WEBVTT - Apple Weighs Using Intel, Samsung Processors 0:00:02.520 --> 0:00:13.480 Bloomberg Audio Studios, Podcasts, radio news. Bloomberg Tech is live 0:00:13.520 --> 0:00:17.319 from coast to coast with Caroline Hyde in New York 0:00:17.600 --> 0:00:20.760 and Eva Though in San Francisco. 0:00:21.720 --> 0:00:23.360 This is Bloomberg Tech coming up. 0:00:23.400 --> 0:00:27.520 Apple holds exploratory talks with Intel and Samsung to build 0:00:27.560 --> 0:00:31.200 their processes for its US devices plus alphabet. Microsoft and 0:00:31.280 --> 0:00:34.000 Xai have agreed to give the US government early access 0:00:34.000 --> 0:00:37.040 to their models for review before a public release, joining 0:00:37.080 --> 0:00:41.479 Open Ai and Anthropic and Pinterest shares jumping after posting 0:00:41.520 --> 0:00:45.680 strong results and revenue forecast. We speak exclusively with the CEO, 0:00:45.760 --> 0:00:47.760 Bill Ready, let's come in and have a look at 0:00:47.760 --> 0:00:49.520 what's happening in terms of these markets. 0:00:49.520 --> 0:00:51.159 A little bit of a tech issue and the start. 0:00:51.000 --> 0:00:52.720 Of this tech show, but we dig into what's happening 0:00:52.760 --> 0:00:55.360 in terms of really the key story that was a 0:00:55.440 --> 0:00:58.960 huge scoop overnight the fact that Apple is starting to 0:00:59.000 --> 0:01:01.960 have conversations with getting production closer to home when it 0:01:02.040 --> 0:01:03.760 means for their own in house chips. 0:01:03.960 --> 0:01:06.160 They designed them, then who fabricates them? 0:01:06.160 --> 0:01:08.520 Maybe less so TSMC, Maybe it's a move more to 0:01:08.600 --> 0:01:11.720 Intel up thirteen percent on this news, looking at Samsung 0:01:11.760 --> 0:01:15.000 Electronics also up eight percent as we understand Apple executives 0:01:15.000 --> 0:01:17.080 have started to look at their production out of Texas. 0:01:17.280 --> 0:01:19.160 Let's get to the man who helps break down all 0:01:19.240 --> 0:01:21.399 these sorts of bits of news, Mark Goverman's with us. 0:01:21.440 --> 0:01:25.480 Mark talk to us about how Apple is rethinking its 0:01:25.520 --> 0:01:26.199 supply chain. 0:01:27.280 --> 0:01:29.440 Well, we've known for the last few years now that 0:01:29.480 --> 0:01:32.600 Apple is increasingly looking to do more in the United States. 0:01:32.720 --> 0:01:35.640 They've put out multiple announcements. They did a six hundred 0:01:35.640 --> 0:01:40.320 billion dollar investment alongside the Trump administration into the US, 0:01:40.840 --> 0:01:44.000 and now they're looking for secondary suppliers to its work 0:01:44.160 --> 0:01:47.920 on manufacturing its main processors. These are the SoCs, the 0:01:47.960 --> 0:01:51.560 engines that powered them back, the iPhone, the iPad, you 0:01:51.680 --> 0:01:54.040 name it. As you know, for well over a decade 0:01:54.040 --> 0:01:56.960 they've designed those processors in house, and they've relied on 0:01:57.040 --> 0:02:01.120 CSMC to produce them in Taiwan. They've been shifting some 0:02:01.160 --> 0:02:05.360 of that manufacturing to a TSMC fab in Phoenix, but 0:02:05.600 --> 0:02:08.120 additional fabs have been slow to come online, even though 0:02:08.120 --> 0:02:10.560 Apple says they'll have one hundred million chips out of 0:02:10.560 --> 0:02:13.200 the area by the end of this year. But they 0:02:13.200 --> 0:02:17.040 also need additional suppliers. Obviously, there's geopolitical tensions there's supply 0:02:17.160 --> 0:02:19.600 chain issues, and there's just the idea that it's not 0:02:19.840 --> 0:02:22.080 smart to have all your eggs in one basket, in 0:02:22.120 --> 0:02:25.480 one geography and in one supplier. So they've had early 0:02:25.520 --> 0:02:29.440 discussions with both Intel and Samsung about using new fabs 0:02:29.440 --> 0:02:32.920 that they're opening the United States to potentially down the 0:02:32.960 --> 0:02:36.640 road produce these A series and M series chips for Apple. 