Hello and welcome to India Today. I am Siddharth Zarabi and joining me today is a very special guest, the Managing Director of the International Monetary Fund, Kristalina Georgieva. Welcome to the show. You are here at the AI Impact Summit in New Delhi, a massive gathering. And given all that you've been seeing and observing, I want your first key takeaways about the AI Summit and what you think it has really delivered so far.
Let me start with Namaste. Thank you very much. I have to respond in kind. It is great to be in India, especially for an event so dynamic and inspiring. What the summit does is to bring not just the world together, but to bring the developing world in India to meet with those that are ahead on this journey.
What India is doing in AI is phenomenal because the goal of India is to democratize AI. What is the biggest risk we face? That AI is going to open the accordion of opportunities too widely. The haves and the have-nots may have a big distance in between. What India is saying is it need not be that way. We can work to make AI a source of benefit for the whole of society. And we, India, are determined to help the rest of the world, especially the Global South, to benefit from AI.
My two big takeaways. One, AI is already here. And the speed with which it is affecting the way we work, we live, we interact with each other is so high, highest in the history of technological transformation. We risk falling behind if we don't lean forward with everything we have. Two, it is a tremendous source of excitement and opportunity. But it also brings risks. And we have to embrace the opportunities of higher productivity, higher growth, and manage the risks, especially the risks of displacement of workers. And that, for us at the IMF, is a very important objective: to help our members grab the opportunities, manage the risks. And here in India, during the AI summit, I learned a lot about ways we can do it.
On the labor front, and you were an economist all your life, you've studied this issue. How significant is this risk? Because for the developing world, labor displacement will be a bigger risk, one would imagine, than other parts of the world.
So, what we know is the following. Already there is evidence that the impact on the labor market is very significant. I mean, AI is like a tsunami hitting the labor market. We assess 40% of jobs globally, 60% in advanced economies, to be impacted by AI in the next years, either enhanced, made more productive, or replaced. And when we study concretely what is happening, we find a very interesting story. Already in the United States, one in ten jobs require AI-enhanced skills, and it pays more. People in these jobs with more money in their pockets, they spend more, and that creates demand for low-skill workers in restaurants, in hotels, in entertainment. The total impact on employment, surprisingly, in areas we studied, is positive. But it is positive because of the increase of low-paying jobs. So, what is happening is we have one part of the labor force getting better pay because of better skills. Another part of the labor force getting jobs, but with less pay, low-paying jobs. And who is squeezed? The middle. The jobs in the middle that are neither enhanced nor created as low service jobs, they get, in relative terms, squeezed.
We also see automation eliminating entry-level jobs. So, recent graduates, they worry: where is the job for me? And this is where attention of policymakers needs to go. We are not helpless. What we can do is, first, redefine education for the era of AI. Learn to learn, and be adaptable in terms of how you adjust to jobs. Two, put in place good social safety nets. So, if people for a while are displaced, they have something to lean on for themselves and their families. There is a very good system, flexicurity in Denmark. I highly recommend that for this world of AI, we create that flexibility in the labor market. And actually, congratulations to India for the labor market reforms. But we combine that flexibility with security, with helping people to adjust. That's a very good prescription that might be needed in the days ahead.
Just one more follow-up on the subject of AI. And this is to do with the AI arms race. These are my words, but you understand this as an economist. The oil economy had haves and have-nots. We've seen that in other spheres. The United States is advocating that the globe should use the American stack. China has a certain stack. India is going down a different path. What is your own assessment of the path that India seems to be choosing? That is to use what exists elsewhere, also a foundational model that has just been announced in India, and build applications for public good in healthcare, in education. Is that the right way for a country like India?
For India, this is the right way. Why? India invested over the last years in digital public infrastructure. It is a foundation for AI to be more accessible to more people, to be applicable in more areas. India created digital ID for 1.4 billion people. And that gave a big boost to the digital economy. Now comes AI. And to make it beneficial for more people, it is appropriate to think of models that are specifically designed for a sector or for a particular activity. And that democratization of AI, is what I started with, is impressive in India. Now, where the world is headed, inevitably there will be different paths. And the issue is, how about the developing world, the countries that risk falling behind? And there, I think the model that India is developing for India may be more applicable than the high-tech corporate model of the United States. What matters today is that we think ahead.
Independent of which path is chosen, what should we be most worried about?
And frankly, I think our biggest worry should be guardrails. We want AI for good, not for evil. We want AI that protects dignity and the rights of humans. And that requires more attention to the ethical foundation of AI and much more attention to what are the guardrails we have to collectively, as humanity, put in place. And there, engagements like this summit are very helpful because they bring the tech world with policymakers, leaders from different countries, and they create space to think through how to go about a transformative technology that brings enormous opportunities. We see this potentially lifting of global growth by 0.8%. Just to give the audience a sense what it does mean: it means that we will have growth globally above the pre-pandemic level. Right now, it's anemic, 3.3%. We can break out of this path of low growth, high debt with AI. But for productivity growth and economic growth to benefit the world, we have to pay attention to the risks and we have to work together as a global community in a world that is more fragmented today than it was before the AI revolution.
But equally, if guardrails are necessary, and one would agree with you, how do you ensure that innovation doesn't get curbed due to regulation, overregulation?
