The rise of the human–AI workforce

If this technology deployment is as widespread as the research already predicts as possible today, everyone will need to understand how to deploy a new interaction model, and not everyone will have to understand all the ways the AI works underneath and behind, just like I don't know how my phone does what it does. I just know how to use it. That's McKinsey Senior Partner Alexis Kriv Kovic. She joins me and McKinsey Partner Anu Madgivkar to help us understand what happens as people increasingly work with agents and robots. This is the McKinsey Podcast, where we help you make sense out of the world's toughest business challenges. I'm your host for today, Lucia Raheli. Alexis Anu, welcome back to the McKinsey Podcast. Thanks for having us. And Kudos on your new McKinsey Global Institute Report, agents, robots, and us in the Karazite Geist. One aspect of the research that stood out for me was that you disrupted the humans versus agents or humans versus robots and construed us very much as in a constructive partnership with these technologies. Alexis, what is it about that partnership that immediately resonates with you or makes you uneasy? I think what's most exciting here is the fact that this technology hits every function and force in any industry you can name. And so the top to bottom scale of the impact here is truly profound. If you're frontline workers, factory workers, if you're white collar workers all the way up through the CEO suite, just the scale of the change we're seeing, it's already here. So the real possibility and the huge challenge for organizations is how on earth do I absorb that much change this quickly? Anu, what about you? We found that the vast majority of skills are really shared. And so this is an opportunity that excites me because it means that by working with AI in some shape of form, we can all super skill ourselves. But it's also a source of uneasiness if you think about it, because what it means is we do need to use AI to super skill ourselves. We can't stay at that same base level of each skill and expect to add a lot of value. So we do need to enable everyone in the workforce to work with AI and get better at those skills. So it's a massive opportunity, but also I think a responsibility of how do you actually upskill people to use AI and be better at what they do? The headlines can really be apocalyptic on the AI vis-a-vis the future of work in job laws. And in this MDI research, you say that more than half of work hours in the US could be automated across you are saying a really broad range of professions, right? So if this isn't about job loss, what is it really about? Well, the research definitely shows that based on currently proven technologies and capabilities that technology possesses, more than half of all current work hours could get automated. But I think what we shouldn't lose sight of is the fact that human beings are viking, even in the current work, a lot of this type of work is cognitive, a lot of it is actually social, emotional, interpersonal, and some of it is actually physical work. But more importantly, as we adopt and use technology in workflows and in business processes and things that people do as part of their day to day work, we're finding that that creates new kinds of work for the people in the loop. So that might involve new tasks to guide or prompt or validate or refine or build on what AI is doing. It may also involve completely new kinds of demand, things that just weren't possible to do. But now we can do them because we have technology. So we can do things at better quality. We can do much more R&D, or we can clean a facility like twice a day instead of once a week. What do you think will define high value human contributions in the next say five to 10 years? There are a whole set of skills that involve critical thinking, so aspects of problem solving, aspects of negotiation or conflict management or managing a team. In healthcare, for example, trauma care or even first aid or complex surgery, we may use robotic and agentic solutions to aid that, but you would still continue to have human beings as vital in those sets of skills, but administration or customer onboarding or communications with clients or customers, those kinds of things could get much more automated. I thought that example of radiology was such an interesting one in the report that radiologists as a profession, the number has actually grown since the acceleration of AI in diagnostic work and so forth. That was a great example. Things become possible when a skill or a quality that's very scarce suddenly gets unlocked. In the case of radiology, the fact that you could use technology to scan images and process data and bring out inferences so much faster just help unlock that supply, make it more available and met what you might call a latent or an untapped need or market for that thing. Even if you go back to the smartphone revolution, there were so many new kinds of needs that we discovered just because they became available, right? Like the whole ecosystem and thinking about social media and marketing in a very different way. In addition, their skills we can already see in the skill index are going to matter more social emotional skills, negotiation skills, coordination skills, process management skills. That existed previously, but we're going to put a premium on them and we want people to spend more of their time in any job or task. Focusing in those areas as AI takes other things off their plate. And then I'd argue there is a new set of skills leaders in particular will need to have to meet this moment, especially things like a voracious learning mindset. Alexis, you lead our people in organizational performance practice at McKinsey and you have talked so much about the trend towards skills based hiring. I'm wondering how that approach fits in here as roles begin to shift and change. I was just with two dozen C.H.R.O.'s last week from leading companies in North America and many of them were discussing this conundrum of. I'm now facing a moment where from a hiring standpoint, I both think every job description needs to be rewritten because some combination of the things I use to look for in the market are really not going to be nearly as important anymore and other ones are going to matter a lot more. But practically what they said is I don't think I'm getting better candidates because now I'm getting candidates who are using the same tools I'm using to try and predict what my tools will look for. And set themselves up. So I think you're going to actually see some amount of us reverting to what you might call old school analog ways like come sit in an office and take I don't know a personality test. You a puzzle. We see folks are asking software engineers to do live code. So all of these evolutions in the distribution of scales across this hybrid labor pool assume that companies actually capture the value of AI. The research puts that value at almost $3 trillion in annual value by 2030. But of course, we also see that most companies are not yet seeing meaningful gains at least at the enterprise level. What needs to change there? The real opportunity sits with how do you take what's right now more pilot and point experimentation in organizations focused on business value and explode those into big at scale. You know, bet the company changed the business points of EBITDA kind of opportunity and I'd argue those are for most organizations at least the starting point for those are few enough that what you really need is the leadership team to line around a value thesis like where is the opportunity here. And then really martial behind it in a way where you can attack at scale and what I mean by that is I think most organizations the biggest opportunities cut across more than one leader's domain they might start in supply chain. but they immediately connect into the front end of customer service and delivery and ordering and processing and back into areas of manufacturing. I know anything to add there. Everybody needs to flex and build their muscles a little bit and get more familiar with how to work with an