The Future of Work Podcast

Good afternoon and welcome to our lunchtime  conversations from the International Labour

Conference here at the Palais des Nations in Geneva.  These conversations are brought to you by the

ILO's Future of Work podcast series. I am Ibon Villelabeitia.  Over the coming days, we will be exploring key    4 00:00:22,960 --> 00:00:30,400 issues shaping the world of work with experts  from across the globe. Today we focus on one  

of the defining forces of that change. Artificial  intelligence is already reshaping work from jobs  

to skills to everyday working life. But the  key question is how this will unfold. Will  

it strengthen decent work, dignity, and share  prosperity or deepen inequality and exclusion?  

The Director-General of the ILO describes  this as a moment of choice. The outcomes are  

not predetermined but depend on governance,  institutions and social dialogue. In this  

session, we will explore what this means for  workers, employers, and economies, and what  

a human-centered future of work requires. To  examine these questions, we are joined by four  

distinguished experts from the ILO and academia.  We have Kostas Papadakis who is a special advisor,  

the Labour Governance and Sectoral Policies  department at the ILO. Thank you, Kostas. We  

also have Janine Berg, a senior economist in  the Research department of the ILO. We also  

have Hannah Leipmann who's also an economist in the  Research department of the ILO. And finally, we  

have Valerio De Stefano who's Professor of Law and  Canada Research, Chair of Innovation, Law and Society  

in Osgoode Law School in York University in the  city of Toronto. Thank you all for being here with  

us. Kostas, let me start with you for the big  picture. As I said, the Director-General describes  

this as a moment of choice for AI and work.  What is the key message behind this framing and  

why is this moment so important for the world of  work right now? Yes, thank you Ibon and thank you  

for the introduction. Indeed, as you said in your  introduction, the key message of the ILO Director-General  

to the International Labour Conference  through his report for the first time dedicated  

to the question of artificial intelligence is that  outcomes, positive or negative, associated  

with this new technology are not predetermined.  Whether artificial intelligence will lead to  

increased productivity and to helping workers  or on the contrary to increasing inequalities  

and insecurities will depend on the choices  we are making and these choices are of course  

going through policies and institutions which  govern the the world of work. Now why isn't  

it – is it important? Well, AI is already shaping  the labour markets, the world of work more broadly.  

And then with this report, the Director-General  wanted to explore the kind of policies that we  

need to have in mind and the kind of institutions  in order to design and implement these policies.  

What kind of policies? Well, we need policies  to facilitate the transition of workers  

whose jobs will be affected. Already  even in in a scenario where there is not,

you know jobs do not disappear at scale, we still  know through the ILO research in particular that  

the effects of this transition will not be  the same for all. You have some groups of workers  

that will be more affected than others. So we have  to design tailor-made policies in terms of  

facilitating the transition. Of course we need to adjust the skills and education systems in  

order to be future ready to respond to the new  skills lens and I guess we will have also  

colleagues discussing about this particular issue.  We need social protection measures to accompany  

those that are undertaking an upskilling and  reskilling, let's say, process because it is  

not – you know these people will not transit to another job from one day to another. They will  

have to have for instance some income protection in the meantime. And of course we need  

to support those enterprises and those  countries that do not have access right now  

and they cannot benefit from this new technology  which has a lot of positive aspects of course. And  

I refer mostly to the micro small and medium  enterprises or to some developing countries that  

do not have access to this AI infrastructure. And of course we need to devise policies that,

I would say, democratize the use of AI  in order to improve services, social services,  

labour inspection services or employment services. Yeah, so those are  

the key messages of the report. Thank you Kostas  for setting out the framing of that moment of  

choice. Janine, now let's turn to the evidence  base and the data. What do we actually know  

today about how AI is affecting jobs both in  terms of quantity and quality? Right. So, our  

research here at the ILO has been mostly focused  on the potential exposure of certain occupations  

to artificial intelligence and what that could  mean in terms of jobs, job losses, but also in  

terms of the transformation of people's day-to-day  jobs. And so in essence it's actually a kind of  

a theoretical exercise because it's really about  what could change and then there's of course what  

has changed what have we seen really in practice.  So let me first talk about what could change. So  

our research shows that 20% of of occupations  across the world are what we call exposed to AI  

