Davos 2024: Generative AI: Steam Engine of the Fourth Industrial Revolution?

Zanny Minton Beddoes: Welcome to our first, of no doubt, very many conversations on the transformative technology of our time. Good to see so many of you here. I gather this is one of the hottest tickets this bright and early, so we've got the crowds in and I'm not surprised.

AI is pretty much on every panel on every shopfront on the promenade. There's nothing that anybody else wants to talk to you and rightly so. It is the transformative technology of our time. Is it the steam engine of the fourth industrial revolution as the WEF says? Is it the printing press? That's another analogy I've heard. Or even fire? So what is it?

What is its impact going to be? How do we harness the opportunities? How do we guard against the risks? Just a small subject we're going to talk about with a terrific group of individuals from all parts of the world and parts of government and the economy and just nobody really needs any introduction.

But very briefly next to me is Senator Rounds, Senator Mike Rounds from South Dakota. Welcome. Next to him, Minister, Al Olama, Sultan Al Olama from the UAE. Next to him, Judy Sweet, a Davos dwyene if I may say so, chairman and chief executive officer of Accenture. Next to her, Arvind Krishna, chairman and CEO of IBM. And last but absolutely not least, Christiano Amon, president and chief executive officer of Qualcomm.

So you have a huge amount of expertise on this panel. Let's get going. And I'm going to start, because you're sitting next to me Senator, with you, which is if this really is the steam engine of our time, let alone the fire of our time, how do you think about AI? And how do you think governments should think about AI?

Mike Rounds: And this is a very timely subject matter for all of us. I'm on the armed services committee and I'm on the intel committee. I look at it in terms of what it means for national defence and what different countries will be doing using artificial intelligence.

Artificial intelligence, whether you're talking about what used to be air, land or sea, now you're talking not just about air, land and sea; we're also talking about space and cyberspace. Artificial Intelligence will impact how we fight wars in all of those domains. It speeds everything up.

What used to be something that took time in the old days, a couple of days to get ready. Now we're talking milliseconds and the country with an army or an armed services that has employed artificial intelligence will have a leg up on everybody else in all of the domains.

Zanny Minton Beddoes: Well, that's a sobering moment to start on. Thank you. But you're absolutely right. So that's one end. Minister, let's turn to you because UAE is a country that has really grasped this technology and now you have built a model, which is one of the more impressive models around, Falcon, as those of you know. You're clearly making a bid to be a very big global player in this space. Explain to us why and what your what your goals are.

Omar Sultan Al Olama: I think if we look at the title of this session, whether it's the steam engine or the printing press or electricity or fire, AI has elements of each and every single revolutionary technology that humanity has embraced in the past and has used to actually leapfrog and develop.

What is it equivalent to? I think it's its own technology; there are certain elements of it that are going to scale up intellect and the brainpower that countries have, as well as their ability to compete on the global front, especially if there are countries that are medium-sized or smaller, compared to the large giants around the world.

In the UAE, we believe in the power of artificial intelligence as well as proactive regulation. So instead of rejecting it, how do we use it effectively? And how do you embrace it? I will use the statements I've used in the past. I think our motto of artificial intelligence is very important and that is a good way for you to understand how important it is.

There's no dictionary in the world that can tell you the difference between the word, "complete" and "finished." But there is a way for you to understand it. If you use and embrace artificial intelligence, you will be complete, if you do not and you are late, you will be finished. And if you reject it altogether, you will be completely finished.

Zanny Minton Beddoes: Is that from ChatGPT or from you?

Omar Sultan Al Olama: No, from me, it's HI not AI.

Zanny Minton Beddoes: That's pretty good. I want to copy that. That's excellent. Writing it down. Julie, we've had the important national security perspective, we've had the complete or finished or both. Tell us when you look around, when you talk to your clients, when you see, where do you think the adoption is fastest? Where is the most difference being made fastest?

Julie Sweet: Well, it's interesting because a year ago, actually, everyone was talking about this, but it was all about how cool someone could write your speech. Right? That was like the whole thing here. Wow, you know, I did my strategy. It's very different a year later, right?

So, we're working on over 700 projects; they're across industry. So a lot of times, people will ask me, well, which industries are going fastest? And I say, actually, that is a recipe for dangerous complacency because, actually, every industry has leaders who are working very quickly in it. And they fall into two camps.

