WEBVTT - AI-Generated Misinformation Is a Top Concern in Davos 0:00:00.440 --> 0:00:02.320 We're going to start the recording in just a second. 0:00:03.000 --> 0:00:04.720 But we're so happy to be here. It's exciting. 0:00:05.000 --> 0:00:07.560 It's very exciting to be here in our first ever 0:00:08.000 --> 0:00:11.240 Bloomberg house, this beautiful space in the city. I think 0:00:11.320 --> 0:00:13.680 it's in the chalet. I think, isn't it the is 0:00:13.720 --> 0:00:16.040 I think what we should be calling. But it does 0:00:16.079 --> 0:00:21.040 feel like the whole world is sort of ascending here, economics, business, finance, 0:00:21.160 --> 0:00:22.119 but most of all tech. 0:00:22.320 --> 0:00:22.480 Right. 0:00:22.560 --> 0:00:25.040 I took a walk down the promenade there and it 0:00:25.160 --> 0:00:28.320 is a little bit like sort of AI trade fair. Right, 0:00:28.360 --> 0:00:32.240 every single shot front is trumpeting something about AI. So 0:00:32.400 --> 0:00:34.680 something we're going to get into a bit of a conversation. 0:00:42.680 --> 0:00:45.000 Right, So we're starting Welcome to in the City, Everyone, 0:00:45.040 --> 0:00:48.800 Bloomberg's podcast, connecting you to the conversations and the stories 0:00:48.800 --> 0:00:50.040 and shaping the world of finance. 0:00:50.080 --> 0:00:52.879 And Franci Laqua and I am David Merritt. And this 0:00:52.960 --> 0:00:57.880 week we are in the mountains of Switzerland in Davos 0:00:57.920 --> 0:01:01.080 as global leaders and quite a few journalists descends on 0:01:01.120 --> 0:01:05.280 this outpoint village for the World Economic Forums annual meeting, 0:01:05.600 --> 0:01:07.600 and we're going to be looking at the challenges facing 0:01:07.640 --> 0:01:10.919 the world, of which there are many in twenty twenty 0:01:10.920 --> 0:01:13.240 four and talk a little bit about how prepared we 0:01:13.319 --> 0:01:14.399 are to face them. 0:01:14.640 --> 0:01:17.680 Yeah, so with us, and we are really delighted and honored. 0:01:17.720 --> 0:01:21.120 Saja Sahidi from the World Economic Forum, a Zimazark, creator 0:01:21.160 --> 0:01:24.320 of the Exponential View of Global platform for in depth 0:01:24.440 --> 0:01:27.960 tech analysis, and the host of Bloomberg original series Exponentially, 0:01:28.000 --> 0:01:32.120 So thank you both for joining us. Sajia, you actually 0:01:32.160 --> 0:01:34.840 published a report looking through the global risks, the top 0:01:34.840 --> 0:01:37.839 global risks in the next twelve months and then longer term. 0:01:38.200 --> 0:01:39.919 Can you run us through your top risks? 0:01:40.280 --> 0:01:43.000 Yeah? Sure, Well, thank you for having me number one 0:01:43.200 --> 0:01:46.920 miss and disinformation and to some extent, we were surprised 0:01:46.959 --> 0:01:49.480 this was down at number sixteen last year and it 0:01:49.640 --> 0:01:52.960 shot up to becoming the number one risk so quickly, 0:01:53.440 --> 0:01:56.960 and of course understandable because we've all seen low cost 0:01:57.040 --> 0:01:59.680 or zero cost technologies in our hands that can produce 0:01:59.680 --> 0:02:01.720 a lot of synthetic content. So a lot of concern 0:02:01.800 --> 0:02:04.400 around that, especially in a year where so many people 0:02:04.400 --> 0:02:07.200 around the world are going into elections following on right 0:02:07.240 --> 0:02:11.960 from that extreme weather, beyond that, societal polarization and obviously 0:02:12.000 --> 0:02:15.200 that is deeply connected to what's happening with miss and disinformation, 0:02:15.480 --> 0:02:18.280 but the economic risks are not that far behind. There's 0:02:18.360 --> 0:02:22.680 still concern about inflation, there's still concern about the potential 0:02:22.760 --> 0:02:25.680 of an economic downturn. That's still number nine in the 0:02:25.720 --> 0:02:29.120 two year timeframe, So looking out over the next twenty 0:02:29.120 --> 0:02:31.000 four months, quite a bit of concern about that. And 0:02:31.000 --> 0:02:33.720 then you look ten