WEBVTT - ICBC Cyber Attack and Call of Duty's Modern Warfare 3 0:00:01.440 --> 0:00:02.120 From Marhard. 0:00:02.279 --> 0:00:06.720 We're Innovation, Money and Power Collie in Silicon Valley, NBN. 0:00:07.040 --> 0:00:10.320 This is Bloomberg Technology with Caroline Hyde and. 0:00:10.480 --> 0:00:11.080 Ed Love Love. 0:00:25.079 --> 0:00:28.480 I'm Caroline Heine, Bloomberg's world headquarters in newlourmed Lovelow, also 0:00:28.480 --> 0:00:30.240 here in New York for the final time this week. 0:00:30.240 --> 0:00:32.479 And this is Bloomberg Technology. 0:00:32.040 --> 0:00:34.640 And coming up. And while we're talking the cyber attack 0:00:34.680 --> 0:00:38.160 of course, disrupting the US Treasury market, the world's largest 0:00:38.200 --> 0:00:40.400 bank falling victim to a ransomware hack. 0:00:40.720 --> 0:00:44.080 We have the story plus stability AI getting new backing 0:00:44.159 --> 0:00:47.120 from Intel. We talk about the cash infusion which comes 0:00:47.120 --> 0:00:49.720 at a critical time for the company and AI at large, 0:00:49.760 --> 0:00:52.560 with none other than Andrew Ung, founder of Google Brain 0:00:52.600 --> 0:00:55.279 and co founder of Course and. 0:00:55.200 --> 0:00:58.000 Call of Duty's Modern Warfare two It brought in one 0:00:58.000 --> 0:01:02.040 billion dollars in ten days. Modern Warfare three perform and 0:01:02.120 --> 0:01:05.640 how is a game about war helping real veterans in 0:01:05.720 --> 0:01:08.800 future employment. We talk to Activision Blizzard later in the hour. 0:01:09.200 --> 0:01:13.200 Now our top story ICBC, a China bank subject according 0:01:13.240 --> 0:01:15.520 to sources, to a hack that's a big one that 0:01:15.520 --> 0:01:18.080 we're watching. Why because it's mega You think about the 0:01:18.120 --> 0:01:19.479 assets that bank controls. 0:01:19.760 --> 0:01:22.640 Then it's going to have ramifications. You're right, it's mega. 0:01:22.680 --> 0:01:25.840 To put it into context, the world's biggest bank ICBC, 0:01:26.040 --> 0:01:27.720 and it's been here, as you say, by cyber attack, 0:01:27.800 --> 0:01:31.240 leaving the Chinese bank unable to clear US Treasury trades, 0:01:31.520 --> 0:01:34.640 forcing employees get this to send settlement details through a 0:01:34.720 --> 0:01:38.959 courier carrying a US basticw mertion Hali Bassett going back 0:01:38.959 --> 0:01:41.080 to basics on technology for us for a moment, I 0:01:41.120 --> 0:01:44.960 mean extreme events, mean extreme moments with a courier, But 0:01:45.040 --> 0:01:49.120 ultimately it really reminds us how all important cybersecurity is. 0:01:49.200 --> 0:01:49.840 Right now it is. 0:01:49.880 --> 0:01:53.680 And remember this is not the first significant cyber attack 0:01:53.760 --> 0:01:57.120 we've really seen on some more critical infrastructure in the 0:01:57.120 --> 0:02:00.480 financial industry. You've seen the same group here really target 0:02:00.600 --> 0:02:05.280 or offshoots of this group target a British fintech firm 0:02:05.360 --> 0:02:07.920 earlier this year when it came to the derivatives market. 0:02:08.120 --> 0:02:09.880 But the reason this is such a big deal not 0:02:10.000 --> 0:02:12.600 just is it the world's largest bank by assets, it's 0:02:12.639 --> 0:02:16.480 also the most large liquid safe market in the world 0:02:16.480 --> 0:02:19.400 that was disrupted for a moment yesterday as we saw 0:02:19.480 --> 0:02:20.440 this attack happen. 0:02:20.520 --> 0:02:22.600 Why you were hanging out on social media this morning 0:02:22.600 --> 0:02:25.640 and going back reading about ICBC, going back to twenty 0:02:25.680 --> 0:02:29.760 seventeen and sharing articles put that research into context with 0:02:29.800 --> 0:02:30.560 the news overnight. 0:02:30.680 --> 0:02:32.320 Yeah, the reason that this is such a big deal 0:02:32.400 --> 0:02:35.600 is remembered that the market for treasuries has really brought 0:02:35.639 --> 0:02:39.240 in to include banks like ICBC, a massive Chinese bank 0:02:39.520 --> 0:02:42.359 in the US treasury market. So some of this is 0:02:42.400 --> 0:02:46.000 a treasury market structure issue. This is relevant to your 0:02:46.040 --> 0:02:49.000 audience at and the technology world because there's a large 0:02:49.080 --> 0:02:52.919 debate on how treasury should be cleared. ICBC has become 0:02:52.919 --> 0:02:55.560 a big clearing counterparty as well as large in the 0:02:55.600 --> 0:02:58.919 repo lending market. These are repurchase agreements that really fuel 0:02:58.960 --> 0:03:01.200 the heart of the financial system and leverage in the 0:03:01.240 --> 0:03:04.120 treasury markets. But there's a big debate now on whether 0:03:04.320 --> 0:03:07.600 treasuries should be centrally cleared. This would be through a 0:03:07.720 --> 0:03:10.760 unit of the dtcc A lot of fintech players are 0:03:10.840 --> 0:03:13.560 very closely associated with the DCCC because they care so 0:03:13.680 --> 0:03:17.560 much about the movement of different assets across the United States, 0:03:17.639 --> 0:03:20.560 let alone the world across different types of assets. But 0:03:20.840 --> 0:03:25.160 if they are centrally cleared, then these cyber problems have 0:03:25.400 --> 0:03:28.920 a more concentration risk. Don't they, And so there's a 0:03:28.919 --> 0:03:33.080 big question about what else is vulnerable besides. 0:03:32.840 --> 0:03:36.760 ICBC here and the people behind this ransomware finding it 0:03:36.920 --> 0:03:39.200 like a business extraordinary set of stories being put out 0:03:39.200 --> 0:03:40.800 by bloom bag nus to any Basset, we thank you 0:03:40.880 --> 0:03:43.200 for the inside track of really what it meant for 0:03:43.480 --> 0:03:46.520 a key Chinese lender. And well, let's talk about AI 0:03:46.720 --> 0:03:48.760 in particular when it comes in the context of China. 0:03:48.800 --> 0:03:50.840 A social intelligence that course a big theme in this 0:03:50.880 --> 0:03:52.960 year's New Economy form. It's been held in sing and Paul, 0:03:53.000 --> 0:03:56.560 but BLOOMGX spook exclusively with Cinnovation Ventures chairman and CEO 0:03:56.640 --> 0:04:00.800 Kaifu li Ai pioneer as he's known, is discuss the 0:04:00.840 --> 0:04:04.760 opportunities that opal intelligence can create and the need for regulations. 0:04:04.840 --> 0:04:05.440 Just take a listen. 0:04:06.520 --> 0:04:09.040 We wanted to be a global company. That's why we 0:04:09.160 --> 0:04:12.640 open source our best model to date, and that's what 0:04:12.760 --> 0:04:15.440 we announced this week. It's a thirty four B model 0:04:15.760 --> 0:04:19.120 that is the highest performance open source model for a 0:04:19.200 --> 0:04:24.160 large language model, exceeding methas Lama too by a significant margin. 0:04:24.640 --> 0:04:27.479 We believe open source will bring the world together. We 0:04:27.560 --> 0:04:30.919 think research is always about building on the shoulders of 0:04:31.240 --> 0:04:33.919 standing on the shoulders of giants, and we want to 0:04:33.960 --> 0:04:35.760 contribute from which we took. 0:04:36.000 --> 0:04:39.039 Obviously, there's a lot of talk about this whole I 0:04:39.040 --> 0:04:42.360 guess race between AI, between US and China here, How 0:04:42.360 --> 0:04:44.719 do you see this all playing out overall? And what 0:04:44.800 --> 0:04:46.200 really is at stake right now? 0:04:46.640 --> 0:04:50.479 I think the likelihood, given where things are, which is regrettable, 0:04:50.920 --> 0:04:56.120 is we have two parallel universes. American companies will supply 0:04:56.800 --> 0:05:00.839 their products and technologies to the US and other countries 0:05:01.080 --> 0:05:04.520 that want to use American products, and Chinese companies will 0:05:04.520 --> 0:05:08.560 build Chinese products for China and countries that will use 0:05:08.640 --> 0:05:09.560 Chinese products. 