Well, welcome everybody to this stakeholder dialogue on defying cognitive atrophy. Thank you for joining us. There are some people who enthusiastically went along to the Board of Peace signing and now can’t escape from there.

But congratulations to those who have done so. Thank you for joining us. This is, I think, one of the most important subjects of the coming years.

I told someone on the way here that this was the session I was moderating and they said, what on earth is that about, defying cognitive atrophy? But I think it will become one of the big topics of the next decade because it’s about really the future of the human mind in the age of artificial intelligence. And I am joined by a great panel here.

I am William Hague, the former Foreign Secretary of the United Kingdom and now Chancellor of the University of Oxford. And I am joined here by three expert colleagues, by Omar Abbosh, who is Chief Executive Officer of Pearson, and by Anna Frances Griffiths (Vignoles), who is the Director of the Leverhulme Trust, and by Karl Eschenbach, who is the Chief Executive Officer of WorkDay.

Now, as I say, this is a vital topic for the coming years. This session is connected to the Centre for the New Economy and Society and its flagship reskilling revolution, which aims to reach a billion people by 2030 with better education and skills and economic opportunity.

I want to remind the online audience that if they’re sharing about us through their social channels, they should use the hashtag WEF26. We’re entering a paradoxical moment in human history. As AI adoption accelerates, cognitive skills are becoming more economically valuable.

Yet we are outsourcing more of those than ever before. So the question for today is how do we ensure people retain the ability to question and reason and judge? in a world where machines will increasingly be thinking for us.

The World Economic Forum’s Future of Jobs report shows that analytical thinking remains the most sought-after core skill among employers. But the latest forum research also shows that human-centric skills are unlikely to be automated due to AI, but they can be quietly eroded. So if you look at the slide that is displayed there, that is the research that shows that mathematical and statistical thinking is easy for AI to replace, but not empathy or active listening, for instance.

But even those skills can be eroded without regular practice and engagement. Core cognitive capabilities, such as judgment and critical thinking, deteriorate over time. And then there’s the issue of whether AI will widen the gap between learners who are taught to question, guide and critically evaluate AI and those who merely consume its output.

Will we divide as humanity between people whose mental faculties are leveraged and enhanced by AI and others for whom it is reduced? So then we have to work out, if this is a problem, how we redesign education, technology, workforce systems for the future to strengthen human cognition. So this small subject is what we are engaged with today.

And let me turn to Anna, first of all, and ask you to think about what do we know about the current situation? It’s a fast-moving situation. We’re short of data on this.

But what do you think we know? What does the latest research tell us about AI’s impact on cognitive development?

So before we get to the impact on cognitive development, we need to really start with what are the risks around a fully AI-enabled education system? Because I think that also leads into, you know, what are we not ready for? So taking the potential of AI as given, there are three problems, I think.

We’re up against evolutionary biology. All species like to conserve energy. Humans are no different.

Learning takes effort. Naturally, people will tend to take the easy route, as we all know. When you’re an adult, that might be fine.

You’re outsourcing your cognitive thinking, as you said. But as a child, if you don’t get the chance to develop those cognitive skills, to develop those neural pathways, we’re in trouble. And I think whatever we do in the AI space has to really be mindful of that, that children will take that shortcut.

So AI in education needs to prompt more effort. And if you think about what it’s doing in the workplace, in a way, it’s designed to reduce effort, right? So it’s a very different way of thinking about it.

The second major issue, I think, is the lack of purpose that we’re seeing in schools. Pupil engagement, as any teacher in the room will know, is a massive issue at the moment, for lots of reasons, the attention economy, other things in their lives. The big one, though, is we painted this picture of an AI-enabled future that doesn’t have many high-skilled jobs.

They’re thinking that their jobs will be very much about manual or low skill, and that’s a really demotivating thing to be thinking. So we need to change the narrative on it. And if I could, I would say that that narrative needs to be more.

The future is incredibly uncertain. And so you need every skill you can get your hands on. And some of those listed on the board, obviously, the human-centric skills are vital, but the cognitive skills need to be developed too.

