On April 29, Nobel laureate in economics Michae l Spence attended the "2024 Zhongguancun Forum - Parallel Forum on Fintech" and delivered a speech on the theme of "The Transformative Impact of Artificial Intelligence on the World Economy" through video recording.
Professor Spencer proposed that generative AI is a major stage in the development of artificial intelligence, capable of communicating more naturally with humans, and is a super general technology. Humans can use artificial intelligence to increase productivity and improve the production and service level of the existing workforce. He also stressed the importance of strengthening the development of public policies and regulatory measures related to AI, advocating for the protection of copyright, and advancing the availability and diffusion of AI within countries and in the global economy.
According to public information, Michael Spencer, born in New Jersey on November 7, 1943, received a doctorate from Harvard University in 1972, and is currently a professor at Harvard and Stanford (Stanford) in the United States, an honorary professor of Qingdao University, and the former dean and current honorary dean of the Graduate School of Business at Stanford University.
Michael Spencer's most important research is on how individuals with information advantages in the market can credibly transmit their information "signals" to individuals with information disadvantage in order to avoid some of the problems associated with adverse selection. In 2001, he was awarded the Nobel Prize in Economic Sciences with George Akerlov and Joseph Stiglitz. In 2017, Lin Chonggeng and Michael Spencer won the 17th Sun Yefang Economic Science Award for "China's Medium- and Long-term Development and Transformation: Thoughts and Suggestions from an International Perspective".
Image source: Zhongguancun Internet Finance Research Institute
The following is a translation of the speech, organized by the Titanium Media APP:
Dear colleagues and friends, I am Michael Spencer, and I am very grateful to the organizers of the Zhongguancun Forum for giving me this opportunity to talk to you as part of this year's enrichment program.
My topic today is the transformative economic impact of the AI revolution, and we are in the early stages of this revolution. GenAI has generated tremendous interest and an equally large amount of investment across the globe. I think there may be an element of over-optimism, especially about the timing of the impact on the large-scale economic, social, financial system, etc. But I have a slightly longer perspective, and I don't think anyone knows how quickly this process is going to happen. But before I get there, I want to focus on the economic and well-being potential of GenAI.
I think it's worth noting that over the last 15 years, there have been a number of important breakthroughs in the field of AI, and not all of them fall into what we called GenAI in the early days. Speech recognition, handwriting recognition, lip reading, etc. are a series of remarkable achievements, followed by a series of major breakthroughs in the field of image and object recognition, and have many applications. We are now witnessing these outcomes being realized.
Despite advances in AI, including image recognition, AI still does not currently have the ability to handle complex, rapidly changing visual environments like humans. This is a related topic about humans and robots and human-robot collaboration, which will be covered in a future discussion.
There are other breakthroughs, such as DeepMind's AlphaFold, which is now part of Google, capable of determining the three-dimensional structure of proteins with a fairly high degree of accuracy. The technique utilizes the amino acid sequences that define proteins and has been successfully used to predict the three-dimensional structure of approximately 200 million known proteins and has been published as an open-source database. Therefore, the global biological community can use it as a tool to improve the efficiency of its own research. Then there's winning a game of Go, and people have different opinions about its importance. But what you learn in the process of developing AI is very important. In many cases, AI has outperformed humans. Human performance is often the criterion by which we evaluate AI.
We'll come back to this later, because AI has a hierarchy from superhuman intelligence to subhuman intelligence, roughly comparable to humans, and they all have their own uses. But why could artificial general intelligence be a very important development? From an economic and financial point of view, there are two or three very important features. I think AI general is like a large language model, specifically, for the first time, it has what you might call "domain switching" or "everyday language." It knows the subject without being told. It has the ability to understand the context of a conversation like a human, no matter how complex the conversation is, and I think it's a step towards artificial general intelligence, even if it's going to take a while for us to get there. So you can talk to the AI about the Italian Renaissance, inflation, early Russian literature, computer coding, and ask it to do math problems, and it will happily communicate with you and respond appropriately. In some cases, it gives surprising insights and clever conclusions.