0:02:37.080 --> 0:02:39.840 Does it have to make sense that it's the AI 0:02:40.720 --> 0:02:44.800 hardware that the US SORRY hardware that this goes into, 0:02:45.160 --> 0:02:47.400 because is there going to be margin pressure here by 0:02:47.400 --> 0:02:48.760 making them in the United States. 0:02:48.800 --> 0:02:51.120 This is also sort of more of a political gamut. 0:02:52.840 --> 0:02:55.760 I mean, in some cases there are politics involved here, 0:02:55.840 --> 0:03:00.160 right Working with Intel will give Apple a better relationship 0:03:00.280 --> 0:03:02.880 or even better relationship they have already with the Trump administration, 0:03:03.560 --> 0:03:07.000 given they've sort of taken ownership through the US government 0:03:07.320 --> 0:03:09.800 of a stake of Intel, and obviously, you know Donald 0:03:09.840 --> 0:03:13.800 Trump has been touting Intel significantly in recent weeks, including 0:03:13.800 --> 0:03:17.200 its stock price and their investment growth given their recent 0:03:17.280 --> 0:03:20.320 market cap. So definitely there's a political factor here. But 0:03:20.360 --> 0:03:23.440 I think the biggest thing is concerns over what happened 0:03:24.040 --> 0:03:27.359 in Taiwan. If there's a situation with China. Obviously geopolitically 0:03:27.360 --> 0:03:30.320 it's not great. Many months ago there were concerns for 0:03:30.360 --> 0:03:33.480 Apple related to tariffs and needing to do more chips 0:03:33.520 --> 0:03:35.120 in the US to avoid those tariffs. 0:03:35.280 --> 0:03:35.880 But I think the. 0:03:35.800 --> 0:03:39.440 Bigger picture overall is wanting additional suppliers for perhaps the 0:03:39.440 --> 0:03:42.560 most critical component in the entire supply chain, even for 0:03:42.640 --> 0:03:46.280 things like displays or speakers. You see Apple using several 0:03:46.280 --> 0:03:49.800 different suppliers to offset supply chain issues. To not have 0:03:50.000 --> 0:03:52.320 all of their eggs in one geo or one basket. 0:03:52.400 --> 0:03:54.760 You never know what could happen. So it's really for 0:03:54.800 --> 0:03:57.200 Apple only smart business to have a backup here. If 0:03:57.240 --> 0:03:59.040 they're not able to get the chips they need out 0:03:59.080 --> 0:04:01.560 of Taiwan, they're not going to have a very good year. 0:04:01.640 --> 0:04:05.600 It's actually probably the biggest risk factor for Apple right 0:04:05.600 --> 0:04:07.960 now is getting all their chips out of Taiwan. So 0:04:08.240 --> 0:04:10.960 expansion from the silicon front is necessary because without the 0:04:10.960 --> 0:04:12.880 silicon the Apple products don't work. 0:04:13.320 --> 0:04:16.280 And it's not just Apple, the entire world relies upon Taiwan. 0:04:16.279 --> 0:04:18.000 About ninety percent of all chips made out of that 0:04:18.080 --> 0:04:21.880 extraordinary reporting As always, thank you Mark German. Now away 0:04:21.880 --> 0:04:25.839 from chips AI heavyweights including Alphabet, Microsoft, and Xai, have 0:04:25.880 --> 0:04:27.680 agreed to give the US government. 0:04:27.360 --> 0:04:29.920 Early access to their models before public release. 0:04:30.279 --> 0:04:32.920 Now, the deal allows the Commerce Department to conduct pre 0:04:33.040 --> 0:04:34.320 release reviews in advance. 0:04:34.600 --> 0:04:37.120 Let's get more with Bloomberg's Maggie Eastland and Adam Washington. 0:04:37.320 --> 0:04:39.880 Now this builds on where we were already going with 0:04:39.920 --> 0:04:41.520 open AI and Anthropic, right. 0:04:43.240 --> 0:04:43.680 Exactly. 0:04:43.720 --> 0:04:47.360 So open ai and Anthropic have already had agreements since 0:04:47.400 --> 0:04:51.000 twenty twenty four with specifically the Center for AI Standards 0:04:51.000 --> 0:04:54.800 and Innovation. That's an office within the Commerce Department that 0:04:54.920 --> 0:04:58.680 knows how to evaluate models. So now you see three 0:04:58.800 --> 0:05:02.039 more firms agree to allow the US to access their 0:05:02.080 --> 0:05:04.920 models early. Now, this does signal that there is at 0:05:04.960 --> 0:05:08.400 the very least appetite from US officials to know what 0:05:08.520 --> 0:05:12.080 these models are capable of before they reach a wider audience. 