We need to make sure our AI systems do that. And then the third, and very briefly, when we’re thinking about what does the evidence show on the system as it stands, we have never, ever had enough investment in public education systems, state-funded education systems, in technology.

I mean, teachers would be grateful if they could get the internet to work, the technology to work, and access tech in an affordable way, and they can’t. So I think there’s a real challenge there, and it speaks to your point earlier, which is there will be schools and individuals using AI tools to fantastic effect. But if you think about those state schools that are not going to use that, you’re getting that divide that you referred to.

So leaving people behind in the equity consequences of some schools and some people being AI-enabled is a major thing that we need to think about. So that’s what I said.

Well, those are very valuable comments because you’ve already identified three important issues there. There’s the impact on the young brain, of the brain taking the easy route, AI-enabled, and not developing the right skills. There is the hopes and ambitions of young people who might be put off things that actually could be a great future.

And then there is the resourcing, the possible divide in education. There could be the highly resourced private schools who actually, it widens the gap between the children going there and the state schools. So already three very important issues, and we haven’t even started on adult education, but Omar, you’re a good person to…

I mean, let me amplify some of the points Anna was saying, and I’ll pick up. I mean, so back to the risks. I mean, so one, I mean, Anna said it.

I talk to professors across universities around the world, teachers, et cetera, and what they’ll tell you is that for a kid to even access AI, they need a digital device. Those things are weapons of mass distraction. And, you know, the sort of things that you’ll hear professors complain about is a kid will be literally standing in front of a professor and ask them a question.

And as the professor starts to answer, the kid goes into TikTok. So that’s not even just a question of manners. It’s like the distraction thing is a huge thing.

That’s without getting to AI. The second thing that I think Anna talked about is the gap between not knowing something and knowing something is learning. And learning requires neuroplasticity.

It requires effort. And actually I think that applies to adults as well. In an era where the half-life of skills are declining, adults are also going to have to learn new stuff.

And so if we don’t go to the trouble of learning, that is problematic. So if you use ChatGPT to grab knowledge. You know you outsource the the research component But then you may not learn enough and so again we can go to the different stages of learning You know in the higher order cognitive elements of that you can use AI to do that extremely effectively if you design it, right?

But that isn’t necessarily what people automatically will go and do and then the third thing I’d like to point out is Humans are incredible incredibly social There is a small proportion of people about 5% who are effective at self-directed learning So in other words sitting in front of a screen consuming digital content and learning There’s a high proportion of those in the tech industry In many many other industries that doesn’t work people need human contact and one of the bigger issues with applying AI in the wrong way In education and in adult learning is you disconnect the learner from the educator and you can erode trust there Because if the teachers are teaching in a particular way with a particular pedagogy and then you use these tools that come up with information That’s perfectly valid and good But in a different way then the student can become skeptical about what they’re learning as well So there are different aspects of this that we have to treat thoughtfully as we apply these tools because we will wisely

So you are say you’re identifying that there is a right there’s going to be a right way in a wrong way To do that there will be a way there are ways you always ate it for up for the use of AI in adult or education and including adult education to enhance Critical thinking yes mental faculties in general, but it would be easy to do that wrong So this is going to be a and and given that presumably we’re going to need much more adult education in this world of AI There’s so many people are going to have to reinvent them themselves and their skills in their working lives This is gonna be a huge issue 100% So and then there’s the workplace so can’t let me bring you in because and comment before I’m sure This is what we’re here for.

Yeah. Yeah

One thing I’d like to elaborate on that you said today if you think about whether it’s it’s our children Or it’s adults in their professional lives Today what we spend a lot of time doing is engaging with technology Whether it’s your phone, whether it’s your screen, whether it’s your computer.

We are actually engaging with technology We are the users of Technology and as we think about the power of AI and as to how it starts to transform our business Something’s going to happen here We’re going to move from a world where we use technology In the front of everything we do to a world where now technology starts to work for us And we don’t even know it.