The second point is ease of use. You don't need any technical training to use it. This is because it is basically communicating in our language, just as we understand in language. As a result, ChatGPT had 100 million users in the first two months, something that had never happened before.
What does this mean? The first observation is that it's hard to find an economic field, a knowledge economy, where there are no significant breakthrough AI applications that could change the entire planet. My view, and the general view, is that there is potential for at least one very large, long-term productivity boost that will have an impact on growth, supply chains, resource constraints on growth.
James Mannick and I published a paper on the economic potential of AI in Foreign Affairs at the end of 2023. But the point is, it looks like a super-versatile technology that can be widely used in the knowledge economy and can be applied almost everywhere. And the knowledge economy is everywhere. We usually associate it with industries such as technology, finance, management, computer programming, etc. But in reality, the knowledge economy is everywhere. There are also important parts of the knowledge economy in hospitals, such as the work of doctors and nurses. In all of these cases, you can find powerful digital assistants that can help both individuals and systems with artificial intelligence (GenAI) capabilities at their core.
Why is this so important? Because of advances in artificial intelligence (AI) and machine learning (ML), we can now create highly intelligent digital assistants that can help us with a variety of tasks, from routine work to complex analysis and decision-making. These assistants can greatly increase our productivity and efficiency, allowing us to focus on more important tasks. In addition, they can also help us process large amounts of data, extract useful information from it, and provide us with better decision support. Therefore, having a strong digital assistant is essential to succeed in the knowledge economy.
Why is this stage so important? In China, at least, there is a temporary problem of insufficient aggregate demand, while many economies around the world are plagued by supply-constrained growth patterns. These problems stem from a combination of factors, including declining productivity, aging populations, a shrinking workforce, rising dependency ratios, labor shortages, the high cost of diversification and fragmentation in key employment sectors, and severe shocks in global supply chains from multiple sources, including climate, pandemics, financial crises, and geopolitical tensions.
Finally, we are also confronted with a phenomenon that I am trying to identify, which I call "the fading of powerful deflationary forces." This has to do with that period of high globalization and rapid growth in emerging economies. It's kind of like what I call the "Lewis tipping point" in the global economy, when all of these factors come together, and the result is that we're facing new inflationary pressures that haven't been seen in decades, rising real interest rates, shrinking fiscal space, and debt distress in some places. In addition, large-scale investments in the energy transition are needed to achieve a sustainable growth model. What would it be if someone gave us the best way to deal with all of these problems, such as the excessive burden associated with aging?
The answer is very high and sustained productivity growth, and our most effective tools are in the field of current and future generations of AI. The question is, will we use these tools? There are a number of steps that need to be taken to achieve this.
To name a few, Eric Brinelson and his colleagues at Stanford University conducted research on the use of artificial intelligence in customer service, a very large field of employment worldwide. Basically, AI has learned how to respond appropriately through thousands of hours of training in customer service, customer interactions, audio recordings, as well as performance metrics, creating a powerful digital assistant that acts as an assistant to customer service agents. They gave the tool to some agents to try it out, but not to others.
The test results are immediately apparent. First, there was a significant productivity increase overall, around 14 percent. Second, it reveals some of the characteristics of AI: if you look at the performance changes of less experienced agents, the impact of AI is even greater, at around 35%.
In hindsight, it's easy to guess that AI actually captures the learning experiences associated with customer service, customer interactions, including what works and what doesn't, and then feeds back in a usable form. As a result, the overall effect is an "escalation effect" that has a significant impact on experienced agents and a greater impact on less experienced agents. But the key takeaway from this story, and many others like it, is that the right model isn't just about automating and replacing humans, it's about powerful digital assistants.
