When people talk about automation, most of us probably imagine a robot arm on a factory assembly line. And, for much of the past few decades, that was a reasonable way to think about automation, because of its focus on replacing human physical labor with machines.
But that image is increasingly obsolete. With the advancement of artificial intelligence technologies, automation is still replacing humans, only it’s now happening the cognitive space as well as the physical one.
Nor is this some remote future vision. When U.S. Treasury Secretary Steven Mnuchin said earlier this year that AI is “not even on our radar screens,” adding that he figured it would be “50 to 100 years” before humans started losing jobs to AI, he couldn’t have been more wrong.
For example, we’re seeing AI technology companies targeting the replacement of what’s estimated to be up to 50 percent of current employees in the finance sector over the next 10 years. We would have considered these types of jobs “safe” from automation only a few years ago.
According to University of Oxford researchers, 47 percent of workers may be at risk of losing their jobs to automation, in particular those in mid-skilled retail jobs, and office workers like cashiers and telemarketers. A recent McKinsey report predicted that a smaller percentage of jobs would be at risk of being completely replaced by machines, but pointed out that the majority of jobs would see some of their tasks replaced by automation.
In other words, we’re all going to feel the impact of AI in some way. And our skills aren’t keeping pace.
The sheer number of both soft skills and technical skills already required by most modern companies is exploding. At the same time, the skills people do pick up remain relevant for a shorter and shorter amount of time. AI only accelerates this trend. We’ve crossed a threshold where the timed obsolescence for skills is shorter than for a single career.
The message: People need to adapt faster than ever. And this could have enormous consequences, including widespread unemployment and devastating disruptions for parts of the global economy.
One easily imaginable scenario: In the United States, there are approximately 3.5 million truck drivers. Suppose a truck company could retrofit a truck for $30,000 to make it into a reliable, safe autonomous vehicle. That would be a one-time cost, and the cost would be less than the annual salary of a truck driver. Once that scenario became possible, the industry would likely overhaul its fleet extremely rapidly.
And what would those 3.5 million former truck drivers do then? What about today’s taxi drivers and Uber and Lyft drivers? In fact, it’s entirely possible that we will still have taxi drivers in the streets protesting Uber when Uber drivers take to the streets to start protesting autonomous vehicles.