meaning that those people perform tasks that  could potentially be performed by artificial  

intelligence. Now that doesn't necessarily mean  and one shouldn't assume that that means that  

one fifth of jobs are going to disappear. It's not  that at all. What it means is that in a particular  

job, we all do certain different tasks in our jobs,  and some of those tasks could be automated. Now  

if the only tasks that one person does could be  automated by that technology then the chances  

of that person losing their jobs is high. But  if the person is doing a task where they do a  

lot of different tasks and just some of those jobs  could be done by the technology then there's more  

likelihood that that person's job would be what  what economists call augmented but really what  

you mean productivity gains. So for example, if I  want to translate something into another language  

and I use AI, I'm more productive because I have  this technology that facilitates my translation.  

So, in short, one fifth of jobs that have the potential  to be translated – the potential to be  

transformed or exposed and this varies depending  on countries. So in the richer countries because  

of the diversification of their economic structure  you have almost a third of jobs that are exposed  

to the technology whereas in poor countries it's  one in 10 jobs that are exposed to the technology.  

So that right there goes to what you're saying  about how this transformation is not equal  

because in some cases you're going to have some  countries that are not at all affected and on  

one hand that could be good because they won't  have job losses but on the other hand that could  

be bad because they don't have the productivity  gains. Okay. And then skipping to what has – what  

do we know in practice. So here the data is a  lot more sketchy because it's hard to  

identify what jobs really have been lost  to AI. What we see instead is more of a slowdown  

in hiring in certain occupations. So you – we've  heard reports for example of software developers,  

customer service representatives,  people that could be replaced, for example,  

by chat bots. Those are some of the occupations  where you hear that there actually have been some  

job losses. But what the more likely effect is is  that there's going to be less hiring. And this of  

course has consequences for young people. It  also has consequences for economy. So we need  

to be doing more work. And that's one of the plans of the ILO and our Research department to  

be doing more research on actually seeing the real  time effects of how this technology is affecting  

jobs. And I guess I've probably taken up my time  so I'll pass the floor to Hannah. Thank you.

That helps ground the discussion and what data  is already showing. Now I'm turning to you Hannah.  

Kostas mentioned skills, and building  on that let's look at skills and adaptation.  

What skills are becoming most important in AI-enabled labour markets and are institutions and  

training systems keeping pace? Yeah, so at the  ILO we just released our new flagship report  

on lifelong learning and skills for the future.  And in this report we analyze online vacancy data  

to understand how skills demand is changing. And  what we find generally speaking is that employers  

increasingly look for workers with rounded skills  profiles. And these rounded skills profiles are  

associated with good working conditions. So  a narrow focus on technical expertise alone  

would be insufficient. And specifically in the  context of AI, it's quite interesting that the  

overall demand for AI-specific skills is still  comparatively low across countries. And we expect  

this low demand to still increase going forward.  But we also think that it's due to the fact that  

many workers, they actually rely on ready-to-use AI tools and these do not require AI expertise  

so much, but they require foundational skills like  critical thinking or digital literacy. And related  

to this we also expect that social emotional  skills will become even more important because  

these cannot be performed by machines. Now I  think the second part of your question was whether  

training systems are keeping pace, and based  on our new report we raised two concerns in this  

regard. So the first concern is that access to  quality learning is still out of reach for many  

adults and it is highly unequal. So, to give one  example, only 16 per cent of all workers participate  

in training or structured learning over the  course of a year. And many informal workers in  

particular, they learn mostly by doing without  additional support or options. So that means that  

there's still a very long way to go before  lifelong learning systems become inclusive.  

And if I may, we raise an additional concern – this pertains to the specific role of  

social emotional skills, because as I mentioned,  these are in high demand and we also find that  

they are associated with significant wage premier.  But our research shows that this is actually not  

the case in the care economy. So for care economy  workers, we do not find these wage premium. And  

this leads us then to conclude that yes obviously skilling, reskilling and upskilling are highly  

important but it is also important that societies  and markets adequately value the skills of all  

workers. Thank you. So we're seeing both  opportunities and adjustments, pressures on  

the skill side. And Valerio, let me bring you  in on the risks dimension. What are the biggest  

risks for workers emerging from AI today? And are  current labour standards and institutions keeping  

up or do we need new approaches through social  dialogue? Well, first of all, let me say that  

this is a very important report that maps  a lot of the existing questions that need to be  

answered at the moment. And so when it comes  to the risks the workers face, again in many  

cases people tend to overfocus on the question of  job quantity, how many jobs are going to be lost to  

AI, but I think that the report does a very good  job of also putting the picture on the quality of  

the jobs that will remain and indeed there are  some concerns that emerge when we talk about that.  