0:05:13.120 --> 0:05:17.000 We have seen more relationships, well more formalizing, more updated 0:05:17.080 --> 0:05:19.960 relationships being signed between the likes of Alphabet and the 0:05:20.080 --> 0:05:23.360 US government. We know XAI has been working within it, 0:05:23.600 --> 0:05:26.359 and we're starting to see it really being deployed in 0:05:26.400 --> 0:05:29.400 the departments of War or Defense, which, of a way, 0:05:29.520 --> 0:05:32.919 termit Maggie, well, broadly, how different is this from the 0:05:32.920 --> 0:05:37.839 way that they operated before. Is this just a formalization, Well, 0:05:37.880 --> 0:05:41.520 it's certainly an expansion of the firms that have agreed 0:05:41.560 --> 0:05:44.479 to allow this center to evaluate their models. 0:05:44.600 --> 0:05:45.960 Now I should note this center it. 0:05:45.920 --> 0:05:49.440 Doesn't do any sort of regulation. So separately, there have 0:05:49.600 --> 0:05:51.920 been reports from the New York Times and the Wall 0:05:51.920 --> 0:05:55.040 Street Journal that the White House is weighing a cybersecurity 0:05:55.320 --> 0:05:59.080 executive order that would include some sort of oversight mechanism. 0:05:59.320 --> 0:06:02.560 Now this news, these agreements could be part of the 0:06:02.600 --> 0:06:05.599 evaluation piece, but it's unlikely to be part of any 0:06:05.600 --> 0:06:09.200 sort of oversight mechanism, even though, of course more evaluation 0:06:09.320 --> 0:06:11.760 of models could pave the way for, at the very 0:06:11.839 --> 0:06:14.800 least new enforcement of existing laws. 0:06:14.839 --> 0:06:18.680 Even most Maggie Eastland always with the latest, Ala Washington 0:06:18.760 --> 0:06:21.600 for us, we appreciate it. Let's look as wels as 0:06:21.640 --> 0:06:24.800 moving in the tech markets. We hire more broadly, and 0:06:24.800 --> 0:06:26.719 that's not one hundred and one point three percent. Guess what, 0:06:26.880 --> 0:06:30.480 yet another record high. But look at the socks. This 0:06:30.560 --> 0:06:34.320 is all about semiconductors continuing to fly higher. We're up 0:06:34.360 --> 0:06:37.279 three point seven percent, yet another record high. We want 0:06:37.279 --> 0:06:40.359 to dig into just the insatiable demand. It feels like 0:06:40.440 --> 0:06:42.320 for tech stock and the bounce back. Lauren Webster's with 0:06:42.400 --> 0:06:45.240 us managing director of Investment Banking focused on technology at 0:06:45.240 --> 0:06:50.359 pypersanla and just break down what is amid geopolitical tensions, 0:06:50.480 --> 0:06:54.760 amid concerns over Iran and the conflict, we're still seeing 0:06:54.839 --> 0:06:56.359 tech managing to catch. 0:06:56.200 --> 0:06:56.760 A bid today. 0:06:57.920 --> 0:07:01.520 Yeah, it's a real turnaround from what we saw to 0:07:01.600 --> 0:07:04.080 start the year in terms of the massive sell off 0:07:04.680 --> 0:07:09.239 driven in part by uncertainty from the introduction of AI 0:07:09.400 --> 0:07:14.760 what that would do to traditional software companies, traditional software stocks. 0:07:15.080 --> 0:07:18.800 In April we really saw the first constructive tape for 0:07:19.000 --> 0:07:23.240 the software indices and moving higher, still lower than the 0:07:23.240 --> 0:07:26.360 broader market, but performing quite nicely. I think this is 0:07:26.400 --> 0:07:29.200 really where we're going to start to see a shift 0:07:29.240 --> 0:07:33.840 from kind of AI panic and uncertainty into AI execution 0:07:34.200 --> 0:07:37.880 and more discernment among investors around who's really going to 0:07:37.880 --> 0:07:40.680 be the winners the beneficiaries from AI and who may 0:07:40.760 --> 0:07:44.280 struggle to perform as we go into this new era. 0:07:44.840 --> 0:07:47.600 It's interesting we have earnings still coming thick and fast, 0:07:47.600 --> 0:07:49.560 and when we're thinking about Paneteer has been the EI 0:07:49.680 --> 0:07:51.480 of the storm in many ways of whether or not 0:07:51.640 --> 0:07:54.440 it's a software exposed stock and AI is going to 0:07:54.880 --> 0:07:57.360 manage to upset it in terms of competition coming from 0:07:57.600 --> 0:07:59.760 the front end model makers or if they're able to 0:08:00.520 --> 0:08:03.600 pass and look, the numbers were again extraordinary, but the 0:08:03.640 --> 0:08:06.880 earnings of down was seeing shopify also LOA. How are 0:08:06.920 --> 0:08:10.680 you discerning who is able to retain a mote? 0:08:11.560 --> 0:08:14.040 Yeah, So there's there's a few things that I look for. 0:08:14.680 --> 0:08:15.480 First is. 0:08:17.120 --> 0:08:20.920 Enterprise adoption, So are you selling more to the SMB 0:08:21.040 --> 0:08:23.760 are you selling more to the enterprise? And enterprise software 0:08:24.280 --> 0:08:26.680 rip and replace is hard to do. These are projects 0:08:26.720 --> 0:08:32.360 that have been budgeted, budgeted years in advance. Implementation timelines 0:08:32.400 --> 0:08:34.880 take a long time. You have a lot of services 0:08:34.920 --> 0:08:38.400 wrapped around that, so you can't rip that overnight. Things 0:08:38.400 --> 0:08:41.160 that are easier to deploy, you may be able to 0:08:41.320 --> 0:08:43.480 turn on a dime a bit more adopt some of 0:08:43.520 --> 0:08:46.439 the newer technology that's coming out of the foundational labs. 0:08:46.480 --> 0:08:47.240 So that's one. 0:08:47.520 --> 0:08:50.080 The second is you mentioned palent Heer, and there's a 0:08:50.120 --> 0:08:55.920 real rotation into the defense tech sector, particularly those names 0:08:56.000 --> 0:08:58.760 that have a software angle to them. So what is 0:08:58.840 --> 0:09:01.920 you know defense tech hardware now software getting a lot 0:09:01.920 --> 0:09:05.560 of focus and I expect will continue to see a 0:09:05.559 --> 0:09:08.240 lot of tailwinds there, given where you start in kind 0:09:08.240 --> 0:09:11.320 of the geopolitical landscape and the money that is pouring 0:09:11.920 --> 0:09:15.760 international security and particularly a new way of doing business 0:09:16.160 --> 0:09:17.439 in defense technology. 0:09:18.120 --> 0:09:20.240 Let's talk about the other theme about as we see 0:09:20.320 --> 0:09:25.400 more efficiency being sought by these companies, Unfortunately that often 0:09:25.440 --> 0:09:28.880 means job cuts. Some of it comes with the view 0:09:29.000 --> 0:09:32.160 that it's because of AI that we're disrupting the labor force. 0:09:32.559 --> 0:09:34.360 Do you believe when we're hearing the latest out of 0:09:34.400 --> 0:09:37.960 coinbase fourteen percent of workers to go, PayPal going to 0:09:38.000 --> 0:09:39.520 be letting go thousands of workers? 0:09:39.840 --> 0:09:40.840 Is this AI washing? 0:09:41.040 --> 0:09:43.960 Is this just a different kind of focus from fintech 0:09:43.960 --> 0:09:44.560 for example. 0:09:45.720 --> 0:09:49.120 You know, there's a couple interesting trends going on there. One, 0:09:49.280 --> 0:09:54.560 you definitely see investors pushing more towards profitability. The ratio 0:09:54.720 --> 0:09:57.560 of growth the profitability has changed over the past year, 0:09:57.960 --> 0:10:01.640 with a focus on profitability matter mattering more than ever 0:10:01.800 --> 0:10:05.000 and certainly looking for ways to make cuts that drive 0:10:05.040 --> 0:10:08.960 that profitability. Businesses are going to do that regardless of AI. 0:10:09.480 --> 0:10:11.319 But yes, AI is also. 0:10:11.040 --> 0:10:15.559 Going to displace certain types of tech workers. But there 0:10:15.600 --> 0:10:18.760 is going to be a new opportunity that opens up. 0:10:18.800 --> 0:10:22.000 One area that I'm really bullish on is the services 0:10:22.000 --> 0:10:27.959 sector that is going to emerge around AI adoption implementation projects. Again, 0:10:28.040 --> 0:10:30.240 some of these are hard to figure out. How do 0:10:30.320 --> 0:10:32.360 I want to put those in place? What are the 0:10:32.400 --> 0:10:35.440 security measures I need to have when I'm implementing AI? 0:10:35.679 --> 0:10:38.040 And that's where you're going to see a big services 0:10:38.080 --> 0:10:40.280 push and that requires some labor to do as well. 0:10:41.360 --> 0:10:45.440 On Webster, PYPASANLA, maybe some music to some of us 0:10:45.440 --> 0:10:47.760 in laborers out there, we appreciate it. 0:10:48.200 --> 0:10:49.679 Now, let's talk about Anthropic. 0:10:49.920 --> 0:10:52.840 It's unveiling new AI agents designed to handle a broader 0:10:52.880 --> 0:10:56.120 mikes of financial services tasks. It's part of the company's 0:10:56.120 --> 0:10:58.720 push to win over wool Street. Now CEO Daria Amide 0:10:59.040 --> 0:11:01.920 will join Jamie din of JP Morgan and an Anthropic 0:11:01.960 --> 0:11:04.880 event in New York today to discuss the move. As 0:11:04.960 --> 0:11:07.880 part of the new offerings, Anthropics says it also is 0:11:08.000 --> 0:11:10.920 enabling claud to work better across third party software like 0:11:11.000 --> 0:11:14.520 Excel or PowerPoint or Outlook from Microsoft, and integrate data 0:11:14.600 --> 0:11:17.480 from partners in the financial services industry such as Dunham, 0:11:17.480 --> 0:11:22.320 Bradstreet or Moody's Coming up. We're here from the CFO 0:11:22.400 --> 0:11:25.800 of GRABS Southeast Asia's leading right hailing and delivery app. 0:11:25.800 --> 0:11:40.520 It's a super app. Sublue Magtech Southeast Asian super ad 0:11:40.559 --> 0:11:44.240 grab top first quarter estimates and shares a jumping after 0:11:44.280 --> 0:11:47.200 a quick dip. This morning, Cfop de Uey joins us 0:11:47.240 --> 0:11:50.079 from Singapore to paint the full picture. What we have 0:11:50.160 --> 0:11:53.640 seen is we're twenty four percent growth in GMV, that 0:11:53.760 --> 0:11:55.800 sort of record. You just seeing that similar sort of 0:11:55.880 --> 0:11:58.400 run rate from revenue growth as well. You've got record 0:11:58.520 --> 0:12:00.520 high users, Peter. What drives that both. 0:12:02.200 --> 0:12:05.040 A lot of very exciting things Caroline. There was what 0:12:05.280 --> 0:12:07.920 usually is Q one a very soft quarter for us, 0:12:08.080 --> 0:12:10.960 just because of Ramadan, the lunar New Year. We saw 0:12:10.960 --> 0:12:13.800 the opposite. It was a very strong quarter. And this 0:12:13.920 --> 0:12:17.160 is despite some of the ad macro economic fuel prices 0:12:17.160 --> 0:12:19.960 that you're seeing towards the March period. But overall it 0:12:20.040 --> 0:12:23.000 was very strong on food, very strong on mobility, and 0:12:23.040 --> 0:12:25.720 also our financial services also where we've clipped over a 0:12:25.760 --> 0:12:28.280 billion dollars in loan dispersal. It was a very just 0:12:28.320 --> 0:12:30.720 strong quarter. And a lot of that Caroline comes to 0:12:31.200 --> 0:12:34.559 just a product that we've been rolling out and affordability 0:12:34.559 --> 0:12:36.600 has been really a critical piece that's really opened up 0:12:36.600 --> 0:12:39.199 the funnel for us that continues to grow the MTUs, 0:12:39.240 --> 0:12:41.880 the users on our platform. We have fifty two million 0:12:42.120 --> 0:12:46.000 monthly transacting users. One in thirteen Southeast Asians are using 0:12:46.000 --> 0:12:49.080 our product now, which is great to see. And this 0:12:49.160 --> 0:12:52.160 has really helped for us to really penetrate the market 0:12:52.200 --> 0:12:52.800 more and more. 0:12:53.480 --> 0:12:55.640 Let's talk about trying to penetrate the market more and 0:12:55.679 --> 0:12:58.720 more when you're also seeing some pressure coming from say 0:12:58.800 --> 0:13:01.080 fuel prices. At the moment, you are a very hard 0:13:01.120 --> 0:13:05.320 a mobility company. How is that going to impact longer terms? 0:13:05.360 --> 0:13:06.880 You think can you do to offset? 