If you're not ready to use it – so you don't have the right data, you're not in the cloud – then they're experimenting with things that come from their applications, their ecosystem. So you know, things like Microsoft Copilot. But if they actually are pretty advanced, then they're doing things right at the core of their operations in every single industry.

So, when you see a pie chart, you say, oh, I'm an energy company and I'm the smallest relative to, say, life sciences, which is going fast; there's no room for complacency because in energy, there are companies who are moving quickly.

So I think, you know, the test for me, in terms of why is this so important is that in the last 30 years, I can't remember a single technology where I could stand in front of a CEO, put up, yes, we still use slides, something that showed every part of the enterprise and a material impact with credibility, where someone would say you're crazy.

No one says we're crazy. That is very different and so there isn't an area, there's not an industry that's not going to be impacted. And we'll get into what does it mean to win?

Zanny Minton Beddoes: I was gonna ask you this but actually, I'm gonna move on and ask Arvind this question, which is, I'm sure you agree with everything that Julie said that every industry is being impacted. Just talk us through – let's look at it via tasks, perhaps or approaches.

As I hear, coding is the area that is really being impacted the most right now. Beyond that, what are the next tasks, business processes? Is it customer relations? Which elements of the kind of overall business are going to be most effective?

Arvind Krishna: So first, it's great to be here and Zanny, thanks for the question. You've got to start with overall productivity. Artificial intelligence, today's form, is going to generate $4 trillion of annual productivity before the end of the decade.

That is incredible economic competitiveness for companies and nations. Hence, the excitement and those who embrace it, I completely agree with the minister and the senator, are going to be advantaged forever. OK, so then you say which tasks?

Absolutely coding is one. We can see 20, 30, 40% productivity for a programmer who embraces AI as opposed to one who doesn't.

Customer service of all kinds, whether you're writing emails back, whether you're answering phone calls, whether you're trying to answer really tough questions, all the way to people call in with tough problems and can you get them through those quicker with a higher satisfaction than without using artificial intelligence? So whole wide area in customer service. And then there is this wide, to build on what Julie just said, there is this wide area called digital labour.

Can you help make your accounts receivable can you help make your HR function? Can you help make your finance function? Can you help make invoicing? Contracting? Supply chain? Ordering? As you go into each of these It is not necessarily job displacement but it is absolutely an impact.

If you embrace AI, you're going to make yourself a lot more productive. If you do not, then probably I agree with the minister, you're going to find that you may not have a job, so you've got to embrace it. And as you go across these areas and it is here and now, this is not two, three years out. You got to get going now.

Zanny Minton Beddoes: You've got to embrace it or you'll be finished but we'll get to the how and what governments do but Christiano, let's just ask of you, from the perspective of the technology itself, what is going to be needed to accelerate the pace?

Because quite often, we have some hype moments about a technology. We think it's going to change the world and then actually it takes longer – the diffusion takes longer than we think. Is that going to happen this time and what needs to be the case for it to happen fast?

Cristiano Amon: No, this is a great question. Look, one way to look at this: the reason we see this explosion of AI and GenAI is really enabled by the ability that you have now access to data and computing power. Really, at the end of the day, AI is a different way to write software and an AI processing is going to be the next way of doing computing.

And I think what we are seeing right now is this is developing at a very fast pace. It started in the data centre and will continue to grow in the data centre but it's going everywhere. I like to describe it, if you think about the history of computing and the best example about the power of distributing computing is you have a computer in the palm of your hand right now, which is a smartphone to fundamentally change society. And that is also going to be true how you're thinking about AI and I think that's some of the things we do.

Our job has been how we can create incredible amount of processing power that we can put on every device so we can put it on your phone. You can put it on your PC, you can put it on your car, and then you can run this everywhere; you can run this on industrial, can run this on on building smart power boxes. You can run this on a metre, you can run this on manufacturing equipment. And I think that is what's going to bring AI to scale. And it's a different way of thinking about computing.

Look, Julie and Arvind gave a bunch of examples on the enterprise and transformation enterprises big. I'm going to give a little bit of some examples, different on the consumer side. We just had a CES show a couple of weeks ago, so we did a demonstration, I think, together with our partner BMW.

