:05:22.240 --> 0:05:24.599 of an IPO in Hong Kong as soon as the 0:05:24.680 --> 0:05:26.800 end of this year. So you're looking at a company 0:05:26.839 --> 0:05:29.520 that was worth closer to ten eleven billion dollars end 0:05:29.520 --> 0:05:32.679 of last year, earlier this year and targeting again fifty 0:05:32.720 --> 0:05:35.400 billion dollars pre money in a private round followed by 0:05:35.480 --> 0:05:37.200 an IPO as soon as the end of this year. 0:05:37.240 --> 0:05:39.880 The momentum very much alive and well. 0:05:40.120 --> 0:05:42.760 On the other side of the world, we've covered in 0:05:42.800 --> 0:05:46.719 great detail the release of Kimmy K three the repercussions 0:05:46.960 --> 0:05:50.159 of that open weight model on markets. You're the IPO guy, 0:05:50.200 --> 0:05:53.560 you're my IPO guy, and I think, what's really interesting here? 0:05:53.920 --> 0:05:57.359 This seems very close proximity. So you have the thirty 0:05:57.360 --> 0:06:00.440 five billion dollar valuation, round fifty billion dollar valuation, and 0:06:00.480 --> 0:06:03.560 you just told us an IPO wasting the wings? Does 0:06:03.560 --> 0:06:07.159 that how it normally works in any jurisdiction? Top those 0:06:07.160 --> 0:06:08.200 things so close together? 0:06:08.680 --> 0:06:11.600 Not normally, And normally we would look at a market 0:06:11.640 --> 0:06:14.440 that you kind of stagger out these fundraises and they'd 0:06:14.440 --> 0:06:16.880 be formal and you'd say a Series A through C 0:06:17.200 --> 0:06:19.200 or D and then go public. But we are not 0:06:19.279 --> 0:06:22.080 a normal market environment when you look at how closely 0:06:22.760 --> 0:06:25.359 tied these deals are. The big question comes back to 0:06:25.760 --> 0:06:27.960 what is moonshot targeting. What are these companies who are 0:06:28.000 --> 0:06:30.960 raising money actually targeting? What are their capital needs? Are 0:06:30.960 --> 0:06:33.800 they raising three and a half billion dollars now to 0:06:33.880 --> 0:06:36.479 maybe make a bit of a sweetheart deal and try 0:06:36.520 --> 0:06:39.560 to kind of bridge to the next round, or is 0:06:39.560 --> 0:06:43.280 the question really scaling valuation, bringing in the right investors 0:06:43.279 --> 0:06:46.560 privately who then can make off pretty darn well. And 0:06:46.640 --> 0:06:49.279 an IPO later on this year We've seen how quickly 0:06:49.320 --> 0:06:51.560 some of these IPO windows can open and shut. So 0:06:51.600 --> 0:06:53.840 the big question coming back to Moonshot or even when 0:06:53.839 --> 0:06:56.480 we talk here in the US about ananthropic, is there 0:06:56.520 --> 0:06:59.320 going to be steady investor demand through the end of 0:06:59.320 --> 0:07:01.880 this year and into next year. What are those valuation 0:07:02.000 --> 0:07:04.080 expectations look like? And really, at the end of the day, 0:07:04.200 --> 0:07:06.280 what are the numbers look like? Because these are companies 0:07:06.360 --> 0:07:08.919 that have high water marks in the private market and 0:07:08.960 --> 0:07:10.560 they need to deliver in the public markets. 0:07:12.000 --> 0:07:14.760 Bloomberg's baby lipshaltz on one of the most read stories 0:07:14.840 --> 0:07:18.720 in the private markets, precursors to something in the public markets. 0:07:18.760 --> 0:07:19.840 Maybe thank you very much. 0:07:20.080 --> 0:07:24.360 Coming up, before the hugging Face attack and hack open 0:07:24.400 --> 0:07:29.000 AI's models access to customer account on a different cloud platform. 