As the amount of digital work increases and the amount of physical work decreases, our notions of employment and work change profoundly. Digital work doesn't require roads and factories; it requires a laptop and an Internet connection--equipment that people have access to in their homes. The need for offices, supervisors and rigid employment arrangements diminishes.
As technology improves, companies should theoretically be able to access in real-time the perfect person for a given job--the one who will do the job the best, enjoy it the most or do it the fastest. All these factors combine in a way that will change the landscape of work. Here's what I think that will look like:
--Within a decade résumés will become less important as we continue to adopt newer, multifaceted ways to measure the quality of a candidate's work. Current hiring processes often involve online research about candidates on sites like LinkedIn and Facebook. Articles, portfolios, presentations and papers by potential job candidates are increasingly found online. Companies like oDesk and Elance rate workers based on past work rather than on what college they attended.
--In a way individuals become their own companies, and they are responsible for marketing themselves, negotiating their rates and deciding which work to pursue. There will be less wage stickiness, and people with unique skills will be rewarded for them. Everyone who's smart and can do good work can have access to jobs. The question then changes from who you know to what you know. A new digital meritocracy is evolving.
--In 1900 60% of Americans lived in rural areas; in 2000 21% did. They moved to where the jobs are. But if people can work from anywhere in the world, then why wouldn't they? So I also predict that the rise of digitally distributed work will reverse this trend of urbanization over the next few decades.
--There will be a global decrease in unemployment. This last point is controversial and counterintuitive. With the growth of any new technology, there's always a fear that it will remove jobs. It's reminiscent of Kurt Vonnegut's Player Piano, which imagines a future dystopia where most jobs have been automated and nobody has access to work. This future makes some intuitive sense with all of the advances in artificial intelligence, however, in practice it hasn't been realized.
If you look at the U.S. unemployment rate since 1948, when Player Piano was written, you will see some unemployment growth in terms of percentage, but also massive employment growth in number of people employed. Certainly automation has led to some jobs becoming obsolete. (For example, in 1948 "computer" was a job that human beings held.) But automation can actually drop unemployment rates in two important and significant ways.
--First, it will reduce the amount of time it takes unemployed people to find jobs. The amount of time it takes a newly unemployed person to find her next job contributes significantly to unemployment rates. Economists call this "frictional unemployment." Companies like Solvate, Elance and oDesk give people instant access to jobs, and over time technology has the power to create a much more efficient market, connecting job-seekers to employers instantly.
--Second, automation has the power to create lots of jobs. For example, machine learning and data mining have become household terms in just the past five years. Consider that Google Translate, which is state-of-the-art, has algorithms that are less sophisticated than algorithms we saw even in early 2000. Their advantage is that they have a humongous parallel corpus, where people have labeled corresponding pieces of different languages down to individual words. And that massive amount of human data collection is what has led to the mass adoption of Artificial Intelligence.
That's the type of AI we're going to see over the next 10 or 20 years. These systems are going to be built on top of tons of human-annotated data, rather than making such human annotation obsolete as some might anticipate.
Google, Amazon and other high-tech companies are actually pushing for new jobs with more human involvement. They can measure the increase in quality that comes from increases in labeled data, which gives them an insatiable desire for labeled data accuracy, and they hire thousands of people to label data to feed information into these computers so that they can automate better. These kinds of jobs didn't exist before. This is the future of work.
Lukas Biewald is chief executive of CrowdFlower.
See Also: