The more powerful a technology is, the faster it disappears from GDP... Hoping that AI will bring sustained productivity and GDP growth is probably a pipe dream.

The marketing department utilizes AI to create more targeted and efficient promotional materials to influence consumer choices. However, if competitors do the same, end consumers will not benefit and GDP will not increase.

Everyone knows that artificial intelligence (AI) is a powerful technology that can have a huge economic impact. The prices of relevant US stocks not only reflect people's confidence in the prospects of technology companies, but also indicate their belief that AI will drive broader economic prosperity. The UK government, obsessed with economic growth, regards the development of AI as a top priority, and at the World Economic Forum in Davos in January this year, everyone wants to hear what the world's AI leaders will say.

We have seen this situation before. In the 1960s, computers were too large and expensive for anyone except for large government agencies and enterprises to use. However, at that time, people's concerns about "automation" were so strong that US President Lyndon B. Johnson launched an investigation into the danger that computer-based technology could "eliminate all jobs except for a few". However, this was not the case. By the 1970s, there were still no signs of a significant increase in productivity, and concerns about large-scale technical unemployment gradually subsided.

In the 1980s, individuals and businesses began to extensively use computers; But by 1990, as economist Robert Solow pointed out, information technology was "everywhere, but nowhere in productivity statistics. With mobile phones, the Internet, expanding hardware capacity and increasing software functions, it is expected to bring a new productivity revolution based on connectivity. Everyone at the 2000 World Economic Forum wanted to hear the opinions of leaders in "information and communication technology". John Chambers, CEO of Cisco, speculates that such technologies will increase the US economy by 5% every year in the foreseeable future, and "by 2010, the Internet will account for half of the gross domestic product (GDP)".

Subsequently, big data, digital economy, machine learning, and now AI emerged. So far, none of these have had any measurable impact on mid-term growth rates. People can say that due to its self-learning ability, generative AI represents not only another stage of technological development. But there are still two reasons why it may not appear in the growth data.

Firstly, a significant portion of economic activity (and possibly still growing) is zero sum competition aimed at gaining an advantage, with no positive impact on growth statistics or human well-being. Using basic Internet search and these mature large-scale language models, lawyers are increasingly able to analyze every possible case before putting forward defense opinions. But if another law firm can also achieve this, the result will be an arms race where neither side has a lasting advantage.

For at least 20 years, experts have been warning that professional services such as law will become the next object of automation as manufacturing jobs continue to decrease, but employment and salaries in the field of commercial law continue to grow.

Similarly, the marketing department can also utilize AI to create more targeted and efficient promotional materials to influence consumer choices. However, if their competitors do the same, end consumers will not benefit and GDP will not increase.

On the contrary, AI is almost certain to bring enormous benefits to humanity for free. The late economist Martin Feldstein wrote in 2017, correctly observing this phenomenon that has emerged in the significant progress of information and communication technology over the past 30 years.

At that time, the processing power and memory of smartphones were thousands of times higher than the largest computers of the 1960s, greatly enhancing functions such as communication, data storage, video and image sharing. However, the share of the telecommunications industry in GDP has hardly changed, which led Feldstein to conclude that "the low growth estimates failed to reflect the remarkable innovations in various aspects, from health care to Internet services to video entertainment, and these innovations have actually made life better over the years."

Similarly, Google DeepMind's AlphaFold protein structure database (which can predict protein structures based on amino acid sequences) will accelerate drug discovery while reducing research costs. Once drugs lose patent protection, their prices will decline towards the marginal production cost, and their contribution to GDP will also decrease accordingly.

If by 2070, a knowledge acquisition accelerated by AI provides us with a miraculous drug that can keep everyone healthy for 100 years, and this drug is produced by fully automated factories driven by cheap nuclear fusion, its role in global GDP will be almost zero. The more powerful a technology is, the faster it disappears from GDP.

At the same time, AI is highly likely to exacerbate the harm caused by previous generations of information and communication technologies to human well-being. Deepfakes have driven the explosive growth of online fraud, AI driven social media algorithms are deepening political polarization, and may lead to a pandemic of mental illness among young people, as pointed out by social psychologist Jonathan Haidt. But these negative factors have also not been reflected in GDP calculations.

Whether good or bad, or simply as a booster for more intense zero sum competition, AI will have a universal and perhaps transformative impact on society. But hoping that AI will bring sustained growth in productivity and GDP is probably a pipe dream.