years out and it's really all 0:02:33.760 --> 0:02:38.800 about climate, so extreme weather, potential changes to the Earth's systems, 0:02:39.560 --> 0:02:43.680 loss of biodiversity, and a lot of concern around loss 0:02:43.720 --> 0:02:48.080 of natural resources. And again they're miss and disinformation. And 0:02:48.120 --> 0:02:51.440 what rises really to the top is other adverse effects 0:02:51.480 --> 0:02:54.480 of artificial intelligence. That's very low down in the two 0:02:54.560 --> 0:02:57.320 year timeframe, number twenty nine out of thirty four and 0:02:57.360 --> 0:02:59.280 then jumps up into the top ten in the ten 0:02:59.360 --> 0:02:59.960 year timeframe. 0:03:00.160 --> 0:03:02.320 And as in, can we even tell when it's fake 0:03:02.400 --> 0:03:05.919 news or fake information and the real one? Given how 0:03:05.960 --> 0:03:07.679 good some of the AI stuff is right now. 0:03:07.840 --> 0:03:09.640 Oh, I think one of the reasons why it's jumped 0:03:09.720 --> 0:03:11.720 up so much is that this past year is in 0:03:11.720 --> 0:03:14.680 the year of generative AI, and it's getting better and better. 0:03:15.960 --> 0:03:18.560 Although and there is an important although it's one thing 0:03:18.560 --> 0:03:20.280 for it to be created it's another thing for it 0:03:20.320 --> 0:03:23.040 to be distributed, and it's another thing for people to 0:03:23.320 --> 0:03:26.239 read it, and it's another thing for people to believe 0:03:26.280 --> 0:03:28.960 in it. And I do agree with the experts who 0:03:28.960 --> 0:03:34.160 contributed to the Forum's Risk Report that misinformation and disinformation 0:03:34.240 --> 0:03:37.480 should be really high off our agenda this year. But 0:03:37.480 --> 0:03:40.240 I think it's also important to recognize that the academics 0:03:40.280 --> 0:03:44.360 are still to pronounce as to whether these things really 0:03:44.480 --> 0:03:47.560 are real, and a lot of the worries we had 0:03:47.560 --> 0:03:51.560 around in misinformation disinformation stemming from as early as twenty 0:03:51.600 --> 0:03:55.960 sixteen seem to have mixed evidence as to whether that's really. 0:03:57.200 --> 0:03:57.560 The path. 0:03:57.640 --> 0:03:59.680 And I think one of the risks that we run 0:03:59.880 --> 0:04:05.520 is not that we don't face a more challenging environment 0:04:05.520 --> 0:04:09.160 as it pertains to truth and fact. It's that the 0:04:09.200 --> 0:04:12.600 AI component is only one part of it. There are 0:04:12.640 --> 0:04:16.160 other aspects in this ecosystem that are weak politicians willing to. 0:04:16.120 --> 0:04:18.960 Exploit these types of stories. 0:04:19.000 --> 0:04:23.080 There are the distribution patterns that occur over social network. 0:04:23.279 --> 0:04:27.880 There's the capacities of media organizations to verify and validate 0:04:27.880 --> 0:04:30.360 this stuff, or even sometimes, as we've seen in the 0:04:30.440 --> 0:04:33.719 last few months, spread the misinformation themselves. So when we 0:04:33.760 --> 0:04:36.080 look at this problem. If we cry wolf and simply 0:04:36.080 --> 0:04:39.440 point towards AI as a source of mis and disinformation, 0:04:39.520 --> 0:04:41.120 will issue the wrong prescription. 0:04:41.839 --> 0:04:44.080 You know, in the podcast we had last week with 0:04:44.160 --> 0:04:46.160 Ian Bremer for the Eurasia, we were talking about their 0:04:46.240 --> 0:04:49.560 risks as well, and he said something which struck in 0:04:49.560 --> 0:04:51.520 my mind. Their top risk, by the way, was the 0:04:51.600 --> 0:04:55.159 United States versus itself thinking about the election and the 0:04:55.160 --> 0:04:57.680 potential outcomes. But he said, in the last twenty years, 0:04:57.760 --> 0:05:01.160 the United States has gone from being the chief export 0:05:01.240 --> 0:05:04.920 of democracy in the world the chief exporter of tools 0:05:04.920 --> 0:05:08.880 to undermine democracy. And he's talking obviously about social media 0:05:08.960 --> 0:05:11.239 and the growth of that is the way that people 0:05:11.240 --> 0:05:13.880 consume their news. Do you agree with that, a z 0:05:13.920 --> 0:05:16.320 eem And does that answer the question of why this 0:05:16.720 --> 0:05:19.719 particular concern is shot right up to the top again. 0:05:19.600 --> 0:05:23.039 Well, the US has exported tools like that, but the 0:05:23.120 --> 0:05:29.080 US also has companies like Google and Microsoft and Meta 0:05:29.320 --> 0:05:33.800 who have election integrity teams who are working on sharing 0:05:34.680 --> 0:05:39.040 with their their private competitors, but also with the public 0:05:39.080 --> 0:05:44.719 sector tools and technologies for video stamping and authentication and watermarking, 0:05:45.279 --> 0:05:48.240 and they have incident responses as well. So I think 0:05:48.240 --> 0:05:51.600 there is an important part of this picture, which is 0:05:51.600 --> 0:05:55.320 is that sufficient though I mean, well, I mean, you know, 0:05:55.400 --> 0:05:57.800 I think that the trouble is that you see the 0:05:57.839 --> 0:05:59.600 fire and then the fire and arrive. But I think 0:05:59.600 --> 0:06:01.520 we can say that when it came to extreme content 0:06:01.560 --> 0:06:06.080 on YouTube, YouTube or alphabet was able ultimately to do 0:06:06.120 --> 0:06:09.000 a great job in in tackling it. And I think 0:06:09.040 --> 0:06:11.320 that it's really telling that a number of the large 0:06:11.600 --> 0:06:15.800 technology platform firms have already announced that they have election 0:06:16.000 --> 0:06:20.960 integrity strategies in place, ahead of actually what I've heard 0:06:20.960 --> 0:06:24.560 from Nation States themselves saying. And so I think, yes, 0:06:24.680 --> 0:06:27.400 of course technology is a problem, but it can also 0:06:27.440 --> 0:06:28.599 be part of the solution. 0:06:29.040 --> 0:06:29.440 Zadia. 0:06:30.080 --> 0:06:32.040 I mean just one other point that I think it's 0:06:32.040 --> 0:06:36.359 also about how people are receiving that information. It's about 0:06:36.400 --> 0:06:39.000 media literacy as well. And we also went a little 0:06:39.000 --> 0:06:42.960 deeper into different countries and in the US, miss and 0:06:43.000 --> 0:06:45.839 disinformation doesn't show up as a number one risk despite 0:06:45.920 --> 0:06:49.960 being in that election year. In India and in Pakistan 0:06:50.160 --> 0:06:52.920 also having elections this year, it's among the top five risks. 0:06:52.920 --> 0:06:56.440 So it's also a little bit about what citizens are 0:06:56.680 --> 0:06:59.880 able to consume, how much media and digital literacy they have, 0:07:00.400 --> 0:07:02.960 and so it pops up much higher in certain parts 0:07:02.960 --> 0:07:03.400 of the world. 0:07:04.240 --> 0:07:06.200 But when you look at, for example, the US election, 0:07:06.360 --> 0:07:10.240 I guess the concern is meddling right from other countries 0:07:10.880 --> 0:07:14.400 and people are worried because this could spread. I mean, again, 0:07:14.400 --> 0:07:16.360 in your research, can you give us a little bit 0:07:16.600 --> 0:07:18.280 of a worst case scenario. 0:07:18.200 --> 0:07:21.800 On the US based risk? Almost all were economic, and 0:07:21.840 --> 0:07:23.800 that seems to be top of mind. But I think 0:07:23.800 --> 0:07:26.640 there's a connection between those two things. Where you have 0:07:26.720 --> 0:07:30.720 economic hardship, where you also have societal polarization. Those two 0:07:30.760 --> 0:07:36.920 things together provide pertile ground for disinformation, for misinformation, whether 0:07:37.040 --> 0:07:40.160 that is state sponsored or whether that's spread through other sources. 0:07:40.480 --> 0:07:41.960 I mean, the problem, Dave, is that I look at 0:07:42.000 --> 0:07:44.560 so I'm on social media a lot, probably too much, 0:07:45.640 --> 0:07:47.520 and a lot of the things I watch I don't 0:07:47.520 --> 0:07:49.800 really I mean, I know because I work at Bloomberg 0:07:49.920 --> 0:07:52.560 that it's fake information, but I don't see the watermarks. Frankly, 0:07:52.640 --> 0:07:55.840 I don't see like in the stuff I consume. You 0:07:55.880 --> 0:07:56.840 don't know how news works. 