0:07:29.320 --> 0:07:31.400 We have the details next. This is Bloomberg Tech. 0:07:39.280 --> 0:07:42.200 The open ai models that hacked the startup hugging Face 0:07:42.240 --> 0:07:45.960 earlier this month also gained access to a customer account 0:07:46.040 --> 0:07:49.920 on the cloud platform model. Bloomberg Tech Managing editor Sarah 0:07:49.960 --> 0:07:53.600 Fry joins us with the details really tricky to follow along, 0:07:53.720 --> 0:07:55.560 but I suppose that's start with the basics. 0:07:56.760 --> 0:07:58.520 This is different to hugging Face. 0:07:58.880 --> 0:08:03.360 This was a sandbox environment, not the modal platform itself. 0:08:03.800 --> 0:08:04.680 What do we need to know? 0:08:05.440 --> 0:08:08.040 Well, I think what we need to understand here is 0:08:08.040 --> 0:08:11.640 that this is leading to an increase in the fears 0:08:11.800 --> 0:08:19.000 over AI agents going beyond what they're asked to do. 0:08:19.760 --> 0:08:22.720 And this was a big concern with anthropic earlier this year, 0:08:23.360 --> 0:08:26.960 and these recent incidents are both with open AI. I 0:08:26.960 --> 0:08:30.200 imagine there's a lot more than this hugging face incident, 0:08:30.640 --> 0:08:32.880 this modal incident, that we need to learn about, and 0:08:32.920 --> 0:08:34.599 we need to understand a lot more. We have a 0:08:34.600 --> 0:08:37.520 lot of more questions, honestly, about how exactly this happened, 0:08:37.960 --> 0:08:41.800 why models didn't notice, how do we prevent it from 0:08:41.800 --> 0:08:46.560 happening again, and what might happen if anything in Washington is. 0:08:46.520 --> 0:08:47.679 A response. 0:08:48.840 --> 0:08:52.120 Models saying that this was a sandboxed environment, so even 0:08:52.160 --> 0:08:54.920 though the models were able to access it, it didn't 0:08:56.080 --> 0:09:00.320 act outside the bands of that testing environment. We don't 0:09:00.360 --> 0:09:02.880 really have a conclusion on the open AI hugging face 0:09:02.960 --> 0:09:05.480 right There was an investigation that was ongoing. I don't 0:09:05.520 --> 0:09:08.200 know that they've released any sort of conclusive report on that, 0:09:08.800 --> 0:09:10.880 But just go back to what you mentioned about government, 0:09:11.040 --> 0:09:13.120 there's something needs to happen next right. 0:09:14.000 --> 0:09:16.400 Right, I mean I think that this is this is 0:09:16.559 --> 0:09:19.360 one of the things that these the leaders in AI 0:09:19.480 --> 0:09:22.520 have have been talking about for years now, the idea 0:09:22.679 --> 0:09:26.800 that you know, an AI may act beyond what a 0:09:27.040 --> 0:09:30.480 human has intended for it to do and do it 0:09:30.520 --> 0:09:33.880 in a way that can be harmful to accompany to 0:09:33.920 --> 0:09:37.400 its customers. On the flip side, you know, AI has 0:09:37.440 --> 0:09:41.120 become very adept at finding security vulnerability, so it can 0:09:41.160 --> 0:09:45.200 be help for cybersecurity in that sense. But this is 0:09:45.240 --> 0:09:48.720 exactly the kind of thing that we're going to need 0:09:48.760 --> 0:09:53.480 to understand a lot more about, because we just can't 0:09:54.040 --> 0:09:58.839 create a solution for these problems unless we understand exactly. 0:09:58.440 --> 0:09:59.080 How they happened. 