0:13:08.720 --> 0:13:08.960 Yeah? 0:13:09.000 --> 0:13:10.480 If you look at the markets that we're operating, we 0:13:10.559 --> 0:13:13.520 operated eight markets in Southeast Asia, not all of them 0:13:13.520 --> 0:13:16.960 are equal. Where certain markets, say the Philippine has been 0:13:17.000 --> 0:13:20.280 the hardest hit where fuel prices has been up two x, 0:13:20.320 --> 0:13:23.080 three x in some cases, when you have the other extreme, 0:13:23.200 --> 0:13:27.800 say Malaysia, Indonesia, where the increase has been very nominal, 0:13:27.920 --> 0:13:31.679 hardly any because just the way the structural macro has 0:13:31.760 --> 0:13:34.640 worked there from a fuel subsidy perspective. So it's very 0:13:34.679 --> 0:13:38.439 different markets to markets. But overall, we're pitching in, we're 0:13:38.520 --> 0:13:40.720 leaning in, and we're helping the drivers. How are we 0:13:40.800 --> 0:13:44.000 doing that we've been In March, we activated a number 0:13:44.000 --> 0:13:48.280 of initiatives where the it's fuel fuel voluchers helping them 0:13:48.360 --> 0:13:51.160 getting discounts at the pump, working with the oil refineries, 0:13:52.080 --> 0:13:58.080 the oil companies. We've also deployed multiple lending programs for 0:13:58.160 --> 0:14:01.280 drivers if they needed. But also at the same time, 0:14:01.400 --> 0:14:04.720 we're continuing to accelerate ev adoption. Actually, it's a great 0:14:05.080 --> 0:14:08.559 opportunity for us to accelerate ev adoption in two wheels 0:14:08.600 --> 0:14:11.120 and four wheels in our markets, and this is now 0:14:11.440 --> 0:14:14.520 an opportunity is to continue to pick up the pace. 0:14:15.440 --> 0:14:18.600 Talking about well picking up the pace, what's been interesting 0:14:18.679 --> 0:14:21.720 is the pace of regulatory change keeps shifting and we're 0:14:21.720 --> 0:14:23.120 seeing that over in Indonesia. 0:14:23.200 --> 0:14:24.400 Could you seem to soothe the. 0:14:24.360 --> 0:14:26.800 Market a lot about what has been a very recent 0:14:26.960 --> 0:14:29.680 change to basically the maximum amount of commission you're able 0:14:29.720 --> 0:14:32.200 to take over in Indonesia to just eight percent. How 0:14:32.200 --> 0:14:34.080 are you able to upset that and still drive for 0:14:34.160 --> 0:14:35.160 ultimate profitability? 0:14:36.800 --> 0:14:40.400 Yeah, So it's putting the perspective our Indonesia, which is 0:14:40.400 --> 0:14:44.800 where the government announced over the weekend the commission structural changes. 0:14:45.640 --> 0:14:49.240 It's roughly about less than six percent of our mobility 0:14:49.360 --> 0:14:53.200 GMB business. Overall, this is for the two Wheels Indonesian 0:14:53.280 --> 0:14:58.080 business and with the changes which we understand where it's 0:14:58.080 --> 0:15:01.760 coming from unexpected. Also at the same time we'll work 0:15:01.800 --> 0:15:03.840 with the government to roll it out. We're still getting 0:15:03.840 --> 0:15:06.680 clarity of what the mechanics are and working with our industry. 0:15:06.720 --> 0:15:10.240 PEE is also across the platform. But if you look 0:15:10.280 --> 0:15:12.720 at what we've been doing with our drivers there, we've 0:15:12.720 --> 0:15:17.040 always been giving benefits. We have medical insurance. We've also 0:15:17.120 --> 0:15:19.400 have incentives for our drivers, so we have a number 0:15:19.400 --> 0:15:22.480 of levers that we could use to actually help neutralize 0:15:22.480 --> 0:15:24.800 some of that. Now it does mean that we need 0:15:24.800 --> 0:15:27.960 to recalibrate some of the business model of two Wheels 0:15:27.960 --> 0:15:30.600 in Indonesia or worth of fair structure. But we want 0:15:30.640 --> 0:15:32.840 to make sure also that we're not just passing on 0:15:32.880 --> 0:15:34.800 to the consumers. That's the last thing we want to do. 0:15:34.840 --> 0:15:36.720 We want to make sure the marketplace can choose to 0:15:36.720 --> 0:15:39.320 be very healthy, but the drivers also benefiting. So we 0:15:39.320 --> 0:15:41.680 have enough leavers in the business Caroline to be able 0:15:41.680 --> 0:15:42.360 to upset this. 0:15:43.600 --> 0:15:46.240 We need to remember your big and fintech. How is 0:15:46.280 --> 0:15:47.520 the consumer feeling right now? 0:15:47.560 --> 0:15:52.040 Briefly, so far, the credit quality in the first quarter 0:15:52.520 --> 0:15:54.800 has been very strong. We haven't seen any impact. 