- 20 hours ago
Global manufacturing is confronting a historic labor crunch: as aging workforces retire at unprecedented rates, AI is emerging as a bridge for knowledge transfer to a new generation. The shift in the industrial sector from automation to autonomy is spawning new career categories, while eliminating old ones. Are we witnessing the death of routine work and the birth of a new industrial workforce?
Category
🤖
TechTranscript
00:00Ready to go. All right. Bonjour et bienvenue, my terrible French, but I'm pleased to be here in Paris at
00:10VivaTech for the second year in a row, my favorite city.
00:14I'm Samantha Glody, the Global Head of Risk Services and Trusted AI at KPMG.
00:20And I'm delighted to be here today with my guests to discuss the industrial scale reshuffle, how a machine collaboration
00:29is transforming work.
00:31Welcome, gentlemen. Great.
00:33Good morning, everyone.
00:34Hello. Nice to be here in Paris and in VivaTech.
00:37Yeah. And we said we're going to have a little bit of fun, so we'll try.
00:41So I just want to start the conversation to really ground us in where the discussions are about AI at
00:48this time.
00:49And I think to date it's largely been around AI as a tool to help people work faster, automate tasks,
00:58increase productivity, but at the individual level.
01:02And KPMG's recent AI Pulse survey made it clear that we're entering a new phase.
01:09We're not just using AI to assist work.
01:13AI is now executing, coordinating, and executing tasks across workflows, systems, and teams.
01:22And that's really changing the conversation.
01:25The question is no longer where do we deploy AI.
01:30It's how do we redesign work, workforce models, and institutions.
01:35And the implications are more than just business productivity, but it's a focus now on skills, trust, governance, and competitiveness.
01:46And so today we're going to have a conversation from two perspectives, from the labour market and employer perspective,
01:53and then from policy and national competitiveness.
01:56So I'm delighted to be here today with Erki Keldo, the Minister of Economy and Industry for Estonia,
02:05who is helping shape economic and workforce strategy in one of the world's most digitally advanced nations.
02:14Welcome.
02:14And Ricardo Barberas, the President of Northern Europe and France at Manpower Group,
02:21who brings a frontline view of how labour demand, skills, and workforce expectations are changing in real time.
02:29So welcome, and let's get into it.
02:33Great.
02:34So I wanted to start the conversation just level set on what is actually changing.
02:40I think people want to think to the future and think about, you know, the future opportunities and implications.
02:47But I think it's really important for employers, leaders, and governments to really understand the position now.
02:55So, Ricardo, I'll start with you.
02:57I understand that you lean toward being human first, but digital always, which I really like.
03:04From what you're seeing in the global labour data, where is machine collaboration already shaping roles,
03:11not in theory, but really in how the hiring demand is changing today?
03:16No, thanks.
03:17It's a very fascinating question, because in a session like this, everybody should expect that I'm starting to talk about
03:23the technology and all this impact.
03:25I will start with other wording and other vocabulary.
03:28I will start with trust, beliefs, values.
03:31So I think that if we forget the word humanity in this discussion, we don't see really the clear picture
03:39that AI is bringing to life.
03:42As Manpower Group, we have this experience, because the expertise is to interview millions of people in 75 countries.
03:49And so we are lucky that we work with different industries, with different type of companies.
03:54And so we are learning in the making.
03:56And there is a research that I can use to answer to your question, that is our human age, the
04:02era of adaptability, because there is no tool, there is no point about the tools without listening to what's happening
04:09today in the organization.
04:11Some data points that I can talk one hour, I cannot, but one data point is employers are saying 70
04:18% we provide the right training for our employees.
04:21The problem is that we have 30% of the employees that are saying, my company is providing to me
04:28the right skills training to be ready.
04:32So point number one, how much there is this mismatch that is creating a risk of trust.
04:38Point number two, 70% of the workers are saying, I have no idea how I can get more productivity
04:45as expected by my company.
04:48So the companies are investing, betting on higher productivity, 75% of the employees are saying, look, it's becoming even
04:57more difficult rather than more productive.
05:01Point number three, that is more interesting for the longer term future, the younger generation, the Gen Z, 70%
05:09are saying, I have today the skills to be ready for the opportunity.
05:1450% of them are saying, I don't know in six months time.
05:19That is including another element for our reflection, that is how much the skills that I have today will be
05:27useful for tomorrow.
05:29And this brings into the labor market, anxiety.
05:33So on one end, the lack of trust, on the other end, anxiety in the middle to have this continuous
05:40need of learning.
05:41If we don't start from this, as Manpower Group, we call it exactly as you mentioned, people first, digital always.
