Last week, the Governor of the Bank of England issued a pretty dire warning about the potential impact of automation. It could, he said, cost up to 15 million jobs over time. And he isn’t just talking about monotonous repetitive jobs like widget-making either; even skilled knowledge-based jobs are at risk as the technology gets ever cleverer. I suspect that he’s being over-optimistic; the speed at which technology is advancing is hard for many to take in, and could lead to many more job losses than that. As one small example, just a few years ago, the idea that vehicles would be able to drive themselves seemed like the stuff of science fiction, but in the near future we will probably be wondering why we ever allowed error-prone humans to control something as dangerous as a vehicle in the first place.
From the point of view of individual companies, the attractions of automation are obvious – uncomplaining machines turning out prodigious numbers of widgets 24 hours a day 365 days of the year has to be better than dealing with people with all their foibles. But I wonder if the outcome for producers of goods and services as a whole, rather than as individuals, is quite so rosy; if we really do lose 15 million jobs, who is going to have the money to buy all those widgets?
Carney was arguing that the government and corporations have a duty to help people manage the change, not least because those job losses are likely to happen before the new jobs (and new types of jobs) are created. That seems reasonable in itself, but my question is about how realistic it is to expect that there will be lots of new jobs created. It’s an essentially unanswerable question. Guided by experience, technological revolutions have always led to new jobs eventually - but there’s no inevitability about something happening in the future just because it’s always happened in the past. Simply assuming that it will owes more to faith than to logic.
There may, as the article suggests, be jobs building the robots and training the AI, but from what I know of technology, I’m not convinced that it won’t advance so quickly that even those jobs will themselves be automated. Why would they not be? (And even if they’re not, is it realistic to assume that everyone can be trained to build robots and train AI?) The logical outcome of every individual enterprise gaining from shedding employees and automating its processes is that all of them lose in the end by destroying their customer base. Isn't that, in essence, what Marx was referring to when he talked about capitalism containing the seeds of its own destruction?
The implication of what Carney was saying is that governments and corporations should take steps to help manage the temporary gap between the disappearance of the old and the emergence of the new. If we assume that there is a ‘new’ to emerge, that is sensible and reasonable – but what if that assumption is wrong? If, in reality, those jobs are going to disappear never to be replaced, how do we as a society adapt to a situation where the demand for people to do work is significantly and permanently reduced?
The answer depends on how we decide to share out what work is available – and to what extent it can be shared anyway; not everyone is likely to be able to be trained as a brain surgeon, for instance. I have suggested previously, not entirely tongue-in-cheek, that people who choose not to work are doing the rest of us a favour; when there aren’t enough jobs to go round, choosing to live on a low income without working could almost be seen as a form of altruism. But paying a significant section of the population not to work is actually one possible, and entirely rational, response to a permanent reduction in the demand for work. Doing more to share the work out is another.It’s ideology, not economics, which drives the idea that everybody should be working, and that those who don’t – because they can’t, or won’t, get a job – are somehow beyond the pale. A future in which machines do most of the work and humans have more leisure time to seek personal fulfilment in other ways might sound like the science fiction I referred to earlier, but if it becomes technically possible, why wouldn’t we choose it? It requires a paradigm shift, certainly – but a paradigm shift may be the inevitable result of 15 million job losses. We really need to be thinking through the possible responses before it happens because perhaps capitalism really will destroy itself.