Like so many other people, I am fascinated by the recent rapid development in artificial intelligence, or more accurately, the emergence of AI’s recent developments into the public’s awareness. One of the frustrations that I’m sure many others share is the struggle of trying to keep up with the rapid pace of change, especially when you are following it from the armchair.
I found a really interesting article on The Atlantic looking at the policy implications of artificial intelligence, in particular how, with the right policy, we can harness the positive potential of AI.
Doom and gloom?
The article quotes a Goldman Sachs report that I’ve seen mentioned before. It estimates that two-thirds of American jobs are at risk from AI. That figure stands at 300 million worldwide, or 1 in 11. That said, the article says that the problem is not AI, but rather the policy response put in place for AI.
One of the policy challenges is going to be handling the speed of change. MIT expert David Autor estimates that three in five workers today are doing jobs that didn’t exist eighty years ago. This shows the potential speed of change and if the seeming onslaught of AI is anything to go by, we might not have eight decades to adjust this time. The result could be a massive wipeout of middle- and low-income jobs, adding to the income inequality which is already at unprecedented levels.
Universal basic income
The Atlantic writer Annie Lowrey suggests that universal basic income (UBI) is the policy answer society needs. I have come across the concept of UBI before, so it was interesting and quite sensible to see it presented as a potential response to the economic pitfalls of AI. Lowrey describes UBI as a ‘social dividend’ and ‘societal insurance’.
Lowrey then posits that the mass deletion of jobs is not just an economic problem but a cultural one: what will people do with their time? But as she says ‘a shortage of remunerative jobs is not the same thing as a shortage of work; there is always something to do’. In a throwaway aside, she mentions Star Trek’s utopian Federation of Planets where crews of starships like the USS Enterprise still have plenty of meaningful things to do, even without the necessity or motivation of an income.
In some ways echoing John Maynard Keynes suggestions from early in the 20th Century that technological innovation would reduce the demands of work, Lowrey suggests that with AI ‘we could all work less and enjoy life more’. However will the powers of the economic status quo accept such a seismic shift, which inevitably involves not just a redistribution of wealth but a restricting of how wealth is distributed. Moreover, Lowrey wonders whether societies like America would accept a shifting role for work in our day-to-day lives and culture. As she says ‘Americans valorise work’ and joblessness and the jobless are seen as a cause of social ills, not a symptom. Somewhat menacing Lowrey closes with the warning: ‘we tolerate inequality and poverty. The problem is not the robots..the problem is us’.
Outside the window
I wrote about policy failing to keep up with the demographic writing on the wall last week; a failure which can be attributed to politicians and the public. While there were conventional policy responses available, which had just been deployed later than they should/could have been, there were also options available outside the status quo. The concept of the Overton Window refers to that policy status quo. Politicians don’t like moving outside that window, especially if there could be a risk along the conventional media or electoral cycles.
Introducing the universal basic income would definitely be moving outside the Overton Window. When you think about how quickly AI might change our lives though, maybe it is time for policy to move beyond the Overton Window. As I said in my piece on demographics though, it takes good and brave politicians to make the case for seemingly radical policy.
When we’re considering these seismic, structure-shaking changes like AI alongside other issues such as demographic change, we should perhaps look at how one potential problem can be harnessed to solve another problem. While economies are grappling with the challenges of ageing populations, could AI make up for declining workforces? In a world where we have plenty, just not equitably distributed, would a universal basic income not just solve the threat of AI but also offset some of the problems with giving everybody a dignified standard of living throughout their lifespan?
These are not oven-ready solutions. Serious thought is needed as well as political bravery. Now is not the time for political CBA. One circle which will need squaring is how the challenges of AI will somewhat ignore national boundaries, and the challenges of ageing populations will see countries needing to collaborate to match up workers with jobs (the ones which can’t be automated). Fascinating issues which will not easily be confronted if we only use the toolkits we currently have. I look forward to trying to keep (or get!) up to speed, especially on those intersections between AI and some of the other major issues facing the world.
Commonplace 