How AI is changing work, skills and economic opportunity

0:05 implementing LJ on pronouncing everybody's name correctly. That's hard to do. it is great to be back at AI for good. I am feeling Laspri de Janev. Are you I would like to be feeling lab breeze de Janev actually but anyway so this is an important conversation and one that I've actually been looking forward to. It impacts really all of us in one way or another and certainly it impacts future generations. It is not a question of if AI is going to change work for the future but how and who is the most impacted by that. You know there are trends that are colliding. Of course it's not just AI. You have demographic shifts that are happening around the world. The youngest workforces are emerging from developing economies. the most established economies have the oldest workforces. I mean, there's a lot at play here. We're going to get into all of that. I know that there's fear. I know there's pessimism. There's also optimism in some corners, but we're going to talk about fact and have that as a jumping off point as we discuss the future of work. So, let's dive right in.

1:12 I want to start on Songbu. So look all we can really know about at this point how AI is impacting is the snapshot of this moment right so from your perspective what are you seeing happening globally in the workforce and how is it impacting the economy >> so well we are expecting 1.2 Two billion young generation will coming to the job market within one and a half you know decades and unfortunately we see only 400 million jobs can be ready if we do not prepare well. So this is the how we see as a challenge for labor market. but another challenge is AI as you know from the high income country perspective more than 25%ish people are really using generative AI these days compared to low income country cases less than 1% of users are using generative AI so then we see the huge gap and on the other hand also we see some some good news or bad news is that regarding the AI we are very concerned that 15% of our current labor in developed world will get some impacted for some automation type of threats compared to only less than 5% of low-income country and developing world can be impacted. So we have some we see some more opportunity in in developing country cases because AI can contribute to some more productivity rather than disrupting their existing jobs from the automation or something like that. So we are very cautiously looking for some some more opportunity for that. So we identified five areas is really promising area we can support to create more jobs in including the AI support. One is the value added manufacturing and health care services. Agriculture still and also tourism will be a big area. The developing world can can invest and develop more for the job creation than the last one is still infrastructure from the public side is really important source of job creation. So we see some more financial support and also some public investment for that area. Thank you for that. And you know I want to turn to Pate because you know this is not the first time that labor has been impacted by an industrial revolution in history in case there in fact there are a number of examples. What makes this fundamentally different in your mind?

4:10 >> Well first of all indeed it's an age-old question. Will machines replace humans? Right? I mean we remember this you know story about the spinning mills etc. and history has always shown that there is there are new types of work appearing. so we are a bit more optimistic about the impact of AI on the labor market globally. I mean a bit similar what we heard from the World Bank. it will however fundamentally transform jobs and according to our most recent estimates we think that a quarter of all jobs globally will or are already being transformed by AI. so that of course is important to link with skilling and upskilling etc. but in terms of fullscale automation at the where we are right now we are talking about a single-digit number 3.3 is our estimate 3.3%.

5:06 Of course we have to see how AI and robotics go together in the years to come that would have a huge impact also in developing countries in manufacturing and labor intensive sectors. but overall we do believe that labor markets will adapt if they are governed inclusively. Yeah. So, Roberto, I I heard your speech at the UN Global Dialogue, and a line in there stuck out to me that I wanted to bring bring up. You're saying that the political debate when it comes to AI policy is outpacing the empirical one. So, I think you're talking broader about, you know, how countries approach AI policy, but let's let's talk about jobs for a second. Is that same thing true? And I guess the question that I would ask you is do we have the luxury of waiting to find out when it's already moving forward at such a breakneck pace?

6:01 >> Thanks. It's a very good question. So first of all also for me nice to be back in AI for good. we are trying to understand and put a test and I I come then to the most important answer on you about do we have the time to do it. two important economic paradoxes that here they really come with the all day complexity. The first is the famous solo paradox that the more technology you throw in no matter productivity goes down. As you know we have seen in the last 20 years decline in productivity in the western world and people say wow digital is coming. Wow. I mean, a platform economy is coming. No matter what, productivity went down. Will it be the same in AI?

6:55 Because I mean, you hear impressive figures about the productivity increase of AI still to be tested. Actually, we see some signs that indeed AI might involve the increase of productivity. which of course this means a lot for jobs. The second paradox which is often quoted is the Jervant paradox. The Jervos paradox says look if something becomes better and then the marginal cost of production goes down not necessarily this outplaces this product or people make less money.

7:40 turning to job means that if things become better there will be more jobs. now if you think about the health care sector what we see is more or less this I mean AI will improve impressively and we have many programs we are sponsoring about advanced screening of population the the work of doctors the work of assistants but not necessarily diminish the need for doctor they simply will work better and once I mean populations screenings becomes the default that everyone will have advanced screening available more will be needed.