One of the powers of AI in the in the professional world and in business is this shift from us Using and engaging with technology to a start to move to the background We don’t have to do that and we can go and focus on some of these skills that you had on your chart up there Freeing us up from those mundane tasks Freeing us up from sitting behind a terminal, a PC, a mobile device To allowing us to do exactly what we’re doing here at the World Economic Forum There’s a reason everyone descends upon this place every single year because as humans we are meant to network We’re meant to collaborate We’re meant to learn from one another and we come here to get that because in our professional lives today the opposite is happening We are working on technology as opposed to technology working for us So I think it’s a really important Dimension of what AI can bring as a benefit to the world of business in the world of work in the future Right, so that that’s a very strong argument I think that’s that’s very clearly put But let me just challenge you a little bit on what the incentives will be for employers to do that because in education at least We know that the you know The purpose of the school or the university is among its purposes to develop the cognitive skills of the of the young people But actually that’s not really the purpose of most employers, isn’t it, or to widen those human skills.

Those employers might be investing huge amounts now in AI. They clearly are. There are billions of pounds that businesses are spending on AI systems.

Somehow they have to get that money back. And is it really one of their concerns as to whether what you’re describing is happening. People can actually enlarge their human skills.

It’s an opportunity to do that. Or is it really an overwhelming opportunity, though, not only to reduce the number of people, but to reduce how much they have to think. And that it’s not really the employer or business’ responsibility to maintain the cognitive skills of their workforce.

Yeah, that is a great question. And let me address it in two different ways. Let me start with the first.

Today there’s a narrative out there in business around the power of AI and the way to justify the cost or the dollars companies are spending on AI technology. It’s a very simple equation that people use. What is my return on investment for the technology, AI, that I’m spending?

Too often the conversation stops there. How much money am I saving? And when the conversation stops there about pure cost savings through the use of this incredibly powerful technology, a couple things happen.

Number one, our employees become very nervous. When it’s all about ROI, in their minds, it’s about how many jobs will be replaced because of technology. And a level of distrust starts to set in between your employees and management.

It’s all about cost. The narrative, William, needs to shift. from cost savings to growth.

If you think about corporate growth over the last 40, 50, 60 years, the greatest driver of corporate growth is productivity gains. The greatest driver of productivity gains is technology. We need to start to have a conversation about how this drives corporate growth.

Every company, including mine at Workday, we have this backlog of things that we need to get done, but we can’t move with speed and purpose and drive business outcomes because it’s not automated. At the same time, we can’t invest enough in our people’s skills. All the things we’re talking about that are on your chart here.

So we need to change the narrative. It’s not purely cost savings, or maybe it is, but what do we do with those cost savings? We reinvest them in the business and people.

Because AI today is talked about as a technology transformation. AI is about business transformation. And when you do business transformation, there’s three pieces.

There’s technology, there’s people, and there’s process. So I advocate for changing that narrative. How do we focus on growth?

The growth comes through humans.

Very good. And this is what we’re here for, to work out these things. But you can understand my skepticism about that.

And you’re saying we have to change the narrative. You’re not saying this is what all businesses will think. But on all these topics we’ve discussed so far, we can see a lot of dangers emerging where we’re going to have to push people in the right.

I’ll come back to Anna, but let me just, Omar, did you want to respond to that? I’m going to build on what

Carl was just saying. So we actually just published some research this week called Mind the Learning Gap. And Carl’s exactly right.

If you’re a business leader, you’re under pressure to deliver results. Your investors want you to invest in AI, but they want the ROI. And so defaulting to using AI for automation is kind of an easy solve.

But I’ll tell you why it won’t work for the long run. First, the automation is simply making what you already have more efficient. So yes, that can drive near term results.

But it’s but only there’s only so much you can do with your existing process you businesses that are successful grow They develop new business models. So so that alone won’t deal with it. Secondly, we’re running out of demographics in the West There are fewer and fewer people.

I mean that that is very very well documented and so actually Augmenting people with AI is the thing and so what the research shows and we looked at the productivity of what happened when IT Happened in the 90s.

There was a productivity ramp then it flatlined AI could give us a next productivity ramp But it requires us to augment people in their work We looked at hundreds of occupations. We looked at the gross value added of those occupations We looked at where AI can augment or not and what level of efficiency you can drive and the unlock in the u.s Alone in white-collar spaces only is between 4.8 and 6.6 trillion dollars by 2034 That at the low end that’s 15% of the US entire GDP So if we do this in the right way, it is an enormously beneficial thing for business and for people

Okay, so we really will also there’s a great prize here. There’s a right way in the great long-run benefit There could be but there’s a lot of short-term dangers and people could take the wrong turnings So let’s think about what advice we would give to employers to schools to governments about and how we win this argument About it.