You get the entire database of the car you put it into the dashboard. You're in your new car; you get a little light on the panel; you ask the car, what is this light? And the car will say, this is what it is, this is what could possibly happen, this is where it's specifically the problem, here's how you can drive it, I think you should set up an appointment, I already checked the availability, I'm going to set an appointment for you. And you'll be able to have a contextual conversation with your car.

And the same thing, you can see how people can use this tool for training and do tasks that you don't know how to do and the AI is gonna teach you how to go step by step. What I like also is how it is changing how we as humans interact with computers.

Today, you have, and I go back to the smartphone example because it's so much a part of our daily lives. You have applications and you are the human, you go in and you operate, touch one app, you get information from one app to another app.

Now you have this engine that is running every time, is running and is processing all the data in your phone; it's your own data; it stays in your phone, and it's going to try to predict every move. It's going to be an assistant for everything. And you're going to see the conversation that Arvind just mentioned about, the productivity increases when you do your job.

There's going to be productivity increases on everything. How you communicate and text between people, how you schedule meetings, how you get access to information and I think we're just at the very beginning.

I will end by saying it was fascinating – I go back to what Julie said – about a year ago and we're kind of showing the art of possible, like we can run large language models on the phone and we have people talking to us about tens of use cases in ideas.

Right now, one year later, we're starting to see thousands and I think that's the speed of development. It's a revolution and I agree with the senator I think is a great opportunity for innovation and technology leadership.

Zanny Minton Beddoes: So Senator, I was watching you having the same reaction that I was when we were hearing that we're going to start talking to our car, which encounter some cars already a little bit but you hear this incredible description about how the world is going to change.

You're in the US Senate. How do you and your fellow senators prepare for this? I mean, you know you are, you're having briefings. There's a lot of conversations but how do you think about what legal regulatory framework, what do you need to do to get ready for this?

Mike Rounds: Well, to begin with, we looked at it and we actually had a Blue Ribbon Commission ordered part of it was for our national defence. And in it, they laid out all of the different areas that would impact our national defence, but as we went through it, there were classified portions of it that were not available to the public.

Some of us had the opportunity to look at it. And as I went through that, I was really enamoured by the really good things that could happen with regard to health care. And at the same time, the very serious issues that we confront in national defence and then also what this will do to our economy as well. And the changes in the economy.

When we got together as a group of surgeons and by the way, I know everybody thinks that if you're Republican and Democrat, you don't talk to one another. It's not true in the Senate. We actually get along very well. We go out to dinner together, we work together.

Chuck Schumer, who is the Majority Leader in the Senate, is a Democrat; I'm a Republican. He actually asked if we would join with him and two other members to do a bipartisan approach on looking at how we integrate AI. And we started out with nine information seminars for members of the Senate and the folks that really do all the work and that's our staffs.

But we got together, we basically said we want to bring in the best and the brightest and we had folks come in and literally visit with us about what they saw as impacts of AI from their perspective. And we had tonnes of different approaches and since we were doing it behind closed doors, they spoke very candidly, some of them agreeing, some of them disagreeing but giving us lots of information about the concerns they had.

What about patents? What about copyrights? How do we incorporate the use of AI and protect the innovators? But how do we also incorporate what AI will do to help them? How do we, how do we regulate it but at the same time incentivize it so that AI development will not slow down?

With regard to the area of healthcare. We believe that it will be transformational in healthcare. We think the vast majority of the American public will buy into AI because of what they will see in the quality of life that can be improved as AI is integrated into healthcare.

If we can do those types of things, then we can truly make a difference and not just the bad things that AI could be a part of but rather the good things that it brings as well will far outweigh the bad if we do this correctly, we learn correctly.

We provide incentives for AI development and once again, in a competitive area, United States is going to do their best to stay at the top of the competitive brackets. But also, that will not harm our own citizens with unbridled AI that does not have appropriate regulatory oversight.

Zanny Minton Beddoes: In a second, I'm going to get you some free unsolicited advice from the three industry representatives here. But Minister, let's first hear from you because, similar question that I asked the senator but tell us a bit about how the UAE's approach to this because you completely finished, you're certainly not completely finished.

