Artificial intelligence is stuck in a tight rut

With the endless promotion of artificial intelligence by analysts, media, and vendors, one can be forgiven for assuming that AI is proliferating in businesses and driving them from far and wide. However, the reality goes beyond simply automating narrow applications – such as credit scoring, upselling recommendations, chatbots or machine performance management – AI still has a limited reach and is beginning barely reaching its full potential as a true augmentation of human intelligence and talent. .

This is what emerges from the recent round table organized by the New York University Center for the Future of Management and the LMU Institute for Strategy, Technology and Organization, which was joined by Daron Acemoglu, professor at the MIT; Jacques Bughin, professor at the Solvay School of Economics and Management; and Raffaella Sadun, professor at Harvard Business School.

“I’m not very keen on narrow applications of AI,” Sadun says. “I’m not keen on the dumb approach of automating a process and pretending you’re on a different technological level.”

At the moment, “AI is pushed too far into automating narrow tasks, where there aren’t a lot of productivity gains and a lot of travel,” agrees Acemoglu. Essentially, “AI is used at a narrow process level as a cost reduction measure. It doesn’t really work very well on a large scale, and it’s really not the best way to use AI. But a lot of managers do. Many companies investing in AI are doing it a bit the wrong way. »

AI will begin to reach its potential when it aligns broader human and machine intelligence, says Acemoglu. “Machines must become a tool for selecting and presenting the right kind of information, to take advantage of human skills, judgment, creativity, social learning and situational learning,” he says. -he. “Very few organizations do this. There is a failure of these organizations, but more importantly, it is a failure in the field of AI,” says Acemoglu. “Managers follow what AI leaders tell them. And AI leaders don’t discuss this type of AI use. They encourage the use of AI at the process level to reduce costs. In that sense, we are in an AI trap.”

Adoption of broader and more meaningful AI must start at the top, where there is still a lot of resistance, Sadun says. “There are too many leaders and executives who see AI as a technology, and not as a lever for organizational change,” she explains. “And they have no idea how the technology is actually used by their employees. People who are really using AI effectively in their business have spent time on the shop floor, seeing how the technology is being used by people. They have an idea of ​​the software and hardware change that needs to be there for people to really take advantage of AI.

Adopting AI must be an organizational transformation. “It takes different skills — not just coding skills,” Sadun says. “It’s much more a question of soft skills. The ability to coordinate knowledge from different pockets of the organization. I am surprised that even some organizations are successful. This is actually a very difficult question.

Companies that use AI to its full potential “consider incorporating new inputs,” Sadun says. They think about training, making sure the technology is actually usable, and can benefit from continuous self-improvement. Basically, these are technologies that think of a human and a machine as a team. »

Leave a Reply

Your email address will not be published.