So Anna, let me come back to you on that in education in particular. What do what do we now need to do?

So I think you’re quite right long term the future might be rosy, but the transition costs are going to be huge And if we look at previous revolutions of various kinds the transition costs have been significant, you know I worked in HR up in the north of England back in the day when there was deindustrialization going on It was heartbreaking watching all these people who were incredibly productive at one moment Basically lose their jobs and therefore need to reskill So there’s a whole piece about adult retraining, but if I can return to the school system, it is different You only get one shot at being five years old, right?

Yes So what we might do in business about a difficult transition can’t necessarily apply to the school system And the other thing I’d like to pick up on two points one in the early years We’ve already got evidence that young children don’t respond well to screens, even interactive screens, in a sense they haven’t got those those neural pathways there yet, they need that human in the loop and we’ve seen degradation of language skills as a result of that, so we have to be incredibly careful with what we’re doing in that space.

And you also said that humans are inherently social, well we are, we come to Davos, right? That’s not what we’re seeing in the data post-Covid, when young teens, again a critical period in their biology where they develop those social skills and those social interactions, they were deprived of it and we’re still seeing the consequences with mental health, even learning outcomes, definitely work readiness, right?

So we need to get it right as we’re going through the education system and I think that means more caution and more regulation about what can be trialled on a system when we’re not quite sure whether AI is great at manipulating and encouraging and motivating students or actually could really be a crutch that’s used in the wrong way.

But what would, let me just

press you a bit then for what these, what does that regulation say or what would be our guidelines as things stand? Because as you say, there are people now in education and AI is now changing, certainly at the university level, speaking from Oxford, the vast majority of students are now making use of AI and indeed we’re encouraging faculty to make use of AI for their own effectiveness, so what are, are there some rough rules about this that we can specify now or is it too early?

No I don’t think it is too early, I mean in terms of the age appropriateness, trialling, not launching into things that you know you don’t know the consequence of, but also we need to be much more conscious that at the end of all of this, we actually want young people to come out with most of the skills sitting in that on that list, even if you’re not using it in your job, you’re certainly going to need it to interpret, you know.

You will even need all the mathematical and statistical skills to know whether the AI is making hopeless mistakes. Exactly, so to write it off and say well this generation can skip that bit I think is a huge risk, so I think actually you know if we go step by step there’s a potential here, but we need to just be very clear what the end goal is and that is the development of skills.

The other thing is that you know education is inherently a collective endeavour. and we want that to continue. We want it to be social, we want it to be collective because when they come out the other end they’re no use to you as employees if they can’t work nicely with others, right?

So you have to combine the

That still has to exist. There is still a teacher standing in front of the class although there may be some AI enabled personalized learning going on as well. I was raising a lot of extreme, sorry, a lot of extremely valid points.

I mean a hundred percent and I’m with you and I think we need to treat this extremely seriously. On the one hand you’ll say well and I’ve seen some educational systems in different countries will literally shut down the use of AI and say like don’t use AI and then people graduate and the employers say what do you know about AI?

And you’ll find employers today will say we

won’t hire someone if they don’t know AI. So employers need to get better at signaling future demand and as Anna’s saying we need local coalitions of lifelong learning ecosystems in education, in government, and in business to set the right standards and the right approaches.

Actually we do know a lot already about what does work and doesn’t work with AI and I think that… Just to tell us about that. So I was gonna go, it’s obviously easier with grown-ups than with small children.

I think we need to distinguish the approaches. So Pearson for example has higher education digital courseware tools with like 10 million students here in the US and we can see with those 10 million students they have millions and millions of interactions every semester with AI study tutors.

And the kids who engage with AI study tutors because they were designed with the learning science that I spoke about earlier, show higher-order cognitive outcomes. They engage in a better way with learning. So there’s a difference between just simply remembering something or understanding it versus applying it and then you get to the higher level of analyzing, critiquing, and creating.

