The most important skills for the next wave of AI jobs

Hi everybody. Thank you so much for coming to the MicroOne Forum. We're so happy to have you here today in our next session of Reshaping the Future of Work, our virtual series with Microsoft AI/EI, the AI Economy Institute. I'm so excited to dive into today's theme on how AI is reshaping jobs, skills, and human capabilities that matter the most. We're joined by an incredible group of speakers. Led by Dr. Mark Esposito, the Chief Economist at MicroOne and Harvard University, who will be guiding today's conversation, as well as Dr. Fabian Stefani, University of Oxford Professor and Microsoft AI Economy Institute Senior Fellow, and Dr. Diana Tuman from Columbia University and a MicroOne expert as well. So we'll leave time at the end for audience Q&A. And feel free to drop questions. There's just a little Q&A chat button on the bottom, and Mark will do his best to get to all of those throughout the session. And we're super excited to kick this off. So let's get started, Mark. I'll pass it off to you.

Great. Thanks so much, Sydney. And I can see this on air. It was Sydney's birthday yesterday. So happy birthday. So at least you have even more people celebrating on top of the ones that have celebrated already. So Guys, first of all, thank you so much for joining. It's quite humbling to see so many of you connected already. Now over 100 people and it's growing, the number of people that are connected now. And what makes this possible is the profile of the two speakers who are basically coming to the series today, Fabian and Diana. Each one of them, they operate with a lot of experience in their own respective fields and domains. And we'll try to use this time to understand a bit more about the work they do. But also try to instigate possible question that you might have. And as Sidney said correctly before, we have a Q&A function and please use it to, to put your question. I'll try to prioritize your questions after we kind of break the ice with Fabian and Diana on the topics that I would like to explore. Fabian, I'll start with you. This is a series that we're curating together with the Institute and Microsoft where you are senior fellow. Now, you've been working a lot on skills, if this is correct, from my understanding of your work. And I guess you have transferred the work when you became a fellow of Microsoft. Just tell a bit more about what motivated you to go from your academic position to working with a technology company as a senior fellow. So what was your thinking behind when you decided to either apply or you were asked to apply, right, into this specific cohort? We met already Morgan Frank from your cohort before. We'll meet two more people. In December, where we're discovering more and more about the incredible contingent that was put together. But tell us a little bit more about why your work and skills was important to be eventually migrated within the context of your fellowship.

Yeah, thanks so much, Mark. That's a really good question. First of all, let me seize this quick opportunity to say thank you for having me, to you and Sydney for organizing this. So you're right, the project that I run in Oxford at the at the Oxford Internet Institute is called the Skills Scale Project. And the name, it says what it does. We investigate how skills and the demand for skills, for certain skills, is changing under the influence of technology. And of course, these days, the technology that we're talking about is artificial intelligence. There are various reasons why I really wanted to join the first cohort of fellows with the Microsoft AI Economy Institute. And maybe just to start with two of them. So one is, of course, the idea of thought leadership. I think that the community that Microsoft has gathered for this institute is really unique in terms of scholars from around the world that in their domain and with the methods that they have studying how the labor market is changing. And I really appreciated that very, very diverse set of ideas and of approaches to this kind of same overarching question. Like there's this technological shock, there's this technology entering our world, and how does this impact the way that we organize work, the way that we organize our economy? And the other one is, frankly speaking, something That is relatable probably to those people that do social data science or computational social science that are somewhat engaged in this field of working with digital technologies or with digital data sources. And that is increasingly the, you know, where we used to harvest surveys and as social science researchers maybe 10, 15, 20 years ago, more and more we of course using the digital breadcrumbs of the digital economy of working with with online job vacancies or with the tools and programming packages that people apply in their daily lives. And of course here, Microsoft is an extremely unique and extremely valuable partner just in terms of, again, the network that people have when you think about LinkedIn data, for example, but also the tools that Microsoft themselves offer and the chance to maybe tap into these rich data sources and see really how are people actually engaging on a daily basis daily day-to-day business with digital technologies like AI.

That's great. Thanks for this, Fabien. Then I'm gonna go over to you and also my first question to you is, you have a fantastic and very, very fascinating profile across so many different domains. And you and I met indirectly through the MicroOne Expert Community where I think we were both serving in the Elephant Project. And I think what has always fascinated me about your profile is the fact that you're trying to translate the conversation at the organizational level. There's a lot in your work around human-centered design. I think you work and mentor organizations. So if you kind of share with us your, your perspective at the moment about whether this technology that we're talking about has improved the quality of work in organization, or whether you find that the balance scorecard between pros and cons is still kind of conflicted, right? And people are still kind of wondering. I'd love for you to tell us what you see from, from the front line of your work in terms of whether the quality of work has really changed, improved, deteriorated, is kind of blurred to tell. So tell us more about this.