When the machine does something better, something will be automated. It's like summarizing the patterns they find in a large dataset (in this case, audio recordings). But that doesn't mean you're erasing humans from the playbook, but rather providing a powerful digital assistant (even for a complete system). There's nothing wrong with comparing digital assistants to human performance in a completely natural way, and it can help assess how far we've come. But it does lead to an automation bias. The reason for this is that once AI reaches the average human level, people tend to think, "Why not replace humans with AI?" and need to think carefully to realize, unless it's considering some specific aspect of automation.
AI exists at all levels. Those superhuman AIs are capable of doing things that humans can't, like recognizing patterns and their meanings in large amounts of data, and doing high-speed calculations at speeds that we can't reach. So, this is a subset, and they are very important. In this area, we will see enhancement, but it will be achieved by augmenting humanity rather than replacing it. It just adds a few things.
There is also a category of things that are in some ways comparable to human abilities. There's another subset of that I'd like to take a moment to talk about, and while it's true that AI isn't at the level of humans that we'd expect in some ways, they can still be very useful. The general direction here is an inclusive growth model.
Some researchers have found that using images of skin cancer to train AI can actually detect skin cancer. Are they as good as the dermatologist you see regularly? If you're prone to this kind of question, the answer is probably no. As a result, my students often don't think it's too interesting. It was a good attempt, but we didn't succeed. But the problem is, that's not the right way to think about it. Probably 85% of the world's population is far away from dermatologists and does not have access to the best treatment. But if AI can detect skin cancer fairly effectively from photos taken with just a regular phone, it can trigger a response from people who are just good enough to take the train to a dermatologist 80 kilometers away, without it's too late.
In areas such as credit, finance, e-commerce, etc., there are many applications related to the inclusion of growth models, and through AI-driven algorithms, the range of services can be greatly expanded. In my opinion, benchmarks should not be allowed to be the determining factor in the application of the economy.
Finally, I would like to briefly summarize that the policy agenda associated with this is extremely complex. Some of these are related to preventing damaging or harmful abuses, such as disinformation, large-scale propaganda campaigns, and fraud, among others.
There are concerns about employment. Are we going to have enough jobs? Personally, it's a long-term debate, but I'm not too worried about the decline in average employment. I don't think we're going to have a problem creating jobs for people, but on a microeconomic level, it's going to be very disruptive. It will change jobs and skills, with some occupations declining while others growing.
These transformations are not easy for people. Governments and policies have a role to play in this regard to mitigate and accelerate these shifts so as to avoid economic losses or excessive anxiety for individuals or families as well as workers.
Copyright protection is a significant issue. The GenAI model absorbs a large amount of data and information from text and creative works in various fields. The original creator of these works needs to consider some form of protection in terms of property rights. This is a very tricky issue and it is still a work in progress. But on the positive side, this issue cannot be ignored. If we want to make a leap in productivity, we need to achieve the diffusion and diffusion of technology within the domestic and global economy, which needs to be driven by government and public policies. I have no doubt that we will have some very advanced industries.
The world of technology and finance is likely to evolve quite rapidly, as will all aspects of management, especially those large companies that have the resources to explore and experiment with new technologies. But the question is, can we democratize technology across the economy and apply it to industries that tend to lag behind in digital adoption?
And what about small and medium-sized businesses that don't have a lot of resources? We really need to address this, not just from an inclusion perspective, because we can't make a leap in productivity without making technology universal. It would be pointless if the impact was limited to a small number of industries that might be ready to embrace the new technology.
We need policies aimed at promoting technology adoption and improving accessibility. Again, this is a positive aspect of the policy agenda. I'm concerned that, at least for now, in the West, the focus of the policy agenda is very much tilted towards preventing negative impacts, rather than accelerating and making sure we get positive ones.
But, in a nutshell, I think the science and technology community, for some reason, is that we provide a set of tools that are largely accessible, widely applicable, and have tremendous economic, social, medical, educational, and other potential value.
It is difficult to predict exactly what the long-term development will be like, but it looks like the potential is huge, so it will be an exciting decade to be part of.
