Every discussion about artificial intelligence seems to alternate between utopia and dystopia. Some believe that the productivity unleashed through automation will lift up all of society, creating a world of superabundance and more meaningful work, while others see robots taking our jobs and an acceleration of trends favoring capital over labor.
In fact, in an article in Harvard Business Review, Accenture’s Mark Knickrehm describes five distinct schools of thought, ranging from both extremes to various shades of gray in between. He suggests that leaders need to reinvent operating models, redefine jobs and include employees in the process of transformation.
Yet that’s easier said than done. Smart leaders know that even small, subtle changes can sometimes result in a backlash. Preparing your organization to leverage artificial intelligence has can be especially problematic because the most profound problems are intensely human. Gandhi, although he was no tech enthusiast, can be a good guide on where to start.
Create A Vision For Tomorrow
Gandhi, by all accounts, was a brilliant tactician and strategist. However, he was also more than a chess player who planned moves a few steps ahead of his adversaries. He had a clear vision of what the future should look like — ”Purna Swaraj” or complete independence from British rule. It was that clear vision that drove his actions and inspired people to follow him.
Martin Luther King Jr., who studied Gandhi closely and was in many ways his disciple, formulated his objectives in a similar way. He wasn’t just fighting for the rights of black Americans, but to “make real the promises of democracy” and to “cash the checks” written into the founding documents of our union.
So the first step to building an AI future would be to form a clear vision of what it’s supposed to look like. Should AI do our work for us so that we can have more time to seek personal fulfillment? Or is it supposed to augment our abilities so that we can become more productive in our work? Or maybe something else?
Listen to just about any AI evangelist today and you’ll hear a different vision. Until we have a clear idea of the future we want, we are unlikely to make it happen.
Prepare The Ground
Gandhi cut his teeth as an activist in South Africa, where Indians were a relatively small minority. He was therefore able to build his principles step-by-step and indoctrinate his followers as he went.
Yet when he returned to India, he made the mistake of trying to lead an entire nation of diverse attitudes and interests that was not yet indoctrinated in his philosophy of Satyagraha. The result was disaster. Instead of peaceful protests of civil disobedience, he got violent riots.
Gandhi would come to call this his Himalayan miscalculation. “Before a people could be fit for offering civil disobedience,” he later wrote, “they should thoroughly understand its deeper implications.” Clearly, he learned his lesson and spent a decade indoctrinating the Indian independence movement in his values.
We already have a number of unresolved ethical issues involving AI ranging from bias in the learning corpus, to classical dilemmas such as the trolley problem. We also lack a clear understanding of what standards AI should be held to. Should algorithms be held to the same level of accountability and transparency as humans or something more?
Today we are already in the process of an AI transformation, with hundreds if not thousands of large-scale implementations. What values should govern these investments? We haven’t even begun to work through the basic issues. Are we making a modern version of Gandhi’s “Himalayan miscalculation?”
Create A Sense Of Shared Purpose Through A Transformational Project
When, on December 31st, 1929 the Indian National Congress declared self-rule by releasing a document modeled after the American Declaration of Independence, it was, much like that earlier document, a symbolic gesture. To further the cause, they needed to take concrete action, but no one could agree on what form it should take.
So the Mahatma returned to his ashram and emerged after weeks of meditation with an answer. He would march for salt. No one was impressed. In fact, to many it seemed like a joke. The British Viceroy, Lord Irwin, remarked dismissively in a report to London that he would not lose sleep over salt.
Yet it proved to be an inspired choice. The Salt Laws were so obviously and fundamentally unjust that the future British Prime Minister, Ramsay MacDonald, had himself denounced them just a year before. They also affected every Indian, whether they be Hindu, Muslim or Sikh, caste or outcaste. It soon became clear that the Salt March was a historic success.
Leaders would be wise to take a similar approach to automation. Rather merely looking to use technology to eliminate jobs and save money, a better way forward would to be to first automate tasks that everyone sees as onerous and free up efforts for more interesting, high level tasks. By focusing on eliminating drudgery first, crucial support can be won early.
We’ve already seen how this approach can be applied in various contexts. Factory workers actively collaborate with robots they program themselves to do low-level tasks. In some cases, soldiers build such strong ties with robots that do dangerous jobs for them they hold funerals for them when they die.
Win Over The Losers
There is perhaps no greater testament to Gandhi than the high regard he was held in by his adversaries. The Boer leader Jan Smuts said of him, “It was my fate to be the antagonist of a man for whom even then I had the highest respect.” Upon his death, Britain’s The Daily Telegraph reported, “The tragic news of Mahatma Gandhi’s has come to his Majesty’s government as a profound shock.”
This was no accident. Gandhi took pains to reach out to those who opposed his goals and sought to form a common purpose, without ever losing sight of his objectives. He wasn’t fighting to win for winning’s sake, but saw his adversaries as partners in a quest for truths that transcended their positions and narrow interests.
In Blueprint for Revolution, Serbian activist Popović calls this “surviving victory” and he stresses that the battle must be won before it starts, by indoctrinating common values and purpose. With AI, it is not enough to simply evangelize the technology. Nobody falls in love with an algorithm. We can’t lose sight of the fact that technology should serve people, not the other way around.
The uncomfortable truth is that, as with any transformation, AI will create winners and losers and some accommodation must be found. Bill Gates advocates taxing robots that take human jobs. Others support a universal basic income. I’m not sure what the right answer is, but we desperately need to achieve a consensus that the problem exists.
What is becoming clear is that as the technological barriers to an AI future fall away, the ones that remain will become more social in nature. It is those that we need to start turning our focus toward now.
An earlier version of this article first appeared in Inc.com