05:48If we don't start from this, it's very difficult.
05:50The last point that I want to mention, AI is just accelerating a structural change that was already present in
05:56the labor market.
05:57Because the topic about upskilling and reskilling is not connected to the AI.
06:01We are talking about it with our clients and partners since 10 years.
06:06But now how to do it at scale is the new name of the game.
06:10Yeah, I think scale is definitely the challenge and you need to be prepared to achieve that.
06:15So, Erki, you lead the macroeconomic policy and national workforce strategy for Estonia, one of the world's most digitized states.
06:24I did a lot of research.
06:25It's very, very impressive.
06:27At a national level, how do you define machine collaboration in terms of economic competitiveness and what actually changes for
06:35your country?
06:37Yes, welcome from my side as well.
06:40Really good to be here.
06:41A lot will change.
06:42That's for certain.
06:44We don't know exactly what and how much in different areas.
06:47But I would maybe go back a couple of years.
06:51We are joking in Estonia a little bit that access to internet is a human right.
06:58So, on that level, to my knowledge, we are still the only country in the world that has all of
07:06our governmental services fully accessible online.
07:09100%.
07:11And if you want to have those services, it shows that you have to have citizens who know how to
07:19use internet, how to access internet.
07:22Moving fast forward, now we're in an era of using AI.
07:26If we want to be a next AI agentic state, the set of skills that people need to acquire is
07:35crucial.
07:36So, usage of AI for us in different areas, it actually means that there is a possibility not to work
07:42less, but to work smarter.
07:44If we're talking about competitive advantages or the growth of economy, that means Estonia is quite small, a little bit
07:50over 1 million people.
07:51We are saying that if we implement AI as a tool to make our people more productive, to make routine
08:01jobs obsolete or to make really, let's say, data-driven jobs, to make all of them be done by AI,
08:13it gives us a possibility for 1 million people to actually have an effect of 10 million people.
08:18This is what we're talking about, this is what we're talking about, this is what we're talking about, the change
08:23of the workforce.
08:24But to make that happen, first of all, people need to know how to use AI.
08:31They need the skills, they need the programs.
08:34That's why we have implemented so-called Estonia.AI program.
08:37We are investing a lot, so our students, our teachers know how to use AI.
08:43We're doing new programs in healthcare, in governmental services.
08:47This is something that we have to build on, layer by layer, structure by structure.
08:53Otherwise, it will be just a buzzword and there will be no actual influence.
08:57So this is where we see the future.
09:00We are building on it.
09:01But the critical value is that the people of your country or your company, they have to have the trust
09:08in the process.
09:09They have to see the transparency and they have to see the end result.
09:13That's what we're implementing.
09:15And right now, we have a survey that half of the Estonian population use AI.
09:24But if you go to the younger generation, Estonia ranks second.
09:28But over 80% of young people actually every day, they use AI.
09:34Now, for the government, it's a question about how to make them use AI more efficiently, smarter.
09:42That's the question for the next, let's say, years to come.
09:46Yeah, well, I think it's a great example of how you're preparing with fluency and literacy and providing access equally.
09:54It's really going to drive greater trust and greater adoption and the recognition of the value from AI.
10:00So it's a great story.
10:01Yeah, and if I may, listening to you, I was recalling a Dutch Nobel Prize, I think in 1975,
10:07that said that inequality is a race between technology and education.
10:13So more the public policymaker can do those kind of initiatives and they can contribute really to what the technology
10:20can bring for all and not just for some.
10:22Yeah, exactly, because one of the potential threats of the implementation of AI or the stronger usage in societies is
10:31that there is a possibility that inequality actually increases.
10:35But to mitigate that risk, as I already mentioned, when we started with the access of Internet has to be
10:43a human right.
10:43Now, the skills, how to use AI has to be universal societal skills in Estonia.
10:49Yeah, this actually emphasizes our potential so much more.
10:55If you don't do that, there's a possibility that, unfortunately, the only inequality increases and dissatisfaction in the society increases.
11:04That's why we need to see the usage of AI as a unified societal skill for everybody.
11:10Yeah, yeah.
11:11So I want to now turn to workforce implications.
11:16What we see with many of our clients and the research that we do at KPMG is that organisations are
11:23definitely excited about the opportunity from AI,
11:26but we see a pretty significant percentage that don't feel confident yet in workforce readiness.
11:33So needing all of the things that you've talked about, because especially now that we're not just using AI for
11:39assistance,
11:39but that it's executing, coordinating and orchestrating across the enterprise.