0:10:00.040 --> 0:10:03.120 Bloomberg Sarah Friar, who manages the team covering all things 0:10:03.200 --> 0:10:05.480 chech out in New York, thank you very much. In vidio, 0:10:05.520 --> 0:10:09.440 Ceo Jensen One's defending open weight AI systems as US 0:10:09.559 --> 0:10:13.840 lawmakers weigh the challenge posed by Chinese startup Moonshot ai, 0:10:13.880 --> 0:10:17.839 whose model rivals those from Enenthropic and open ai it's 0:10:17.880 --> 0:10:21.719 open weighted model. Huang met with politicians in Washington on Tuesday, 0:10:22.080 --> 0:10:25.679 saying afterwards that open weight systems are quote really important 0:10:26.080 --> 0:10:30.319 for the vibrancy of the whole industry. Meanwhile, more than 0:10:30.400 --> 0:10:33.360 one thousand workers from top AI firms have signed a 0:10:33.360 --> 0:10:37.600 petition calling on the US government to help quote deliberately 0:10:37.800 --> 0:10:41.320 pace AI development. The signatories of the letter include the 0:10:41.360 --> 0:10:45.240 CEO of Anthropic and the chief scientists from Open Ai 0:10:45.600 --> 0:10:49.360 and Meta Superintelligence Lab. Let's get more Bloomberg's AI editor 0:10:49.640 --> 0:10:52.439 Seth Figerman's with us. This has been playing out on 0:10:52.520 --> 0:10:56.840 social media too, as more high profile names become literal 0:10:56.880 --> 0:11:00.640 signatories to it. Let's start with the idea out like, 0:11:00.720 --> 0:11:03.200 what is it that this group of those in industry 0:11:03.320 --> 0:11:06.760 researchers working on AI are trying to affect. 0:11:06.840 --> 0:11:08.800 What is the result that they're looking for. 0:11:09.760 --> 0:11:12.400 Yeah, you know, it's really a moment to step back 0:11:12.400 --> 0:11:15.360 and take stock of. They're calling for basically a technical 0:11:15.440 --> 0:11:20.600 and policy framework the governments or an international group could 0:11:20.720 --> 0:11:26.240 use at some point if necessary, to quote deliberately pace 0:11:26.480 --> 0:11:29.800 AI development. They're not specifically calling to slow it down, 0:11:29.880 --> 0:11:33.080 though that seems to be the implication. And you know, 0:11:33.120 --> 0:11:37.160 the backstory seems to be both a growing awareness that 0:11:37.360 --> 0:11:40.120 AI might one day be able to kind of augomate itself, 0:11:40.160 --> 0:11:42.600 that they're building technology that can kind of conduct its 0:11:42.600 --> 0:11:45.240 own research and development and start to kind of super 0:11:45.320 --> 0:11:48.080 charge and accelerate that development in ways that could potentially 0:11:48.120 --> 0:11:50.800 create runaway Yeah, I progress that we may not be 0:11:50.840 --> 0:11:52.920 there at the moment. And then, as you were just 0:11:52.960 --> 0:11:56.760 talking about, it comes days after the first of now 0:11:56.880 --> 0:12:00.160 two different but related breaches from Opening Eyes Technology you 0:12:00.200 --> 0:12:01.720 that have raised heightened concerns. 0:12:03.000 --> 0:12:06.040 Let's use Anthropic as a case study. They've they made 0:12:06.040 --> 0:12:13.000 a social post on Tuesday yesterday explaining their people's participation. 0:12:13.800 --> 0:12:16.680 You know, they are one of the leading frontier labs 0:12:16.840 --> 0:12:20.079 working on advancing AI, and they're saying, we're okay with 0:12:20.440 --> 0:12:23.080 joined this effort to let's call it slow it down. 0:12:23.720 --> 0:12:25.480 You're right, and in some ways I think it's actually 0:12:25.520 --> 0:12:28.240 borrowing from some of anthropics earlier rhetoric. I mean you 0:12:28.320 --> 0:12:31.200 might recall that Jack Clark, one of anthropics co founders, 0:12:31.240 --> 0:12:33.880 you