In my new book, Silicon Collar I looked at automation and impact on jobs in 50+ work settings. On my innovation blog I have been excerpting weekly from those settings — you can see some of them linked below. In each case, practitioners are applying AI, robotics, wearables and other technologies to automate tasks.
The simplest job I profiled was that of a garbage collector in our county. He used to dangle at the back of the truck, pick up bags at each house and throw them into the compactor. The job did not need any formal education or much training.
To be able to automate that job we had to retrofit the garbage trucks with a robotic arm. Each of our houses had to be delivered 2 sturdy, large new carts — 95 gallon size for the regular garbage, 65 gallon one for recyclables. It is estimated over a half a million carts were delivered to residents. The drivers had to be retrained to operate the arm. Each household had to be trained how to position the carts. We had posters and billboards with instructions in neighborhoods, brochures in the mail and public service announcements on TV, social media and the Web. Quite a significant investment in time and money.
Informal polls showed 2/3 of our residents opposed the move. Some found the carts too cumbersome to store and drag to the street, others felt guilty about the potential job losses. We went ahead anyways. Unlike the city of Toledo, OH which gave up after a year, we have persisted and our streets look neater, our recycle rates have improved and we see other benefits.
However, our use case was relatively straightforward given our mostly sunny weather and suburban neighborhoods. The Volvo Group is working on an all-weather truck with a collection robot at back and a drone to communicate with sensors on the garbage cans and guide the robot to each can. Think of the incremental capital commitment that will need.
Now think about automating more elaborate jobs like some below which require multiple skills during the course of a workday — hand dexterity, frequent walking, visual acumen, judgment calls, social interaction, ability to sense hot and cold and many others.
Just for one occupation — accountants, the O*Net (the country’s primary source of occupational information) lists 13 skills. For garbage collectors, they list 6.
Try to find machines which can duplicate these skills, preferably from a single source vendor not a “Frankensoft” solution which require many custom manufacturers and systems integrators. The reality is most automation can replace a task or two, not complete jobs.
Next think through what regulators would need to certify such automation in healthcare, transportation, aviation and many other industries. Finally factor the ROI analysis corporations would need to conduct, the comfort consumers would need to develop to adopt such machines. In this article, Slow-Motion Automation I describe some of the “circuit breakers” to societal adoption of automation.
Now factor that the 50 settings I looked at for the book barely scratch the surface of the 840 occupations the Bureau of Labor Statistics currently tracks. An update to that is due in 2017. Many new occupations are emerging every year. FastCompany expects new ones in the next decade like neuro-implant technicians and urban farmers.
Still feeling pessimistic and that we are moving soon to jobless societies due to machines doing all our jobs?
I will make you an offer to cheer you up. I am taking bets on specific job loss scenarios in the next decade or two. In fact I have a dinner scheduled on my calendar for September 1, 2026. I will pick up the tab if in that year 20% or more of all car sales/leases in the U.S. have fully autonomous capability. Otherwise my friend who I have the wager with will buy the nice meal.
Ping me if you want to wager something similar. I am that confident humans will be doing all kinds of work even decades from today and that machines will be our peers. not our replacements.
In the meantime, enjoy these excerpts from the book.
cross posted at Medium