And the kids who engage with AI study tools we can… Help them get to those higher order. So we know that that can work I mean Anna of course is right.

You the the human brain has a wide range of neurodiversity When and and in the formation stages of a child’s life I think there’s still a lot more to learn in terms of what really works and what doesn’t work We know we’ve seen some evidence that that people with ADHD for example actually do benefit with an interactive screen format And they struggle absolutely to sit there and focus at text on a piece of paper So I think we need to be wise about the steps, but I think the answer isn’t you know Shut it off or turn it off.

I mean that that will not prepare people for a future in employment

But what do we do there? Are we heading for a world where we say for young people right? There’s part of your day where you’re really gonna use a you’re gonna be using these great things And yeah, you’re just describing and you’re gonna be using AI to the fullest possible extent So you are ready for the future of that and it will and it will help a lot of people with their cognitive However, there’s the other part of the day where you actually you’re not using there.

You don’t have access to that You are you that you’re gonna have a book and you’re gonna be in a classroom And you are going to be playing sports and you’re and you know So we’re literally gonna divide the day up.

So you’ve got the full range of your mind. Oh, is this I mean

Anna knows this like the public educational systems around the world the stretch financially for sorts of reasons Eight billion people are about to feel the effects of AI coming into the economy. I mean, it’s gonna be transformational There’s no question the people who can help them manage that transformation are the hundreds of millions of educators around the world So so a Pearson we believe that there’s at least three areas that you want to happen in public education One of the foundational skills, so yes indeed maths literacy and incredibly importantly learning to learn You can teach people how to learn we know those techniques learning science has figured that out We don’t actually teach it Widely in the educational system, but those are foundational skills in a world where the skills keep changing you need that Secondly a durable skills and that’s what I was talking about with signals with employers in the US places like Singapore there’s a big focus on career and technical education where employees are signaling clearly what sort of things they want like AI skills, that’s okay, to build that into the curriculum.

And then you have all the human motivational interest skills, team skills, these things, that’s another dimension. So yes, I think curriculum overall, we need to think carefully about how do we evolve it in a way that sets people up for success in their future lives.

Listen, I am clearly not an educator, right, like you two, and don’t understand the education market and how we go about it to train students of all ages. But I think today, we’re maybe over-rotating about the impact of AI. Because eventually, it will peacefully coexist with humans in our education system.

Today, it’s front and center, right? Think about the internet, way back when. What’s the internet gonna do to learning?

And you guys actually now leverage it as part of your curriculum. I think it’s just gonna become part of how we learn. And I actually think of AI as being the great equalizer.

Because a lot of the things we all have to learn, the general skills we need to be productive in society, our knowledge will now come to us in a very simplified way. AI will serve that up to us, allowing us to go focus on more specialized tasks, all the things you have on your chart here. So I actually think AI will just become part of our daily lives.

Like I said, move from working with technology to technology working with us. And freeing us up from those mundane tasks in the business world that we do every day at our PC, at our terminal, reviewing everything. That goes away.

And we shift the focus from technology, and we focus now on amplifying human potential, not replacing. It’s all gonna come together. But today, it’s just so raw.

It’s so powerful. We’re nervous it’s gonna replace us. But over time, if we do this correctly, like we’ve always got right in our society, I think it just becomes part of our everyday life.

It goes to the background. We leverage it. And we actually advance education.

We advance knowledge. Because the generalized knowledge skills that we always have to focus on, everyone goes to college, you have to take all these core curriculums. You might not have to do that in the future, because that’s going to be served up to us and we don’t have to learn it.

Again, just an outside perspective on the education system and how to think about it.

No, it’s a really inspiring vision that you’ve got the narrative, but do you accept that could all go wrong? You know, there could be this great scope to be more human, to augment our scope, and probably most of the people who come to the World Economic Forum are really going to be well-equipped to do that. But they might be employing a lot of people who then, actually, we don’t need their skills anymore.

We don’t think they… So, there is a great danger here as well, isn’t there? Of not being a great equalizer, but actually of creating even more inequality in the world.

We could go either way.

There’s danger if we don’t invest in human skills, and we go completely to the background, right? Right. You talked about how powerful the mind is.

If you compare it to AI, if we’re not exercising the mind, we’re not getting smarter. How does AI get better each and every day? It looks at data, it trains data, and the more it does it, the more powerful it gets.

It’s the same with us. The more we exercise our mind. So, what are we going to exercise it on in the future is really the question, because we’re not going to have to do what we do today to exercise our mind.

We’re going to think completely different about the future. We’re going to specialize, and I think in the education system, again, educators, so I’m probably not the right person to say this. Today, the more education you get around all of the things on your chart, the more better off you’re going to be in the future serving our society.

Today’s specialization, you go to school to be a software developer, 30, 40, 50% of that can be done by AI today, and we’re in the very early stages. People go to school to be a lawyer or an attorney. I will tell you, today, we can use AI to read contracts.

We can use AI to negotiate contracts. That can all be done. So, certain specialization skills that we go to school for today will kind of go to the side and will become more generalist in the future.

Right, yes, a future of more generalists as perhaps an even wider subject, but I think it’s probably true. But here we’re getting towards, we’re developing a bit of a platform here of things we would recommend. You know, employers have to invest in human skills, people have to learn how to learn, and there are ways already where the use of AI can enhance critical thinking.

We’re going to need some caution and be age-appropriate and retain some important aspects of today’s education while using this. So we are identifying things we need to do, and in a few minutes we’re going to ask if there are any questions around the room. But what else am I missing here?

If we’re preparing a list here of what needs doing, what haven’t I got on my list so far, Anna?

I would agree wholeheartedly with the vision of the positive future, and I think you’re right on many counts, but bringing us back down with a rather big bump, I was smiling at your reference to a textbook.

So many schools in… Well, a book, I said. Yeah, a book, a textbook.

In my previous life as Professor of Education we struggled with textbooks and books because actually schools can’t afford to have textbooks and books because they get degraded and you have to replace them regularly and they’re very expensive, right?

Internet. Yes, internet-enabled education would be wonderful, but you’d be surprised when you go into so many classrooms what they’re using. And they’re not using the internet partly because, partly your point, if you let children loose in a classroom with an internet connection it’s an absolute disaster, which means actually we need to invest in software and devices and however integrated that are locked down, that are appropriate for purpose, this is not the same as the way we’re working in businesses, which is actually to open things up much more.

So just as we want the AI to make people put in more effort in schools, we’re taking it away in the workplace, we need it more locked down in schools and probably more open in the workplace. So I just think we need to be very mindful of that big distinction.

Yes, there’s a clear difference here between education and employment.

I’ll give you some angles of things that we can. The genie is out of the bottle on devices where you can go and get any knowledge that is out there. I mean, as Carl said, these LLMs have basically sucked up all human knowledge and you can go and ask about it.

So, the traditional format in education was the teacher would bring you into a classroom, give you the learning, and then send you back home to practice it in homework. Homework, you can now cheat at 100%. So, the format clearly has to change.

Again, the teaching system knows what to do. There’s this thing called flip classroom. So, actually what happens now is you send the kids home and say use as much AI as you like, learn the content, and then come into the classroom and we’re going to assess you.

And you’re going to work with your colleagues and you’re going to do problems together right here in the classroom. So, that deals with the lockdown browser problem that Anna was talking about. So, that actually is fine, but we need to make that flip happen across the whole educational system.

Some of the positives of AI we know about. In the 1980s, it was shown that if you give a human a personal tutor, they could improve their learning acquisition and knowledge by two standard deviations, the two sigma problem. That means an average person goes right to the top of the class.

It’s a huge transformative thing. The problem is you need wealthy people who can pay for personal tutors. Now, as the cost of AI comes down, down, down, which the hundreds of billions of capex is causing, you can put a personal tutor that can travel with you, knows you, has a history with you, knows where you struggle, knows how you learn, and you can bring that around with everyone.

That will happen. That is an incredibly positive thing. Now, we need to do it right.

In the workplace, we launched something just a few months ago with Microsoft called Communications Coach. It turns out that a lot of people want help on things like how do I better communicate, but it’s embarrassing to go and ask others. Communication Coach will literally, after you wrap up a team’s call, it will say to you, hey, William, I love the way you got your points across, but think about this range of expression.

