Needed: Humans To Break Artificial Intelligence Out Of Its Silo – Forbes
AI is nothing without the people behind it.
AI will fill many roles, but perhaps its greatest potential is in its ability to reach across or break through organizational silos. At this point, however, with most implementations in pilot stages, labs or focused on single tasks such as chatbots, AI has yet to break out of its own silo. For AI to be a truly revolutionary force across organizations, it needs to be liberated, in a very human way.
That’s the word from Sharmila Chatterjee and Zoran Latinovic, both with MIT, speaking in a webcast hosted by MIT Sloan Management Review. Chatterjee, who conducted research into the factors driving successful AI, finds a common denominator across successful AI projects: they leverage human intelligence to the max. “We need to keep in mind that human intelligence is more powerful than human-driven AI,” she says. “It is collective intelligence — AI and human together.”
In her research, Chatterjee found purely AI-driven engagements don’t succeed too well in the market. Human interaction scored higher than technical capabilities in areas such as “customer satisfaction, word of mouth, and repurchase intent,” she explains. “It’s not the scope of services, but rather the value mindset, the human interface that plays the dominant role in building positive relationships and customer retention.”
UPS is an example of a company that is blending AI and human skills to deliver results. “The company uses a network planning tool app to increase operational efficiency,” Latinovic explained. “A customer’s package is categorized by a zip code, weight, and volume, then given a bar code, scanned, and loaded for delivery. The app organizes packages by final destination while also considering the type of parcel and time of year. For instance, pharmaceuticals are not routed near the desert as extreme temperatures can affect the potency of certain medications.”
However, the real value from this AI-driven app is that it augments human intelligence, versus attempting to replace it. “It empowers human engineers to make better decisions,” says Latinovic. “When a package reroutes, the app notifies a UPS engineer. The engineer then looks at various options, evaluates them, and takes action. The app also serves as a check on the engineer’s choice. This not only saves UPS time and money, but it’s also a boon to customer satisfaction.”
AI helps break down silos in three ways — by enhancing communication, customization, and coordination. “AI’s promise is to help create more connected, coordinated systems, both inside and outside the organization,” he says. Communication should be “ongoing and iterated. AI has the capacity to revolutionize communication and deepen relationships. When used correctly, AI applications have immense upside potential.”
In terms of customization, “with the help of the AI, companies collect data about customer needs, wants, and preferences, and then they apply it to customize their offerings, create personalized experience, and improve the customer experience. This can completely transform companies’ marketing strategies with a focus on identification with customers and seamless customer engagement.”
Finally, AI enhances coordination, as “AI-powered platforms can improve the ease of working together across functions within an organization and with partner organizations outside,” Latinovic says.
Chatterjee and Latinovic make the following recommendations to successfully blend human and machine strengths:
Plan: “Top management has to believe in an analytics or insights-driven organization. Then one has to, of course, examine the organizational workflows and processes, assess the potential for incorporating AI into those workflows and processes, develop a phased timeline for organizational adoption of AI strategy.”
Put ethics first: “Ethics is a really important aspect of AI that we are just beginning to grapple with,” Chatterjee says. This includes data protection and privacy. “You cannot just add ethics at the end. At a minimum, just awareness that there are ethical things to be considered is a really good start.”
Think change management: Similarly, “change management cannot be an add-on at the end,” Chatterjee says. “Starting from the very beginning, the changes that have to be made along with the adoption of the new technology need to be on the table. Legacy processes will not work with a new technology. The processes need to evolve. The workflows need to evolve.”
Cultivate grassroots support and culture: “Organizational culture cannot be underestimated,” says Chatterjee. “The grassroots, the employees have to absolutely buy into that culture. And how do you promote that? Nothing succeeds like success. Pitch sort of small successes and then try to evangelize the organization around it to try and make people see the benefits that they are getting.”
Acquire and develop the human capital, expertise, and skills: This is perhaps the most essential ingredient of all. AI may take humans out of the loop, but it takes human ingenuity and imagination to provide AI its value.