Good morning, from Bloomberg's Economy Forum, a gathering of some of the world's most influential leaders and changemakers. It is taking place in India with discussions on AI, on trade, on energy and the forces reshaping the world economy. I will be joined by PwC's Global Chairman, one of India's largest pharmaceutical companies, the cofounder of Moderna as well as Brookfield Asset Management. The Prime Minister has warned that interconnected systems like global supply chains, trade and technology are becoming instruments of coercion. Prime Minister Modi delivered his address earlier and he said chokepoints are emerging along decade-old trade routes compounded by the growing impact of AI. He also said India stands apart with a different approach.
Artificial intelligence is changing the way we work in every sector. In several countries, aging-related challenges are also increasing at a fast pace. In this situation, one of the biggest questions facing the world today is that of trust, of confidence. The confidence that supplies will continue uninterrupted in times of need. The confidence that promises made will also be kept. The confidence that partner countries will stand by us in times of need.
The comments come at the same time as accounting and advisory firms commit to AI strategies, despite the technology's top developers warning of potential safety risks. Joining us now is the Global Chairman of PwC. Thanks so much for coming on. We can't ignore AI, it is coming and it is coming for jobs or strategy. How do you look at it?
First, thank you for having me, always a pleasure to be here. The way we are looking at AI today is as a change for the businesses that we have today. Not only what AI can do for our people, but also for the next generation of products and services that we are seeing. Because today, the companies are also struggling with decisions that they have to make. Overall, I would say that AI is an opportunity story.
It can be the algorithm that helps me do my expenses to something that thinks for me. Is there something in the value chain that you say this is what we need?
That's exactly the case because for some of the mundane activities and tasks that many people need to execute or take care of, you don't need to spend the money. So you have already models you can use depending on the need. The key thing is to understand what model to use for what application. But it is true that it is evolving really, really rapidly.
Almost too rapidly for comfort? How many jobs does it forgo, or does it create jobs instead?
Somehow over the past 18 months we've looked at the impact of artificial intelligence on jobs. And what we're finding is that of course some jobs are going to be impacted, whether it's rules-based jobs or jobs at a recurring level, but generally what we've seen AI do is be more productive, so they have the ability to deliver the important data much faster and with better value. The second thing that we also see is that now human skills are becoming very, very important. Judgment, collaboration. We are finding that third, we are now finding new job categories about who is going to provide the safety for AI.
How many of those jobs will there be? Safety, we don't know. Not a lot of people really regulate it. Is changing the application the same as regulating?
We need different type of skill set when it comes to the implementation of AI and providing safety from AI would be a different set of people that would be required. It is hard to predict how many jobs would be created. The more you have a rapid evolution of AI, the more these new job categories are going to be more important.
How much will it disrupt your industry?
When it comes to the value we deliver for the client, but for us to be able to solve, the value for us is everything.
When you look around the world, where is business doing better than other places and how much of it is actually pricing? How much is it integrating the AI model?
Somehow the problem that we have around artificial intelligence, the decisions to be made around artificial intelligence, geopolitical issues we are currently facing are impacting pretty much everybody around the world. But the impact is very different. We've seen that depending on where companies find themselves, it could be the cost of energy or the cost of supplies. We shouldn't forget about the supplies the people need. But companies are being very resilient and finding different ways to manage this crisis that we are all living with. But when it comes to artificial intelligence, especially when the cost of capital is increasing.
Any supply chain disruption can be fixed with the right amount of money, so it pushes up inflation. It makes the economy a little bit more uncertain.
It impacts the cost of supplies but also the cost of resilience. Many companies are building resilience in the supply chain. To be able to do that across additional categories, what we have seen over the past six to nine months is that companies are adapting to the environment, making the right decisions.
How many chief executives come to see you and say they worry they are spending too much on AI, either building or helping to build infrastructure which we don't have a test case exactly?
It is a variant of the question because most of the executives are not telling me they're spending too much. They know they have to spend because nobody wants to be behind when it comes to the level of AI. The answer that we give them is that AI is not a technology for automation. It is business automation. They have to use AI to change how they run their businesses. It is a different game.
It is really hard, but the return is going to be great if they are able to do that.
You are exposed to different sectors. What sector do you think will have the biggest challenge in trying to transform the business through the lens of AI?
The challenges are more based on location, access to capital but also access to technology. And what we are seeing is that services-based sectors are moving more and more into manufacturing, what we are calling the physical AI, not just about what is within the enterprise but also in the physical world of artificial intelligence. Use cases are coming up.