And in many cases even when those concerns about  job quality are voiced it's only about privacy  

and the question of data governance, which it's quite important but it's not the only  

point in the picture. And again the report  does a good job of explaining what other risks   

workers face. First of all there's the question  of how managerial prerogatives are expanded by  

AI and magnify the power and authority that exists  in workplaces. And this increases the pace of work,  

increases surveillance of work, which creates risk to occupational health and safety, which in  

many cases are not treated  in the public sphere. Work becomes more intense  

and more demanding, and this intensification  is driven by the technology. So this is one side  

of the picture that needs to be brought up. The  other thing is non-discrimination, because these  

systems are mistakenly perceived to be neutral  while they incorporate biases that exist in  

society in the data sets and in the decision tree.  So they reverberate biases and discrimination in  

society. There's risks to collective rights  because some of this surveillance can and  

is being used to prevent people from unionizing  or organizing in trade unions. And also at  

the same time, AI – and this is another thing that  I found very interesting in the report that the  

report points out that in some cases the AI system  that replace human decision making strip workers  

and managers of authority and they actually,  the workers need to re respond and answer to  

a sort of technocratic and technological authority in the workplace. So there is that  

dimension. And when it comes to the question of  standards, ILO standards already do a very good  

job of establishing a framework of principles that  can be followed, but there are some limitations,  

especially when it comes to the question of  algorithmic management. And when it comes to  

algorithmic management the ILO can do a very good job  of incorporating social dialogue in the devising  

and production of these systems before  they are introduced in the workplace. So this  

is where the key dimension but the tripartite  dimension of the ILO can play a very good role.  

Thank you. I think after this first round we have  a clearer picture of both the opportunities and the  

risks emerging from AI. Let's move on now to the  diagnosis and to policy and responses. How can  

we ensure that AI-driven changes in the world of  work lead to decent work outcomes across different  

countries and groups? And I'm going to ask the  same question to the four of you. I don't know  

who wants to go first. Maybe Kostas? Yes. Thank  you. With pleasure. I will give a generic  

response. Okay. Which is a response that runs  throughout all five chapters of the report.  

I think in one word we need good governance. Good  governance will lead to decent work outcomes.  

And let me break it down into three components,  this good governance. Let's say the three  

dimensions, which we can identify throughout the  report. The first is an active governance,

meaning we know that in every technological  disruption there are winners and losers. Okay.  

For instance, electrification and mechanization  in the early phases of the industrial revolution  

led on the one hand to increased productivity  and profits but at the same time to suboptimal  

working conditions, including let's say extreme  cases of child labour or more intense working  

conditions and working hours. It was only after appropriate, let's say, rules of the game were  

adopted, policies for fair competition, policies  for – labour policies – that the costs  

and the benefits of technology and increased  productivity were more fairly distributed, and  

the workers managed to profit. So that is the  first component. The second component of good  

governance is inclusive governance and there  I will agree with Valerio on the importance  

of social dialogue. We know we have – I mean  the DG says this in his report – but we know also  

from earlier technological disruptions that  social dialogue is important. Social dialogue taking  

aboard workers' and employers' views in designing  and deploying technology is important with  

a view to coming up with more legitimate, more workable, I would say, policies not only  

in order to protect the rights of workers but  also to improve organizational performance.  

So this is –and there is as the DG mentions in his  report – there is clearly a business case for social  

dialogue especially in this area. And third  the last component, last but not least, we need  

international governance. We need multilateralism. AI and the business models organized around  

this new technology do not recognize  borders. No country alone can address,

let's say, its impacts positive or negative. And of course not all countries are faced with  

the same challenges. We know that there are, for  instance, digital divides that need to be addressed.  