0:05:09.800 --> 0:05:11.520 So the reality is the. 0:05:11.520 --> 0:05:14.560 American companies and Chinese companies are probably not going to 0:05:14.600 --> 0:05:17.640 compete all that much in the same marketplace, but they'll. 0:05:17.480 --> 0:05:21.159 Probably have to compete on chip makers. Right, So the 0:05:21.279 --> 0:05:24.320 US Administration has said that some of the higher tech 0:05:24.520 --> 0:05:27.039 chips cannot be sold to Chinese companies. What does it 0:05:27.120 --> 0:05:28.760 mean for your company? 0:05:29.160 --> 0:05:29.320 Right? 0:05:29.720 --> 0:05:32.919 We saw this coming, so we actually stockpiled a lot 0:05:32.960 --> 0:05:35.880 of Nvidia chips. So we're in good shape for about 0:05:35.880 --> 0:05:39.000 a year and a half, and I think the jury 0:05:39.080 --> 0:05:41.279 is out whether in a year and a half China 0:05:41.360 --> 0:05:44.880 can make equivalently or nearly as good chips. 0:05:45.279 --> 0:05:47.560 Time will tell you're not just targeting China but also 0:05:47.560 --> 0:05:49.880 global markets. And just want to how do you view 0:05:49.960 --> 0:05:53.000 the whole regulatory landscape right now when it comes to AI. 0:05:53.920 --> 0:05:57.200 I think there clearly needs to be regulation because AI 0:05:57.320 --> 0:06:00.400 can be used in ways that are harmful. I also 0:06:00.440 --> 0:06:05.080 think different countries will have somewhat different regulations, but hopefully 0:06:05.080 --> 0:06:09.720 there's enough commonality in them, so regulation is necessary. But 0:06:09.800 --> 0:06:13.680 also I think we shouldn't expect regulation to alone to 0:06:13.720 --> 0:06:18.480 be enough because a lot of these issues hallucination, misinformation, 0:06:19.320 --> 0:06:23.040 harm can and should be measured by technology, and we 0:06:23.080 --> 0:06:27.200 should let the technologies find a way to contain the technology, 0:06:27.839 --> 0:06:30.680 but not completely rely on them either. So it's really 0:06:31.320 --> 0:06:35.960 technology to provide the guardwail rail and the detection and 0:06:36.320 --> 0:06:40.720 removal of harmful and illegal content, but working side by 0:06:40.800 --> 0:06:42.320 side with regulations. 0:06:43.480 --> 0:06:44.800 At what point, I mean, do you think that it 0:06:44.839 --> 0:06:48.279 could actually disrupt innovation in some way? Are you concerned 0:06:48.279 --> 0:06:50.880 that the regulations could actually hurt innovation? 0:06:51.440 --> 0:06:55.840 I think excessive regulation will hurt innovation, and I think 0:06:56.040 --> 0:06:59.520 we as a technology community are willing to take some 0:06:59.680 --> 0:07:03.320 hits and be slowed down somewhat for the good of humanity, 0:07:03.720 --> 0:07:07.320 but not so much so that will completely stifle innovation. 0:07:08.320 --> 0:07:10.240 How difficult is it? And this is one of the 0:07:10.240 --> 0:07:12.920 conversations that we try and have with policymakers. First of all, 0:07:12.920 --> 0:07:17.600 policymakers probably don't regulators don't understand how quickly or how 0:07:17.640 --> 0:07:20.720 slowly this comes compared to someone like you and your expertise. 0:07:21.080 --> 0:07:24.120 So what's the biggest mistake that they could do now? 0:07:25.000 --> 0:07:28.560 Well, I think a moratorium would be a huge mistake. 0:07:28.720 --> 0:07:33.040 It's not enforceable and it's not going to be effective. 0:07:33.080 --> 0:07:36.080 You have to let technology move forward. I think things 0:07:36.080 --> 0:07:41.400 that they should do are to promote letting technologists, fixing 0:07:42.120 --> 0:07:46.200 hallucination and other problems. And I think they should use 0:07:46.360 --> 0:07:52.880 existing laws to punish those who misbehave for example, for 0:07:53.000 --> 0:07:56.360 spreading misinformation. We have laws in every country on that already, 0:07:56.720 --> 0:07:59.480 just use them. 0:08:00.120 --> 0:08:03.480 Exclusive interview with Sinnovation Ventures chairman and CEO Kai Fu 0:08:03.560 --> 0:08:06.680 Lee at the Bloomberg New Economy Forum in Singapore, and 0:08:06.760 --> 0:08:10.240 he referenced in video Caroline. It's the subject of the 0:08:10.280 --> 0:08:13.520 Techwatch column today, talking about how its earnings ratio is 0:08:13.600 --> 0:08:16.480 less