In closing, allow me to point out that the advancement of artificial intelligence is far from over. I think we're going to see more amazing breakthroughs that happen in an expanded range of possibilities. For example, we seem to be at the beginning of the intersection of artificial intelligence and biomedical science. If this is true, it will accelerate the ongoing life sciences revolution. So in general, people can spend hours of their time on these things, but I hope I've amply demonstrated that AI is going to be a positive driver of the global economy, especially the U.S. economy. I think the other big player in AI is obviously China.
The world is in a challenging time, with slowing economic growth, problems making life more difficult, and reducing economic performance. In my opinion, AI is a bright light that pushes us in the opposite direction.
Thank you so much.
The following is the original text of the speech:
Greetings, colleagues and friends, I'm Michael Spence. I'm grateful to the organizers of ZGC forum this year for giving me the chance to speak with you as part of this year's very rich program.
My subject today, is the transformative economic impact of the revolution in artificial intelligence, in which we find ourselves in the early stages. GenAI is generated a huge amount of interests globally, and an equally large amount of investment. I think there may be elements of excessive optimism, especially about the timing of the arrival of large impacts on economies, societies, financial systems, and so on. But I have a slightly longer time horizon. I don't think anybody knows how fast this is going to occur. But I want to focus on the economic and welfare potential of GenAI before we get there. I think it's worth noting that there were a set of important breakthroughs in the past 15 years in artificial intelligence, not all of which fit the category head under GenAI we had in the earlier days. Speech recognition, handwriting recognition, lip reading, which was extraordinary set of achievements, and then a very big and important set of breakthroughs in image and object recognition with a huge host of applications, that we are now seeing coming to fruition.
In spite of the progress of the AI, including image recognition and so on, AI don't yet have a human like ability to process very complex, rapidly evolving visual environments with no latency. That's a relevant subject for later discussion that has to do with the relationship between humans and robots and human robotic collaboration.
There were other breakthroughs, Alpha Fold, product of Deep Mind, now part of Google in determining the three dimensional structure of proteins with reasonable accuracy. Using the amino acid sequence that defines a protein that work was successful and has now been used to predict the three dimensional structure of approximately the 200 million known proteins, and then published as an open source database. So the biological community globally can use this as a productivity enhancer in their own research.
Then there was winning the games of Go, which many people have different opinions about the importance of that. But the learning that went along with developing an AI was extremely important. In many cases, now, AI exceed human performance. Human performance is usually the benchmark by which we assess AI.
We'll come back to that because there's a kind of hierarchy of AI from superhuman to subhuman with roughly at a par with human, and they all have their uses. But why is GenAI potentially such an important development? From an economic and financial point of view, there are two or three characteristics that are really important. I think of GenAI as large language models and just to be concrete for the really the first time has what you might call a domain switching capability or an ordinary language. It knows what the subject is without being told. It has a human like ability to understand the context of a conversation, no matter how complex, which I think is a step in the direction of artificial general intelligence, even if it takes us a while to get there.
So you can talk AI about the Italian renaissance and inflation, early Russian literature, computer coding, and ask it to do math problems, and it doesn't have any trouble going along with you and to respond appropriately. In some cases, with sort of surprisingly insightful and clever results. Second is accessibility. You don't really need technical training to use it. And that's because it essentially talks our language and understands in the same way we understand using language. So ChatGpt had a hundred million users in the first 2 months that just never happened before. What does it all mean?
First observation is that it's very hard to find a part of the economy, the part of the knowledge economy that where there isn't an important set of potentially transformative applications of the whole planet play of breakthroughs in AI and so my belief and one that is widely shared is that there's at least the potential for a very huge, extended productivity surge, which will have impacts on growth, supply chain, supply constraint patterns of growth. James Manic and I made this argument, and wrote a paper on foreign affairs late 2023 on the economic potential of AI. But the main point is, it looks like a super general purpose technology that can be used in the knowledge economy, pretty much everywhere. And the knowledge economy is everywhere. We associate it with sectors like technology and finance, management, computer coding, et cetera. But in fact, the knowledge economy is everywhere. There's important parts of the knowledge economy in hospitals and what doctors and nurses do and so on. And in all of these cases, you can find powerful digital assistance, both to in individuals and the systems that have as at their core, i.e. the GenAI capabilities.