You're perhaps the opposite, thanks to AI. Just tell us, you're harnessing capital, data, you now have the actual LLM itself, what's your strategy for ensuring that your country is benefiting and is at the forefront of this new technology?

Omar Sultan Al Olama: I think when it comes to artificial intelligence, the technology itself is at a disadvantage compared to any other technology that humanity has gone through in the past. There's a lot of negative stigma because of the movies that we watch as children, the science fiction books, the Terminator and iRobot and others about AI taking over the world and destroying humanity.

It doesn't mean that these risks are not risks that could possibly happen. But the possibility of them happening are very low today, compared to the possibility of the positives of the technology that we can deploy as the central nerve as well as the industry leaders have said on the panel.

In the UAE, we look at AI with let's say two views. First, is we need to be a responsible AI nation. So our motto for deploying AI is building a responsible artificial intelligence nation which extends our brain. The second is we believe that there's a lot of ignorance when it comes to the technology, especially in government.

So, when you look at the industry players, we have a lot of expertise at hand and we are actually developing the technology.

When it comes to government, our first reaction is we don't understand it, let's put a tape around it and let's ban it. We want to have a different approach. If we look at certain things like climate change, right there are very similar challenges when it comes to climate change in AI.

It's a matter of time, it crosses borders and there's a lot of different things that we need to do and we don't have the human capacity to do it. So we need to deploy AI to combat climate change. And we also need to deploy technology to combat the risk of AI and the use of AI as well.

So what we believe is in the UAE, first we're going to deploy AI and things like, for example, oil and gas, combating climate in let's say, trying to minimize traffic and hindrance on quality of life.

Things that are non-controversial that improve quality of life positively and then work with our partners, like US and other countries around the world, to ensure that the negative ramifications of the technology can be combated because they will cross borders and it's only a matter of time before we have a catastrophe.

Zanny Minton Beddoes: Why did you decide that as a country you needed to have your own AI capability? Why didn't you just you know, go to Open AI or...?

Omar Sultan Al Olama: So, we are working with the likes of OpenAI and IBM and many other partners that are on this panel and others as well. But the fact of the matter is you are always dependent on these players if you have 100% of your capabilities, your technological capabilities coming from them.

The other thing is we have a lot to add. Think about the UAE, we have 200 nationalities living in a densely populated set of cities, which makes our dataset quite unique, which will allow the AI capabilities that we developed to be truly global, especially in healthcare, especially in certain things where today there is a lack globally.

The second is we have cutting-edge infrastructure, which allows the velocity of data that comes in to be used very quickly and for us to actually deploy much faster than others. And our regulatory regime allows for us to move much quicker than countries that have more bureaucracy, right?

So we can bring a lot to the table. Why don't we help work with our partners but at the same time, we have to provide value to them by having our own capabilities and then trying to leapfrog as a country and work together.

Zanny Minton Beddoes: So it's a leapfrog opportunity?

Omar Sultan Al Olama: Yes.

Zanny Minton Beddoes: It's a leapfrog opportunity. Julie, the senator talked about how it was important to get the regulation right. Let's offer the senator some advice on what areas and where you think having seen this technology and seen your clients think about it. Where would you advise the senator that the Senate should focus and not focus?

Julie Sweet: Well, I probably take a step back and start with what the minister said is, when the leaders don't understand it, they just try to regulate it. Right? So whether you're a government or a company, the single biggest differentiator between whether your goals succeed or not, is your leadership.

Do you actually really understand the technology and that is very different than prior technology revolutions where yes, most C-suites today understand their business as a technology business but you actually have to understand it at a very deep level because if not, there are a million use cases as Cristiano said, they're stacks and a lot of great videos but you have to operationalize it.

And when you operationalize it, you certainly have to do it in a responsible way. Let's remember, there wasn't a responsible PC or responsible cloud; there is responsible AI for a reason, right? And so I think, whether you're a government or a company in education and actually understanding it so you can make the choices and learn actually how to operationalize it.

And my passion is around talent because this will be a great technology if we can bring people along the journey and we have to be able to reskill them. And we will not be able to reskill unless we think very differently about talent both as a government and as a company and we spend a lot of time not because it just drives our business or that but because fundamentally, it's very basic.

Well-paying jobs for people solve a lot of problems and AI will create a lot of new jobs but you won't be able to take the current people and put them in the jobs unless you partner together government and companies on reskilling.