0:07:56.840 --> 0:07:59.600 Well, I'm saying more now is people saying images of 0:07:59.680 --> 0:08:02.640 things and instantly saying, well, that's just a I or 0:08:02.680 --> 0:08:05.440 that's generally that's fake. There's a kind of there's a 0:08:05.480 --> 0:08:10.080 skepticism now led on everything and news reporting, journalism that 0:08:10.240 --> 0:08:12.640 wasn't there, particularly things like images, which in the past 0:08:12.720 --> 0:08:15.800 people would trust, and now that trust is really eroding. 0:08:15.840 --> 0:08:18.000 Well, I mean the Pope wasn't worrying about. 0:08:18.480 --> 0:08:21.400 Right, Puffer, we sure about this? 0:08:22.080 --> 0:08:24.200 I mean, you're no, franc I would say if anyone. 0:08:24.960 --> 0:08:27.560 But I think that that that takes us to a 0:08:27.600 --> 0:08:32.280 really really important risk, which is that by by flooding 0:08:32.640 --> 0:08:34.960 the phrase was flooding the zone, right, With flooding the 0:08:35.040 --> 0:08:37.719 zone with material that might or might not be believable, 0:08:37.800 --> 0:08:40.080 or might or might not be true, what you end 0:08:40.160 --> 0:08:43.319 up doing is weakening the trust in the few institutions 0:08:43.760 --> 0:08:47.000 that are are arbiters of that factuality. Now, I think 0:08:47.000 --> 0:08:49.800 Bloomberg and some of the other media businesses generally do 0:08:49.880 --> 0:08:53.520 quite a good job, but the ops now a nice 0:08:53.559 --> 0:08:58.040 cookie earlier on. But the onus is going to be 0:08:58.120 --> 0:09:01.280 on you and your peers as well to increase your 0:09:01.320 --> 0:09:06.000 capabilities of identifying what these issues are, raising the level 0:09:06.040 --> 0:09:09.960 of factuality across this information ecosystem, because if we start 0:09:10.000 --> 0:09:14.760 to distrust the likes of the story newspapers, because they 0:09:14.800 --> 0:09:16.920 do make mistakes from time to time, and because the 0:09:17.240 --> 0:09:20.760 zone is full of synthetic material, I think we then 0:09:21.040 --> 0:09:23.080 do make ourselves susceptible in the way that some of 0:09:23.080 --> 0:09:27.240 the populations that Salia alluded to are more susceptible to misinformation. 0:09:28.200 --> 0:09:29.960 Is it because it's such a big political year this 0:09:30.040 --> 0:09:30.440 as well? 0:09:30.480 --> 0:09:33.480 Like people, I mean, everyone's talked about this, the Great 0:09:33.640 --> 0:09:37.880 Democratic Exercises, there's twenty twenty four, huge amounts of the 0:09:37.880 --> 0:09:39.960 world are going to the polls. Is that why this 0:09:40.120 --> 0:09:42.040 is such a hot topic right now? 0:09:43.520 --> 0:09:45.559 I think that's certainly some of what we've picked up 0:09:45.640 --> 0:09:48.720 from the experts that the fifteen hundred or so experts 0:09:48.760 --> 0:09:51.240 that fed into this, as to the reason why this 0:09:51.360 --> 0:09:53.960 raise up to the top. But I think the longer 0:09:54.080 --> 0:09:58.720 term sort of erosion in trust in truth and objectivity 0:09:58.800 --> 0:10:01.720 in facts, I think they're it goes beyond that particular 0:10:01.760 --> 0:10:05.400 election year. It can affect everything from you know, health 0:10:05.440 --> 0:10:07.400 and I think we've seen that through the pandemic. It 0:10:07.400 --> 0:10:10.920 can affect what we believe in terms of social justice, 0:10:10.960 --> 0:10:13.680 It can affect what people believe on so many different topics. 0:10:13.760 --> 0:10:16.320 So I think it affects more. And of course climate 0:10:16.400 --> 0:10:19.160 change is another big one where there needs to be 0:10:19.240 --> 0:10:23.439 some agreement around science and objectivity in facts. 