Absolutely. And it's quite interesting because the, the headlines, I would say, that you see in the news are quite contradictory to what we're seeing in the in the workforce. And I'm not going to say that jobs are not being cut, because they are, and we are seeing that headlines. But it's important to think of it from the perspective of, um, the historical aspect as well as organizational design and transformation and the mindset that we as professionals can carry forward. And if we learn from sort of the Industrial Revolution, where we were in the past, and today we can think of this time as the AI transformation, is really no different in sort of the 20,000-foot view. And to answer your question in short, yes, we are seeing tremendous benefits. We are seeing tremendous benefits, but we're also seeing a divide in the workforce in terms of how people are thinking and how they're leveraging the technology. And what we can see is that those that can see the transferability of skills are adapting much better and faster towards this transformation and particularly using their skills in a transferable way. So individuals, for example, to bring it forward in a storyteller way, instead of thinking of yourself, I'm an engineer, and putting yourself in a box, to say my skills and values are around maybe human-centered design, maybe it's to leverage technology in a certain way, maybe it's around leadership, Whatever those skills are, they don't have to be confined to a particular vertical, which is where maybe our parents and grandparents felt comfortable. And I think that's where we're seeing the transferability of skills, and those who can do that in a way that is effective are able to leverage the technology like AI in the most effective way and continue learning, which is another key element in today's environment. In a positive way.

Great. You make me think of the follow-up question I want to ask you, Diana, right? And I'm trying to build a bit of a web in the conversation I'm hearing between Fabian and yourself. And so I'll try to now bounce you over to Fabian on this. Fabian, I think there's a paper about substitute or replace or supplement that you have been part of. And I guess the question should never be a binary, right? There are circumstances, I guess, where technology has replaced. And that's part of the industrial evolution of technology. And I guess there are circumstances where technology is being complemented. Now, if you're looking at the skills that have been complemented the most and the work you're doing at Oxford, right, so where you say, yes, this is not necessarily a binary question, there will be a bit of both, but clearly there are skills that have been better complemented than others, right? What are those skills? This is one of the questions where Even myself working a bit more on the governance of technology, I love some time to know whether or not in the narrative between augmentation or skill complementing, there are specific skills that are better designed to be complemented. And I'd love for you to walk us through that.

Yeah, sure. Thanks so much for giving us the chance to advertise a bit of our research as well. First of all, I'd love to— I'd say that I fully agree with Diana. I think it's very, very important, as she nicely said, to get out of this vertical. And think not so much about occupations but think about skills because that is not only the basis for our research but I think that's really a really important mindset for everybody who is maybe skeptical or slightly worried about this whole process of change because ultimately it comes down to acquiring skills and maybe also letting go of certain skills. I probably assume that The latter is less commonly thought of and it's also probably more problematic, particularly to the people that kind of think like, I'm so good at this specific skill, right? I really mastered this skill. Now I really have to hand it over to the machine and the machine does that. Well, for some skills that's the case. And that is basically the starting point for the paper that you mentioned, Complement or Substitute, where you had these contrasting camps that said like, okay, there will be automation to some extent and on some markets. Markets, but those are very special markets, I have to say. We already see that on freelancing markets, for example. We have already observed over the last years since the introduction of large-scale, popular large language models that some of the roles— I call them roles, I wouldn't call them jobs because they're very neatly defined, centered around certain tasks— have declined, but both in popularity but also in monetary value. So translation work, for example, or simple summarization work. And for these skills, it's clearly the case also if we think about our daily engagement with the technology, that the technology does have an edge. If we only think about that specific skill, like translating a document or summarizing a document, the AI models are probably much better than any skilled human in doing that. On the other hand, though, there's a camp that says, okay, but what happens to the truly unique human features? The skills such as resilience, such as teamwork, for example, or digital literacy, you know, as a human understanding how the technology works. And in a nutshell, what we did, we looked at 3 economies: the US, the United Kingdom, and Australia. And we looked at what firms are demanding. That's— I need to say that as, you know, being fully transparent. So we're looking at the demand side of the market. But what we see is pretty astonishing because on the one hand side, we see that there for certain, for AI roles, so the roles where at least one AI skill is being required, something like that could be ranging from prompt engineering to machine learning development. We see a significant increase in the demand for human-centered skills. What this means is that people that are engaging or asked to be engaged with the technology are much more likely to be also asked to bring in skills like resilience, like ethics skills, like digital literacy, for example, like being a good manager, being a good team worker. Our logic behind this is to think about this in comparative advantages, right? In the sense that you always have a bundle of skills. Any role that you have is always a combination of different skills. If the technology gets really good in doing something, you might let go of this. But in relative terms, then you can focus much more on the other part. The other parts not only become more popular, what we see in our study, also become more valuable. So there's even a wage premium for, let's say, data scientists not having AI ethics skills, having ethics skills, knowing how to ethically work with data or thinking about the biases that data might bring to you. And I think that is something really, really interesting, you know, thinking about this in relative terms. So letting go of something could increase the value of the other thing, of the other skills that you're unique with, right? That you're not competing with against the machine.

