Scalable learning vs. scalable efficiency in the automation age

Scalable efficiency

According to Hagel, most institutions react to market pressures by getting more efficient at ever-larger scale. As a result, managers focus on defining work as tightly specified tasks that can be routinely performed in the same way -- and thus automated. One side effect of this model is the elimination of jobs.

John Hagel

To be sure, some of the proponents of the scalable efficiency model also talk about reskilling people whose jobs have been eliminated through automation. But Hagel argued that most new skills don't last long in times of rapid change, so people will need to be continuously reskilled -- a commitment many companies are unwilling or unable to make. Other companies use gig economy models -- hiring people on an as-needed, project-by-project basis -- to make fixed labor costs variable.

The downside of the scalable efficiency model is the loss of trust and loyalty, as workers -- and even leaders -- are left wondering how long their jobs will last.

"The institution is going to win, and the worker will lose as we become more efficient," Hagel said. But he said, at some point, the loss of human capital catches up with institution.

Scalable learning

Hagel contended that the institutions with the greatest chance of succeeding in the future will be driven by scalable learning. In this approach, leaders focus on how the organization can learn faster at scale. This type of learning is not about watching e-learning videos or training programs to do a job more efficiently.

Scalable learning involves creating the infrastructure and incentives that will make everyone think about creating new value for the business, rather than on learning just enough to do the job at hand.

According to proponents of the scalable learning model, by making value creation the focus of everyone -- from the workers on the factory floor to the front-line sales teams to the maintenance staff -- organizations will get better at identifying problems that stand in the way of success and be more likely to think about new ways of working.

Learning versus knowing

The scalable learning model only works when leaders believe value can be created by people at all levels of the organization.

In the scalable efficiency model, leaders are expected to have answers to all the questions -- and if they don't, they get replaced. In the new model, smart leaders solve problems by asking better questions of employees, customers, business partners and so on. Leaders following the scalable learning model also have to admit when they don't have an answer and be willing to ask for help.

"Our view is that, if you ask powerful questions, exciting questions, it excites the passion in [employees] who could make a difference," Hagel said.

Start small

Hagel warned the audience not to expect overnight success with a scalable learning model. "If you believe transformation is a rational process, about collecting the right data and presenting it to the right people, you have already lost," Hagel said. "Transformation is fundamentally a political process, not a rational process."

Before doing anything, he said it is important to identify and neutralize the "enemies of change." Hagel said he has never seen a senior leader get up and announce they are an enemy of change. Rather, they go back to their office and conspire how to subvert the new program.

The next step lies in identifying and strengthening the champions of change. But don't give them too big a budget, or it will just make their project a target of other departments, he said.  "Never underestimate the power of the immune system and antibodies that exist in every institution today."

Be a better human

One approach for growing this scalable learning model may lie in finding ways to align the interests of workers and AI, said Benjamin Pring, co-founder and director of Cognizant's Center for the Future of Work. He suggested AI can help empower front-line workers to find new ways to improve the enterprise.

One place to start may be getting people to think about beauty -- not an attribute most enterprises put into their mission statements. But, as Pring noted, Apple's success is often directly attributed to Steve Job's decision to hire Jony Ive to bring beauty to the smartphone. Jobs kept going and thought about bringing beauty to the factories, because he reckoned it would inspire the workers that made them.

Benjamin Pring

Not all enterprises are this enlightened today, Pring quipped, showing a picture of the Cognizant expense-tracking application. "It is ugly," he said. "But we all know that things sucking is the real mother of invention. It is a simple equation that the future of work is to make things suck less."

Instead of fearing that AI will automate them out of a job, workers could be asked how AI could make each of their jobs suck less, he said.

"Societies have always adapted to changes in tools, but individuals along the way haven't," Pring said. He has no illusions about the impact of AI and automation. He said he expects technology to automate a big swath of worker jobs, just like it did with agriculture over the last 100 years. Big business will process more insurance claims and loan applications with far fewer people, because robotic process automation  can do it far more efficiently than humans.

But this may not be a bad thing. Much of human work has amounted to the impersonation of very robotic work, Pring said. Now that the real robots are showing up, they will do this type of work much better than humans. The human response is to "double down on what makes you a good human being instead," he said. "Don't be a bad robot."

Alluding to Peter Drucker, Steve Ardire, an AI startup consultant, said, "Efficiency should be delegated to machines, while effectiveness is a human pursuit."

That's all well and good, assuming companies don't want to just settle for the robot.