0:10:24.360 --> 0:10:28.360 AI is everything from I guess if you're you know, 0:10:28.440 --> 0:10:30.960 googling something like color matching, if you need to get 0:10:30.960 --> 0:10:34.000 a foundation, to people saying actually, you don't put the 0:10:34.120 --> 0:10:37.000 human you know, in the middle of this technology anymore. 0:10:37.000 --> 0:10:39.280 And so we could literally end up like Terminator two 0:10:39.720 --> 0:10:41.920 or three or I don't know how many terminators they did, 0:10:42.000 --> 0:10:45.920 like where are you? And how ugly AI could become 0:10:45.960 --> 0:10:47.120 for humankind. 0:10:47.600 --> 0:10:50.559 Well, I think we do that partly because it's more 0:10:50.600 --> 0:10:53.000 interesting to have lurid conversations like that than to talk 0:10:53.040 --> 0:10:56.959 about questions like workforce transformation. As we start to automate 0:10:56.960 --> 0:11:00.520 certain business processes. I mean one is it feels offices 0:11:00.559 --> 0:11:03.319 and you know the other films consulting firms, So we 0:11:03.440 --> 0:11:05.640 need have to to pick you pick your stories. But 0:11:05.960 --> 0:11:08.240 I think one of the things that's happened with with 0:11:08.360 --> 0:11:11.360 AI is that, compared to other technologies that I've tracked 0:11:11.400 --> 0:11:15.760 from the Internet and beyond, there is very very strong 0:11:15.920 --> 0:11:20.360 osterroots adoption of AI tools like chat, GPT in organizations, 0:11:20.360 --> 0:11:24.000 even if there isn't an official policy, and increasingly the 0:11:24.120 --> 0:11:27.160 CEO is the one who's racking the cage saying what 0:11:27.200 --> 0:11:29.520 are we doing with the generative AI? And that wasn't 0:11:29.520 --> 0:11:32.120 the case with the Internet, and it wasn't the case 0:11:32.120 --> 0:11:34.640 with mobile or cloud computing. And I think that creates 0:11:34.640 --> 0:11:37.440 the conditions for a perfect storm in the next year 0:11:37.520 --> 0:11:40.760 or two for firms to really build lots of AI 0:11:40.880 --> 0:11:45.480 tools because their frontline employees want them and the bosses 0:11:45.520 --> 0:11:48.160 at the top want them, and that that will compress 0:11:48.200 --> 0:11:50.120 the amount of time it will take for these these 0:11:50.120 --> 0:11:53.280 things to become quotidian within large firms. 0:11:53.000 --> 0:11:55.319 And then of course with us as customers of those businesses. 0:11:55.960 --> 0:11:58.079 But do we do we have for Axlvis a number 0:11:58.280 --> 0:12:00.960 of how many jobs will be lost? I was speaking 0:12:01.000 --> 0:12:04.160 to the Norwegian Sovereign Wealth fun head, the biggest sovereign 0:12:04.160 --> 0:12:05.360 wealth in the world, and he was saying, you know what, 0:12:05.360 --> 0:12:08.000 You're going to be twenty percent more productive when you 0:12:08.080 --> 0:12:11.040 use AI. I was like, twenty percent, But then you 0:12:11.040 --> 0:12:14.360 don't really know the timeframe. Again, are we over optimistic 0:12:14.400 --> 0:12:16.319 about how fast the transformation happens? 0:12:16.960 --> 0:12:20.040 Yeah, so I think the pictures of it mixed. Overall, 0:12:20.160 --> 0:12:22.880 what we found is that roughly a quarter of all 0:12:22.960 --> 0:12:26.000 jobs could go through some kind of structural churn, both 0:12:26.120 --> 0:12:30.040 increasing and decreasing over time in the next five years. 0:12:30.120 --> 0:12:33.080 So we did this last year up to twenty twenty seven. 0:12:33.440 --> 0:12:35.280 The second thing we did, though, is try to go 0:12:35.360 --> 0:12:40.640 deeper into where are llms actually most likely to be 0:12:40.679 --> 0:12:43.280 impacting different roles. And there's a set of roles that 0:12:43.720 --> 0:12:47.480 will be displaced because so much of the tasks within 0:12:47.520 --> 0:12:52.120 those roles are essentially repetitive language based tasks, and so 0:12:52.200 --> 0:12:55.319 they can be displaced. There are roles which will be automated, 0:12:55.360 --> 0:12:58.760 and that's the big firms, where there will be product 0:12:59.240 --> 0:13:03.280 increase in productctivity and increase in efficiency because we