Zanny Minton Beddoes: So, we're gonna come to skill in a second and what kind of skills people will need to thrive as well but one question for you first, you talked about the C-suite and and whether CEOs are prepared and understand this technology and if I may ask of the other two of you too, when you talk to your clients, what overall level of understanding is there?

I mean, does the average CEO and there are many of them here, really understand this technology? Do they know it's big and they need to do something about it? What's the level of understanding do you think?

Julie Sweet: I think there's a level of understanding that they need to learn more. We have 150 people signed up here to do workshops, not attend panels because they understand they have to know more.

But right now, you have a lot of focus on the value, a lot of focus on the talent. The C suite needs to understand better the actual technology and we're seeing them do the work. So I don't think there's a gap in belief. But this is still really, really early and it's going to take a while but it's moving fast.

Zanny Minton Beddoes: Would you agree with that?

Arvind Krishna: It's moving really really fast. You would say that this technology is moving at about 10 times the pace of the previous big one. If I look at it, semiconductors, internet, AI, I will put it kind of in that category. And if you look at the rate and pace, it is incredible.

But if you sort of step back, go back to the point that the minister and the senator made, how do you put guardrails while allowing innovation to happen? That's kind of the dilemma. Because if you just put guardrails, that's bureaucracy, that's red tape. So how do you allow innovation to happen? And the advice that I would give is, it's really hard to kind of regulate the technology itself. Because if you do that innovation stops.

So regulate the use cases, so the more risky the use case, the more regulation there should be. That's an approach that can be taken and has been taken in prior technologies to all the developers of these models accountable from a legal sense. So if they are misusing data or misusing something or letting it get applied badly, hold the developers accountable from a legal sense.

And three, because it is an economic advantage, foster an open ecosystem, not a closed ecosystem. If you can kind of get regulation that allows for those three, then I think you're going to satisfy industry but you're going to give the guardrails that both the Senator and the Minister asked for.

Zanny Minton Beddoes: And you mean an open international ecosystem or open national ecosystem?

Mike Rounds: I think that this is where people get very confused. Can you tell me a digital technology that you can keep to inside a physical boundary? The two things are.

Zanny Minton Beddoes: The is where, the senator might disagree here because some of what the US is doing with its chip export.

Arvind Krishna: A chip is a physical good.

Zanny Minton Beddoes: It's a physical good but it is designed, part of the goal of this is to prevent acceleration of this in this area.

Mike Rounds: Yes but can you allow others or are you stopping others from accessing a cloud service from over the internet? Of course not. So if you're going to allow digital technologies, you're going to allow them then there are certain activities of a nature the senator talked about, where they do not want to access it remotely, they need it physically.

Other than military purposes, there's very few you can think about that require that level of physical locality. So I think it actually satisfies both but digital technologies are really hard to contain to a border.

Zanny Minton Beddoes: Cristiano, I'm going to come to you in just a second but I want to get back to the senator here because this is a really interesting one because the US is, in effect, using a physical means to try and control the development of digital capabilities with the chip.

Mike Rounds: It is so sensitive to us that we remain a leader in terms of the high speed technology available in the most advanced chips that we measure our spread from us versus our near peer adversaries in a technology period of months.

How many months ahead do we believe we are in the development of AI capabilities and by simply restricting the availability of chips and in fact, the most advanced chips we know it's not a long-term success but it is a short-term success while we can slow down other development while we proceed as best we can to maintain our competitive edge.

Other areas: networking is one area that we have a real advantage and it's one that we can never allow ourselves to be in second place on and so that's the reason why we do have the restrictions on it. We know it's not a long-term fix but it gave us a few more months in terms of our technology edge.

Zanny Minton Beddoes: Cristiano, you, in some ways, are impacted by this too. Do you sense, do you agree with the senator that these kinds of fixes are short-term ones buying the US time?

Cristiano Amon: In the present moment we're not impacted because we're not focused on the data centre we're really focused on building that capability to phones to PCs, to industrial devices, to cars and everything that is happening outside the data centre.

Look, our expertise is not on national security. We're really focused on building computing technologies. What I will say is this technology is very broad, it's very pervasive, and it's going to have long-lasting applications. I tend to agree with the senator.