You've just come back from China and I imagine as a chief executive exposed to the Chinese sector or the U.S. sector, you feel better if you are.
What is the best technology to be used and where and when. For all of us, we have to make capital decisions, we have to decide where to invest. We have to make different choices, and choices are always good for executives because we must try for optionality.
Do Europeans have it tougher because they don't have any of their own models?
Not necessarily. Having the models is not a reflection of the adoption of AI. We don't need to have large language models from Europe to prevent Europe from developing artificial intelligence.
Thank you for joining us for a good conversation on supply chains and also a lot on AI. The Fed Reserve is currently speaking here at Bloomberg's Economy Forum about monetary policy in the age of AI. Let's listen in.
All of the buildup is not coming from an inflation premium. That is not where it's coming from. It is a factor of real growth. If you have real growth and real productivity, you're willing to pay more in terms of your interest costs and people know that and they will ask for it. Real growth has an effect on real yields and on top of that, you have general fiscal concerns around the world, governments kind of sitting here at least in advanced countries, not in such a great state. We are seeing that in France right now. Some of this is just driving up yields in some countries and then the competition from AI is a serious concern. Even if the money flows into the U.S., so demand goes up but there was inflows from abroad, that kind of cancels out the price effect. That money is leaving some of the countries. So even if you have competition and no inflows, U.S. yields are going to go up. If you have higher demand, you're going to see higher yields in those countries. But the AI build I think is definitely having an impact. If Google issued a 40-year bond and it is paying 100 basis points over U.S. Treasuries, it's not like they are going to probably default. That is affecting the way people are looking at things. So anyway, that is it.
There's economists who say the U.S., maybe the rest of the world is entering an era of higher yields and needs to adjust. You agree?
If this is all because of higher growth, that is a good thing. I'd rather have higher growth and higher yields than lower growth and low yields. That seems like an obvious one to me. What we are concerned about is if it is driven by inflation expectations, fiscal concerns which I think is really a problem in France right now. Could be in the U.S., people are starting to wake up to our fiscal situation is not that good. And eventually, bondholders are going to demand a higher yield to get them to buy that debt. If you're living with higher yields because of higher growth, to me that is a good thing. If you are living with higher yields because of greater fear of inflation, that is not a good outcome.
I want to ask you about the fiscal policy situation and how that affects your role, your work at the Fed. But first, when yields go up, they do tend to do the job of a little bit of a rate hike. And the picture that we have now, you think that is enough, or is there more room for the Fed to raise interest rates?
I have long argued that the 10-year rate is determined by global supply and demand. But changing the overnight interest rate on deposits, that is not driving the 10-year yield much. And even if I do it, it is not 10 years out. So I will tell you the 10-year is being driven by global factors, global demand, also supply and U.S. Treasury, and it is not so much what I do. So when yields go up, is that going to impact extra sensitive sectors like housing and auto? Of course it is. Does that tighten off demand for those products and put downward pressure on prices? It should. It is not our intent to do it, but that has got to be one of the natural spillovers from this kind of surge in yields at the 10-year and higher. So it is doing some of the work.
Global, highest levels. In France we are seeing that investors are seeing maybe there is a ceiling for how much debt France can issue. The U.S. Treasuries are quite different, they are seen as a safe haven in times of turmoil and risk. The U.S. has an exorbitant privilege. Do you see that privilege running out?
There is a lot of academic research in the U.S. that that safety liquidity premium on U.S. Treasuries is gone. It is gone. So that was always the following. People would pay more for U.S. Treasury because it is safe, liquid, easy to trade compared to other government debt, so that higher price they're willing to pay. But if you no longer feel like it is a safe asset.
That was Chris Waller. You can continue to keep watching that conversation on your Bloomberg Terminal. We will bring you another important conversation on the biotech and pharmaceutical landscape with Moderna's co-founder and chairman. That conversation is up next and this is Bloomberg.
This is "The Pulse." Here is where the Bloomberg New Economy Forum is happening and a lot of the conversations happened with the Prime Minister and then the talk is what's happening to India in the future. But during this, shares have surged more than 600 percent fueled by a breakthrough in experimental cancer vaccine. Promising late-stage trial results showing the vaccine reduces the risk of melanoma returning or spreading. Shares rallied last week following reports of a public-private partnership to accelerate vaccine development. I'm delighted to be joined by Noubar Afeyan, a co-founder and chairman of Moderna, founder and executive chairman of Flagship. Thank you for joining me on Bloomberg. I have a million and one questions about AI and science. Is this really the one thing that could change the fortune of humans in terms of what we find vaccines for?