That's, that's fascinating, Fabian. I, I love the way you, you phrase also the confinement of the study when you said it's about the demands. So there was definitely an economic dimension that was probably determined the study itself to the term, to a point that you could measure it. But I love the fact that you're almost like helping me to go in the next direction with you, which is when I'm going to ask you to think more about the portfolio skills, right? Considering almost like if you where a finance manager and you're thinking about this as a portfolio, right? How would you see this? But before I go in there, I wanna take the very same question, but I wanna now basically translate it over to the work you do, Diana, right? So this morning I was doing a session here in Florida and a participant asked me a question. It's like, okay, so if I have to consider who inside of the organization is better designed to navigate this change that AI is clearly imposing upon us, right? Who would you think those people would be? And, and again, I, I'm limited. My training is in economics, so I tend to see things sometimes more from perspective of, of economic, uh, of economic narrative, less from the perspective of organizational narratives, from a functional perspective. And I'd love to ask this question to you. I wish I could ask you before the session, right? So if you think about who in the organization is better designed to help navigate this, this extremely liminal space where we are learning as we are doing. Is there any particular profile or skills that you have identified that to some degree is better designed for the, so like, set of versatility that is required to navigate this? Would that be a department? Would it be across the organization? Would it be about seniority? What is your take on that, Diana?

Yeah, and I think Fabian did a fantastic job in describing the complexities and letting go of certain aspects. And, and what he did so well is, is talk about the, the human mind as well. When we think about organizational change, or really kind of any change, we're also having to deal with the nuances of, of organizational psychology and just the human aspect of navigating change, which is never easy. And when you ask the question of who does it well, it's, it's not per se any kind of I would say, function or, or job title. I would say it's individuals who are, who do well with ambiguity, those who navigate well in uncertainty, those who are able to, I would say, um, recognize that and have a mindset, a growth mindset. So individuals, when we talk about a learning ecosystem, a learning environment, those who embrace those who do well with uncertainty, those who are able to approach conversations around integration of AI into the workforce as something as a positive and think of it as unbundling of tasks. And think of it from a perspective of, you know, I think it was Kahneman that came up with System 1 and System 2 thinking. And this lands a perfect lens through which I could say in a workforce that I'm going to leverage AI capabilities as systems to thinking outside of me, which allows me to really hyperfocus on some of the skills that Fabian outlined. So if you view this from a scarcity mindset, what are the human skills that are becoming scarce in the workplace? And it's exactly what he outlined. It's around critical thinking. It's around our ability to leverage the tools and advances that AI provides and build them into the workforce in a way that complements and augments the repetitive tasks that used to exist that took up so much of our time, which now allows me, gives me the freedom to then focus on those scarce tasks that we as humans are so good at or can be good at. So I think it's more looking at it through that lens rather than to focus on things like this particular job, because really anybody can be a leader within their own, within their own capacity, which is really a great way to think about it, which is why you're seeing so many organizations becoming more flat, which is why, because again, we're being able to recognize that anybody could be a leader within their own ecosystem.

Yeah, that's fascinating. Then you make me think about, um, kind of the Renaissance man, right, that has this, this polycentric structure, right, rather than the single competence, which I I think it reinforces a lot what, what you folks have been talking about, no matter whether it was coming from the point of research that Fabian was bringing forward or your point of practice directly working in the organization, which brings me to a question that you inspired now in the formulation, right, Diana, to Fabian. So Fabian, let's say that in a conversation with, with the organizations or leaders, we do advocate for a skills portfolio rather than single competencies. And let's imagine that the argument we're using is exactly what Diana said, right? We do this because navigating dimensionality across different levels of the workflows creates just a better integration. Do we have any evidence from the work you're doing that somehow the economic value of a portfolio of skills is higher than a single competence? And I'm asking this because I think sometimes in these days we find out that sometimes the perception of what we think feels righteous might not necessarily be supported by the empirical data. And the opposite is equally true, right? So sometimes research helps us to, to navigate the difference between the assumption versus basically what is proven to some degree within the boundary of the study. So what is your, your work telling you about this? Do portfolio skills generate a better economic value?