Why is this so important at current stage?
There's at least a transitory problem of insufficient aggregate demand in China, whiel much of the global economy is suffering from a supply constrained patterns of growth. These are the result of a lot of things coming together, declining productivity, declining productivity, aging, declining labor forces, higher and rising dependency ratios, labor shortages, in key large employment sectors, expensive patterns of diversification and fragmentation in global supply chains that are the result of severe shocks with multiple sources including climate pandemic, financial distress, geopolitical tensions.
And finally, we have a pattern that I've tried to identify, which I call the fading of the powerful deflationary forces that were associated with that period of hyper globalization and very rapid emerging economy growth. It's like what I call a kind of Lewis Turning Point in the global economy when all of those things come together. The effect is that we have new inflationary pressures that we haven't had for three or four decades, rising real interest rates, declining fiscal space, debt distress in a number of places.
And alongside of that, the need for very large investments in the energy transition, in pursuit of sustainable growth patterns. What's the best anecdote that we could possibly have if somebody gave it to us to deal with all these things, excessive burdens on the young associated with aging and et cetera.
The answer would be, a period of very high, sustained productivity growth, and the most powerful tools we have to engineer that are very prominently in the area of the current and future generations of artificial intelligence. The question is, are we going to take advantage of this? A number of steps are required to do that? Let me mention a couple of examples.
There was a study done by Eric Brinelson and his colleagues at Stanford of the application of artificial intelligence and customer service, application or industry sector, which is a very large employment sector globally. Basically, AI was trained on literally thousands of hours of customer service, customer interactions, audio recordings, along with performance measures. And it learned how to respond appropriately, which created a sort of powerful digital assistant as an assistant to the customer service agent. They gave it to a subset of the customer service agents and knocked others and tested it.
The two conclusions emerged immediately. One, there was a very large productivity increase overall on the order of 14%. And the second one, tells you something about artificial intelligence is if you looked at the impact on performance of the less experienced customer service agents, the impact was even larger on the order of 35%.
With the benefit of hindsight, it's fairly easy to guess that what the AI is doing basically is capturing the learning that is associated with customer service, customer interactions, what works and what doesn't, and then delivering it back in a usable form.
So the net effect of that is a kind of leveling up effect, getting a noticeable impact for the experienced agents and a much bigger impact for the less experienced and impacted. But the main point of that story and many others like it is that the right model, notwithstanding a very powerful tendency to believe that this is really just all about automation and getting rid of human beings. The right model is the powerful digital assistant.
Some things will be automated when the machines are better at it. Like summarizing patterns that they finds in very large collections of data, in this case，audio recordings. But that doesn't mean you've written a human being out of the script, as opposed to giving the human being or even a full system a powerful digital assistant. Digital assistants are benchmarked against human performance is a perfectly natural way. There's nothing wrong with this to get to assess how much progress we've made. But this does lead to a kind of automation bias. And the reason is that once AI passes the average human performance. The tendency is to think why don't we get rid of the human and just use the AI. It takes a little bit of careful thought to realize that's probably not the right answer unless you're thinking about very partial aspects of automation.
AI’s come at all levels. The superhuman ones can do things that human beings just can't do in sort of recognizing patterns and their implications in vast quantities of data that are involved. High speed calculations of at a level that we just can't accomplish. So that's a subset. They're very important. And in that area, we will see augmentation, but it will be augmentation not by replacing humans. It will just adding something.
Then there's not a set of things that they do on a par with humans. There's another subset that I want to just spend a minute on which AI don't really quite measure up to the kind of human performance that we would hope for, but are still potentially very useful. The general heading here is inclusive growth patterns.