And this necessitates international cooperation. Without international cooperation  

some countries would not be able on the one hand  to mitigate the negative impacts or on the other  

hand to profit from the advantages that this  new technology brings about. And of course in  

multilateralism and international  governance I include also the role of the ILO – the  

ILO needs to be future-ready in order to be able  to support the constituents in whatever policies  

they decide to design and implement at the  national level. Thank you. The ILO needs to be  

future-ready. Who wants to come next? Maybe Janine  or Hannah? Yeah. So we know for example – I mean as  

we've been talking about – that the effects of AI  are not going to be equal across groups, across  

occupations, sectors, countries. So there's all  these different breakdowns of how people are going  

to be affected. And one of the real reason – I mean  one of the big motivations for doing the ILO's  

numbers on occupational exposure is to give policy  makers that information so that they can identify  

who are the groups of workers and what are the  occupations that are most at risk. And with that  

type of information then you can develop proactive  strategies rather than being reactive. And so what  

we want is really this proactive thinking by governments and social partners about how  

to respond. And so we can also think of,  you know, yes we need to have more, you know, if  

there are going to be some job losses, and there might be, what could be some new jobs  

that be created? What are the sectors where we know  that we need to be investing and could be a source  

of employment? We know that there's shortages in  the care economy across the world, poor countries  

and rich countries. We know that there is  a need for the green transition and a lot of  

work in that area. So if you start thinking  about developing, you know, shifting some of these  

occupations and some of the investments in countries to these areas, this could be a source  

of job creation. And of course, people need  to be supported during this transition. How can  

we have social protection systems that are robust  enough to support workers during that transition?  

People need some sort of income security when  they're doing training programmes, for example. So  

that's one thing. The other issue is about  the job quality aspect. So as I mentioned, a lot  

of jobs are going to be transformed, not necess – you know, so they might not, you know, they won't  

be necessarily lost, but they could be transformed.  And this could be positive or this could be  

negative. And this is where this whole issue of  governance and choice is so important, because  

for that transformation to be possible, it's really  important to have a process that is participatory , 

that does use social dialogue so that you can get  workers' perspectives along with the employers'  

in how the technology can be integrated into the  workplace. Workers know their jobs best and so  

when technology is adapted with the workers in a  participatory design process, you're more likely  

to have technology that is more productive,  you're more likely to have a situation where  

working conditions are improved. So, those  are some of the things that can be

done. And we know also that, you know, some of, in  some countries with stronger industrial relations  

systems and stronger collective bargaining, a lot of collective bargaining agreements have come  

out in the past few years where there has been  explicit safeguards and involvement of workers  

in that process that can be beneficial. But I mean I think the real message is this message  

about being proactive and not being reactive, and thinking, you know, what can we be doing  

to help manage this transition. Thank you. Hannah, you want to come in on that – proactive? On being  

proactive? Yes. I think that may be one of the  conclusions of this whole discussion, no? But  

I will respond again from the perspective of  lifelong learning, given this new flagship report  

that we just published. And because in the  era of AI, it's obviously important that workers  

can acquire new skills. And this needs to be true  for workers of all ages and also for workers with  

different levels of formal qualification. And  based on our research, we argue that this really  

requires an ambitious systemic approach, and  that has several implications. So first of all,  

it requires acknowledging the many different ways  in which people learn, which clearly go beyond the  

traditional classroom, to make sure that workers'  skills are recognized and also that they have  

access to adequate qualifications. The second  implication of this ambitious systemic approach  

is more conceptual in nature. So lifelong  learning systems should be designed not only with  

narrow objectives in mind like employability or  productivity, but they can also be used in a way  

to advance broader objectives like decent work  through innovation, active citizenship or social  

cohesion. And then finally more from the  perspective of policy, we define the  

building blocks of successful lifelong learning  systems. So financing is one of the elephants in  

the room. We need sustainable financial  solutions that are understood as a shared  

responsibility. And then related to what Kostas and Janine said, we obviously also need a strong  

social dialogue. And Valerio? Yeah, I think that one of the things that need to be faced is that  

at the moment workers are not involved at all  in the design and the production, let alone the  

introduction of the systems that actually are used  in the workplace by them or over them. And this  

creates, as Janine said, many shortcomings  in how these systems operate. We know that  

most of AI pilots in businesses fail after  a short implementation, precisely because they  

are introduced top down on workers without any  feedback from them. And actually workers are the  

ones that know how the job is actually done. So there is a lack of involvement that  

also is a waste of time for employers and a drain  of resources for employers. So what we actually  

need is fostering the dialogue between capital and labour, employers and workers  

to come together and have a say in the  technology that is implemented in workplaces.  