will 0:13:03.280 --> 0:13:05.800 all be able to, you know, find that twenty percent 0:13:05.880 --> 0:13:07.920 extra time and space because we'll be using some of 0:13:07.960 --> 0:13:11.000 those tools. And then there are roles that will remain 0:13:11.760 --> 0:13:15.319 untouched because they're going to be very much about physical 0:13:15.360 --> 0:13:18.440 activity and labor, whether that's in retail or in care. 0:13:18.559 --> 0:13:21.559 Maybe they will become more efficient because planning tools will 0:13:21.559 --> 0:13:24.800 be better, but the core of the role is going 0:13:24.840 --> 0:13:27.600 to remain fairly untouched. So we sort of ran through that, 0:13:27.720 --> 0:13:30.319 and I don't think that we can conclusively say how 0:13:30.360 --> 0:13:33.880 many roles that's going to be impacting, but what is 0:13:33.920 --> 0:13:36.520 clear is that given the nature of these types of roles, 0:13:36.840 --> 0:13:38.560 there's going to be a win for a lot of 0:13:38.600 --> 0:13:41.480 developed economies, and it's going to be more disruptive in 0:13:41.520 --> 0:13:44.640 developing economies. Again, it depends on what type of task 0:13:44.679 --> 0:13:45.680 and role somebody is doing. 0:13:45.960 --> 0:13:47.840 I mean, as I was saying, when you walk around 0:13:47.840 --> 0:13:52.800 with the you know, the AI badging in all the companies, 0:13:52.840 --> 0:13:54.800 I mean, does this feel like and you're a veteran 0:13:54.880 --> 0:13:58.160 of Davos as em and YouTube fancy, does this feel 0:13:58.200 --> 0:14:00.439 like the AI DeVos where this is. 0:14:00.360 --> 0:14:02.360 Going to be totally dominant this conversation. 0:14:02.200 --> 0:14:07.000 Well, I think the Davos represents the forces that are 0:14:07.120 --> 0:14:10.880 shaping the world in some way. And of course we 0:14:10.920 --> 0:14:13.400 are in a world where it's the technology companies that 0:14:13.480 --> 0:14:16.959 are the commercial businesses that drive things in the way 0:14:16.960 --> 0:14:19.520 that thirty years ago it was the oil businesses. So 0:14:20.320 --> 0:14:23.800 I think it's not surprising that you see quite a 0:14:23.800 --> 0:14:25.520 lot of AI companies. You know, there are a lot 0:14:25.560 --> 0:14:28.880 of prime ministers here, and we know that society has 0:14:28.920 --> 0:14:31.640 woken up to the fact that technology is not just 0:14:31.680 --> 0:14:35.200 about productivity and efficiency. It's also about the impact it 0:14:35.240 --> 0:14:38.640 has on people's lives as citizens, their sense of belonging, 0:14:38.680 --> 0:14:41.560 their sense of worth and actually how well they feel 0:14:41.600 --> 0:14:45.720 they're doing. So I'm actually pleased that the AI discussion 0:14:45.800 --> 0:14:48.880 is not one that's restricted to boardrooms but is also 0:14:49.040 --> 0:14:51.200 taking place in cabinet rooms as well. 0:14:51.440 --> 0:14:54.480 Just thinking about, you know, the image that Davos has 0:14:54.520 --> 0:14:56.600 as a convening around the world. It is about the elite, 0:14:56.880 --> 0:14:59.840 isn't it? And over the years, questions are asked about 0:15:00.200 --> 0:15:03.080 Davos man or woman and the relevance to the real world. 0:15:03.120 --> 0:15:05.640 And on this question AI, if it really is going 0:15:05.680 --> 0:15:08.320 to cut fifteen percent or twenty percent of jobs that 0:15:08.400 --> 0:15:12.200 has a huge impact on the average person who works 0:15:12.200 --> 0:15:14.720 for an organization. Is this debate, is this discussion of 0:15:14.760 --> 0:15:17.280 the elite talking to themselves about these efficiencies? 0:15:17.280 --> 0:15:18.440 Is it going to be? 0:15:18.920 --> 0:15:21.560 Is that just going to drive home that stereotype of 0:15:21.560 --> 0:15:22.640 this bubble up here in. 0:15:22.720 --> 0:15:28.000 I mean, I would love to, but my impression is 0:15:28.000 --> 0:15:30.760 that I just was one sentence that it isn't I 0:15:30.960 --> 0:15:33.800 rarely in the last year or two have had conversations. 