know, had put out a lengthy post weeks or 0:12:33.880 --> 0:12:36.480 maybe a couple of months ago calling for the idea 0:12:36.640 --> 0:12:39.600 that maybe we would want to pause button or a 0:12:39.640 --> 0:12:43.040 mechanism to pause AH development, stopping short of saying we 0:12:43.080 --> 0:12:46.199 needed it now or it's definitive, but kind of introducing 0:12:46.679 --> 0:12:48.880 that idea, and I think that thought has been percolating 0:12:48.960 --> 0:12:51.880 a lot now in recent weeks, whether it's Demos Hasabe 0:12:51.920 --> 0:12:55.040 is kind of calling for some sort of international framework 0:12:55.440 --> 0:12:58.720 or a finn Red type organization to and review models, 0:12:58.800 --> 0:13:02.800 or a more explicit to slow or pause development. Again, 0:13:02.840 --> 0:13:04.480 there's a lot of ifs ands or butts here, but 0:13:04.559 --> 0:13:07.400 it seems like the industry is increasingly captivated by this 0:13:07.480 --> 0:13:10.000 thought that there needs to be some mechanism to slow, pause, 0:13:10.120 --> 0:13:13.280 or rethink the pace of development to enseth. 0:13:14.200 --> 0:13:17.480 The petition is urging the US government to do something, 0:13:17.520 --> 0:13:20.280 to take some kind of action in conjunction with something 0:13:20.320 --> 0:13:23.320 that's happening around the world any way in different jurisdictions. 0:13:23.640 --> 0:13:25.160 Do we have a sense of what that would look 0:13:25.240 --> 0:13:27.840 like from the US government? Are we talking about Congress 0:13:27.880 --> 0:13:29.720 taking action and executive order? 0:13:30.160 --> 0:13:31.280 What's the mechanism? 0:13:31.559 --> 0:13:34.000 Yeah, certainly in that petition they're not spelling it out, 0:13:34.080 --> 0:13:37.560 but coming you know, weeks after Demis and Sam Altman, 0:13:37.600 --> 0:13:39.400 I have kind of been advocating for more of an 0:13:39.400 --> 0:13:42.559 international coalition. I would think of it in that mold, 0:13:42.720 --> 0:13:44.320 you know. I don't know if that's just the Trump 0:13:44.400 --> 0:13:49.480 administration lending and support or particular piece of legislation from Congress. 0:13:49.760 --> 0:13:51.480 But I think it's the idea of the US working 0:13:51.480 --> 0:13:54.319 with part of an international body to begin to rethink 0:13:54.360 --> 0:13:55.920 how we govern this technology. 0:13:57.840 --> 0:14:00.680 Being bog Sephigman, who is our AI editor here at 0:14:00.679 --> 0:14:04.240 Bloomberg News on a really critical important read story. 0:14:04.280 --> 0:14:04.600 Read it. 0:14:04.720 --> 0:14:08.560 Rachel Met Shringafari reporting on the Terminal online. Okay coming 0:14:08.640 --> 0:14:12.080 up so far reported its second quarter earnings earlier this morning. 0:14:12.200 --> 0:14:15.160 CEO Anthony Notto's back with us on the show Next. 0:14:15.160 --> 0:14:20.200 There's lots of questions around growth, the outlook, Neo Banks, 0:14:21.000 --> 0:14:23.680 tech in finance, and the FED. 0:14:24.360 --> 0:14:25.800 That's next. This is Winburg Tech. 0:14:39.760 --> 0:14:42.040 It's time now for talking tech. I'm your hira Anan. 0:14:42.200 --> 0:14:42.680 First up. 0:14:42.760 --> 0:14:45.840 Executives and investors tied to the AI supply chain. We're 0:14:45.880 --> 0:14:48.720 among the biggest insider sellers in the second quarter of 0:14:48.800 --> 0:14:52.240 benefiting from the sharp and volatile rally of tech stocks. 0:14:52.360 --> 0:14:56.800 Those include executives of companies like Nvidia, Coreweave, and Twilio. 