If you worry that robots will render human workers obsolete, the coronavirus pandemic offers ample reason to worry more: Surely companies will take the opportunity to replace demanding, imperfect people with machines or artificial intelligence wherever possible.
That said, I’m hoping the crisis will also create an opportunity to negotiate better terms with our algorithmic overlords.
The pandemic has made me reconsider the sanctity of work. Just look at the situation at America’s meat processing plants, where hundreds of mostly immigrant workers are getting sick. Workers have been brutally divided into two classes: those whose employers protect them from the virus, and those who lose their jobs or their unemployment insurance if they refuse to labor under dangerous conditions.
If certain businesses — say, the next generation of meat plants — can’t reopen safely and profitably with humans, they can and should do so with robots. Some jobs just aren’t good enough to protect. Until now, among the biggest obstacles was the transition cost of going from badly paid humans to machines. But if companies disrupt their workflow by actually shutting down production to save lives (as they should), then they will have paid much of the cost. They can train the robots using data gleaned by thoroughly surveilling some model workers.
It’s not just meat packing plants. Robots run by algorithms will do a lot of jobs, replacing the human workers who in many cases will have trained them. People will probably welcome the brave new world, particularly if it’s more hygienic. They’ll buy their groceries by simply taking them and leaving the store (now available from Amazon), they’ll let robots vacuum their floors (already quite popular) and deliver their purchases (not quite yet). I confess I’d prefer a self-cleaning, self-driving car so I don’t have to share space with a human driver, for both our sakes.
Accepting such a transformation, however, means subjecting ourselves to the companies that make the algorithms, such as Amazon and Google. They will essentially hold all the data, control the supply chain, and have more direct power over our consumer and even political behavior than ever before. And a lot of people seem to be OK with that: Surveys suggest that they trust the tech companies more than the federal government.
The question, then, is what will happen to the enormous jobless underclass that such an accelerated shift to automation will create. This is where I think the sheer magnitude of the coronavirus crisis might actually help, for three reasons.
First, when so many people are suddenly and violently thrown out of work at the same time, it creates a sense of solidarity that a slow, insidious process such as offshoring does not.
Second, the jobless are not perceived, and do not perceive themselves, as at fault for their predicament. This is a natural disaster, beyond their control. They will be more likely to claim a political voice, because they will feel entitled to one.
Third, and perhaps most important, real change will look newly possible in light of the unprecedented measures the government has already taken to combat the crisis. Congress conjured trillions of dollars to bail out mostly businesses, which means it could do the same for people. Granted, that’s not what happened in the 2008 financial crisis, when the government primarily rescued the banks — but that experience could also be invoked to demand a better deal this time around.
So this crisis brings with it an opening for collective action. How it will work remains to be seen. Pitchforks aren’t effective against social-media algorithms, and Amazon might not care if some large fraction of the population can’t buy its stuff, as long as the rest keeps buying more. But if the winners of the AI revolution want to avoid the business disruption of an actual revolution, they should be prepared to negotiate a new and very different deal.
Cathy O’Neil is a Bloomberg Opinion columnist. She is a mathematician who has worked as a professor, hedge-fund analyst and data scientist. She founded ORCAA, an algorithmic auditing company, and is the author of “Weapons of Math Destruction.”