0:15:33.960 --> 0:15:35.920 There's one exception when I was in a room full 0:15:35.920 --> 0:15:39.000 of CFOs where the only chief financial officer is the 0:15:39.000 --> 0:15:41.920 only thing we talked about they talked about was efficiencies. 0:15:42.200 --> 0:15:44.240 But in every other conversation that I've had, whether it's 0:15:44.280 --> 0:15:48.440 policymakers or business leaders, they're really really aware of questions 0:15:48.440 --> 0:15:50.280 like reskilling and displacement. 0:15:50.640 --> 0:15:52.000 But you have a much better picture. 0:15:52.040 --> 0:15:53.560 I'm Davis community glad you said that. 0:15:53.640 --> 0:15:58.280 I think there's really a positive trajectory here because I 0:15:58.280 --> 0:16:01.760 think we all recall twenty fifteen, twenty sixteen Rise of 0:16:01.800 --> 0:16:04.840 the Robots. Everyone's talking about how the integration of robots 0:16:04.840 --> 0:16:08.160 into manufacturing could be so disruptive to workers that are 0:16:08.240 --> 0:16:11.000 in factories, and I think we've seen that the reality 0:16:11.080 --> 0:16:13.360 play out there where yes, there has been integration of 0:16:13.440 --> 0:16:17.320 a lot of that technology, but there also hasn't quite 0:16:17.320 --> 0:16:19.560 been in the same way that people thought it would 0:16:19.560 --> 0:16:20.239 be disruptive. 0:16:20.440 --> 0:16:21.640 Now, because of. 0:16:21.560 --> 0:16:24.920 Some of that conversation, we kicked off in twenty twenty, 0:16:25.000 --> 0:16:28.120 this ten year initiative on the Reskilling Revolution with a 0:16:28.240 --> 0:16:32.440 number of the ministers, the companies that are really in 0:16:32.480 --> 0:16:35.280 a position to actually provide some of that reskilling and upskilling, 0:16:35.600 --> 0:16:37.720 and that kicked off a movement where we said we'll 0:16:37.720 --> 0:16:40.440 get there in ten years, reach a billion people, and 0:16:40.600 --> 0:16:43.040 we're going to be announcing this week that we're crossing 0:16:43.080 --> 0:16:46.040 the six hundred million mark with that effort so across 0:16:46.400 --> 0:16:50.040 thirty economies, across I think one hundred and fifty different companies, 0:16:50.320 --> 0:16:52.360 so a lot happening there. And now, of course this 0:16:52.440 --> 0:16:54.480 year the conversation is going to be how do we 0:16:54.560 --> 0:16:57.080 orient that that specifically to the jobs that are going 0:16:57.120 --> 0:16:59.480 to be disrupted because of artificial intelligence, and how do 0:16:59.520 --> 0:17:02.400 we take some of that reskilling towards the new roles 0:17:02.440 --> 0:17:03.920 that will be created because of AI. 0:17:04.280 --> 0:17:05.560 Can I pivot a little bit to some of the 0:17:05.560 --> 0:17:08.400 geopolitics you mentioned There are a lot of prime ministers 0:17:08.840 --> 0:17:12.440 and wildly, I mean it struck me seeing the presence 0:17:12.480 --> 0:17:16.240 of the Ukrainian delegation here and you know, this forum 0:17:16.280 --> 0:17:19.720 making headlines over the weekend and an attempt to put 0:17:19.760 --> 0:17:22.960 forward a proposal potentially for peace. I mean, that's a 0:17:23.240 --> 0:17:26.600 you know, one of the biggest geopolitical problems basing the world, 0:17:26.920 --> 0:17:29.680 and the Ukrainian President will be here this week making 0:17:29.720 --> 0:17:35.040 an address. What role can this convening make in solving 0:17:36.280 --> 0:17:38.280 auguably one of the most attractable problems in the world 0:17:38.320 --> 0:17:38.800 at the moment. 0:17:38.880 --> 0:17:42.320 So do for us? I mean, we look at the 0:17:42.320 --> 0:17:45.320 theme of this meeting, rebuilding trust. It is a particularly 0:17:45.359 --> 0:17:48.919 polarizing moment. We believe that people have to talk to 0:17:48.960 --> 0:17:51.199 each other, and so a big part of what this 0:17:51.280 --> 0:17:54.960 meeting is about bring together people from across the world. 