Short answer, yes, they do. So this is really a spot-on question in the sense because it encourages us again to think in skill bundles. And that is, from my point of view, not only a natural way, I mean, it's not the way that usually people think about this. I was surprised how little people think actually about this because often, you know, we have it particularly in the public discussion. We have this translation, oh, we see AI excels in something, in some specific task. Task. And then all of a sudden we make this huge jump to say, well, now everybody in an occupation that is maybe centered around the task or contains the task can go. They're all jobless, right? So AI is becoming really good in, I don't know, debugging code. Oh, now all software developers will have to leave, right? And that is the tremendous, tremendous, and I think also kind of a misleading jump because you never apply skills in isolation. Never. Unless you are kind of a machine, right? Then you do. Or maybe in very, very special markets, like again, the freelancing market might be something unique in the sense because you might have just a document that needs to be translated from Portuguese to English and it's only that one skill, but it's still there, still in the freelancing market. And that's one of the markets that we studied for the study that I'm getting to right now. It's always a compendium. It's always a bundle, right? You always, you have to communicate your results. You have to work with a team to get to these results. You not only have to work with the technology, but you have to debug the technology, maybe combine it with a different type of technology. And what we did is, that was one of our first studies actually for the Skillscale Group, that's already 2 or 3 years ago. We looked at the freelancing market and looked at the value of individual skills and looked at the value of bundle of skills. And what we see is, first of all, that it's almost always the case that the bundle of skills, certain permutations, certain bundles of skills are more valuable You know, the sum is larger than the individual parts. In that sense, it's always more valuable than the individual skills. But one specific group of skills that is extremely valuable in terms of complementarity are AI skills. Our skills center around programming languages, our skills centered around machine learning, our skills centered around data analysis. No matter how you combine them, we see that the value of AI skills strongly depends on the way that you combine them and increases strongly if you combine them in the right way. And this type of versatility, the fact that, you know, AI skills, this broad domain of AI skills is applicable to graphic design, but also to admin work, but also to software development, is in our study one of the key determinants, explains a lot of the variance in the value of AI skills. So again, coming back to your question, yes, absolutely in for the portfolio theory of skills. And that is also the way to to think ahead, right? To think ahead when you say, I like the portfolio idea, by the way, I love what you suggested, because it's very much to that. It's like, okay, I have this stock, you know, in my portfolio that might not be performing well for whatever reason, maybe, maybe because of technology, and I have to let go, but I might want to buy something else in, right? I might want to invest in something else. But it has to fit my portfolio, right? It's not— if I'm a green investor, if I invest in low-carbon technologies or companies, I'm not all of a sudden going to buy something that's, you know, completely contradicting my investment strategy. And that's the same on the labor market. You want to buy the skills in that fit to the skills that you have in your portfolio.

Yeah, Fabian, fantástico. I really think that that is what really generates return on social capital, right? Because it's really, if you think about portfolio and you're looking at the return, this is on the social capital that we have in the organization. And it only happened through this distribution across skills. That could constitute the portfolio. So I'm happy that your work, for what you have been working on so far, equally demonstrates this is not just a nice perception that would be nice to fulfill, but it seems to be also indicating strength, right? And so I'm going to try to use the very same question this time to you, Diana. And I know already questions are flocking in, and I will definitely leave the last 15 minutes or so for the questions coming from the audience. We have been So far, over 155 people always online. So this is amazing. Diana, I have a question. So imagine that we now have asserted and ascertained that, yes, a portfolio of multiple skills is important. We don't want this to be single competence. We want it to be more distributed. And people that have that tendency to be able to navigate ambiguity to some degree are probably better to lead this effort. And let's assume that we now have integrated technology as our priority because we have created a space between between humans and technology without having antagonism. If somebody asks you, Diana, is there any blind spots that I should be considering that maybe I cannot think about because I cannot know what I don't know, right? From the work you have done with the organization, you've been coaching, consulting, mentoring, right? Is there anything, areas of vigilance that sometimes we don't think about because we're so excited about the conversion or the adoption of the digitization, right? But we're not necessarily thinking about the fact that organizations are dynamic, right? They're not rigid structure. They are their networks in many ways. So what would you say in terms of areas of vigilance where we should consider some attention towards the blind spots?

Yes, absolutely. And you're right, organizations by design are fluid systems. And there is— what's different about this AI transformation is because it's happening so rapidly. And organizations are one unit of this transformational change that needs to occur. Institutions are different, educational systems, So what we're really talking about is we're talking about this holistic organizational design challenge that is ongoing in our industry at the moment. Organizations as a whole need to think about ways of upscaling as a capabilities to— in order to be successful, I think they need to create an organizational ecosystem in a way where individuals are viewed or given capabilities and are given tools and resources for upskilling in this environment that is in line with the needs of where the organization is going in the next 3 to 5 years. And right now that doesn't exist. Right now it's sort of, is really fragmented the way we think about where the organization design is headed to where the institutions and educational organizations are at the moment and the way that they view and design curriculum, for instance. Because really, fundamentally, it used to be very top-heavy, and the way it's still today. It hasn't really evolved outside of very few nuanced organizations, institutions. It used to be you go to school, you go to grad school, you get a degree, you get a job, you get a career, you stay in that career, and then you know who you are. That no longer applies. I think the shelf life of those kinds of mindset is very much very much becoming extinct. And I see both of you guys nodding because that's very much true. The way that we need to think about our, our learning is very much fluid. It's very dynamic. It's very hybrid. It's very modular. It continues to evolve. That needs to be, that needs to coincide and needs to live. And I think organizations as a whole need to have a much more active voice in how they that's being done, and they need to have a seat at the table, which I think, you know, Fabian, is what you're doing in part. And there needs to be more of that happening. There needs to be this huge think tank of ideas, and organizations need to reframe and say, these are the capabilities, these are the skills, this is what is needed and demand, and this is how we're going to train our skills. So people are of the mindset, my job is this, and I bring this, and I do this task. No. I bring this value based on these skills and I don't need to be so concrete and defined by these parameters.