11:44At the same time, employees are using AI extensively in their personal life, at work, and often ahead of or
11:53faster than organisations have prepared for training,
11:57for the governance and for the controls.
12:00And so it creates this tension between, you know, you want to scale adoption, but you also don't want to
12:06widen the skills gap, reduce trust, lose control.
12:09So, Erky, with Estonia, obviously, as you've said, investing so early in education and talent pipelines,
12:17how do you avoid a divide between people who can collaborate with AI and those who can't?
12:22And I know we talked in the backstage about access, so I'd love to hear a little bit more about
12:28that.
12:29Yes, I a little bit already mentioned it, that the access of using those tools is really, really important.
12:35And that's why already last year we started the pilot program in our high schools.
12:41So we gave access to open AI sources platforms for students, for teachers.
12:49Now we're piloting, we are teaching them how to use.
12:53And with this pilot year, we will take feedback, how to renew the program, how to be all the time
13:00present.
13:01So to say, because we know AI at the moment, it's developing so fast.
13:08Any normal regulation doesn't come even near that.
13:12So as a country or as potential new programs, you always have to learn, you always have to adapt.
13:19So this is the question for not only the today's workforce, but for the future as well.
13:26Critical assets are ability always to learn more, ability to adapt.
13:32This is something we have to teach on a young age because I've seen and I've met with a lot
13:39of students in Estonia
13:41and a little bit mentioned already, they are not sure if what they study today will actually be beneficiary for
13:49the future.
13:50So this is something we, as a government or as political leaders or bigger entrepreneurs, we have to think about
13:56it.
13:57What are the solutions?
13:58What are the toolkits we are offering to our people, to our younger generation, to adapt, to want to learn
14:04more?
14:05This is something we do.
14:06That's why the next step will try to, I guess, in every Western world country, there's problems with healthcare.
14:14Access to healthcare, there's not enough doctors, it's expensive, so on.
14:18The next step will be implementing the usage of AI in healthcare.
14:22I was told that an average two hours per day, an Estonian doctor fills different paperwork still.
14:29Patient data, not on paper, is it on their computers, this can be all done by AI.
14:37It means that the doctors will have more time to learn new skills, take more patients, they won't be so
14:47burned out.
14:48So this is actually how you implement AI smartly.
14:51You need rules, you need safeguards, you need the overall infrastructure.
14:56But the number one thing is actually the set of skills that you give to people.
15:00This is something that the government has to enable.
15:02That's something we have to empower.
15:04Yeah, it's really important.
15:05And so, Ricardo, you've highlighted that AI skills are now the hardest to find.
15:10But you've also said that AI should enhance the human experience, not replace it.
15:15So what does AI skilled really mean to you in practice?
15:19And how fast can the workforce realistically catch up?
15:23This is an interesting question because when we look at the discussion on AI that started very boldly a couple
15:31of years ago,
15:32and as Manpower Group, we are always offering our view, as I said, at Davos.
15:37It was clear, for instance, two years ago, that AI will save the world.
15:43So everything will change and everything will be different.
15:47And this year, the reaction is completely different.
15:51Companies are starting to realize that investing in AI point-to-point doesn't have an impact.
15:58All the companies are struggling with the same challenges.
16:02Where is the adoption?
16:04Where is the productivity?
16:05Where is the scalability?
16:07And where are the ethical standards to use it in a good way?
16:11All the companies are struggling with those four things.
16:14So the real challenge is how to manage the workflow differently, how to manage the skill set of today and
16:22tomorrow differently,
16:23to leverage AI for the different phases that we are going to have with AI.
16:29This is what we see with the companies.
16:31And they are just looking to the single example of a productivity gain,
16:36but this doesn't move the needle of the company investment.
16:39What we see today is that, okay, we can use AI for replacing, to be more efficient.
16:47They do a routine job that nobody wants to do anymore.
16:50We are starting, and many companies are already starting to move from the phase of doing instead of the human,
16:56to have the deliver value inside predictable data analysis.
17:02Those are the things that AI is already starting to offer to the companies.
17:07The future is the third element that is exactly what you were mentioning.
17:11So the co-creation between human and the AI.
17:16And this is where we start to see, talking with the companies across the world, three very clear paradoxes.
17:24Yeah.
17:25The paradox of the productivity when you compare automation versus augmentation.
17:32There are already studies that are showing that just automation is decreasing productivity.
17:38Why?
17:39Because there is a need of human that control, verify, monitor the automation investment.
17:45The augmentation means the combination of the machine with a human is creating 24% of better productivity.
17:55So paradox number one, paradox number two, the finance director communities are starting to understand where is the value of
18:05what we are doing and where is the cost.