Some researchers have determined that using images of skin cancer to train AI can actually detect skin cancer. Are they as good as the dermatologists that you visit regularly? If you were prone to this kind of problem, the answer is probably no. So my students normally think that's not very interesting. It was nice try, but we didn't make it. But the problem is that it's not the right way to think about it. There probably is 85% of the world's population that doesn't live near a dermatologist and can't get the first best version of the treatment. But if AI is reasonably good at detecting skin and cancer from images that are just taken with an ordinary mobile phone, you could very easily have a significant increment to the preventive health care by giving people signals that are good enough to trigger a response and get them to get on a train and go 80 kilometers and visit a dermatologist before it's kind of too late.
There's a lot of applications in credit and finance in e-commerce and so on. That have to do with inclusiveness of growth patterns where you can extend the range of service enormously with AI-powered algorithms. And I think it's important not to let the benchmarking be the determination of what the economic application is.
Finally, I'm going to conclude by saying a few words. There's an enormously complex policy agenda associated with this. Some of it is associated with preventing destructive or damaging misuse, e.g. disinformation, campaigns, fraud at a massive scale.
There's concern about jobs. Will we have enough jobs? On my side, this is a longer argument, but I’m not too worried about the net average, a problem with job loss. I don't think we'll have a problem generating work for people, but at a more microeconomic level, it will be very disruptive. It will change work and skills. There will be some occupations that decline and others that grow.
And those are not easy transitions for people. And there is a role for government and for policy in easing and making and accelerating. Those transitions in a way that they don't produce both economic damage at the individual or household level, at the level of the worker on excessive anxiety.
There's a big issue of copyright protection. The GenAI models just vacuum up enormous and vast quantities of data and Information of writing and creative work in various areas. And the original creators of that work need to consider some kind of protection in terms of property rights. This is a very hard problem and it is a work in progress. But it can’t be simply ignored on the positive side. If we're going to get the productivity surge, we need accessibility and diffusion within countries and eventually in the global economy. And that takes government and public policy to get there. And I don't have any doubt that we'll have very advanced sectors.
The tech sectors finance will probably move fairly fast. Various parts of management will move quickly, especially among the bigger firms that have the resources to explore and experiment with this. But the question is, are we going to get it across the economy and set into sectors that tend to lag in terms of digital adoption?
In small and medium sized businesses that don't have the massive resources available? We really need to address that problem, not just from an inclusion point of view, but because we won't get the productivity surge. If the impact only comes in a few may sectors that where it's very likely that it'll get adopted anyway.
We need policies that are designed to increase to make diffusion work, to make accessibility easier and so on. And again, that's the positive side of the policy agenda. And I'm worried, at least in the west at the moment, the policy agenda is very heavily weighted toward what you might call preventing the negative side, and not so much weighted toward accelerating and ensuring we get the positive side.
But to summarize, I think we have been given for whatever reason by the scientific and technological community, a set of tools that are accessible for the most part can be made widely available that have a huge economic and social and medical and educational and other potential.
It was in the early stages of intense exploration and experimentation. It's very hard to know exactly how this will play out over time in detail. But it certainly looks like the potential is enormous, so it'll be an exciting decade to be involved in all of this.
Finally, just let me mention that the progress in AI is far from over. I think we will see more breathtaking breakthroughs that come within an expanded set of potential. For example, we seem to be in the early days of the intersection of artificial Intelligence and biomedical science. If that's true, it will turbocharge the revolution in life sciences that's already underway. So bottom line is one can spend hours on these things, but I hope I've made a pretty convincing case that this is going to be one of the positive driving forces in the global economy, certainly in the United States. And I think the other major player in artificial Intelligence is obviously China.
It's in a pretty difficult world with slowing growth and lots of things that are making life more difficult and reducing economic performance. This seems to me the bright light that's pushing us in the opposite direction.
Thanks very much.（本文首发于钛媒体APP，作者｜颜繁瑶，编辑｜刘洋雪）
For more macro research dry goods, please pay attention to the official account of Titanium Media International Think Tank: