There is a rising realization in the world that we are currently just treading water with respect to the creation of new jobs, and that AI and robots are eventually going to take our jobs - as in pretty-much *all* of our jobs.
Gone are the days when "the professions" were immune from these changes - advances in AI and machine learning mean that doctors, lawyers, accountants, interior designers, engineers - even supreme court justices - every job, including that of "robot programmer", is under threat from thinking, self-learning machines that will soon know everything, and know exactly what to do (statistically, or by rote) in any given circumstance.
One of my cigar buddies sells robots for a large robotics company. He specializes in developing specialized robots for manufacturing plants. I asked him one night to tell me how they go about replacing human workers with robots. He said they put up a video camera on a pole above the guy whose job it is to stick the rear window on the car moving along the assembly line, then tape his actions for six months. The one time the guy does his task in 39 seconds instead of his average 53 seconds (he needed a bathroom break, it was the last car of the shift, he had a hot date...), they use that routine as the baseline for desiging the robot.
Then they go to work. They build the prototype. Then they tweak the routine. Pretty soon, double-digit seconds have been shaved off the routine and the human worker isn't even in the same league anymore. And the robot, which does not require bathroom breaks, pensions, bonuses, or Christmas parties, is working 60/24/7 and replacing not one, but potentially four or even six or more employees. Then they install four thousand of them and lay off fifteen thousand workers.
I know what you're thinking - there's some things that robots can do well, but there's a lot of things that they can't, and it will be a long, long time before they can match or outperform humans in these tasks. Construction, food preparation, agriculture, mining, manufacturing... while many of these jobs can be automated, my job absolutely cannot be taken by a robot. I'm safe.
Sorry, but that argument is deeply flawed. Thanks to accounting conventions and tax laws dating back centuries, a robot doesn't need to be better - or more efficient - than a human being at a task to make a business more profitable. It just needs to be 34% as good, or 11% as good, depending on that business's accounting and amortization policies.
Yes, you heard that right. As a business owner, while I need to expense (with rare exceptions) my human payroll, when it comes to robots, I can amortize the capital expenses related to buying them across three, or even ten years - or, in some jurisdictions, if I can successfully make the argument, to the end of their effective working life. Even better, I can claim the interest related to the purchase, and possibly even qualify for grants or tax credits for a large percetage of the purchase. Why? Because I'm deploying new technology and creating 'greater productivity'.
What does this mean for a business owner looking to replace humans with robots? In a nutshell, good things. Firstly, it means that *the* critical business measure, EBITDA (Earnings Before Interest, Taxes, Depreciation, and Amortization), is potentially going to go way up, relative to the old business model, which employed humans, and created greater expense. The return on equity? Also way up. ROI? Way up. Revenue per employee? Way, way up. Coffee, training costs, health and liability insurance? Way down.
Pensions? Payroll taxes? HR and recruitment costs? Dramatically lower. The negatives? Almost none - other than the effort required to find the capex.
So... will these ancient, boring, accounting rules be the cause of Armageddon? Or will they be the catalyst for massive positive social change, and potentially the driver of massive future profits which will enable legislation to be passed in support of a universal basic income (UBI) for all?
I really don't know. But I don't see the business case for robots and AI going away. For them to lose this battle, some very fundamental (and extremely unlikely) changes would need to be made to scores of centuries-old accounting laws and policies worldwide, across some 200+ countries. I don't see that happening in our lifetimes - but I do see the robots taking virtually all of our our jobs in that timeframe.
While I applaud Elon Musk for thinking about this issue, I don't see the creation of regulations around AI doing anything at all useful when it comes to stopping companies from using more and more AI and robotics. As I've outlined above, current financial policy and tax laws greatly favor the use of robots over humans. Unless that changes, no amount of regulation will have any effect, other than making people feel like something is being done.
My suggestion to politicians is - start working on implementing a country-wide universal basic income policy *today* - because that is going to be easier to make happen than the alternative - a globally-adopted set of new accounting or tax rules. There is limited time left to do this - right now, there is nervousness and some anger, but it isn't widespread. Let's not wait until the situation worsens to act.
Disclaimer: I am a partner in a venture firm that is out to disrupt venture capital using AI and machine learning to drive investment decisions. Which puts me firmly on the side of the accountants, the robots - and the future. And, in a few years, potentially out of a job.
Graph credit: Graetz and Michaels, “Robots at Work" - taken from the Brookings Institute article located here- which manages to interpret data from the Graetz and Michaels study rather too positively, I think.