than the estimated growth rate and talking about valuation, 0:08:17.240 --> 0:08:19.360 which is an uncommon situation me in But I don't 0:08:19.360 --> 0:08:21.160 know if you saw the news twenty four hours ago 0:08:21.280 --> 0:08:23.720 that they're basically the stock rose because it's working on 0:08:23.800 --> 0:08:27.000 three new AI chips specifically for China, and there Kyfu 0:08:27.080 --> 0:08:29.640 Lee was basically saying, in Video is the only game 0:08:29.680 --> 0:08:31.520 in town right now. Remains to be seen if China 0:08:31.520 --> 0:08:32.440 can make its own chick and. 0:08:32.400 --> 0:08:34.800 The fact they need stockpiled a load ahead of some 0:08:34.840 --> 0:08:37.560 of these restrictions. But going back to whether in video 0:08:37.640 --> 0:08:39.800 is expensive or not, what is it training at thirty 0:08:39.840 --> 0:08:43.600 five time sales, which by is like the most expensive 0:08:43.640 --> 0:08:45.600 on the s and P fie hundred. But many analysts 0:08:45.600 --> 0:08:47.520 are saying this is actually relatively too and that's a 0:08:47.520 --> 0:08:49.720 big right. Meanwhile, look, we just want to talk about 0:08:49.720 --> 0:08:54.040 another publicly traded company, Meta. The company is close, we understand, 0:08:54.160 --> 0:08:56.480 reaching a deal that will bring his mixed reality headset 0:08:56.559 --> 0:08:59.360 to China through a partnership with Tencent. Now calling to 0:08:59.440 --> 0:09:03.280 sources good manufacturer version of a headset using matters hardware, 0:09:03.280 --> 0:09:06.000 specifications and software with Metro and Tencent didn't respond to 0:09:06.040 --> 0:09:10.040 request for common Interesting the Metro actually reaccessing the Chinese 0:09:10.080 --> 0:09:12.679 economy from New York. There's a boom meg technology. 0:09:22.480 --> 0:09:26.760 Stability AI is raising new financing led by chipmaker Intel 0:09:26.800 --> 0:09:30.160 Stability raising just under fifty million dollars in the form 0:09:30.280 --> 0:09:33.600 of a convertible. Note that, according to sources, the cash 0:09:33.640 --> 0:09:36.000 infusion coming at a pretty critical time for the company, 0:09:36.160 --> 0:09:40.640 which has recently shed more than half a dozen senior employees. 0:09:40.679 --> 0:09:42.280 A big story in the world of AI. 0:09:42.400 --> 0:09:45.000 Let's stick with the world of AI and bringing Andrew 0:09:45.120 --> 0:09:48.959 um who founded Google Brain and deep Learning to AI 0:09:49.280 --> 0:09:52.400 and is currently chairman and co founder of course Serah 0:09:52.480 --> 0:09:55.160 and of course an adjunct professor at Stanford in the 0:09:55.160 --> 0:09:59.240 field of AI. And welcome to the program, Andrew, and 0:09:59.280 --> 0:10:02.200 thank you for your time. I want to start with 0:10:02.440 --> 0:10:05.120 the reason that you're so talked about right now in 0:10:05.160 --> 0:10:08.080 the news and the media, online on social media. And 0:10:08.120 --> 0:10:11.320 it's a view that you hold that big tech is 0:10:11.400 --> 0:10:15.480 talking up the threat, the existential threat of AI for 0:10:15.520 --> 0:10:18.960 a particular reason because you outline your thesis to us. 0:10:20.720 --> 0:10:23.720 So there's been a lot of fear about AI creating 0:10:23.760 --> 0:10:27.560 and the existential threat to humanity. But I just don't 0:10:27.600 --> 0:10:30.000 get it. I don't see any plausible path by which 0:10:30.200 --> 0:10:33.840 AI will make humanity go extinct. But the notion that 0:10:33.840 --> 0:10:36.439 they could just treace a lot of fear and unfortunately 0:10:36.880 --> 0:10:40.559 just creates cover for lobbyists as well as some regulators 0:10:40.880 --> 0:10:43.720 to pass laws that I'm concerned will be very cycling 0:10:43.880 --> 0:10:46.480 on innovation and open source, and so I think of 0:10:46.480 --> 0:10:48.000 this in the interest of some of the big tech 0:10:48.000 --> 0:10:52.560 companies to let these regulations be passed to shut up 0:10:52.640 --> 0:10:55.520 open source that some large companies would very rather not 0:10:55.640 --> 0:10:56.160 compete with. 0:10:56.720 --> 0:10:59.000 We've just had the AI summon in the UK, where 0:10:59.040 --> 0:11:01.839 many would say perhaps it upt some of the fear 0:11:01.960 --> 0:11:05.440 mongering our regulations as they're being formed the right ones. 0:11:07.160 --> 0:11:09.560 I'm very concerned about the regulations being formed both in 0:11:09.559 --> 0:11:12.440 the United States as well as in the UK and Europe. 