You know, over time, we will see this technology everywhere. But I want to go back to some of the two topics you brought up. One was the regulatory framework and the other one is the speed of this technology within the enterprises.

There's a lot of discussion about responsible AI and guardrails. I want to maybe talk about something else and touch on one of the things that the senator mentioned. The respect of intellectual property is very important for you to have an innovation company, the United States benefited from that, Europe benefited from that.

And there was a company who just really focused on this, I think 2023, where we're the number one company in the United States in patent applications. I think Arvind was a very close second as well. Because it's both companies is based on innovation. And this is a novel area when you think about the role of patents in AI.

So what the senator mentioned is extremely important: how can we protect the value of human invention and intellectual property as we enter into this new world because at the end of the day, we want to maintain the innovation-based economy that is extremely important.

The other point on a regulatory framework is something that Arvind mentioned. And I can't emphasise enough the importance of keeping the platforms open. And that is a very important topic because what AI is doing and as I mentioned before, it's the new computing platform and it's disaggregating different systems. You can run an AI on a phone and you can go to an application on the phone or you can go to the cloud, it doesn't matter. So it's not about one closed ecosystem, it's about open ecosystem.

And this whole conversation that we had about CEOs looking at this technology as a necessity for their companies to stay competitive. We need to make sure that everybody has the right to innovate and not just a few companies have the right to innovate. So, the importance of keeping the platform's open, it's a very important step in the regulatory and I would just finish my point on your question about how fast the CEOs are reacting to this.

I think Julie mentioned on this, Arvind mentioned this. I think the overall industry understood before AI, before this whole conversation of GenAI, the digital transformation is going to be required in every industry. GenAI is just accelerating that, it's accelerating it by a lot.

I think there was a recognition in the industry that I need to be digital. Now, with GenAI it's a necessity and is now the understanding between I'm going to be in business or not going to be in business. That's a positive thing and I think it's going to drive a lot of development growth in innovation in the industry.

Zanny Minton Beddoes: I would as a media organization absolutely second you on the importance of IP and copyright in this. I mean, this will be and you mentioned yourself senator too this will be a big issue.

I want to do one more round with you guys. And then I'm going to hopefully get some time for a couple of questions from the floor. But let's Julie, you mentioned skills, what skills people will need. I remember moderating panels like this when we didn't talk about generative AI but we did talk about the digital transformation and what skills would be needed in the digital transformation. Let's kind of update that conversation now.

And okay, now we're going in the world of generative AI, it's all coming much faster. Those of us who have kids, young adults, there's a lot of questions, what kind of skills do you need to thrive in this world? You must think about it minister from a government perspective, when you're thinking about what do you want your young UAE citizens to be learning? What kind of skills do you want them to have to thrive in the next decades in this world?

Omar Sultan Al Olama: Absolutely. I remember attending a panel in Davos a few years ago, where they said, if you don't learn how to code, you're finished. And today, in hindsight, we discovered that ChatGPT and many of these tools can actually do that for you.

I think the only constant is change. And we need to build a populace that is able to adapt to change that's able to embrace technologies and it's able to also always be agile and have a robust set of skill sets. And be curious about the future.

What we did in the UAE so just quickly to touch upon some of the things that we did. The first is to combat ignorance in the government. We launched a programme with the University of Oxford, where we trained senior government officials to understand what artificial intelligence is, what ethics of AI are, how do you use it responsibly, what is good and bad use cases as well.

We have over 400 officials across the government of the UAE that have graduated from this programme but they're leading the charge, so you know, we have more awareness in the government.

We launched another programme, which is on 29 October of every year, we have a day that we celebrate technology and coding and AI and all these tools. So we said there are two ways for us to teach people about artificial intelligence. Either do a pull approach, they can come to this hall and we do a seminar on artificial intelligence. Or we do a push approach.

So on that day, every single person in the UAE gets an SMS on their phone that says, technology and artificial intelligence is the future. Start your journey to understand what the future is and how you can play a role in it, click the link. They click the link, they write a line of code and they understand how easy it is and then they can create a journey towards actually learning with artificial intelligence, how it's impacting their lives.

If they're going to be an architect, a doctor, even just a normal citizen, what AI is going to do for them and over 180,000 people in a country that is not very big have actually gone through that journey and the numbers are just snowballing into bigger numbers.