8:24 So if the two paradoxes or whoever solves this paradox wins a Nobel prize. So you can subscribe to the race. those paradox simply show and the evidence of the number show that it is a bit early to say the direction but the direction could be that some of the fears about AI stealing my job are a bit outplaced. And when I see the statistics in Europe I most of the working population doesn't think AI is the job.

8:57 actually 70% today they think AI improves my job but there's indeed anxiety that is building up and then I come to the final answer that I owe you this anxiety is captured by politics so will it be enough time to observe the data and have a rational approach to this well it's a race against time >> you know when you you mentioned platform economy I I think it's important to take a step back here that yeah okay AI has the potential and is already transforming work but work has been in the process of transformation for a while. I would say the pandemic really expedited the move toward obviously remote work, right? More and more people can work from anywhere. They call them digital nomads. They call it the gig economy. And Betta, you your organization just kind of confronted that headon in June. Talk a little bit about this latest convention and and how it could potentially be helpful for workers as we move forward in AI.

10:03 >> Yeah, thank thanks for that question. Indeed, last month, I think we've really seen, a major breakthrough in the negotiations between governments, employers, and workers organizations and coming up with a new convention, a new international labor standard on decent work in the platform economy. The first ever type of instrument that is now of course open to ratification. We had difficult negotiations. I I don't hide that because there have been very different views. and fundamentally and I think this is also a lot what we hear here at the summit is the the tension between regulation and flexibility.

10:43 what do business what does business need to to innovate to increase productivity and in turn of course also to increase jobs or to to create more jobs decent jobs we we say. but at the same time what are the needs in terms of workers protection and the convention is really we believe a very good landing place where we've been able to to also have a definition on what is the employment relationship in the platform economy you know the nomads are they actually workers in the sense of the labor law >> and you know what are the whites of workers in terms of their privacy in terms of understanding algorithm why they are being potentially dismissed for instance. Yeah. But at the same time, we also need to put emphasis on the enabling environment for businesses to innovate and to to adopt AI systems and other technologies and in particular small and medium-sized enterprises and in particular in in the global south. I want to just have you define decent work for a second because I know it's a big part of ILO's mission for people who maybe don't quite understand and and in the context of what we might be facing with AI, how do you protect decent work?

11:59 >> So the the definition of decent work essentially rests on on on four pillars. One is for everyone to have access to productive freely chosen employment. the second is to have to be protected by the fundamental principles and rights at work and and labor standards in in in general. to have access to social protection by the way also a very difficult question for governments how to regulate this in the platform economy. Who pays into the pension funds of the future? So social protection and of course the fourth pillar very important from our perspective is social dialogue. the ability for workers and employers to negotiate and come up with solutions.

12:41 >> Thank you for that. And Sbu, I want to talk. I have a few topics I hope to get to. We only have 12 minutes. This is going to go by so quickly, but this is all really relevant and exciting. And as I was saying backstage, at least in the United States, although I'm sure it was happening in other places, too. We just went through university commencement graduation time you know and there were a number of very highlevel speakers who got booed by the graduates Eric Schmidt being one of them I don't know if he's in the room sorry but but it was true you know so when they mentioned in a positive way the you know the revolution that AI is bringing that stems from one thing alone which is this fear that there are no entrylevel jobs waiting for them and that their futures are going to be negatively impacted by that.

13:33 >> Take it away, Sbu. I mean, is is there any right now empirical evidence to go back to Roberto's comment that that's already happening? Well actually we I see some real challenge as you mentioned that in in de in the high income country and developed world that especially the college graduate are really concerned and still really struggling to find the new jobs these days especially for tech world for example.

14:07 But also we see some more some you know even the opportunity in developing world again because those type of very high-end STEM jobs still very valid in developing world they we see a lot of shortage of that. So maybe there's some gap and mismatch between the developed world and the developing world. So that is our the you know some observation currently the what is happening in in the market. But on the other hand we also see some threat to be honest in even in developing world such as the BO the business process now outsourcing the business cases but at the same time we are really worried about the that disruptions but the trend is not f that drastic. it it is slowly impacting but also at the same time business adjusting to the new norm and new trends. So what I am expecting with some positive mindset is that even in developed world and the many you know high level potential the workers are from university and college definitely will adapt and find some better way to get a better job if they can really get ready to fully utilizing the current AI innovation by themselves but not everybody. So I want to you know the flag that the more important thing is that the society and individual should get really ready just piggyback on this new wave of AI. So unless they are really get ready for taking all of that trend and wave it will be really hard to really adapt.

16:08 You know, I also feel like these conversations forget a huge part of the global population. There are two billion people that work in the informal economy. Biate, you know, what do we know about what potential impact this could have, positive or negative, and you know, how do we make this conversation more inclusive because that's a quarter of the planet. >> Yeah, absolutely. I mean as I said before a major challenge is to enable in particular small medium-sized companies in the global south to adapt to AI so that we avoid having further fragmentation and and deeper digital divide and then these companies are often informal. so it goes back to what I said earlier, how to create an enabling environment for business also to formalize, not just to embrace technology, but also to to formalize.

17:03 And there are some good innovative strategies by the way where technology can also help lower the entry levels in terms of registration and and etc. or for instance incentivizing workers to to go to social protection systems and thereby you you formalize the work. So in in that clearly has to go hand in hand to address the also the decent work deficits for the more than two billion men and women in in the informal economy and while I'm on the risk of exclusion I mean of course a major part is as we already heard from the world bank is you know the exclusion of young people also elderly people you know how do you adapt through lifelong learning systems but also women we have seen through the research we have done that the exposure is in much higher in clerical work and in other female dominated sectors. so I think there's also that gender dimension that we need to look at for sure. And I actually have a pop quiz for the audience just to make sure you're all paying attention. Who was the first US president to propose a universal basic income?

18:16 Anybody shout it out? Anybody here know? was Richard Nixon, believe it or not, a Republican. That certainly wouldn't be something you'd hear today. But so the whole premise of of a UBI is essentially like what happens if there's just not enough work or enough adequate work, right, for people to meet their needs? Roberto. >> Mhm. >> what happens if the AI impact means there won't be enough decent jobs? We're being speculative, but h when then do governments when does policy come in to create a stop gap?

18:58 >> Well, I mean a system like the system of the European Union is already quite close to the universal income. First of all, the health care guaranteed and good level of guaranteed health care for everyone. pensions and assistance is guaranteed for everyone and some countries are even experimenting or introduce a form of universal income. The problem is this system is an extremely expensive system that lives on value added. So you have seen that the European growth and competitiveness in the last 20 years has not really going in a stellar way compared to other parts of the world where the similar system does not exist.

19:52 And the the explanation is very simple because systems like this they are eating up a lot of value added and you need a lot of value added in the economy to actually sustain systems like this. So the point is Europeans at the end of the day are jealous when they see others which have other stellar figures of productivity but they want to keep their social welfare system. at the same time they are desperate to find a way towards economic growth in order to maintain the system in equilibrium.

20:30 That's why I said before we should not treat and I said it yesterday all these things as the end of the history but the beginning of the history because this is a new industrial revolution which is not just by the way the eye it's also the robots don't forget the robots you have seen some of the robots I saw a robot pouring the wine now you >> that was my favorite robot >> yeah you might laugh at it or say it's my favorite robot but this will replace a waiter and at the same times you might say there are millions of people aging without assistance why not having a robot. So the society changes and changes profoundly. What the society cannot afford and this is the real issue is to become poorer because if the society becomes poorer then the system cracks and what is concerning of the present model that makes it is instable is that very few companies own the value added of this and this is not in equilibrium. I mean so this is not the system right now in equilibrium. Will the system find itself its own equilibrium or there will be the need of a very important political corrections time with >> there have been proposals on the table. I think most of them have come from European parliamentarians but about creating an AI tax that if you displace work then you know because of AI you owe more money back to society. I mean are any of those things viable really? Look, I I'm totally against this kind of things because I mean it it's basically you have a system to improve productivity and you tax it more. I mean I think the point is how to make sure that every single company, every single small enterprise benefits from technology. This is the real issue. 85% of the European economy are small enterprises. They are just now touched gently by this but they will be in the future heavily touched by this by the robots by the eye by whatever now if they can internalize this growth s will go well and also things will go well because I mean people will have a longer lifetime much longer so there will be many more opportunities I mean it's a different world and we have to be prepared for this but if system the system goes was wrong and this wealth distribution does not work and cannot work through taxation then we have a problem and I don't believe it could work simply through taxation a system like this so we only have a minute and 48 seconds left but what is one decision from each of your perspectives that you would like to see made today that you think make this conversation a more positive one 5 years from now.

23:37 >> For me, it's adaptation. it's not so much the doomsday scenario. It's how do we adapt labor markets in the macroeconomic sense. how can we help workers to adapt? How can we help enterprises adapt in particular smaller ones? So that is is key. It's not the end of work and there's something inherent decent in work and meaningful to our human existence. so it's really that question to what do we move towards too?

24:09 >> So one single thing I would like to recommend is that just please help our youth get ready for your AI. So we have to educate and teach them get really fully ready for the AI relevant jobs in the future. So the hope is always on youth group. >> Last word >> right in 35 seconds I can only agree with my sense. So I think I mean getting the benefits to a broader part of the population, managing the doomsday narrative in a in a way that I mean yes I mean these are there are many horrible scenarios but there are many good scenarios and managing it well this balance of narrative will be very important. Well, we could continue this for at least 10 more seconds and and beyond. But anyway, Viet, thank you so much. Su Roberto Fiola, thank you for this great conversation. Thank you all for listening and for being here. And enjoy the rest of AI for good.