In September, Oxford University's Programme on the Impacts of Future Technology released a report "The Future Of Employment: How Susceptible Are Jobs To Computerisation?" A PDF for the paper can be downloaded free from Oxford.
It's an important analysis of the current prospects for what has been called "technical unemployment" (the loss of jobs from automation). This paper focuses on automation based on computerization. That is, of course, the most immediate threat to people's jobs, but in the future there may be other types such as nanotech which will also contribute to replacing human workers.
Many view automation as something which impacts factory workers and others who are viewed as having more physical jobs rather than more mental jobs. Below are some quotes from the Oxford report which shows this is not as true as one might think:
Fraud detection is a task that requires both impartial decision making and the ability to detect trends in big data. As such, this task is now almost completely automated
In health care, diagnostics tasks are already being computerised. Oncologists at Memorial Sloan-Kettering Cancer Center are, for example, using IBM’s Watson computer to provide chronic care and cancer treatment diagnostics.
Sophisticated algorithms are gradually taking on a number of tasks performed by paralegals, contract and patent lawyers
Although the extent of these developments remains to be seen, estimates by MGI (2013) suggests that sophisticated algorithms could substitute for approximately 140 million full-time knowledge workers worldwide.
Also, sophisticated hands-on jobs are also being automated. For instance:
...General Electric has recently developed robots to climb and maintain wind turbines, and more flexible surgical robots with a greater range of motion will soon perform more types of operations
With improved sensors, robots are capable of producing goods with higher quality and reliability than human labour.
Meanwhile, automation is becoming financially more advantageous:
Technological advances are contributing to declining costs in robotics. Over the past decades, robot prices have fallen about 10 percent annually and are expected to decline at an even faster pace in the near future
There are some occupations with requirements which make them currently low-risk of automation. But even these can only be considered safe within a limited time period.
occupations that involve complex perception and manipulation tasks, creative intelligence tasks, and social intelligence tasks are unlikely to be substituted by computer capital over the next decade or two.
Their conclusions are stunning:
According to our estimate, 47 percent of total US employment is in the high risk category, meaning that associated occupations are potentially automatable over some unspecified number of years, perhaps a decade or two.
Of course, this doesn't mean all 47% will necessarily lose their jobs in any amount of time. Nor does it necessarily mean the time frame can't be longer than they estimate. But we do know that unemployment rates can be serious problems even below 10%.
There is a long-term debate in economics whether the loss of jobs from automation will lead to a crisis, or whether there will always be enough new kinds of jobs being created which will make up for the loss of the other jobs. It's hard to imagine new jobs can compensate for automation of jobs on this scale. That seems to be backed by this factor mentioned by Oxford:
we find that a substantial share of employment in service occupations, where most US job growth has occurred over the past decades (Autor and Dorn, 2013), are highly susceptible to computerisation.
Often, those arguing that automation will take away more jobs than the number of new kinds of jobs created point to an accelerating rate of technological advances as the reason new types of jobs will eventually be unable to keep up. For those who are not convinced by this, let me suggest something else. The ability of humans or machines to perform tasks which are parts of jobs involves certain capabilities for actions, sensing, analysis, etc. Over time - regardless of the pace of technlogy - we accumulate more kinds of technology. Some of these technologies make it possible for machines to do some of these actions, sensing, analysis, etc. So, we accumulate machine task capabilities. When these technological capabilities are first developed, it is expensive to produce a machine to operate with these capabilities in a work environment. Over time, refinements in the technology and market dynamics cause the prices to fall, making the machines more financially available to companies. As time goes by, an increasing number of task capabilities become possible for machines at prices which allow companies to profitably use machines rather than people. While new occupations may be created over time, unless these jobs require new kinds of task capabilities, those jobs are no more restricted to humans than the old occupations. In order to prevent the number of jobs done by people from falling, it's not enough to create new occupations, it's necessary to create new occupations with either old task capabilities which machines can't do or new task capabilities which machines can't do - and those task capabilities which machines can't do must be task capabilities which the people who have lost their other jbos can do. It's not enough that there are a few extraordinary people somewhere who can do the task capabilities, they must be doable by vast numbers of non-extraordinary people if these people are to stay employed.
This is not a rant against new technology. Automation has the potential to be a great thing for humanity. It could provide people with more leisure and more prosperity. Even if we were to find that having machines do work instead of people caused more problems than it solved, in a rational, democratic society we would then simply start decommissioing the machines and let people do the work again. The dynamics are different in a society based on private business and big money influence on government policy.
In the abstract, automation could be used to spread the wealth to all members of society, and the jobs that remained for people to do could be split between all working people. If there were 100 people in the workforce and machines were doing everything except for 3000 hours worth of work each week, then everyone would work 30 hours. If the machines later were doing everything except 2000 work-hours per week, then everyone would have a 20 hour work week. But that's not how it works in a private business system. Your boss gets a new machine and doesn't need you anymore - he throws you out on the street. You apply for a job elsewhere. That boss isn't going to want to pay you a living wage for 20 or 30 hours a week. They'll make their employees work 40 hours a week. So each person who is able to find a job is working more hours than the number which would allow everyone to have a job. More and more people won't be able to find a job at any number of hours a week.
Eventually, the number of jobless will get so high that consumer spending won't support an expanding economy. An economic downturn will occur. But a significant part of the unemployment will not be from the traditional business cycle. So, it will become harder to come out of these downturns. Automation under the current system will mean lack of paychecks, hardship, desperation and despair, rather than prosperous leisure.
It's not essential that readers of this diary be 100% certain that there is no way to prevent automation in a private business economy from causing permanent large-scale unemployment and poverty. What one needs to ask, is how much likelihood is needed for us to take action to insure that it does not happen. Readers don't have to believe machines will ever replace 100% of human workers - or even 75% or 50%. A 10% unemployment rate would constitute a serious economic crisis - and that 10% doesn't have to be solely caused by automation. There's always a few percent of the workforce without jobs at any particular time, so automation would only have to take away a few more percent to crash the economy. So, all you need to believe is there is a sufficient likelihood that automation with add a few percent to the number of unemployed for a length of time long enough to drop the economy into a hole it can't climb out of. Furthermore, readers don't even necessarily have to believe that machines will literally take jobs away from people - if one simply sees that there is a sufficient likelihood that the availability of machines which could do the work people now do will force working people to accept lower pay and benefits (maybe longer hours) in order to keep their jobs - that could be reason to work for social change.
It's time we asked which of these two paths we want automation to take us in.