0:14:57.560 --> 0:15:00.840 Plus the largest US power utility, next Era Energy, is 0:15:00.920 --> 0:15:04.160 joining forces with Brookfield to build a one hundred billion 0:15:04.160 --> 0:15:07.960 dollars data center campus in Kentucky. The privately funded efforts 0:15:08.000 --> 0:15:12.440 will transform a cold war uranium enrichment facility into a 0:15:12.520 --> 0:15:16.240 data center campus with a generating plants, adding to growing 0:15:16.320 --> 0:15:21.040 efforts to add new electric capacity. And Canada is scrapping 0:15:21.120 --> 0:15:24.160 a tax plan that would have forced entertainment companies like 0:15:24.280 --> 0:15:28.320 Netflix and Disney to pay more after receiving pushback from 0:15:28.480 --> 0:15:32.600 US officials and Hollywood studios. Streaming platforms above a certain 0:15:32.640 --> 0:15:36.160 size will still need to reinvest a percentage of their 0:15:36.200 --> 0:15:39.080 revenues into Canadian and indigenous content. 0:15:39.400 --> 0:15:39.600 Edit. 0:15:40.360 --> 0:15:43.640 Thank you Jahira so far reporting second quarter results before 0:15:43.640 --> 0:15:46.960 the opening bill, the company posted record revenue one point 0:15:46.960 --> 0:15:49.840 two billion dollars, but also raise its full year revenue outlook. 0:15:50.200 --> 0:15:52.720 The bit that's unchanged is the earnings guidance look at 0:15:52.760 --> 0:15:56.080 shares down eight point four percent right now on track 0:15:56.160 --> 0:16:00.800 as it stands for its biggest drop since Appril. There 0:16:00.840 --> 0:16:03.920 had been a deeper decline earlier in the session. Joining 0:16:03.960 --> 0:16:07.760 us now is Sofi CEO and Tinodo record revenue and 0:16:07.840 --> 0:16:11.280 the outlook for the balance of this year. It's really interesting, 0:16:11.400 --> 0:16:14.960 like projects a lot of confidence. Let's just address the 0:16:15.000 --> 0:16:17.560 stock move and the story of the quarter. You know, 0:16:17.640 --> 0:16:19.840 maybe the street felt like you needed to take the 0:16:19.840 --> 0:16:21.920 bottom line with you in your confidence. 0:16:22.480 --> 0:16:25.240 Yeah, I think you're right. The quarter was incredibly strong. 0:16:25.360 --> 0:16:28.120 It was our nineteenth consecutive quarter of more than a 0:16:28.160 --> 0:16:30.240 rule of forty. It was actually a rule of seventy 0:16:30.560 --> 0:16:33.600 with forty percent revenue growth to one point two billion 0:16:33.640 --> 0:16:36.800 dollars and over a billion dollars of cash revenue, and 0:16:36.800 --> 0:16:39.160 our EBADAM margin was thirty percent and we had a 0:16:39.160 --> 0:16:40.560 really strong nir income margin. 0:16:41.080 --> 0:16:43.520 I think the street likes the trend in the business today. 0:16:43.680 --> 0:16:46.720 Credits performed well, our new products are getting great adoption. 0:16:47.200 --> 0:16:49.520 We're seeing the flywheel of being a one stop shop 0:16:50.040 --> 0:16:54.480 driving success and increased products per member. So really positive 0:16:54.520 --> 0:16:58.440 story fundamentally in Q two in Q three week, expect 0:16:58.440 --> 0:17:03.160 that fundamental story to remain intact and continue strong member growth, 0:17:03.200 --> 0:17:06.760 product growth, and product per member growth with good profitability. 