0:11:12.800 --> 0:11:15.480 One distinction, i feel like there's more of the AI 0:11:15.600 --> 0:11:19.120 existential risk fear in Europe than in the United States. 0:11:19.600 --> 0:11:23.160 But I'm concerned about regulations on both sides. So the 0:11:23.160 --> 0:11:25.920 White House had released and executive order just a couple 0:11:25.960 --> 0:11:29.000 of weeks ago in which it's starting to put in 0:11:29.040 --> 0:11:33.240 place reporting requirements and maybe other requirements about models that 0:11:33.280 --> 0:11:36.400 are very powerful, and I'm very concerned if we'll end 0:11:36.440 --> 0:11:39.560 with one a path more open source software or at 0:11:39.640 --> 0:11:43.000 least past regulations that make it so difficult for people 0:11:43.040 --> 0:11:45.680 to write open source software, free software for anyone to 0:11:45.679 --> 0:11:49.080 build on, which has been an amazing contribution to the world, 0:11:49.160 --> 0:11:53.120 and it's a key for how innovators and startups compete 0:11:53.120 --> 0:11:53.560 and build. 0:11:54.200 --> 0:11:55.880 There was a lot of interest from our audience about 0:11:55.880 --> 0:11:57.840 you coming on, and I think probably the most common 0:11:58.160 --> 0:12:01.640 question was for you to ex how you see the 0:12:01.679 --> 0:12:06.440 intersection of deep learning, machine learning, neural networks impacting jobs. 0:12:06.720 --> 0:12:11.000 You know, this debate of displacement versus being an assistant 0:12:11.280 --> 0:12:13.920 based on all of the research and development currently happening. 0:12:15.600 --> 0:12:19.000 So it turns out that AI doesn't automate jobs. I 0:12:19.040 --> 0:12:22.439 know that the whole discussion has been worre air aultimate jobs, 0:12:22.480 --> 0:12:24.960 but I think a more useful analysis is not to 0:12:25.000 --> 0:12:28.080 think of AIR as automating jobs, for it's air automating tasks. 0:12:28.400 --> 0:12:30.880 So what my teams do is we actually take jobs, 0:12:31.360 --> 0:12:35.240 break them down into tasks. For example, are radiologists REUSEICX, 0:12:35.360 --> 0:12:38.200 raised communicates with patients, galicipation history is meant to s 0:12:38.200 --> 0:12:40.640 more junior radiologists, and on and on and on, and 0:12:40.679 --> 0:12:42.920 then to take all those tasks and figure out where 0:12:42.960 --> 0:12:45.920 are the AI opportunities. One of my friends every Brunos 0:12:45.960 --> 0:12:48.480 and an others had pioneer this technique, and it turns 0:12:48.480 --> 0:12:50.760 out that I think on this spot, AI does automate 0:12:50.800 --> 0:12:54.440 a tiny number of jobs in whole, but the majority 0:12:54.440 --> 0:12:57.080 of the time you find that for someone's job, maybe 0:12:57.160 --> 0:12:59.520 I can make them, you know, twenty percent more productive 0:12:59.520 --> 0:13:02.840 of and more productive, which means that they are still needed. 0:13:03.640 --> 0:13:06.880 But with AI they can be more productive. I don't 0:13:06.920 --> 0:13:09.600 mean to minimize the suffering of this much smaller number 0:13:09.640 --> 0:13:12.439 of people whose jobs will go away. But it turns 0:13:12.480 --> 0:13:15.400 out also that if you look at most waves of innovation, 0:13:16.000 --> 0:13:19.400 companies end up pursuing growth more than cost savings because 0:13:19.400 --> 0:13:22.319 you can only save so much money's cost savings is nice, 0:13:22.480 --> 0:13:26.360 but the growth, the revenue growth has no limit, and 0:13:26.440 --> 0:13:28.240 so most companies end up. I think we'll end up 0:13:28.280 --> 