Artificial Intelligence & The Future of Knowledge Work
Tom Davenport has been thinking about working for a long while. He and Larry Prusak may fairly be said to have defined knowledge management for business in their 2000 book Working Knowledge: How Organizations Manage What They Know. Just five years later Thinking for a Living focused with practical force Peter Drucker’s 1999 insight that “the most valuable asset of a 21st-century institution … will be its knowledge workers and their productivity.”
Davenport and Julia Kirby, in their recent article Beyond Automation, Harvard Business Review, June 2015, have now joined the debate about the impact of artificial intelligence on the future of knowledge work.
Fifty years ago, long before Google taught cars to drive themselves and started buying up robotics companies, the Ad Hoc Committee on the Triple Revolution warned President Lyndon Johnson:
“The cybernation revolution has been brought about by the combination of the computer and the automated self-regulating machine. This results in a system of almost unlimited productive capacity which requires progressively less human labor.”
Happily, the Committee’s grim forecast of mass unemployment proved wrong. However, as Frank Levy (MIT) and Richard Murnane (Harvard) wrote in their 2013 paper Dancing With Robots, “computers have changed the jobs that are available, the skills those jobs require, and the wages the jobs pay.”
Two of Professor Levy’s colleagues at MIT, Erik Brynjolfsson and Andrew McAfee, have also been thinking about the impact of technology on employment. As they wrote in Race Against the Machine (2011), “Our technologies are racing ahead but many of our skills and organizations are lagging behind.” In The Second Machine Age (2013), they extended the argument:
“There’s never been a better time to be a worker with special skills or the right education, because these people can use technology to create and capture value. However, there’s never been a worse time to be a worker with only ‘ordinary’ skills and abilities to offer, because computers, robots, and other digital technologies are acquiring these skills and abilities at an extraordinary rate.”
Davenport and Kirby write:
“If this wave of automation seems scarier than previous ones, it’s for good reason. As machines encroach on decision making, it’s hard to see the higher ground to which humans might move.”
They define three eras of automation. In each, machines take over from humans another level of work, because the machines are faster, cheaper, and … yes, better:
- The dirty and dangerous (19th c.)
- The dull (20th c.)
- The decisions (21st c.)
But Davenport and Kirby do not stop the analysis, nor do they despair:
“Instead of seeing work as a zero-sum game with machines taking an ever greater share … we could reframe the threat of automation as an opportunity for augmentation.”
They cite MIT economist David Autor on “the immense challenge of applying machines to any tasks that call for flexibility, judgment, or common sense” and quote Autor’s paper on Polanyi’s Paradox: “tasks that cannot be substituted by computerization are generally complemented by it.”
Augmentation is not a new theme. Douglas Englebart founded the Augmentation Research Center more than fifty years ago, in 1963, with the aim of building software and systems sophisticated enough to be flexibly useful to humans. (He invented the computer mouse too.)
Here at Neota Logic, we design our software to both automate and augment knowledge work—in the form of expert systems that transition smoothly from one role to the other as the task warrants. When the facts and the rules are clear and consistent, systems will automate. When the context is uncertain and the judgment required is more subtle, systems will augment. They will apply expertise to gather relevant facts efficiently, identify rules that may apply, and present the problem to human experts for resolution.
We also agree with Davenport and Kirby that the future holds important and rewarding employment for those who engage in “constructing the next generation of computing and AI tools. It’s still true that behind every great machine is a person—in fact, many people.” Building Neota Logic expert systems does not require programmers (thanks to the no-code tools we create), but there is exciting and challenging work for subject matter experts and for knowledge engineers to help the experts codify what they know.