Self-driving vehicles could upend the transportation sector and eliminate a million or more jobs. Algorithms that decode M.R.I.s put a whole medical subfield at risk. And the list of professions and sectors soon to be obsolete grows steadily by the day.
New technologies are rattling the economy on all fronts. While the predictions are specific and dire, bigger changes are surely coming. Clearly, we need to adjust for the turbulence ahead.
But we may be preparing in the wrong way.
Both history and psychology tell us that our capacity to predict the future is limited, while our capacity to believe in such predictions is unlimited. We have always been surprised.
Rather than planning for the specific changes we imagine, it is better to prepare for the unimagined — for change itself.
Preparing for the unknown is not as hard as it may seem, though it implies fundamental shifts in our policies on education, employment and social insurance.
Take education. Were we to plan for specific changes, we would start revamping curriculums to include skills we thought would be rewarded in the future. For example, computer programming might become even more of a staple in high schools than it already is. Maybe that will prove to be wise and we will have a more productive work force.
But perhaps technology evolves quickly enough that in a few decades we talk to, rather than program, computers. In that case, millions of people would have invested in a skill as outdated as precise penmanship.
Instead, rather than changing what we teach, we could change when we teach.
Currently, all the formal education most people will receive comes early in life. Specific skills may be learned on the job, but the fundamentals are acquired in school when we are young. This sequence — learn early, benefit for a lifetime — makes sense only in a world where the useful skills stay constant.
But in a rapidly changing world, the fundamentals that were useful decades ago may be obsolete now; more important, new essential skills may have arisen. Anyone helping a grandparent navigate a computer has experienced this problem.
Once we recognize that human capital, like technology, needs refreshing, we have to restructure our institutions so people acquire education later in life. We don’t merely need training programs for niche populations or circumstances, expensive and short executive-education programs or brief excursions like TED talks. Instead we need the kind of in-depth education and training people receive routinely at age 13.
In addition, we must recognize that economic upheaval at the macro level means turmoil and instability at the personal level. A lifetime of work will be a lifetime of change, moving between firms, jobs, careers and cities.
Each move has financial and personal costs: It might involve going without a paycheck, looking for new housing, finding a new school district or adjusting to a new vocation. We cannot expect to create a vibrant and flexible overall economy unless we make these shifts as painless as possible. We need a fresh round of policy innovation focused on creating a safety net that gives workers the peace of mind — and the money — to move deftly when circumstances change.
Finally, we can learn from the economist Joseph Schumpeter’s prescient analysis of entrepreneurs. He noted that for new innovations to spread and improve our lives, there will always be creative destruction. For new firms and sectors to arise, some of the old ones must die.
Even if we should be humble in predicting that self-driving vehicles will upend the trucking sector or drone delivery will decimate supermarkets, we can be confident that some creative destruction is coming.
Our current policies — and impulses — are to resist such destruction. If a large manufacturer is set to close, subsidies and other policies kick into action to prevent that shutdown. But while we may save a factory, ultimately we hinder the rise of new technologies; rather than propping up incumbent firms we ought to enable innovation to take its course.
If that idea makes you uneasy, it is probably because our current policies do nothing to protect the most vulnerable from the costs of all this destruction. We resist letting factories close because we worry about what will become of the people who work there. But if we had a social insurance system that allowed workers to move fluidly between jobs, we could comfortably allow firms to follow their natural life and death cycle.
In the 1990s, Denmark began adopting what has been called “flexicurity,” combining policies that promote a flexible economy — allowing creative destruction as needed — with those that promote security for workers. The Danes have also emphasized lifelong learning, giving workers income support as they transition between jobs and circumstances.
By contrast, the current approach in the United States could be called “flex-nosecurity,” which hardly seems the appropriate way of preparing for an economy of rapid change.
There are surely many other ways of preparing for upheaval. We should broaden the current conversation — centered on drones, the end of work or the prospect of super-intelligent algorithms governing the world — to include innovative proposals for handling the unexpected.
One problem is that social policy may seem boring compared with the wonderfully evocative story arcs telling us where current technologies might be heading. How can the minutiae of unemployment insurance compete for attention with movies describing the birth of Skynet, the diabolical neural network in the “Terminator” series?
Yet even science fiction teaches humility.
Take “Star Trek.” The future it imagines is wondrous to the point of bordering on the impossible. The laws of physics as currently understood are circumvented so that ships can travel faster than the speed of light. Unfathomable technologies are routine. People can be disassembled atom by atom and transported somewhere else, keeping their memories and consciousness intact. Any kind of food can be instantly replicated.
Even the inventive “Star Trek” writers peering into the future, though, could not imagine a completely self-driving Starship Enterprise. While at times the Enterprise appeared to have some autopilot capacity, it routinely relied on a navigator to pilot it and was even equipped with a view screen that looked suspiciously like a car windshield.
The safest prediction is that reality will outstrip our imaginations. So let us craft our policies not just for what we expect but for what will surely surprise us.