Author Adair Turner is the Chairman of the Energy Transition Committee and served as the Chairman of the UK Financial Services Authority from 2008 to 2012

The more powerful a technology is, the faster it disappears from GDP... Hoping that AI will bring sustained productivity and GDP growth is probably a pipe dream.

The marketing department utilizes AI to create more targeted and efficient promotional materials to influence consumer choices. However, if competitors do the same, end consumers will not benefit and GDP will not increase.

Everyone knows that artificial intelligence (AI) is a powerful technology that can have a huge economic impact. The prices of relevant US stocks not only reflect people's confidence in the prospects of technology companies, but also indicate their belief that AI will drive broader economic prosperity. The UK government, obsessed with economic growth, regards the development of AI as a top priority, and at the World Economic Forum in Davos in January this year, everyone wants to hear what the world's AI leaders will say.

We have seen this situation before. In the 1960s, computers were too large and expensive for anyone except for large government agencies and enterprises to use. However, at that time, people's concerns about "automation" were so strong that US President Lyndon B. Johnson launched an investigation into the danger that computer-based technology could "eliminate all jobs except for a few". However, this was not the case. By the 1970s, there were still no signs of a significant increase in productivity, and concerns about large-scale technical unemployment gradually subsided.

In the 1980s, individuals and businesses began to extensively use computers; But by 1990, as economist Robert Solow pointed out, information technology was "everywhere, but nowhere in productivity statistics. With mobile phones, the Internet, expanding hardware capacity and increasing software functions, it is expected to bring a new productivity revolution based on connectivity. Everyone at the 2000 World Economic Forum wanted to hear the opinions of leaders in "information and communication technology". John Chambers, CEO of Cisco, speculates that such technologies will increase the US economy by 5% every year in the foreseeable future, and "by 2010, the Internet will account for half of the gross domestic product (GDP)".

Subsequently, big data, digital economy, machine learning, and now AI emerged. So far, none of these have had any measurable impact on mid-term growth rates. People can say that due to its self-learning ability, generative AI represents not only another stage of technological development. But there are still two reasons why it may not appear in the growth data.

Firstly, a significant portion of economic activity (and possibly still growing) is zero sum competition aimed at gaining an advantage, with no positive impact on growth statistics or human well-being. Using basic Internet search and these mature large-scale language models, lawyers are increasingly able to analyze every possible case before putting forward defense opinions. But if another law firm can also achieve this, the result will be an arms race where neither side has a lasting advantage.

For at least 20 years, experts have been warning that professional services such as law will become the next object of automation as manufacturing jobs continue to decrease, but employment and salaries in the field of commercial law continue to grow.

Similarly, the marketing department can also utilize AI to create more targeted and efficient promotional materials to influence consumer choices. However, if their competitors do the same, end consumers will not benefit and GDP will not increase.

On the contrary, AI is almost certain to bring enormous benefits to humanity for free. The late economist Martin Feldstein wrote in 2017, correctly observing this phenomenon that has emerged in the significant progress of information and communication technology over the past 30 years.

At that time, the processing power and memory of smartphones were thousands of times higher than the largest computers of the 1960s, greatly enhancing functions such as communication, data storage, video and image sharing. However, the share of the telecommunications industry in GDP has hardly changed, which led Feldstein to conclude that "the low growth estimates failed to reflect the remarkable innovations in various aspects, from health care to Internet services to video entertainment, and these innovations have actually made life better over the years."

Similarly, Google DeepMind's AlphaFold protein structure database (which can predict protein structures based on amino acid sequences) will accelerate drug discovery while reducing research costs. Once drugs lose patent protection, their prices will decline towards the marginal production cost, and their contribution to GDP will also decrease accordingly.

If by 2070, a knowledge acquisition accelerated by AI provides us with a miraculous drug that can keep everyone healthy for 100 years, and this drug is produced by fully automated factories driven by cheap nuclear fusion, its role in global GDP will be almost zero. The more powerful a technology is, the faster it disappears from GDP.

At the same time, AI is highly likely to exacerbate the harm caused by previous generations of information and communication technologies to human well-being. Deepfakes have driven the explosive growth of online fraud, AI driven social media algorithms are deepening political polarization, and may lead to a pandemic of mental illness among young people, as pointed out by social psychologist Jonathan Haidt. But these negative factors have also not been reflected in GDP calculations.

Whether good or bad, or simply as a booster for more intense zero sum competition, AI will have a universal and perhaps transformative impact on society. But hoping that AI will bring sustained growth in productivity and GDP is probably a pipe dream.

Author Adair Turner is the Chairman of the Energy Transition Committee and served as the Chairman of the UK Financial Services Authority from 2008 to 2012