Fascinating. Diana and Fabian, I love— I'm enjoying this so much. I think that the lucky one is always me because I'm the one learning from this amazing conversation that we're having. So I have my final question, the same one for both of you, and then I'll ask you to join me in looking at some of the questions coming in the Q&A. Exactly the same question. So Fabian, I think you have, you also have studied gender disparity in generative AI employability, right? If I'm not mistaken. And so what are the factors that are preventing this? Are they structural? Are they perception? Are they behavior? Are they industry-specific? So if you had to think about what you have found out in your research, right, that are structural barriers to, to basically be the higher rate of adoption of generative AI employability for gender, right? For women. What would be in your, in your specific experience, the lesson that you have taken from this? So what can we learn from the work you've done? And I'm gonna ask you exactly the same, not only because, you know, you are a successful woman working in kind of the tech sphere, but I think you also meet other, other women in the workplace that might maybe share their own version of the story. So I'd love to have your sense on this, and then we're gonna get to the Q&A. Abin.

Should I start? I guess that's probably the right question to say ladies first, right? In terms of non-barriers.

No, go ahead. Have you go first.

Okay, okay, okay. So I have to again be fully transparent as a researcher here. The work that I'm talking about or will be talking about now is not published yet. We're hopefully putting out a working paper soon in the next week. So there is an old story and there's a new story from my point of view. The old story is there are structural barriers, unfortunately, when comparing the success and also technology adoption of men and women, even outside the domain of AI. We are working hard, many people are working hard to tearing down these barriers. Some of them are more difficult, you know, and like, you know, there's still a motherhood pay gap, for example. So it's harder for women also in terms of technology adoption or building their skills because the career has been fragmented by motherhood? Those are questions, those are very important discussions, but I'm not going to touch upon them because I think here's a novel aspect. Because one of the novel aspects that we found is that it's a lot about risk perception when it comes to these technologies. So what we see is that there are these hard factors like age, of course, like education, like the type of work that you do, you know, there is still the tendency of men working more, it's more frequent that we work more frequently in STEM domains, which are of course more exposed to the technology, you know, something like engineering, mathematics, technology, than the occupations that conventionally are, you know, where we find women in more frequently. But what we also see is that particularly for the youngest cohorts, so 18 to 35, the perception matters as much as digital literacy. I'll give you one example. We looked at data in the UK where the gap in terms of personal use of GenAI is around 5 percentage points. It's not big. If I remember the numbers correctly, it's around 16% for women and about like 20-21% for men. Once we look at the younger cohorts, and look at those people that are concerned that AI has a negative impact on mental health, this gap increases to 44 percentage points. It becomes humongous. And the reason for this is because the concern about mental health does not really translate for men in that age group, but it translates significantly for women into lower usage of this technology. And I think that is something really crucial to bear in mind because we can work on all these hard factors and we should work on all these hard factors. Don't get me wrong, people should have physical access to technology and also the digital literacy to understand the technology. But what our study is going to show is that some of these risk perceptions, particularly for younger females, matter as much as the hard factors. And what to do about it? Well, I mean, to some extent they're right. I mean, it'll be hard to say like AI is is without any harms with regard to climate or the labor market or mental health, they are correct. So first of all, I guess the most intuitive approach would be to say like, let's reduce some of these harms. And then if there's still a residual which is a tilt in perspective, then let's think about like, why are women more critical? Well, because maybe conventionally they always had a more critical, you know, stance when it came to potential harms to others and society, while men focus more on themselves Women tend to, I mean, I know it's a very vast generalization, tend to focus more on society. And maybe we want to design these technologies to be more inclusive, to be less harmful. And that's a, you know, that's a really, that's, that is a real one of the many indications why we should be doing that. Right.

So same question to you, Diana. And again, Fabian has helped us a lot, right, to spell out a lot of factors already, but your experience on this and, you know, it's interesting that maybe his study was referring to the UK market. At you operating maybe in a different geography, right? So what has been your, your experience on that?

It is quite fascinating because if you uplevel that question, first and foremost, it deals with human perception and tolerance of change, really anything. And there is a really heavy human component when it comes to understanding and tolerating and embracing any kind of change. And when you think of it, and, you know, a lot of the things that you mentioned, Fabian, are absolutely correct, is that on a human level, we perceive change as an unknown. And even evolutionary speaking, we tend to avoid change at all levels as humans. So, and there's so many change models that we can point to, from Lewin's change model to Kotter, to whoever, whichever theory you are fond of, you will understand that change efforts at large fail. A huge percentage of them fail because of that human element. And for females in the professional setting, a lot of things that happen to prevent us to embrace change, or women to embrace change, have to do with aspects around— maybe it's psychological safety, maybe it's fear of the unknown, but to really, to sort of bucket it into categories. Individuals, in order to embrace and to lead and to be effective in any kind of change, need to understand it on a personal level. So they need to understand, why is this beneficial for me? What is my role in it? How do I effectively navigate it? They need to have the psychological safety and the tools to do that. Then they need to understand it on a, on sort of the next dimension, which is the team in my, my close ecosystem, in my close framework. Why does this matter? How do I play into the larger vision? Who is my support? What are the tools, the capabilities? What are the benefits of doing that? And then within the organizational context, so the organization, you need to have your champions from above, you need to have the, the frameworks in place, you need to have the mechanism to support this. And then you, you're moving outside of the organizational context contexts, which is where you have this broader framework and environment. And this is why this, this again, broader ecosystem that I described of change agents needs to sort of come together to make this a successful effort.