0:17:55.119 --> 0:17:59.400 We have over one hundred countries represented, nearly sixty heads 0:17:59.440 --> 0:18:03.840 of state, two hundred plus ministers across various portfolios, sixteen 0:18:03.920 --> 0:18:07.160 hundred businesses. There's a reason I think that everybody's coming 0:18:07.200 --> 0:18:09.359 together because they have the same sense that we need 0:18:09.400 --> 0:18:12.119 to have that conversation. I think a second element is 0:18:12.200 --> 0:18:14.640 there has to be alignment around some of the things 0:18:14.640 --> 0:18:17.560 that need to be done. And certainly on the geopolitical side, 0:18:17.600 --> 0:18:20.119 there will be quite a few private meetings, including the 0:18:20.160 --> 0:18:22.600 meeting that was hosted by the Swiss and the Ukrainian 0:18:22.640 --> 0:18:25.160 governments here in Dabos, but on a number of these 0:18:25.160 --> 0:18:28.080 other topics, So what we just discussed AI and reskilling, 0:18:28.160 --> 0:18:31.040 there has to be an actual alignment around Okay, what next, 0:18:31.080 --> 0:18:33.560 what's the next two years sort of plan? Where is 0:18:33.600 --> 0:18:35.760 it that the focus areas should be. And then the 0:18:35.800 --> 0:18:38.480 third element is I think for any of that trust 0:18:38.680 --> 0:18:42.400 to be rebuilt, there has to be a focus on action. 0:18:42.480 --> 0:18:44.560 There has to be delivery, there has to be accountability, 0:18:44.560 --> 0:18:46.920 and there has to be transparency and that is where 0:18:46.960 --> 0:18:51.640 actually reporting out on the proposals that were made last year. 0:18:51.960 --> 0:18:54.040 How did that go? How have we delivered against that? 0:18:54.080 --> 0:18:54.760 I think that's key. 0:18:55.359 --> 0:18:56.920 I guess when we were talking about trust and the 0:18:56.960 --> 0:18:58.920 fact that you know the need to talk to each other. 0:18:59.440 --> 0:19:02.879 Is AI an enabler of people coming together or is 0:19:02.880 --> 0:19:03.119 it not? 0:19:04.160 --> 0:19:06.840 I think it is an enabler. And we all have 0:19:06.960 --> 0:19:09.800 had this experience because if you've ever used Google Translate, 0:19:09.920 --> 0:19:15.200 that is AI. Because it's become so commonplace, we don't 0:19:15.240 --> 0:19:18.119 think of it as AI. But when machine translation, the 0:19:18.160 --> 0:19:22.760 idea that I can communicate to someone who is in 0:19:22.800 --> 0:19:25.240 Thailand and doesn't speak much English and I don't certainly 0:19:25.280 --> 0:19:29.080 speak no taie is remarkable. And so we've already seen 0:19:29.160 --> 0:19:33.320 we've already benefited from the dividend of AI in that 0:19:33.760 --> 0:19:36.800 in that way, now we have a new set of tools, 0:19:36.800 --> 0:19:39.520 and I think there's this sort of ironic workworld where 0:19:40.000 --> 0:19:43.280 we can all produce perfect email sales pitches because we've 0:19:43.320 --> 0:19:45.880 all used chat GPT, But at the same time, none 0:19:45.880 --> 0:19:47.680 of them are being read by humans because they're being 0:19:47.720 --> 0:19:52.679 read by my chat GPT enabled email clients. And you 0:19:52.680 --> 0:19:54.800 know that all the beauty is all the beauty is lost. 0:19:55.080 --> 0:20:01.440 I do think that you know, people do remain social creatures. 0:20:01.680 --> 0:20:03.800 They come to Davols. 0:20:03.840 --> 0:20:05.840 It's not the easiest place to come to, it's cold 0:20:05.880 --> 0:20:08.439 for most of us, and they come because there are 0:20:08.480 --> 0:20:11.120 things you can do when you're with other people. And 0:20:11.520 --> 0:20:15.560 AI has the capability to allow us to do that, 0:20:15.640 --> 0:20:18.399 to create the space to spend time. 0:20:18.160 --> 0:20:18.680 To do that. 0:20:18.920 --> 0:20:22.640 It could also be used to create solipsistic digital virtual 0:20:22.680 --> 0:20:25.560 worlds where we just you know, sit in our headsets 0:20:25.560 --> 0:20:27.680 and never meet another human And in that in some 0:20:27.720 --> 0:20:30.919 sense that's our that's our choice. But I think, you know, 0:20:31.440 --> 0:20:34.119