18:07And so they are starting to put together the cost of networking, the cost of IT infrastructure, the cost of
18:13the licenses to understand where to come up with additional value.
18:17And the third and last paradox that I see is the paradox of the risk.
18:22And this is where the companies, especially the ones that are really taking care of our long-term presence as
18:28organization economically and socially,
18:31they are taking care of the risk that is reputational, is legal, is financial.
18:35And these are the three things that the company have to deal when they think about how to create the
18:41right combination between human and machine.
18:43Yeah, if I may add that as a society, we have to give people reassurance that AI will not replace
18:51you.
18:51That's not the question.
18:53AI is really sophisticated, but still a tool.
18:56AI is meant for us to help us, not to replace us.
19:00And it's a question about what we are seeing, that AI can make our, as I mentioned at the start,
19:07make us so much more productive.
19:08It's a question about AI can make obsolete different mid-level management jobs, analyses that take a lot of data,
19:19time and everything, effort for a human.
19:21But it will take seconds, minutes for AI.
19:24And this means we can actually work smarter, create more value.
19:29It means we will get richer because, sorry, we live in a global world.
19:34It's a question about we are competing with everybody now.
19:37It means we need competitive advantages.
19:39We can't be afraid of a really sophisticated tool because nobody else is.
19:44And it's a question about how to use it more smartly than others.
19:48And this is actually where we can find a competitive advantage.
19:51Because if you look throughout the world, Europeans, we're really smart.
19:55We do a lot of strong basic research.
19:58Where we lack, sorry to be honest, but we tend to overregulate.
20:03We are not so entrepreneurial that we were decades ago.
20:08So this is where we actually want to emphasize or want to empower our people more.
20:13We have to deregulate.
20:15We have to trust our people more.
20:17Of course, we don't want the wild west of digital tools.
20:21But we want to trust more people that we trust.
20:25They have more freedom.
20:27And with more freedom, they actually create more value.
20:29Yeah.
20:30And I really believe if you have the right balance of regulation globally that's aligned and the right guardrails in
20:37there,
20:37it does allow that speeding up of innovation because you're also speeding up the governance and making the regulation make
20:44it feel safe.
20:45So I agree with that.
20:47Yes.
20:47Our mindset overall is, I like to emphasize it in Estonia as well, I see we don't run the economy.
20:53Government's responsibility is to build the best business environment that can be.
20:58Yeah.
20:58So we want to be the enablers, not the overregulators.
21:02Yeah.
21:03Yeah.
21:03So I want to move on.
21:04We're starting to run out of time, but we've got so much to cover.
21:07I want to talk about the operating model shift.
21:10So we're not now just talking about replacing jobs or automating certain tasks, but really looking at how work is
21:17being reorganized and operating models are changing.
21:20But with that, we see a competitive gap.
21:24You know, the leading organizations who can orchestrate at scale versus those who are just experimenting at the edges.
21:31So, Ricardo, do you see that companies truly are redesigning operating models or are they still stuck in pilot mode?
21:38And what do you think separates those leaders in making AI actually work at scale?
21:45Thanks.
21:46Because let's start from some data, because we have also some survey that we do regularly as an Empower group.
21:51One is a predictive intention of hiring in many countries in the world.
21:55And we have 40% of the employers in the world that even in this quarter, they are saying that
22:01they are willing to hire.
22:02And at the same time, we have a nearly talent survey that is checking the level of scarcity of the
22:09talent.
22:1070% of the employer in the world are saying that they don't find exactly the skills that they need.
22:15So, for me, those are two researches that are showing that the discussion around some job will disappear and there
22:23will be no job for human.
22:24It's not data-driven, at least.
22:26It could be emotional, sometimes ideological, but it's not data-driven at this stage for what we know today.
22:32So, first point.
22:34Second point, indeed, let's talk about, are we talking about the job title or are we talking about skills?
22:42Because, for instance, the Draghi document, the famous Draghi document of one year and a half ago, rightly was talking
22:48about skills and competencies.
22:50He has never been talking about, in the document, about jobs.
22:53Yeah.
22:54Because it's what is helping Europe to think differently about this tsunami that AI is bringing and easier to stay.
23:01So, definitely, to answer to your question, yes, something has to happen, is about learnability, is about upskilling and reskilling.
23:13The Draghi document, again, is mentioning that only 30% of the European adults do training at least one per
23:24year.
23:24So, we are talking about this upskilling and reskilling and long-life learning.
23:28The data is showing that this has not yet come to execution in the countries in Europe.