That's great. But then, look, I think I want to stay with you as we move into the Q&A. What I think I'm going to try to do is to bundle questions that maybe they will work well with one formulation of question. And the question I want to ask you, they're inspired by Nicolette and Jaime or Jamie. That is, they're asking this question. I tried to put them together, right? So how do you quantify the non-quantifiable skills? And Nicolette is mentioning empathy, grit, creativity, right? So I think this is a question that is important. There is a lot of these incredibly powerful skills that are difficult to be quantified. You were mentioning before Kahneman on System 1 and 2, so the behavioral side of it. And I think Nicolette's question naturally has a follow-up, which is the one from Jaime or Jaime. So what will be the role of HR, right, considering that we are in the presence of probably a significant shift of purpose semantics, meaning of the function of skills or jobs inside the organization? So how do you see this eventually evolve in your own experience?

Absolutely. And I think just as you mentioned, this, the skills that are described here would fit really well into the System 1 thinking of Kahneman's model. And they are, I would describe them as the skills that are actually going to be needed more and elevated when we look at that scarcity mindset in terms of the critical human skills that we can bring into the organizational context. And absolutely should not be lost because as artificial intelligence and AGI is soon coming to, you know, on the forefront, We as humans need to be cognizant and understand what are our strengths and capabilities and where can we augment and elevate our presence within the workplace. And those skills are exactly at the forefront. We need to be able to communicate. We need to be empathetic. We need to— those are the core components of leadership that absolutely must be present and cannot be lost, which really segments into your second question around around HR. And the more that HR— again, HR is where we are, where we started. It's sort of, again, putting something into a box and defining it. But what are the capabilities? What are the skill sets of somebody who is an HR professional? What can— what do they do? What value do they bring? I would sort of reframe that question in terms of not that I am an HR professional, but what value do my skills bring to the workplace and how do they transfer across those verticals? How can they be augmented or amplified in the age of AI? Is it organizational design? Is it teaching individuals how to work side along agents? Things like that. Those I would ask people to be more creative and think those questions through that lens.

Yeah, great. Um, Fabian, there's a question that, um, Sydney helped me to, to synthesize from Aman, who has written a quite, quite interesting and equally complex question. And to simplify that in, in this, you know, do you ever see— and, you know, you are, you're a scholar who was working on the study of this technology, right? And there's, there's both an economic, there's a social dimension to it, there are many dimensions, right? Do you ever see this as a reversal at any point in time? And I'll try to use— I was having dinner with a friend of mine. We have written a book together on digitizing the emerging economies with Cambridge University Press. And he said to me, well, we eventually did nuclearize because we decided even if we could, we didn't want to, right, move forward. Right. So, so we do have the ability to think that maybe there could be a degree of reversal. From the technology the way we see that. And I think this is because maybe we realize that there are tasks that we want to delegate back to humans for any kind of reasons. Is it because we are concerned that this could become a bubble which is asymmetrically distributed across the world? Therefore, it could create uneven distributed future for too many people, right? There are many reasons for that. But in the nature of the work you do, do you see a space for the conversation about reversing this acceleration? To ever be in place, or do you think we are past that point?

I think there's certain things that are out of control for us. I wouldn't put myself in a position here to say I have the most comprehensive overview of the entire literature, economic literature that's going on around AI. But I think this comes to— we don't have to necessarily limit this to AI. We could talk about technological change in general. And I I think here a certain pattern emerges, and that is we can definitely administer certain aspects of change, but I think some of the most fundamental logics cannot be changed. Or let me put it differently, if we wanted to change them, it will be potentially very, very harmful for us. Let me give you one example, because the thing is The advancement of any technology, the proliferation of any technology, crucially depends on certain market dynamics. It depends on all sorts of economic aspects that firms innovate, they have a certain incentive to innovate, they want to sell services and products, people buy these services and products, but for the firm to do that, they need the technology, they need skilled humans to do that. And if we wanted to change this overarching dynamic, I think we would be— the only way to do that would be really to break that system in a sense. And I guess there are various reasons why we would not like to break the system of innovation, scientific research, and competitive markets or competition. But certainly some aspects of this can and should be administered in a sustainable way. You know, it's the same thing when talking about energy. You know, the, the, of course, the, the market will go for for the most efficient, productive, and cheapest way for producing energy. And similarly, this applies for other technologies like AI. But that might not be the most sustainable form. Coal and oil were these types of energy. And we are, I guess we're all on the same page that they come with huge negative externalities, to use economic language, meaning they pollute our ecosystem, they're harmful to people, they have a lot a lot of bad sides to it, apart from the fact that the market, of course, is pushing towards them. And similarly, we could think about AI, and I think we can and should administer some of these aspects. So long story short, there is no reversal of the big— from my point of view, from my perspective, there's no reversal of the big technological trends. But whether AI on the road there will be harmful in terms of towards the labor market, the environment, or mental health, that is our task as humans to govern properly and, and, and, and reduce these in terms of effects.

That's great. Fantastic. Thank you for sharing this. I mean, really appreciate, we know, the kind of the discovery, right, of this inquiry that we're building together. So folks, stay with me a few more minutes as we try to go through this question. It's amazing that now questions naturally, they are flooding us, right? So there's a question that Priya is asking, and I'll try to ask that to you, Diana. And I'll try to do it by also matching a question that was coming on the human resource department, which I think was asked by Mesa later on, right? So I'll try to put them together. So think in terms of basically the role of HR that has been historically sourcing skills from institution, university, education. And now they go through this incredibly stormy period of time where things do not work according to the blueprint anymore. The protocol is broken, right? Or the house of cards has fallen, if you want to use a more dramatic metaphor. So how do you eventually help them navigate this in terms of question about where should I get my talent? Where do I get the skills? How do I redefine a human resource department where the primary function has equally changed in the sourcing of candidates? And so I think there's a question about recruitment, you know, because you've been within proximity with Microwon that Microwon has been trying to help a lot with the use of AI in recruitment. So I think it's really about building the narrative rather than necessarily reinforcing the old paradigm. So what is your sense of this in terms of the effectiveness of sourcing talent and therefore the skills necessary for that differently from how we did it so far? So what would you say to question coming from, from that kind of, you know, point of view?

I think it's too many, too many things around AI. I think the right way to think about is it, is to sort of turn it around to say, just because something has been done a certain way for the past 50 years doesn't make it the right way. And I would invite people to think creatively, to sort of throw away the blueprint of how things used to be done and ask yourself the question of, how would I do things if I had to start from scratch and just start all over again? And to emphasize the fact is maybe no longer is applicable, maybe what used to be done no longer fits the blueprint, and that's okay. So in this new age, in this new reality, what would you like to see happen? And this is where human-centered design is actually paramount, because what you're doing is that you're not taking what used to work and you're saying, how can I augment this? How can I change this? But no, instead you're looking at the problem. And you're solving that. So if that is the problem, maybe instead of saying, well, recruitment used to be my focus point, or how do I recruit people from this regular way of things, maybe instead I need to think of how do I create an ecosystem in partnership with institutions like MicroOne, with, with institutions that provide modular learning with others to create this new environment that works in this new paradigm. So maybe that's a new way of thinking about things. Maybe it's about creating a new way instead of focusing on what didn't work. Maybe it's about just forgetting that and thinking, what can I do differently? If I had to recreate it from scratch, what do I want it to look like? And that's okay. That's the thing, is that maybe that's That's okay. Maybe we can just think differently.

So that's great. Thanks so much. And I've been noticing questions coming also from the MicroOne community of experts. And folks, in that case, as you have access to us nonetheless through this live channel, please maybe redirect them directly so that we can expand it to the community as well. I would love for this to happen. I'm talking about Edward, your question. And I guess any other of you that is part of the MicroOne community, please folks, let's do that as well within our channels. Dublin, I'm trying to wrap this up, folks, and I realize that question came at speed, right, when we started the Q&A. I'll try to blend again a question that is a bit about Priya and Alexander, right, that asked in together. So I think there's a question I'll try to phrase this slightly different. We prepare a lot or we talk a lot about complementarity, augmentation. We identify a taxonomy of skills. And we kind of spend a lot of time in recognizing that, yes, skills have changed and we need to make them much more distributed, interoperable and all of that. But we also assume in a certain kind of zeitgeist, right, which is where things are happening right now. Imagine that we're now trying to look at the limitation of our present imagination and we're thinking, well, how do I know that this will be future ready or that will have a balance of future readiness in the future? How do I know I know that what I'm doing right now in prioritizing certain kind of skills will carry some level of value also into the future, right? What gives me that sense of reassurance? And, and I know the question is limited because I might be using a language that was legitimated by how in the past historical data and profile could undertake certain kind of distress or tests. And today you can say things are changing so fast, but we still need to have a way of gauging the risk risk when we're making this decision, right? So what has been your sense of it? And it doesn't have to come necessarily from the research. It can also be your, your sense of hunches on this. So how do we define the future readiness considering that things are changing so fast?