0:15:55.320 --> 0:15:55.600 Now. 0:15:55.640 --> 0:16:00.239 Having said that, we're also very cautious with this macroeconomic situation, 0:16:00.440 --> 0:16:02.960 so we're doing a lot of stress testing. We're putting 0:16:02.960 --> 0:16:05.600 a lot of new variables into the input into the 0:16:05.640 --> 0:16:09.560 model itself, where there it's fuel prices increase, potential inflation, 0:16:09.760 --> 0:16:13.360 also the pot unemployment. All those factors are putting in 0:16:14.160 --> 0:16:16.200 so that we are a little bit more cautious and careful. 0:16:16.200 --> 0:16:19.800 But so far, actually we haven't seen any from the 0:16:19.920 --> 0:16:23.880 stress testing anything that's actually detrimental to the loan quality 0:16:23.920 --> 0:16:27.000 of our book. If anything, the loane disperson continues to 0:16:27.040 --> 0:16:29.800 be very healthy. But also we're very cautious also in 0:16:29.840 --> 0:16:32.360 certain countries where we might tie up the funnel a 0:16:32.360 --> 0:16:34.360 little bit more also just to make sure that our 0:16:34.440 --> 0:16:36.640 risk appetite continues to be very balanced. 0:16:37.440 --> 0:16:38.760 We alwaso balanced with us. 0:16:38.880 --> 0:16:41.480 It's very light late or say early in the morning 0:16:41.520 --> 0:16:43.880 for you that we appreciate you live from Singapore. 0:16:44.240 --> 0:16:45.760 CFO of GRAB coming up. 0:16:46.200 --> 0:16:48.320 We speak with the Pinterest CEO Bill Ready after the 0:16:48.320 --> 0:16:50.080 company beats quarterly sales estimates. 0:16:50.080 --> 0:16:51.720 To this is bring back tack. 0:17:01.400 --> 0:17:04.400 Chess Pinterest best day in a year. We're currently up 0:17:04.440 --> 0:17:07.560 eleven percent. The company reporting first quarter sales the top 0:17:07.600 --> 0:17:12.040 down assessments and forecast revenue exceeding all street expectations too. Already, 0:17:12.240 --> 0:17:14.359 pinterest CEO joins us now exclusively. 0:17:14.400 --> 0:17:16.800 It's great to speak with you, Bill. What's driving the growth? 0:17:18.119 --> 0:17:21.040 Thanks for having me on, So you know, we're really 0:17:21.119 --> 0:17:24.000 encouraged by the growth on the platform. It's eleven straight 0:17:24.080 --> 0:17:27.879 quarters of record high users, ten straight quarters of double 0:17:27.880 --> 0:17:30.159 disser growth in users, and at the core of that 0:17:30.320 --> 0:17:33.160 is that we've really turned Pinterest into an AI powered 0:17:33.200 --> 0:17:38.240 shopping assistant that is really really winning with users, especially 0:17:38.240 --> 0:17:40.240 gen Z that's now more than half our platform and 0:17:40.280 --> 0:17:45.200 our fastest growing demographic pinteresses where gen Z goes to shop, 0:17:45.680 --> 0:17:49.200 and that's been really great for users, and increasingly advertisers 0:17:49.200 --> 0:17:51.160 are leaning into that and you see that reflecting our results. 0:17:51.560 --> 0:17:55.040 So ad targeting improves, we can see more efficiency. But 0:17:55.359 --> 0:17:58.920 look at getting eighty billion monthly searches, as you say, 0:17:59.000 --> 0:18:02.960 record hive users. What is the AI strategy here longer term? 0:18:03.000 --> 0:18:05.719 Because already you're seeing the efficiency in a flywheel. 0:18:06.440 --> 0:18:09.320 That's right. So we've shared that we have more than 0:18:09.359 --> 0:18:13.080 eighty billion searches per month on the platform, which would 0:18:13.080 --> 0:18:15.720 have been unbelievable to say even a few years ago. 0:18:16.320 --> 0:18:18.800 And we've done that by really making it a visual 0:18:18.840 --> 0:18:21.760 first platform. So much of shopping is a visual journey, 0:18:21.760 --> 0:18:23.280 and so we are really focused on that. And I 0:18:23.320 --> 0:18:25.720 think in the world of AI, you've had a lot 0:18:25.720 --> 0:18:28.320 of discussion around general purpose winners, but you're also seeing 0:18:28.320 --> 0:18:30.280 specialization play out, and you're seeing that play out and 0:18:30.520 --> 0:18:33.359 sort of consumer versus enterprise. But even within consumer, there 0:18:33.359 --> 0:18:34.800 are a lot of different use cases and we've been 0:18:34.840 --> 0:18:38.240 really focused on shopping and visual search and discovery, and 0:18:38.359 --> 0:18:40.879 we're using AI to power that really based off the 0:18:40.920 --> 0:18:43.480 curation signal on our platform, and we're seeing that really 0:18:43.520 --> 0:18:47.919 really resonate with users. And O those eighty billion searches 0:18:47.960 --> 0:18:50.120 per month, more than half of them are commercial, which 0:18:50.160 --> 0:18:53.840 is a much more significant skew towards commerciality than most 0:18:53.880 --> 0:18:56.400 chatbots would have and that you would see in sort 0:18:56.440 --> 0:18:59.720 of historical general purpose search. So it really has become 0:18:59.720 --> 0:19:02.000 a great shopping destination for our users. 0:19:02.440 --> 0:19:05.080 It's interesting those chatbots had been sort of some of 0:19:05.080 --> 0:19:08.160 the investor anxiety that this is where competition. 0:19:07.800 --> 0:19:09.840 Is coming from in visual search. It is Google, It 0:19:09.880 --> 0:19:12.720 is opening on the others. But how dependent are you 0:19:12.800 --> 0:19:14.919 on are the frenemies? 0:19:14.960 --> 0:19:18.520 Shall we say, for the large language models, you're training 0:19:18.560 --> 0:19:21.760 your own, but how much you're using underlying models. 0:19:21.520 --> 0:19:24.720 Of others, that's right, So we primarily run our own 0:19:24.840 --> 0:19:27.119 compact fit for purpose models. I think this is one 0:19:27.119 --> 0:19:29.119 of the things that you will see over time is 0:19:29.160 --> 0:19:33.119 there's been so much discussion around AGI and general purpose models, 0:19:33.240 --> 0:19:35.600 but they can be really expensive to run, and we've 0:19:35.600 --> 0:19:38.520 demonstrated it through our own compact fit for purpose models 0:19:38.640 --> 0:19:42.600 as well as leveraging open source and retraining that off 0:19:42.600 --> 0:19:44.959 of our unique data set that we're able to see 0:19:45.280 --> 0:19:49.080 comparable or better results at oftentimes less than ten percent 0:19:49.080 --> 0:19:51.680 of the cost. And in terms of shopping, we've shared 0:19:51.680 --> 0:19:54.560 that we're able to get to thirty percent better relevancy 0:19:54.560 --> 0:19:57.240 of recommendations on shopping for our users than what we'd 0:19:57.240 --> 0:20:00.679 see from leading off the shelf proprietary model, which just 0:20:00.720 --> 0:20:03.960 really gets to the uniqueness of pinterest as a shopping destination. 0:20:04.359 --> 0:20:07.520 And it's important to note that what makes our models 0:20:07.560 --> 0:20:10.399 so powerful, it's not just the models themselves. The AI 0:20:10.480 --> 0:20:13.720 doesn't have taste or style. Humans have tastes in style, 0:20:13.880 --> 0:20:16.320 and so people come to Pinterest to figure out their taste, 0:20:16.600 --> 0:20:18.359 to figure out how they want to put outfits together, 0:20:18.560 --> 0:20:20.160 and we're able to learn from that and make better 0:20:20.200 --> 0:20:22.040 and better recommendations that are aligned. 0:20:21.760 --> 0:20:22.440 To a user's taste. 0:20:22.480 --> 0:20:24.720 And the result of that, as usually say things like well, 0:20:24.760 --> 0:20:27.200 Pinterest just gets me. And that's something that we think 0:20:27.240 --> 0:20:30.159 we're doing really uniquely here at pinterest and where we 0:20:30.240 --> 0:20:32.960 just have totally different signal based on what our users 0:20:33.000 --> 0:20:35.560 do on the platform, and that's what lets us take 0:20:35.960 --> 0:20:39.200 much smaller but really fit for purpose models and deliver 0:20:39.240 --> 0:20:40.280 better results for users. 0:20:40.600 --> 0:20:42.440 It's not just a signal on your platform. 0:20:42.440 --> 0:20:45.159 Though you've been doing M and A talk to us 0:20:45.160 --> 0:20:48.600 about TV scientific and while you're going into performance driven 0:20:48.680 --> 0:20:50.120