AI tool or agent, but whether it really will be transformational in terms of the potential unlock is a question. And so it's probably the t shaped approach or maybe a series of t's right where you do need that horizontal capability building, but then you absolutely need to place a few important bets around end to end processes to really change and reimagine and think about doing very differently. You also have to think about where is the market going to lead me and what parts of my profit, who are most sensitive to this disruption, whether it's an internal possibility or a competitor threat or frankly just what the market is telling us customers will want. So if you're in retail, it's entirely conceivable that your customers want to do agentic commerce or if you're a bank, it's possible that your mid to large sized client might start wanting to use agents in the interface with, you know, the finance function and the bank right. It's interesting to think about in the future, if this technology deployment is as widespread as the research already predicts as possible today, even if we don't yet see it in the workforce. Everyone will need to understand how to deploy a new interaction model and not everyone will have to understand all the ways that AI works underneath and behind just like I don't know how. My phone does what it does right, I just know how to use it right, but when you're asking, particularly with things like agents, when you're asking them to do bodies of work, we will need to learn how to validate that provide the right judgment redirect as on who's described work iteratively test and learn. And now there's going to be a much higher expectation of experimentation, which comes with it real judgment to know where did that work, where did that not work, what part of what I tried didn't work so well for me and that's just a very different. Cultural day to day expectation what you're describing is almost managerial so suppose I'm leading a hybrid team. Where some of my colleagues are agents and they're learning continuously and they're working on flagging lead 24/7. How does my leadership of that team need to change? They will fuel the ability for me to work continuously at a different scale across a broader array of things. I think the whole paradigm or the frame of reference that many organizations and managers have about productivity levels or KPIs or optimizing workload quality and output. These are the typical questions or dilemmas for managers and there's a certain frame of reference they have that's based off. What do you know work to look like and I think all these things are going to come up and be questioned in pretty fundamental ways if you have an agent solution that's able to you generate 5000 reports overnight. You will come up against a new bottleneck because the human capacity to kind of review those reports won't exist right. Every time you're up against a new bottleneck I think there will be innovation and technology that then thinks about addressing that bottleneck right and our benchmarks of what is good or what is attainable will constantly change over that time so I think there's going to be a lot of that flux and change and. Part of being a good manager will be just having an appetite for that and being resilient through it and being able to be invested and creative through that process. What if we were to redesign education and training from scratch on the basis of this research what would we stop teaching and what would we double down on. It's tempting to say I can do so many things and you know why should we learn how to do them with the analogy to that is because I have a calculator does that mean I spend no time just developing a certain ability to think about numbers or just because I have a chat GPT that can write something for me does it mean that I spend no time actually trying to write myself right because. I think there is some element of very foundational cognitive ability or physical ability that comes just from doing something and you're going to have to think quite carefully about what part of that you need to preserve versus what part you don't and you may also I think over time have less focus on very specialized. types of learning and more emphasis on more transferable foundational generalizable types of skills and capabilities because the workforce will be in flux and if you spend five years going deep on just one narrow area or learning one particular skill or certification that may not be relevant by the time you're in the workforce careful on who you're going to dissuade everyone from PhD programs. One thing I think is really exciting about this moment of re-skilling and upskilling is how AI itself can allow us to do this so much better so a lot of the feedback on workforce learning is twofold first you can get to a tailoring of one I've ingested. You know all of the data points about how you perform and I have built a tailored point of view of where you have skill gaps or opportunities and how you can close those. And then second I can introduce them in the moments when you're doing the work I just saw that you have 30 minutes free do you want to use this time to send out these five notes as follow up to customers this is a common practice for people. You know developing these sorts of relationships I have taken a first pass based on your tone of voice and what best practice suggests will lead customers like this to respond. So many are calling this an existential moment for leaders or managing partner for North America or could you recently describe this as a CEO legacy moment a maker break moment in that context which risks worry more moving too fast. With agents or moving too slowly and why moving too slow I think speed is a strategy in and of itself and the biggest risk companies faces because there's so much ambiguity so much unknown and unproven that they will wait for more clarity before they make bets and move with a belief that they don't want to waste. Money and it's an incredibly dangerous position to be in because by the time you get enough clarity to know definitively where to go you'll be so much further behind. And of course there are all of the considerations and concerns people have about speed whether it's you know risk management right or whether it's an understanding of ethics and compliance in the new context of AI or whether it's regulation or whether it's public education about how to use AI responsibly. On all these fronts we're going to have to move faster to tap into this opportunity looking ahead say 10 years from now what would convince you that humans and AI have built a partnership that truly works across the board for employees for organizations and for society. Oh my gosh 10 years ahead this augments an unleashes human potential not replaces it and that we're on the other side of the uncertainty and the fear curve and we're in a place where just like a smartphone or a laptop or the internet it's a positive part of our daily life and the things we grapple with are how to maintain the right controls around that the right. Equity for folks to have appropriate access but what we don't worry about is should it exist or not and is it a force for good I think for me it's about this new way of thinking about work as something we really enjoy and find more fulfilling because work is so critical to the human project right but if we can use AI or if AI has the effect of enhancing the quality of work the experience of work. And then the other thing I would love to see is whether the benefits of AI have been democratized has everyone really felt them have been seen transformational effects and wider access and better quality I think that would be real success on a little success thanks so much for joining us today thank you thank you. Thanks so much for listening to the McKinsey podcast I'm Lucia Riley and I'm Roberta Fasaro find us on McKinsey dot com we'll have a transcript of this episode up shortly and download the McKinsey Insights app where you can find this podcast in other helpful content updated daily if you enjoyed. We'd love for you to leave a rating and a review. We'll see you next time.