14 April, 2021
The present is a significant point of inflection for enterprises in their digital journeys. Even as cloud and SaaS promise to free business from the shackles of physical immobility, few recognize what it takes to deliver excellence in an environment where the customers were expecting instant response 24x7, every single time. Ubiquitous app culture, for everything from telecom to banking to the next-door grocery store, promise customer intimacy hitherto unknown. However, scaling with excellence in this digital ecosystem is not the same as scaling yesterday – throwing more people capacity at the problem.
It’s time for the latest entrant into the workforce – AI. AI has already made its way into our homes, our cars, our mobile phones and cameras and computers, and even our beds. While the big tech is busy developing technologies that can spot bugs in production environments, the small techs are busy selling minuscule devices that spot bed-bugs in hotel rooms – with AI. AI is surely moving from good-to-have, to transformational technology
Yes, this new hire too takes time to learn, and understand the subtleties of your enterprise’s ways of working, talking and problem-solving. And like all new hires, this one is going to commit some mistakes too when it’s on its learning curve. Yes, it will help you keep your lights on at a fraction of the cost when the humans are resting and gearing up for the next working day.
The popular, though layperson, narrative of AI shows it in mixed light. While futurists were furiously convincing the world of a time when AI would accompany humans in their space odysseys, drive cars for them, and even help them feed and entertain their pets, early Hollywood releases starting from The Matrix up until the later ones like Westworld, Skynet and Ex-Machina warned us of a time when AI would take over humans turn us into its slaves.
Perhaps our trust issues with AI started with such visions. Curiously, most naysayers don’t even realize when they are using AI. Unlocking our phones by pointing our faces to the front cameras in anticipation of being recognized, or rating streaming services on the basis of how well they could recognize our taste, or interacting with intelligent chatbots for how well they could understand our problems when no support executives were available to lend an ear – it’s all AI at play. But when questioned, our distrust in AI surfaces, and not particularly for the right reasons.
It’s up to business leaders as to whether they will let this new hire work on autopilot without trammeling it with their own control issues. This is not to advocate a blind trust in AI systems that can make manifest grave disasters – like instances of racial bias in hiring, or unmonitored engagement generating algorithms that amplified biased and politically-threatening voices in its system. Naturally, such stories will seed fears in the minds of the C-level. Are we expecting AI to become what we couldn’t and perpetuating our own unattainable vision of a bug-free world and offloading it onto this new hire?
Perhaps we are giving it a tough time on its quarterly performance reviews. AI deployments need to be evaluated after they have had a chance to sit through a business cycle and not obsessing over what algorithms are responsible for those two errors in the last ten problems. Business leaders must move on from numeral to comparative when assessing mistakes –from how many to what percentage of errors.
Our instinct of pitting AI against the incapacities of the human is what needs to change. AI deployments need to be evaluated on the basis of their learning performance and how their accuracy evolves over time, and whether their performance truly calls for an overhaul. It is only then, that this new hire will be able to reveal what it is capable of. Until then, AI will remain in an uneasy dynamic – where our own trust issues will continue to impact its performance and our gains within the rules of digital business.
The famous physicist Allen Bartlett’s claim of our inability to understand the exponential curve deserves mention here. We cannot address exponential grown in business needs with linear addition of capacity. AI is not the future – it is the present. And the ability to leverage it to solve today’s problems will separate the enterprises that will thrive in the brave new world from those that become mere markers of a world that was.