I guess we can't, and that, that's part of the, that's part of the picture. So Diana already said that those that can bear uncertainty are much rewarded. I think that is something that I fully agree, agree on. I think there is this nihilistic part in this answer where I say, in the long run, why do we care? A lot of the doomsday projections that are being made are about the distant future. In 50 or 100 years, maybe this is going to happen. I would ask, do we really want to invest a lot of time thinking about this while there are some obvious challenges changes that have zero uncertainty that they are happening right now. And this is often my frustration when I hear these talks about technological unemployment, mass unemployment because of technologies. I don't think this is going to happen for very scientifically based reasons in my own studies. But let's assume there might be a scenario like this still in 50, 100 years. Why do we invest so much time on thinking about this rather than looking at the problems that we have right now? And in addition to that, I think there are two things that we can, that we can certainly, um, do. One is kind of a bit of an, uh, yeah, an, an evergreen. Might be a bit of a boring answer, but the skill to reskill will remain important, right? So, so the, the ability to adapt, adapt to a changing environment is, is certainly something that I, that, that I strongly recommend. There's not a single other skill, not meta-level skill, that I would say we definitely have to invest in the long run. But there's also something that saves us in the short run, in the medium term, and that is things develop much slower than people think. We see this disruption of technology. So we see all of a sudden the jump on the technology side. And I'm saying that explicitly, on the pure technology side, we see the jump that a certain technology like AI is capable of doing extraordinary things.

Things.

But that is purely the technological side. There's no social aspect included in this. What I'm trying to say here is this technology still makes mistakes. It still needs to be incorporated and handled by a human being. It needs to be governed by a human being. It needs to be further developed by a human being. And even then, on the next level, the firms are not going to react immediately to this, right? HR departments need to adjust to new roles. They need to write new— they need to, you know, offer new positions. The economy as a whole then on the top level is so slow moving, even in light of these dramatic technological changes that we have loads of time to use this momentum still. So even as AI has become good in debugging code, the demand for programming, for programmers, for working with code, for example, the roles might change, but the demand for this will still continue because there are many, many, many other social aspects that are not tailored in or that are not put into this discussion when we look at technological change. Yeah.

Fabian and Diana, I have to tell you, I ran several virtual series. This is the first time where I know I will not be able to finish the Q&A. And it speaks highly of, first of all, you two. They are here for you. Second, that we probably need to dedicate even more time time to engage with our communities of reference, not only within MicroOne, but also in our respective, you know, respective places where we operate. Because I think there's so much curiosity that is so important to create the sense of convergence of the common ground of dialogue, right, that create the discourse that I think we deserve to have on conversation that sometimes feels to be overly simplified by just an economic narrow of metrics, right? And I think we should definitely make it much more humanized in this case. So I'm going to invite those of you that will be disappointed to stay in touch. I'm sure— and I'm speaking on behalf— that Fabian and Diana will not mind if you are asking them questions related to the work they're doing. And so please forgive me if I'm not able to give you some face time on your question, but please understand this was incredible in terms of the amount of questions that we received. To Citra, who is connecting from Indonesia and it's 3:55 in the morning, Citra, as you're part of the Micro One Expert, please send us that question in the channel on Slack. And look, I'm sure Diana will not mind maybe provide you with an answer. I will equally give you maybe an insight if necessary, but do not please, do not stop engaging. I really appreciate that. I don't have the time to address your question. So I'm going to do this. I'm going to be rude. But first of all, folks, thank you so much. This has been amazing. We had an amazing number of people always online. It makes me think that we probably need to go deeper and continue to work on this by making sure that the questions that people are asking are going to be addressing the context of a boundary of perspective, a polarity of reference that we can keep on exchanging. And that can only be done both by the practice and the research at the same time. It is where I think we create our legitimacy by theorizing and implementing the real world, right? So that we're understanding this calibration that is required to just get to a better integration of this technology as part of a form of augmenting our life in many ways. So I really appreciate on a personal level the work that you respectively do in your own fields. I hope this is not the last time we're going to see each other and that we're going be able to continue to work together. So Fabian, I'm going to start with you. Thank you so much for the work you're doing at Oxford, but equally with the Institute and Microsoft. And I look forward to continue to work with you in any possible ways. Diane, I'm very happy that we're sharing MicroOne as an environment where I have an easy access to you, at least on Slack. Fascinated by the work you do, and I just love that we learn so much from your own experience firsthand in working with organization on this daunting task that are not necessarily easy to navigate, but that nonetheless engage— they require engagement, right? And that level of engagement clearly transpires from the work you do. So folks, thank you so much for this. This has been amazing. We'll stop the call, and I won't see you again because that Zoom will actually close. But this will be recorded. We'll have a way to share with our network. And I invite all of you who were with us throughout the entire session to stay in touch, to write directly to Fabian and Diane if you have a question. And thank you again for being part of this. Truly appreciate it. For thought leadership, and I'm really happy that we met, although virtually. Thank you again, folks. I really appreciate it.

Thank you.

Thanks so much. Take care.