We are entering a brave new and complex world of smart-machines replacing humans. This tension between man and machines is not entirely new, however. The creation of new technologies—from washing machines to assembly line robots—has replaced and displaced human workers for over a century. The difference is that robots and machines can’t truly learn, communicate, or evolve.
Today we are seeing more “intelligent” robots that are capable of using technology to perform traditionally human jobs and tasks. We are becoming part of them and they a part of us. Think about IBM. First Big Blue won at chess and then Watson at Jeopardy. Think about all the single-player games or e-commerce. Amazon wiped out thousands of bookstores. Expedia helped shutter the travel agency industry. Consider the actual mechanical robots constructing the cars we drive. All of these robots can exchange data in real time about real people.
Companies like LinkedIn, Monster.com, Glassdoor, and others are finding jobs for those who have appropriate resumes. These systems routinely send out spiders, a form of robot, to scrape the Web of the data in resumes and job listings, and bring them back and link them with individual profiles from applicants. This raises the question: could there be a class of robots designed to help people keep and find jobs that might otherwise be taken by robots? Introducing a new class of robots into the market would allow us to apply theory and extract information about the power of humans paired with robots to help keep humans employable—including building robots, versus the more common scenario of robots replacing humans.
Instead of passively standing by and watching jobs go to automation, especially entry level jobs, there could be a way to fight back by working collaboratively with automation. Consider a class of apps called “butlers,” which are specifically designed to understand job seekers’ strengths, and helps them to either better compete for existing jobs or join into pairings with automation to increase their competitiveness.
The U.S. could become an expert leader at determining the optimum mix of what machines and humans can do independently and together. What can certain mixes and pairings do better than separately, or better than overseas labor, or overseas robots? Perhaps one day there will be human versus robot talent indexes and firms will emerge as agents for robots and human/robot teams. And perhaps hybrid police will emerge who can detect virus-makers before they strike.
Recently, my organization, National Laboratory for Educational Transformation (NLET), began working with a new team of researchers to explore the frontiers of people, training, and job matching by regions. Robot butlers will be needed here for sure, because despite the unemployment numbers going down in general, there are still a persistent seven million unemployed or underemployed people in this country. There are also a persistent five million open jobs, many of them skill-based or middle-skills jobs that require specific training. The butlers could help identify a range of training opportunities to quickly prepare people for work before the investment is made in replacement robots.
Butlers could help construct resumes or experience graphs for those who are unlikely to have resumes or be on networking platforms like LinkedIn. They could also help employers read their employment trends as new job demand is forming. This is necessary because those creating occupational training for the rapidly evolving job market cannot make those predictions.
There are also questions of math and economics. To build a robot, either to replace human workers or to support them, how many jobs are created? What will be the ratio of new jobs created to make robots and butlers compared to the jobs eliminated by automation? One thing we do know is that job training in an ever-changing world will require smart training that can empower people to adapt quickly to new types of work. Historically, that task has fallen to community colleges providing such regional training or employers doing it themselves. Right now neither is working efficiently.
NLET is gathering up interesting players across the country who have tried to work on the more sophisticated pieces of the puzzle, but have had trouble raising the level of funds or grants to do so. This is a problem of scale. Until an Amazon-size player or a National Science Foundation decides that training people for new kinds of work is as important as consumer sales or genetic research, we may not get the tech-empowered employment solutions that our economy needs.