Contestant Ken Jennings appeared on the popular trivia game show Jeopardy in 2004. The year before, the show’s rules changed, stating the winner of each episode would proceed to the next episode until they lost. The longest winning streak up to Jennings’ debut was eight episodes and by the time Jennings was defeated, he’d won an unprecedented 74 Jeopardy episodes.
However, not even Jennings could compete against IBM’s Watson, a computer designed to interpret Jeopardy questions, formulate an answer and ring the buzzer before any opponents. After Watson’s wide victory over Jennings, IBM continued to improve Watson’s technology, turning it into a diagnostic tool for oncologists, assisting them in choosing the best therapies for their cancer patients.
IBM and other companies continue to develop these machines, characterized as “deep learning” computers, to best humans in games such as Jeopardy and chess, demonstrating their ability to critically apply and use information in ways challenging to even humans. The trend in computer advancement supports the oft-cited prediction that machines will continue to replace human jobs.
Grocery stores’ self checkout registers are no longer novel, McDonald’s offers the option to place your order via touch screen, remote video monitoring at businesses replaces the need for on-site security and manufacturing continues to displace line workers as new machines (along with cheaper offshore manufacturers) take their places. Economists predict continued disruption in the labor market once the self-driving car becomes available.
However, recent automation developments usually displace lower skilled, blue collar workers. Now economists and academics warn that with the advent and advancement of deep learning technology, not even higher level, white collar jobs are safe from automation. Jerry Kaplan, a Stanford professor, explains, “Automation is now blind to the colour of your collar.”
Recent studies have predicted that 47 percent of the American workforce have jobs with high risk of being automated, while 35 percent of the British workforce and 49 percent of the Japanese workforce face the same potential outcome. Early in 2016, the Obama White House released a report stating that a worker making less than $20 an hour would have an 83 percent chance of eventually losing his or her job to a machine, while people who make as much as $40 an hour still face a 31 percent chance of machine caused job loss.
Luckily, not at all experts are discouraged. David Autor, an MIT economist, points out in the past technology ends up creating more jobs than it destroys. Bank Tellers were despaired when the ATM was first introduced, but instead of eliminating jobs, ATMs created more demand, necessitating more branches with tellers, who now had different responsibilities focused on customer service and sales.
James Bessen, another economist at the Boston University School of Law, did a study where he found that jobs which made use of computer automation to improve productivity actually increased employment in those occupations. By looking at the US workforce over 30 years, he found that jobs, such as graphic design, which had incorporated more technology, enabled workers to be more efficient and improve in manual areas of their jobs. Computers often require workers to learn or focus on new skills once a task has been automated. Bessen explains that computers usually reallocate, instead of replace workers.
Encouragingly, it seems that the trend of technology spurring job growth seems to be true in some early uses of artificial intelligence. Legal clerks and paralegals used to spend countless hours reading and researching during discovery phases of legal cases. Now AI software can analyze huge volumes of data quickly and affordably, and makes judges more willing to allow time for discovery. The overall number of paralegals increased with this technology. The same kind of trend was seen with the uptick of online shopping and digital marketing, which increased the demand for retail locations due to the reinforced desire to shop.
The mystery of how work will change with continued automation remains. Autor, the MIT economist, illustrates the similarity of today’s world on the cusp of self driving cars and last century’s introduction of the automobile. People then worried about horse-related jobs, but completely new jobs and industries, like motels and fast food, sprung up to meet the demands of the new motorists. While new demand can’t be predicted with certainty, individuals can take advantage of new opportunities as they emerge alongside automation.
By Rick Delgado, Freelance Automation Writer