23:34So, maybe the wake-up call is, what should we do to create this system that can generate an understanding
23:42of the skill needed and co-create predictable analysis that we can act upon?
23:48We can act upon as public policymaker.
23:51We can act upon as public and private.
23:55But we can imagine certification, for instance, we represent as Manpower Group, an industry that can play a role.
24:02We cannot save the world, but we can contribute.
24:05Just thinking about certification of some skills, competencies along the chain of the careers of the workers.
24:12Yeah.
24:13You should to certify.
24:15So, what I mean is that we should forget an old model of working.
24:20Yeah.
24:21And also, the companies have to reinvent.
24:23For what I know, there is only one company that has put together HR and technology as a function in
24:28the U.S., Moderna.
24:29Maybe there are other today.
24:31This is the first one that came to my mind some months ago.
24:33We are still working in the companies with the McKinsey organizational chart of 100 years ago.
24:39So, maybe it's an opportunity to review how we can organize big operations, thinking about starting from bottom up, the
24:46skill, and creating the organization from there.
24:48Yeah.
24:49Okay.
24:49So, I really want to get on to our last topic, which is around a pretty big discussion of trade
24:56-offs between growth and trust.
24:58And I think about that a lot every day in my job.
25:00And I think it's the hardest part to figure out.
25:03You've got to balance speed with governance, openness and control, and innovation and trust.
25:10And then not just in theory, because I think it shapes how employers make decisions, how governments regulate, and ultimately
25:17how quickly society will be willing to move with AI.
25:21So, Erke, if you could talk about, at a national level, is trust becoming a competitive advantage or a constraint?
25:29Definitely an advantage.
25:31I would even go further.
25:33Trust is a basis, a foundation that people will use those systems.
25:38Because when I went years back, if we want, as a government, to people use digital services, I would say
25:46there are two key elements.
25:48First of it is full transparency.
25:50People need to know how their data is being used.
25:54They can check it.
25:55They know their data can be used only by a court order because it's our data.
26:00It's our personal data.
26:01And we need to be sure that it's fully protected.
26:04And that's the second pillar is the cybersecurity part.
26:08Because there's so much sensitive data, we can overlap it to usage of AI.
26:13We want security.
26:14We want, if we do anything in a digital world, that we want to know that our data, our knowledge
26:20can't go on our wrong hands, so to say.
26:23So, trust is essential and trust comes only with knowledge and full transparency.
26:31Both without, there can't be implementation of digital services or different AI-agentic new, I don't know, possibilities.
26:44What can we even not think of it right now?
26:48Well, I think about this every day, and it's music to my ears that you agree with me that trust
26:53is a competitive advantage, because I really believe that.
26:55But I know we're close to time, so we've had a great discussion about what is actually changing in AI,
27:03workforce implications, the operating model shift, and that trade-off between growth and trust.
27:08I just want to close with one lightning round question to both of you.
27:13What do you think is one capability that every leader or country must build in the next two to three
27:20years to succeed in the world of machine collaboration?
27:25Yeah, I think one, completely rewrite the organization.
27:29Two, prepare strong reskilling and upskilling at scale in a sustainable way.
27:35And point number three, we didn't talk about this today, but it's very relevant, completely reimagining the leadership model that
27:43we are used to be versus the future leadership model.
27:47Yeah, if I would have to pick one, I would pick adaptability.
27:52As I a little bit mentioned, that already people, younger people already feel that what we are learning today might
28:00not be useful after 10 years or so.
28:03So, it means things will change, always change.
28:08Time goes on and on every day and every month.
28:12So, the adaptability.
28:13So, you have to be ready to learn something new basically every year.
28:20You need to adapt with a really, really rapid change.
28:23Because what has happened in the past years, we are seeing the transition goes faster and faster.
28:31So, the possibility or the ability to adapt with those really rapid changes is a key factor of future success.
28:40Yeah.
28:41Just a tweet supporting what you said.
28:44There is a company that in the world, all the recruiters that have to hire for this company, ask one
28:49question.
28:49What have you learned in the last six months?
28:52What have I learned?
28:54I've learned that I have to continue to be a student.
28:56I'm amazed actually at how different my job has become in very recent months.
29:04I was interviewed in an article about six months ago about what I'm doing with AI and now it is
29:09so incredibly different with the agentic capabilities that I have.
29:13So, I have my little buddies, my digital teammates and it's really changing what I do every day and I'm
29:19learning more than I think I have since I left university.
29:22So, I'm pretty excited about it.
29:25Great.
29:26Okay.
29:27Well, thank you everyone.
29:28Enjoy the rest of Viva Tech and we'll hope to see you around.
29:31Thanks.
Comments