0:13:28.840 doing more. 0:13:28.720 --> 0:13:31.400 Of that, and companies need to reskill. And that's where 0:13:31.400 --> 0:13:34.480 Coursera comes in. How much demand have you seen? How 0:13:34.559 --> 0:13:37.920 much ability to satiate that demand have you been able 0:13:37.920 --> 0:13:38.200 to give? 0:13:39.840 --> 0:13:44.280 So I think coursertvet privilege to serve many enterprises wanting 0:13:44.320 --> 0:13:48.000 to reskill, their import their teams. In fact, just about 0:13:48.000 --> 0:13:51.040 a week ago we launched a new course, genitive AI 0:13:51.080 --> 0:13:54.920 for everyone. I think with this fear and uncertainty as 0:13:54.960 --> 0:13:58.360 well as massive upside, we wanted to teach something online 0:13:58.400 --> 0:14:01.480 to help everyone learn to use genetive AI in your 0:14:01.480 --> 0:14:04.240 own day to day life, for personal things or for work, 0:14:04.559 --> 0:14:06.640 as well as figure out how does impact your job 0:14:06.720 --> 0:14:10.560 or your business, So genetally for everyone. Launching Coosera just 0:14:10.559 --> 0:14:12.840 over a week ago, it was actually the fastest rain 0:14:12.880 --> 0:14:15.240 calls on the Polsera this year, with I think forty 0:14:15.320 --> 0:14:18.440 three thousand learners signed out in the first seven days. 0:14:18.559 --> 0:14:19.600 So I think there's a lot. 0:14:19.440 --> 0:14:23.560 Of interest to learn about GENITIVII and to understand how 0:14:23.560 --> 0:14:26.200 to use it yourself, how this may affect your job, 0:14:26.320 --> 0:14:28.440 and how does may affect your business and your industry. 0:14:28.520 --> 0:14:31.800 Vertical Andrew, this week we've been digesting Rock and the 0:14:31.840 --> 0:14:34.920 work that you Musk's done in generative AI. 0:14:35.160 --> 0:14:36.200 What is your assessment? 0:14:37.960 --> 0:14:40.320 You know, I saw it dropped from about not played 0:14:40.360 --> 0:14:42.600 with it myself yet. I think this is a sign 0:14:42.680 --> 0:14:46.200 of the rapid explosion of AI technology. Just last week, 0:14:46.320 --> 0:14:49.480 open Eye had a developer conference. That's exciting even I 0:14:49.560 --> 0:14:52.160 must release drop. I think you just had a typeoly 0:14:52.240 --> 0:14:56.200 on that released another open source model. So this explosion 0:14:56.320 --> 0:15:00.520 of technology tools and what would happen is with better tools, 0:15:00.560 --> 0:15:04.560 it opens up opportunities for businesses and startups to build 0:15:04.600 --> 0:15:07.120 a lot more applications that just were not possible before. 0:15:07.880 --> 0:15:11.960 And so I think that we're better tools the application layer. 0:15:12.320 --> 0:15:14.760 You know, not the media tends to talk about tools 0:15:14.760 --> 0:15:17.120 because you tech out here. That's exciting thing people want 0:15:17.120 --> 0:15:19.360 to hear from the two in companies. But two in 0:15:19.440 --> 0:15:23.040 companies will be successful only if the applications built on 0:15:23.080 --> 0:15:25.560 top of them are even more successful and generally even 0:15:25.600 --> 0:15:27.360 more revenue so they can afford to pay the. 0:15:27.400 --> 0:15:28.320 Two in companies. 0:15:28.720 --> 0:15:30.760 So while I think companies are in the year are 0:15:30.880 --> 0:15:34.720 brilliant and AMD and Intel for them to succeed, someone 0:15:34.760 --> 0:15:37.520 has made even more money to keep paying for these gps, 0:15:37.640 --> 0:15:39.440 and so I'm excited the applications. 0:15:39.520 --> 0:15:43.680 We optimism of course, Sarah chairman, co founder and as 0:15:43.760 --> 0:15:46.360 well as a wealth of other titles, professor, We thank 0:15:46.400 --> 0:15:56.800 you for your time. The Revenue Intelligence platform Gone and 0:15:56.800 --> 0:15:59.920 it's launching a new AI powered forecasting tool to help 0:16:00.200 --> 0:16:03.560 the CFOs CROs out there, basically four cost revenue bringing 0:16:03.560 --> 0:16:06.240 the understanding game from get this more than three hundred 0:16:06.240 --> 0:16:09.560 buying signals across more than three billion customer interactions as 0:16:09.560 --> 0:16:12.680 you captured on Comm's platform. I'm at Bendorf, you to say, 0:16:12.680 --> 0:16:15.400 CEO co Foundra Gong joins us. Now, So how much 0:16:15.440 --> 0:16:17.360 more efficient at predicting does this make people? 