That is one question that will take a million and one minutes to answer. I do think the current version of AI that takes into account huge amounts of data and creates models is particularly useful to understand how nature works at the cellular level. You can look at the last 50 years of progress in technology with essentially a black box and that is going to lend itself to some kind of modeling, some kind of understanding at its core and it's hard to see what comes out of that but it's also hard to imagine not a huge amount of disruption comes out of that. We will be able to understand these, we will be able to pick patients that respond to drugs much earlier. We will be able to find early signs of pre-disease states. In future we will have pre-diseases and we will pre-treat them so we don't have to enter the medicine paradigm. That's a few things this fundamental explosion will launch.
We already have this or will it keep getting better and better?
It will keep on getting better. We are very primitive versions today. There's an interesting discrepancy in the way frontier companies talk about AI in science and AI companies talk about it. For us what's missing is the data to train models. For them they think the models they've trained plus a little bit of data will unravel and reveal all the basics of life. That I don't think is possible but I think it will take a period of massive investment in data generation centers, centers where we are measuring by the billions and trillions of cells and how they behave so that we can get a better understanding and then the payoff will begin to come in.
This would be built by and paid by which companies? The tech companies or is it big pharmaceutical?
I think the tech companies have realized the place they can build the value from AI intervention in a positive sense of the word so they will spend money on this. It's going to fundamentally affect sovereign intelligence of a country as to how much knowledge they can generate not just acquire. You can read books that eventually are on the internet, now the knowledge will be held in very few places and even getting that capability to places like India, to generate these models because you cannot really count on these models being available quite the way they are based on internet data that we can train open-weight models on.
Do you worry about the countries that have less access to some of this biotech being further damaged?
There is a hope that instead of having to re-create, they can start racing with this. But they have to have fundamental access. There will be partnerships. I think the examples we are already beginning to see, that is essentially a 600-person team enabling autonomous discovery, so these are AI-based hypothesis generation engines, experimental protocol design. When you put all that together and turn the crank you start making discoveries we don't even understand. If you could test them and verify them for real you can make big leaps not just progress, so we will enter an era of leaps in many directions. Then the question will be what is the right business model to do this safely and responsibly, not only create value.
How far away are we from a full cancer vaccine?
Talking about what Moderna has recently announced which is essentially personalized, individualized therapy. People call it a vaccine because it's similar to how COVID worked but it is different in that you have to have the disease to be able to use it.
So how far away are we. We announced with our partners a phase three trial that's read out and we have shown this in a secondary. So we will present that data at a national meeting in two weeks. Thereafter we expect more data that will be generated from cancer trials with kidney cancer, lung cancer.
We are live on Bloomberg.
How are you, sir. That's funny.
But not a cancer vaccine.
If you think about it, preventing ever getting cancer, we are not close to that. Because to do that you need to understand shared antigens, shared signals you can use ahead of getting disease. What we are doing to give the audience a sense is taking individual patients' biopsies, sequencing their DNA, computationally with AI determining pieces of the DNA. A piece of mRNA just like the COVID vaccine injected into your arm, and the immune system does the rest. That in melanoma has shown significance, in eight other cancers it's being tried. We get that treatment potential that's pursued to include 50%. Then we can think about can we learn so much about these diseases that we can get ahead of them. I think if we can treat, prevention is a reachable ambition. For now that's what we're working on.
We talk about regulation and I don't know if it's self-regulation from tech companies, not a lot of people believe in it, or some kind of body. Does it not slow down the scientific research if you crack down on the AGI?
I don't think so. Recently I had occasion to have discussions in private sessions. But I have a particularly interesting place because I come from a highly regulated field which is medicines. Now looking at AI entering the realm where it's not dissimilar risks. And if you go back, it used to be medicines were not regulated like this. And then people basically unknowingly released a drug that caused major birth defects and then there was overregulation or at least a search for regulation. I'm afraid we are heading in that direction and I want to clarify the thing to regulate is not large language models but agents.
Thank you so much, the co-founder and chairman of Moderna. Plenty more throughout the day.
Good morning. I am Francine Lacqua in New Delhi for the Bloomberg Economy Forum. I know building Biocon from a rented garage almost 50 years ago. It is one of the world's biggest producers of insulin and she joins me now. Great pleasure to speak with you. There's a lot of patents that are expiring right now. What does that mean for Biocon? How much of an opportunity and how do you see growth the next couple of years?