The final thing we did was we actually deployed AI education within our schooling. So from grade five onwards, people learn how to code. From grade nine onwards, kids in schools actually learn what artificial intelligence is, what are the ethics of it and they also understand whatever their career path is going to be, how AI is going to affect it.

Zanny Minton Beddoes: Wow, senator, what do you think? SMSs to every citizen of the United States?

Mike Rounds: What they're doing is phenomenal and they're in a position to be able to do it very, very rapidly. We can't do that in the United States but we can be open to innovation and we can be inviting.

One of the challenges that we have right now is immigration and we have a lot of people that would like to come to the United States, who are very, very capable of being real winners in terms of the development of AI and we need to be inviting to them.

I tell my colleagues, I said, can you imagine the world today, if we had not allowed Albert Einstein to immigrate into the United States? The same thing goes on right now with regard to AI and there are lots of people out there who want to participate and in the free world.

We need to make sure that those individuals have an opportunity to go where they want to go. And we want to be able to compete by having an inviting workplace for those individuals as well – as well as across our entire country – allowing for the education of individuals regardless of if they're in a rural state or in an urban community to allow that to happen.

South Dakota. I'm in the middle of the country of 880,000 people in the entire state but we've got Dakota State University, which is specializing in cybersecurity and artificial intelligence.

In fact, Arvind and I have spoken about that and we want that for young people that might not otherwise have an opportunity to grow the talents that they have. We have to take advantage of these young people in the talents they've got and allow that to grow. And so what you're doing is one major step forward and I think it's very forward looking.

Omar Sultan Al Olama: So, just to comment on that, we are working very closely with the US, so we share the curriculum that we did with Oxford with my counterpart in the US is leading the AI charge as well.

I don't think that is an apples-to-apples comparison; the US is a continent, compared to the UAE – it's the size of one of the states in the 50 states of the US, so I don't think you can imitate exactly what we're doing but we can learn from each other.

Zanny Minton Beddoes: And what was one thing that you think you can teach or what can others learn from you, even if their size is much bigger?

Omar Sultan Al Olama: So, we've been doing this since 2017. I've been appointed in 2017, we learned a lot. We also learned as well, there's a lot of hype relating to this. And governments tend to over regulate at times because of the unknown and the ambiguity.

But if we work together, we're able to see where we learn from each other and what we can do. I'll give you a simple example. In 2017, the whole conversation was about self-driving cars. We don't regulate them, the world's going to end because these cars are going to crash. And they're going to kill people and all these things. And it was a matter of time, so we need to regulate it now.

Did we regulate self-driving cars? Not yet. Did it change the world? Not yet; it will change the world in the future. But the goalpost keeps moving and governments lose interest very quickly.

We need to ensure that we actually follow through in what we want. Second, we don't jump into the hype, but we actually look at what responsible regulation means, what responsible deployment means and how we can do this together collectively because I don't think a single country can get it right.

Zanny Minton Beddoes: That's really interesting. Julie, you mentioned skills, you brought it up. One of the things that I think we probably did less well in the first era of globalization was to think about what skills people needed to thrive, particularly in the US and other advanced economies.

That's one of the reasons that there was this backlash against trade, we didn't actually equip people to have the skills they needed. What are the skills now that you think people need? Is it coding? Is it, as the Minister says, an ability to keep seeking the new? What makes people flexible to do that? What kinds of skills do you think when you see your clients? What are the skills that the employees are thriving with?

Julie Sweet: Well, let me be very practical. We have 740,000 people, we hire 100,000 a year. We ask one question to every person regardless of whether you're a coder or you're a strategist or you're a doctor or you're working in HR. We ask one interview question to everyone. We say, what have you learned in the last six months?

We don't care if it's how to bake a cake but we have to have people who like to learn. And that question is incredibly insightful because what you're teaching in 2017 about AI is really different.

In 2019, we had 500,000 people. We started a programme called TQ. We said whether you worked in HR, a strategist in the mailroom, you had to have certain basic technology skills, you had to know what cloud was, you had to know what AI was, you need to know what an agile organization are.