0:16:19.840 --> 0:16:23.040 Our testing have shown that it's twenty one percent on 0:16:23.120 --> 0:16:28.680 average more effective and more accurate than traditional methods. Traditionally, 0:16:28.680 --> 0:16:33.120 companies has been aligned on statistical models in some internal systems, 0:16:33.640 --> 0:16:37.520 using the voice of the customers and customer sentiment. With AI, 0:16:38.080 --> 0:16:39.440 it creates much better results. 0:16:40.480 --> 0:16:43.280 I mean, who are we comparing this with? Is this 0:16:43.440 --> 0:16:48.120 like sort of lead generation chasing al CRM and salesforce 0:16:48.480 --> 0:16:50.080 or is this kind of more like data bricks and 0:16:50.120 --> 0:16:52.800 the Data Lake product that they're doing out there with prediction? 0:16:52.880 --> 0:16:56.440 Who are you trying to be here? It's neither. 0:16:57.160 --> 0:17:01.680 Traditional companies have been using spreadsheets to understand how much 0:17:01.720 --> 0:17:05.040 they're going to sell this quarter or next quarter, using 0:17:05.080 --> 0:17:10.920 a variety of sources anywhere from internal systems, weather forecast, 0:17:11.920 --> 0:17:17.960 traditional sales records, and mostly sentiment of the sales departments 0:17:18.240 --> 0:17:21.360 the best judgments, which isn't very good, especially like in 0:17:21.400 --> 0:17:26.200 these challenging times with the macro and geopolitical situation. What 0:17:26.600 --> 0:17:30.480 GONG does. It taps into conversations of companies with their 0:17:30.520 --> 0:17:34.240 customers and it listens to what customers are saying, are 0:17:34.240 --> 0:17:38.560 they likely to buy, what's their sentiment, what are the 0:17:38.600 --> 0:17:41.840 budgets that they have, and turns all this into insights 0:17:42.280 --> 0:17:45.760 that are immediately fed into predictions, so companies can understand 0:17:45.800 --> 0:17:48.560 their changes to the forecast in near real time. 0:17:49.240 --> 0:17:50.400 What's your mote hair? 0:17:50.600 --> 0:17:53.360 You can obviously integrate salesforce data within GONG for example. 0:17:53.800 --> 0:17:55.920 How do you ensure that some of these big pairs 0:17:55.960 --> 0:17:58.399 don't unable to replicate what you've just built? 0:18:00.480 --> 0:18:05.160 Well, everybody can can try to replicate anything, but we're 0:18:05.240 --> 0:18:09.440 sitting on over like three billion customer interaction. We've been 0:18:09.480 --> 0:18:14.520 training the models for years specifically for that particular case, 0:18:14.800 --> 0:18:16.720 and it's not something that is easy to replicate. 0:18:17.600 --> 0:18:20.840 So, I mean, the aim is to improve the accuracy 0:18:20.840 --> 0:18:21.680 of forecasting. 0:18:22.320 --> 0:18:24.639 How does it work? How does the how does AI 0:18:24.920 --> 0:18:25.639 help you do that? 0:18:27.359 --> 0:18:27.560 Well? 0:18:27.600 --> 0:18:32.640 You need to understand scan through vast amounts of customer information. 0:18:33.720 --> 0:18:36.280 For every customer that the company has, what are they 0:18:36.320 --> 0:18:38.520 saying or whether they're not asking If they're not asking 0:18:38.520 --> 0:18:41.440 about pricing, maybe that's a negative signal. If they're talking 0:18:41.440 --> 0:18:45.360 about like a frozen budget. That's a negative signal. All 0:18:45.400 --> 0:18:48.600 of that is understood by the AI to create a 0:18:48.680 --> 0:18:51.359 model that actually predicts what the projection is going to be. 0:18:51.400 --> 0:18:54.560 So you look at customer by customer bottom up and 0:18:54.640 --> 0:18:56.640 understand how likely they are to buy, how much you're 0:18:56.760 --> 0:18:59.399 likely to spend, and then you combine all of that 0:18:59.480 --> 0:19:04.240 into prediction. The importance is that companies can now use 0:19:04.240 --> 0:19:08.960 that either to upgrade, update their forecast, to plan their 0:19:09.040 --> 0:19:12.080 hiring and their spend Accordingly. 