Decoupling Income From Work
On of the fundamental question raised by AI has been trying to grasp the impact it will have on our society organization, among which regarding the work balance and the general work load. As computed by the World Economic Forum earlier this year, the “4th industrial revolution” will be probably the first one destroying more jobs than it creates in the next decades. Useless to say that we have to think this through very carefully. Fortunately, eminent figures in the AI world such as Chris Eliasmith, director of the Centre for Theoretical Neuroscience, Moshe Vardi, Professor of Computer Science at Rice University, or Andrew Ng, the great founder of Google’s “Google Brain” deep learning project, are starting to ask the right questions. In a nutshell, as data scientist Jeremy Howard asked during a panel discussion “Do you want half of people to starve because they literally can’t add economic value, or not?” Assuming you are thinking not, there is an answer that is gaining traction. Its name is universal basic income.
Let’s first look at the economics. Jobs can be divided following two axes — routine and nonroutine, cognitive and manual — creating 4 different clusters.
As we can see, until 1990, all clusters were steadily growing. From that very year, both routine clusters started decreasing as routine became easily replicable by some kind of machines. As most of these two job clusters were constituting the middle class, partly created by Henry Ford, the fact that only 2 clusters remain illustrate our split-in-two society where discrepancies are getting bigger and bigger each year.
Another study, focusing this time the US market and pushlished by the White House, has put the probability at 83% that a worker making less than $20 an hour in 2010 will eventually lose their job to a machine. Even workers making as much as $40 an hour face odds of 31%. It should be clear by now that nothing humans do as a job is uniquely safe anymore. From hamburgers to healthcare, machines will be doing it better, faster, stronger…
Let’s now focus on the automation of non routine jobs that AI is about to trigger. Google self driving cars are a famous example of potentially eliminating millions of jobs within a short period of time. Here is a less known example with a much more damaging impact: Amelia is a super AI call center employee currently being tested. Of course, she speaks 20 languages and she is improving over time. In one company putting her through the test, she successfully handled one of every ten calls in the first week, and by the end of the second month, she could resolve six of ten calls. Potential long term damage? 250 million people out of a job, worldwide.
Here are the questions that Google cars and Amelia world triggers: if a machine can do a job instead of a human, why would anyone want the human to do it ? Alright, but isn’t having a job the only way to obtain an income ?
There are tons of arguments in favour of a universal basic income such as immunizing against the negative effects of automation, the power of choice, recognizing that unpaid work has great societal value, the multiplier effect, decreasing the risks inherent in entrepreneurship and the sizes of bureaucracies necessary to boost incomes. Also many studies have been written and real experiments have been made such as the pilot project in Namibia and it is even now in the beginning stages of possible implementation in countries like Switzerland, Finland, the Netherlands, and Canada.
Of course, I am just listing them out when every one of these arguments, studies and experiments should deserve a dedicated blog post. But for now, let’s simply ask ourselves, what’s the purpose of the technologies we are creating ?
What strikes me the most is probably that the idea of basic income cuts across all party lines. From the tea party to the extreme left, we are hearing calls for basic income. Those on the right love its potential to shrink the size of government, while those on the left love its potential to once and for all put an end to poverty. It reached an extent to which Scott Santens, a great advocate for basic income writes great punch lines such as :
« Basic income is not “left” or “right”. It’s forward. »
Universal basic income is just one very straight forward way of implementing wealth redistribution. It might also have negative effects on the economic such as creating massive inflation but I believe it would be worse a serious trial at a local level before AI creates great damage that we can not roll back.
