There is growing concern about the potential power of AI to develop itself and to dominate humanity. Elon Musk was the best known of the thousands of signatories of an open letter last month calling for a pause in artificial intelligence research, but at the same time that he was issuing the warnings, Musk announced plans to ramp up his own funding of AI to create TruthGPT, a rival to the stupendously successful ChatGPT.
Musk might well say that he will hold back on development if everyone else will do so. As long as his rivals continue to push forward with AI, he has no alternative.
The problem of enforcing a moratorium on AI research runs much deeper than the particular interests of the firms involved. One of the biggest future artificial intelligence powers, China, would likely never sign – or adhere – to such a ceasefire.
In the meantime, we just need to learn how to live with it. An interesting example was provided last week by the energy supplier Octopus, when it claimed AI was doing the work of 200 people. Customer satisfaction from emails written by AI delivered 80 per cent customer satisfaction compared to 65 per cent from those written by humans.
Staff have not been made redundant but have instead been switched to other tasks, such as answering phones.
A key point from this specific example is that it supports the more general finding by the top economist Phillippe Aghion that the effect of automation on the number of workers in a firm is, on average, positive. It creates more jobs than it destroys, even for unskilled workers.
If a firm is reluctant to embrace new technology, it becomes much more vulnerable to its more innovative competitors. Measures which appear to defend jobs in the short-term eventually lead to a substantial loss of employment.
This result is of course highly relevant to the remaining disputes across the public sector, which are as much about resisting changes to working practices as they are about pay increases.
Each new wave of technology reduces the value of some skills or even makes them completely redundant. At the same time, it creates new opportunities, jobs which were not even dreamt of by previous generations.
There are good grounds for believing that the net effect of technology on jobs is either positive or, at worst, neutral. It does not create high levels of unemployment.
The past 150 years have seen a massive amount of technological advance. But there has never been a significant wave of unemployment due to innovation. There have indeed been periods when unemployment was high, in particular the 1930s and late 1970s/early 1980s. But these recessions were emphatically not caused by rapid changes in technology.
Over this century and a half, the three leading economies, taking the period as a whole, have been the US, the UK and Germany. The average rate of unemployment in all three has been close to 5 per cent, a figure which most would accept is more or less full employment.
More jobs will eventually be lost if we resist rather than embrace technology. There is a global marketplace and countries such as China and India will not stand still.