It will land better. It gives very specific feedback to you. Our GED learners in the US, these are people who are adults who are now later in life going for a high school diploma, have said to us, I love my AI tutor because it’s not embarrassing to ask these questions.

There are a lot of positive scenarios

where you can use these tools in the right way, but it has to be done in the right way. Right, well, okay, so plenty of positive things. I’m gonna take, see if there are any questions and comments from around the room.

I think we have microphones. Yes, we do have it. So there’s one here, yes, Sam.

Just say who you are and plow in.

My name is Mohammed Jalfar from Kuwait. In this race between the organic brain and the inorganic brain, is there percentage in asking the people who are developing the inorganic brain to invest in ensuring that the original brain, at the age of five, when it’s most vulnerable, continues to be built up through sleep, through proper nutrition, through interaction, through exercise, so that the artificial brain doesn’t overcome the original brain that created it?

So the question is, could we ask the Nvidia’s of this world to pay attention, time and money into protecting, yes.

Right, we’d have a global fund or a new tax. I’d be interested, as a businessman, I’d be interested in such a fund. I think it is important.

Right, okay, any views on that, Anna?

I do, I mean, I don’t think we ask nicely. I mean, I think that is actually fundamentally what safety and regulation looks like. What we do in classrooms, particularly in young children, I’m separating them out from adults because it is different, we need to be incredibly careful.

And in any technology, we need to think about risks and downsides. And just a free-for-all in the education system could be a disaster. And the other thing is the scaling issue.

Coming back to your study of two standard deviations. When the study was replicated in the UK, for real, across the nation, it never produces that. and so that’s both the opportunity for AI because the quality if you scaled it would be there but it’s also the risk that a lot of these things when they get put into the actual real system with children who are not likely to end up at Davos it’s much harder people who find cognitive effort hard really struggle and I’ve yet to see the evidence that AI can turn that one around and that’s your point about preserving the health of that brain.

Okay so a fund okay we’ve got we’ve got we’re adding to our ideas. Yes sir over there.

Hi I’m Jonathan Haidt author of The Anxious Generation and I want to strongly support the position taken by Anna. AI is coming into childhood and it’s not as though we don’t have a precedent for this. Social media came into childhood and always on on your phone came into childhood and the result is the greatest destruction of human capital in human history.

Hard to compare to World War I and World War II but in terms of the devastation across the world and the ability to pay attention, think, be happy, be not anxious. So childhood is so important to get the skills right and now we’re about to do it with AI. I totally understand the benefits of a specific chat specific tutor that I see but what I can say up front I think what we can all know without even having to wait for the evidence is that when you hack the attachment system, when you have chatbots, when kids are developing relationships with AI before they’re 16 or 18 the results are likely to be devastating.

The tech companies already hacked our attention and now they’re hacking our attachments the results will be much much worse. So I would urge and I guess my question for the two who are more optimistic here is would you endorse a general attitude of basically keep it the hell out of elementary school, keep it the hell out of childhood except for very specific things that have been proven to work.

What I’m afraid of is these companies are just pushing it in, pushing it in, making claims not based on valid research and we’re going to end up repeating the disaster for Gen Alpha that we did for Gen Z.

Well that’s a very important question. So I’ll make a comment I mean

it’s a very tough set of questions. I mean the pragmatist in me says look the genie’s out of the bottle whether we like it or not

And I mean if you want to challenge a whole model of shareholder capitalism that drives these companies I mean we could talk about that like that that is a conversation But but that that is I like to see the world as it is and so I mean for a long time Silicon Valley has exported technologies that have utility or convenience or you know huge uptake and And and some of the founders will kind of not think too carefully about the consequences And and and and those people are Very successful in our in our world.

I think to Anna’s point I think I and I’ve been speaking about this for a long time You know that long before I’d got to Pearson every technology has a side effect I mean the ships have shipwrecks planes have plane crashes and the internet has cyber warfare And and and but but humanity has progressed with technology so applying it right is the thing and So I’m with Anna, but I don’t want to throw like Digital devices in the hands of kids at schools at all I’ve got two young boys.

Well. They’re not that young anymore, but I saw actually some of the things that you’ve spoken about Jonathan But at the same time not giving a kid an iPad literally cuts them out of the workforce At some point at some point So so so the kids will have to learn how to interact with the technology because so much work happens in that way so Designing the transition that the key moments building on real-world experiments So and I said the point about you know can you really build a two to the third the answer is not yet actually you can’t Yet, but the thing that an AI tutor in principle could do that most human tutors would not be able to do is diagnose in detail the thousands of learning objectives that you meant to have gone through and have picked up on and Figure out where you got stuck in the past and then and solve and build that so so I am positive that we can use These things for good, and we have to do it thoughtfully and wisely

Now we’re going to stop in four minutes, but did anybody else want to, or we can continue on this point, or, I just want to make sure. Yes, let’s get one more question, but we are going to stop in four minutes, so just a quick point. There you are.

The child that doesn’t have the iPad before the age of seven, I guarantee that’s the child you’re going to want to hire. That child is the child developing the connection between the hand and the brain. That’s the child taking the physical risks.

And so I just don’t understand how we can come out with those kind of statements.

Yes, sorry, just to clarify, actually, I wasn’t talking about smaller children.

So what age are we talking about?

So even in your concept, there is some age differentiation here. A hundred percent. You’re talking more about the AI coach is more an adult thing, isn’t it?

The AI coach, I think, can apply, I mean, for sure with adults, I mean, a hundred percent. I think with late teens as well. But the reality is every human is different.

Our brains are different. And so some people mature and achieve some of the social skills earlier than others. And I think we have to be mindful of that in our design of the evolution of the educational system.

But I think what I’m just saying practically is we can cut it out of the school, but then they’re going to go home and play with it. And so we just need to be thoughtful about how these things can be used properly.

Last comment for about one minute. Karl and Anne, did you want to make a comment just to – we’ve got three minutes left, and I don’t want all the last word to go to Omar. Anne?

So, I mean, we’ve covered most of the issues. I think two points, I mean, one is this age sensitivity is coming through, you know, loud and clear. And I think that needs to be taken very seriously.

And to your point, Jonathan, you know, we’ve had monopoly power in the past that has been tamed to some degree. We’re not talking about destroying capital or companies. We’re talking about safety regulation in certain settings.

And I don’t think we should give up on that, frankly.

Yeah. And also I’d say it’s in business, kind of your point, the genius left the bottle. You need to find a way to embrace the technology.

Most people today in the workforce believe they’re competing. against AI. They really do.

They really think that’s why they’re worried about their jobs. The truth is, in the workforce, an employee is not competing against AI. They’re now going to be competing against their peers who are leveraging AI.

So I think we need to educate our workforce. We’re doing that at Workday about how to use AI, the benefit of it, and what we’re going to do with those cost savings to reskill them to become more social, to be more networking, to be more collaborative, to have empathy.

I personally believe, now in my 37th year of work, we have lost the social skills in the enterprise, and we’ve got to get them back. And maybe think about it slightly different. How can AI let us bring the human back into work, as opposed to technology driving work

outcomes? Right. Well, I think we’ve had a final question.

I mean, I think what’s loud and clear from this conversation is there’s no positive future AI scenario without a strong focus on human development. I’ll leave it at that.

Exactly. I think that is exactly right. I know you’re done, and I

highlight for you. We have to remember one thing. Technology only enables change.

At the end of the day, people are required still to drive change to implement technology. It comes back to being human-centric in

everything we do. Very good. Well, I think we’ve identified some very important things that need doing, and that investment in human skills is going to be an absolutely critical part of whatever happens in the future.

There are some views more optimistic than others in our panel, but then I think it was Professor Erik Brynjolfsson at Stanford who said, Professor of AI, the next decade for humanity could be the best ever, or it could be the worst.

And what happens on this topic is going to be one of the determining factors. as to whether it’s the best or the worst. So we’re part of what is going to be a much bigger discussion.

Thank you very much to this brilliant panel for their participation. And thank you all for coming. Thank you.

Thank you, William.