0:17:07.040 --> 0:17:09.760 We did raise our revenue expectations for the year and 0:17:09.800 --> 0:17:12.679 the back half of the year, but we left earnings unchanged, 0:17:12.760 --> 0:17:15.640 primarily because we want to make sure that we keep 0:17:15.680 --> 0:17:18.359 investing in the business to maintain these high levels of 0:17:18.400 --> 0:17:21.520 revenue growth for a longer period of time. And in 0:17:21.520 --> 0:17:25.320 addition to that, we're also now anticipating two rate increases 0:17:25.720 --> 0:17:27.200 as opposed to the beginning of the year when we 0:17:27.200 --> 0:17:31.000 anticipated two rate decreases, and that creates some uncertainty, and 0:17:31.280 --> 0:17:32.800 so we decided to have a little bit of a 0:17:32.800 --> 0:17:36.480 cushion as it relates to earnings if something like that unfolds. 0:17:36.480 --> 0:17:38.760 But we couldn't be more confident in the business and 0:17:38.840 --> 0:17:41.880 the long term profitability of the business. It's just we're 0:17:41.920 --> 0:17:43.840 not going to take earnings up in the near term 0:17:43.920 --> 0:17:47.159 to try to drive market favorability. That will be short 0:17:47.280 --> 0:17:50.000 term beneficial, but not long term prudent. 0:17:50.800 --> 0:17:55.280 Right, let's talk about the FED. Why not? It's FED Day. 0:17:55.280 --> 0:17:56.959 No one wants to talk to the tech guy on 0:17:57.000 --> 0:17:59.119 FED Day, Anthony, to be honest with you, and I 0:17:59.240 --> 0:18:01.000 just want to partake in it like everyone else. 0:18:01.040 --> 0:18:02.359 So you explained it. 0:18:02.440 --> 0:18:06.679 Your assumption reflects one two hikes rather than cuts. You know, 0:18:07.240 --> 0:18:10.119 put the sort of the numbers and monetary policy joggon 0:18:10.320 --> 0:18:12.480 to one side. How does that change how you think 0:18:12.520 --> 0:18:15.760 about the world, about how the economy grows, about the 0:18:15.800 --> 0:18:17.520 consumer in the back half of this year. 0:18:18.320 --> 0:18:20.399 Yeah, I mean, if the FED is raising rates, that 0:18:20.440 --> 0:18:22.520 means we have a strong economy and we're trying to 0:18:22.600 --> 0:18:28.159 maintain stable inflation in an environment which rates going up 0:18:28.200 --> 0:18:30.439 one to two percent will perform very well on the 0:18:30.480 --> 0:18:33.320 revenue line, which is why we're still forecasting strong revenue 0:18:33.560 --> 0:18:36.040 and I think ultimately we'll perform really well on the 0:18:36.040 --> 0:18:39.119 bottom line. We're already in our guidance in you know, 0:18:39.240 --> 0:18:42.600 providing guidance of a net income margin on an incremental 0:18:42.600 --> 0:18:46.240 basis of thirty percent, which is our long term profile 0:18:46.320 --> 0:18:49.000 where we can get to. So it doesn't really change 0:18:49.000 --> 0:18:51.960 that much. It just eliminates some of the upside that 0:18:52.040 --> 0:18:55.480 one may have called today compared to a different environment. 0:18:56.200 --> 0:19:00.280 I think a stable economy with unemployment below five percent 0:19:00.000 --> 0:19:03.679 and inflation around two to three percent is a great 0:19:03.720 --> 0:19:06.600 backdob for our business. And even better backjob for our 0:19:06.600 --> 0:19:08.400 business is if rates were declining. 0:19:10.440 --> 0:19:13.560 Final question on the economic side, you know, what is 0:19:13.600 --> 0:19:15.720 the health of your customer look like? And not just 0:19:15.720 --> 0:19:17.440 the health, but like, what are the behaviors of your 0:19:17.440 --> 0:19:20.480 customers in this environment. We've gone over in the past, 0:19:20.520 --> 0:19:23.879 You and I the demographics and who you're serving, So 0:19:24.280 --> 0:19:25.600 how are they behaving? 0:19:26.440 --> 0:19:29.320 The customer's behavior is very strong. We see very strong 0:19:29.400 --> 0:19:32.760 levels of spending. We annualize that at about thirty billion 0:19:32.840 --> 0:19:36.399 dollars of debit spending a year. We're seeing very strong 0:19:36.440 --> 0:19:39.880 trends in credit, which we reported continuing, we're seeing good 0:19:39.920 --> 0:19:44.720 inflows into our investment business assets under management standpoint, and 0:19:44.760 --> 0:19:48.000 we're seeing people take out more products. There definitely is 0:19:48.119 --> 0:19:51.000 a desire to reduce the cost of debt. And we 0:19:51.040 --> 0:19:55.760 saw record levels of origination across our unsecured personal loans, 0:19:56.000 --> 0:19:59.200 our student loans were financing, and our home equity lines 0:19:59.240 --> 0:20:01.960 of credit and home equity loans, and that's a secured 0:20:02.000 --> 0:20:04.359 loan that has a lower interest rate. So and we 0:20:04.480 --> 0:20:07.000 recently launched small medium business and we see a lot 0:20:07.000 --> 0:20:08.840 of demand for small medium business. 0:20:08.520 --> 0:20:09.359 Loans as well. 0:20:09.440 --> 0:20:12.560 So from our vantage point, the economy looks very strong. 0:20:12.720 --> 0:20:15.800 Inflation isn't causing consumers to hold back. We're not seeing 0:20:15.800 --> 0:20:19.520 an uptick in unemployment. We're not seeing any deviation on 0:20:19.560 --> 0:20:20.400 the credit side. 0:20:21.080 --> 0:20:22.000 Let's talk about AI. 0:20:22.280 --> 0:20:24.639 When we went through the big bank earnings, there's debate 0:20:24.720 --> 0:20:28.040 right on how much AI is really reshaping financial services 0:20:28.320 --> 0:20:32.800 and the investment that's required for so far today, where 0:20:32.880 --> 0:20:37.199 is AI having an impact, you know, across underwriting, customer service, 0:20:37.280 --> 0:20:39.680 fraud detection, something that's real right now. 0:20:40.280 --> 0:20:42.560 Yeah, the first thing where it's having an impact is 0:20:42.600 --> 0:20:45.760 actually on code production. We have a unit of work 0:20:45.800 --> 0:20:49.800 within our engineering group called a squad, and historically a 0:20:49.880 --> 0:20:53.640 squad consists of seven people. Today a squad is going 0:20:53.680 --> 0:20:56.080 to consist of four people that can produce the same 0:20:56.119 --> 0:20:58.760 amount of code as seven people, So that's a huge 0:20:58.760 --> 0:21:01.320 productivity sailings from AI capabilities. 0:21:01.400 --> 0:21:04.399 It also helps with testing. On the cost side of. 0:21:04.359 --> 0:21:08.240 The equation, we're using it to automate fraud investigations, We're 0:21:08.280 --> 0:21:11.159 using it to automate dispute resolutions, We're using it to 0:21:11.160 --> 0:21:14.520 help with customer service. On the underwriting side, we're not 0:21:14.600 --> 0:21:17.919 really using it to predict the ability for people to pay, 0:21:18.119 --> 0:21:20.960 because we underwrite to their actual cash flow based on 0:21:21.000 --> 0:21:24.040 their actual income and their expenses. I do think that 0:21:24.200 --> 0:21:28.800 AI is helpful in verifying some potential outliers or finding 0:21:28.840 --> 0:21:32.400 them that could be fraudulent activity first party fraud where 0:21:32.400 --> 0:21:36.480 an individual has no intent to repay their obligations. And 0:21:36.520 --> 0:21:40.159 then on the overall value prop of SOFI side the equation, 0:21:40.280 --> 0:21:43.520