0:19:12.040 --> 0:19:15.639 You mentioned geopolitics making it harder to forecast. Now we 0:19:15.680 --> 0:19:17.720 only have a minute left, but I just want to 0:19:17.760 --> 0:19:20.640 ask you, how are your workforce currently in Israel? How 0:19:20.720 --> 0:19:24.040 is the company able to ensure protection of this time? 0:19:25.359 --> 0:19:28.719 Well, overall, the impact on the business has been minimal. 0:19:28.800 --> 0:19:32.720 We have about ten percent of people being drafted. But 0:19:32.760 --> 0:19:35.639 everybody's working extra hard and mostly it's like very difficult 0:19:35.640 --> 0:19:39.280 for people. Everybody is still traumatized by the atrocity of 0:19:39.320 --> 0:19:43.000 October seventh, and there are still over two hundred hostages 0:19:43.080 --> 0:19:46.000 with babies. So these are difficult times, but the spirit 0:19:46.040 --> 0:19:51.399 is super high. The amount of volunteering and citizen activity 0:19:51.680 --> 0:19:54.200 is super high, and. 0:19:54.080 --> 0:19:55.360 We thank you for your time as well. 0:19:55.400 --> 0:19:58.280 I'm at Bendoff CEO and co founder have gone. 0:19:58.320 --> 0:20:07.920 Thank you, Welcome back to Bluemo Technology. I'm Caroline had 0:20:07.960 --> 0:20:08.440 and Neil. 0:20:08.280 --> 0:20:11.440 And Imed Ludlow in San Francisco. There are more. Oh 0:20:11.480 --> 0:20:14.119 my goodness, have fallen for it again? You got me. 0:20:14.400 --> 0:20:16.960 I'm here in New York City. Here are some technology 0:20:17.040 --> 0:20:20.200 stories moving in the markets. Let's start with Trade Desk, 0:20:20.440 --> 0:20:23.760 a two day chart for a visual representation of what's 0:20:23.760 --> 0:20:26.359 a big story. Down eighteen percent in the session, on 0:20:26.440 --> 0:20:28.680 track for its biggest drop since May of twenty twenty two. 0:20:28.960 --> 0:20:31.680 This is an ad technology platform. It gave a week 0:20:31.720 --> 0:20:34.359 forecast for the current period. It also hit the names 0:20:34.440 --> 0:20:37.400 like Meta Snap that have such AD dependent businesses. 0:20:37.400 --> 0:20:38.160 But it's a big move. 0:20:38.320 --> 0:20:40.720 It's got us thinking about the health of the advertising 0:20:40.760 --> 0:20:43.200 business overall. A big theme for the show's day is 0:20:43.200 --> 0:20:45.680 also going to be Alphabet, the parent company of Google. 0:20:46.080 --> 0:20:47.399 We know that some of the pitch I is going 0:20:47.440 --> 0:20:50.639 to testify next week. That's more around the app store story, 0:20:50.640 --> 0:20:52.720 but in the European Union this carries about to tell 0:20:52.800 --> 0:20:55.440 us there is also a lot of regulatory scrutiny as well, 0:20:55.720 --> 0:20:58.240 and thinking what does Google do? Next to kind of 0:20:58.280 --> 0:21:01.320 revamp its business in that market. Here I am in 0:21:01.320 --> 0:21:02.280 New York City, and. 0:21:02.320 --> 0:21:05.280 Let's talk about the EU for a moment, because look, 0:21:05.320 --> 0:21:08.360 Google has been pressured by the European Union regulators. It's 0:21:08.359 --> 0:21:10.399 made an effort to beef up the quality of its 0:21:10.400 --> 0:21:11.160 search engine. 0:21:11.280 --> 0:21:12.600 At least that's the word. 0:21:12.920 --> 0:21:15.640 According to the internal documents by the US Justice Department, 0:21:15.920 --> 0:21:19.919 DJ claims Google's failure to willingly make improvements that it 0:21:20.000 --> 0:21:23.200 proves it's illegally maintaining and monopoly. This comes them in 0:21:23.280 --> 0:21:26.200 news at Alphabet CEO, so which I will testify next 0:21:26.200 --> 0:21:29.560 week in a separate anti trust fight with Epic Gains. Boy, 0:21:29.680 --> 0:21:31.680 that's what going on. So Fordan can help us break 0:21:31.720 --> 0:21:34.680 it down. So what are we on the ninth week 0:21:35.040 --> 0:21:38.320