At the moment, most of these things are decided by  tech companies and tech people who have absolutely  

zero idea of how a job is done and how it actually  works. And this is what is creating a lot of waste  

for everyone and negative externalities for  societies. Thank you. And now we're going  

to move to our final – our third and final round and  I'd like to ask each of you to look ahead. I'm  

also going to ask you, if possible, you can stick  to one minute for this final message. And the  

question is, what is the most important priority  to ensure that a human-centred future of work  

remains achievable in the age of AI? I don't know  who wants to start first, but please go ahead and  

final message in one minute if possible. Okay. So  I think it should be strengthening social dialogue  

and collective bargaining over the design and  introduction of algorithmic management at work,  

again for the benefits of workers, employers  and societies alike. It's really tricky, because  

a hundred different aspects come to mind, and I think the  DG's report to the ILC picks up on many of them.  

But I will select a point that Janine already  alluded to. So there's now a debate whether  

entry-level hiring has stalled in certain white  collar occupations. We still need more evidence to  

be able to answer that question conclusively.  But what's sometimes missing in the debate for  

me is the realization that the entry- level  workers of today will be the mid-level and higher- 

level professionals of the future. And so I think  it can be very much in the interest of enterprises  

themselves to have a longer-term planning  horizon and to invest in this human potential.

So I actually don't think there's one important  priority. I mean if I had to pick one and maybe  

it would probably be the dialogue but but it's just like the the technology itself. People  

think it's easy to use. It's shiny. It works  well. And so you you put it in simple solutions,  

you'll get great results. And it doesn't work  that way. And it's the same with policy-making

as well. You have to have a lot of different  approaches to achieve the ultimate objective  

of improving well-being, you know, improving  human welfare. And so that means you need to  

be, you need to be investing resources in job  creation. You need to be investing in improving  

working conditions. You need to be investing in  in reducing the digital divide. There are so many  

different fronts that need to be tackled, which  is why I think governments and social partners  

need to be really working together to develop comprehensive plans and enacting them, and  

of course dealing with the financing issues. But  if I had to sum it up in one I would say dialogue  

between all these different parties and that  recognition that yes it's going to be a hard –

it's a long haul and it's a hard haul and we have  to go forward with it. Yes. Thank you. Well,  

from a governance perspective, first of all, I  agree it's very difficult to choose a priority,  

but I would say from a governance perspective,  I think the priority is twofold, composed of two  

interrelated elements. The first is having  a solid knowledge basis on the benefits and  

on the risks of this new technology. Obtaining and sharing this information  

on the risks and opportunities will be  extremely important for designing and implementing  

policies, the policies which we mentioned  before. So we need to identify both the  

opportunities and the risks, and we need to do  it all the time, given the pace of developments  

and evolution of this new technology. So – and the  second component, which is related to the first one,  

always again involve those directly affected  by these decisions, by these policies, 

workers and employers. And I would say even addressing, inviting also the experts  

to talk to the workers and the employers. Because in addition to the social partners participating  

there, they have to have the capacity not only  they have to be invited but they have to have the  

capacity to participate effectively in these  policy-making processes. So I would say those are the  

two most important components. And if we have  both information and the capacity to participate  

effectively, I believe both the legitimacy and the efficiency gains that we  

can win are opening very positive and new horizons.  Well thank you very much to the four of you for  

this extremely rich discussion. If there is one  message from today's conversation, I think it's that  

the impact of AI will depend on the choices  we make today through inclusive dialogue. Thank  

you for joining us at the International  Labour Conference in Geneva. And be sure to  

tune in again as we continue to explore the major  transformations shaping the world of work today.  

You can follow us online @ILO on  X, Tik Tok, YouTube, @ILO.org on Facebook,  

Instagram and Threads and Blue Sky and  also @InternationalLabourOrganization on  

LinkedIn. Thank you for listening and we look  forward to welcome you again soon. Goodbye.