We introduced the fear of the rise of machines and artificial intelligence (A.I.) as reviewed by the editors of The Economist in our prior post, where we ended up asking the question: What will it mean? We’ll parse through what economists and others are saying in today’s post, which attempts to answer the larger question of whether smarter machines are causing (or poised to cause) mass unemployment.
Machines today are imposing on even the highest tech jobs, such as those produced by Enlitic, a startup involved in deep learning in the medical field, which has produced a system for scanning lungs for abnormalities. In a test against three expert human radiologists, Enlitic machines were 50% better at classifying malignant tumors. Another of the company’s machines which examines x-rays to detect fractures outperformed human experts, and the firm’s technology is already being deployed in 40 clinics. That’s just one example of how white collar jobs can now be automated.
It turns out that what determines whether a person can be replaced by a machine – thus becoming highly vulnerable – is “not so much whether the work concerned is manual or white-collar, but whether or not it is routine,” notes the editors. Thus, a highly trained and specialized radiologist may in fact be in greater danger of being replaced by a machine than his own executive assistant.
Among the most vulnerable, 47% of U.S. jobs are said to be at “high risk” of potential automation. A 2013 tally published by Carl Frey & Michael Osborne on job susceptibility to computerization found the following had at least a 50% probability of being replaced:
- Telemarketers (99%)
- Accountants and auditors (94%)
- Retail salespeople (92%)
- Technical writers (86%)
- Real estate sales agents (86%)
- Word processors & typists (81%)
- Machinists (65%)
- Commercial pilots (55%)
Among the least vulnerable:
- Recreational therapists, dentists, athletic trainers, clergy, chemical engineers, editors, firefighters, actors, health technologists and (of course) economists.
Clearly, a substantial risk exists across a broad swath of the employment spectrum. Some, like Sebastian Thrun of Stanford, say this is only the tip of the iceberg. Martin Ford, a software entrepreneur and author of “Rise of the Robots” warns of the threat of a “jobless future,” noting that most jobs can be broken down into a set of routine tasks, and are thus increasingly vulnerable to A.I. and machines.
As we noted in our prior post, these sorts of job-obliterating threats have been around since at least the Industrial Revolution, when the Luddites protested against machines and steam engines that they felt would destroy their livelihoods.
Such declarations have reappeared regularly since, in the 1930s-40s, in the 60s, and most recently with the advent of personal computers in the 80s. Invariably, the progress of technology has always ended up creating more jobs than it destroys. Once something can be done more quickly and cheaply, it is. But that in turn “increases the demand for human workers to do the other tasks around it that have not been automated.”
We’re running long, so we’ll conclude our thinking in our third and final post on this topic. Stay tuned…