So this is basic digital literacy that every company has to have, every person and that meant that traditionally when GenAI came on, we already had all of our people because we're using it and how we run our people understand the concepts when we say, we now want you to work differently using this technology. They're not starting from "Can you explain AI?" So, you have to have people who want to learn and build a learning culture.

We invest a billion dollars a year in training. And you have to be willing as a part of that to upskill because people in our workforce have not received this training in their education.

And then finally, we have to partner with governments to change basic education. It's not going to help now but we need to think 10, 20, 30 years ahead and education in every country – we're a global company – has to change. The UAE – that's something that we can learn from the UAE.

Zanny Minton Beddoes: Arvind, what have you learned in the last six months? It's a great question, I'm gonna ask everyone this.

Arvind Krishna: So the reason that we get this question on AI more than other technologies, is it in fact rooms like this? It is the first technology that goes after white collar work or what I would call the lower half of cognitive work.

All right, so acknowledge that it's not a particular domain skill. I completely agree with Julie. It doesn't matter whether you're a physicist, mathematician, a computer scientist, a doctor, a writer, it means that the lower half of cognitive work gets taken over by GenAI, it implies that you've got to learn critical thinking. That means critical thinking, regardless of which domain you're in, becomes the skill that is far, far more needed.

Of course, it means that you need to keep learning and continuous learning, I think is a hallmark. A statistic there – half-life of skills used to be 30 years is now seven years. So if you think about that, that means that an average career, you got to change five times your overall skills.

A practical suggestion. We have 250,000 people. We took a week last August, we gave a task or a challenge, people could go modify those challenges. Everybody had to form teams – 1, 2, 3, 5 people, your choice – go, use our generative AI platform to go see how far you can go, have a competition, as a result of which 160,000 people came out well trained at the end of a week but in a fun exercise.

I think every company, every organization can do something like that and it gets your people hands-on experience on a task that they care about. I think that those are simple observations, which you could do but which get the workforce trained and then comfortable as opposed to fearing the technologies.

Zanny Minton Beddoes: Cristiano, the last word is yours. What skills do you think people should focus on?

Cristiano Amon: I think the issue is, as Arvind mentioned, everything that is a repetitive task, it's a task that it's very easy to predict as you process data, AI will be able to do that and help you.

So, I think people need to really focus on how they can remain creative, how they can remain focused on what the human can do, which is going to be driving innovation and how people can be ready to embrace those technologies so they can be more efficient because, at the end of the day, it's not about it's going to lessen your thinking about generational code, it's not about you're just going to have less people.

The key thing is you can do a lot more with the people you have. And I think that's what this bright future is enabled by this technology. Computing and the evolution of computing we have seen is not bad. Like, if you just look at how much technology so far have a big role in democratizing technology and make it accessible. One of the things we're doing about trying to bring AI to every single device is really to democratize the technology and I think that's a great opportunity.

Zanny Minton Beddoes: Unfortunately, I'm, going to get into terrible trouble if I open to questions because we only got one minute left so I'm not, so forgive me. But I am going to end by asking you all to say whether you agree with the premise of this whole session, which is that AI is the steam engine of the Fourth Industrial Revolution.

I mean, is it that scale of technology? Where are you – it's fire at the most transformative, to the tractor perhaps at the least –and where are you on that scale? Julie.

Julie Sweet: Yes, it is the steam engine.

Zanny Minton Beddoes: It is the steam engine.

Julie Sweet: Absolutely.

Omar Sultan Al Olama: In itself, yes but even more so.

Zanny Minton Beddoes: So, this time is not hype.

Omar Sultan Al Olama: It's a field of computer science. So the hype is on certain use cases.

Zanny Minton Beddoes: But the actual computer science, no. Senator? Is this the most transformative thing you've ever seen?

Mike Rounds: When you see cancers being cured? When you see Parkinson's and Alzheimer's being cured or significantly reduced. Then people will start to understand just how life changing this is. Yes, this is like the steam engine.

Zanny Minton Beddoes: Cristiano.

Cristiano Amon: Yes, I agree, it's like electricity.

Zanny Minton Beddoes: It's like electricity. Arvind?

Mike Rounds: Absolutely. Steam engine is a great example.

Zanny Minton Beddoes: That is the place to end on it – this is not hype, even by Davos standards. This is not hype, this is real and we've learned a lot about it. Thank you all very much.