Tracy Gusher from KPMG talks about scaling AI, and common misconceptions about use of machine learning in business
During the AI Summit in San Francisco, Traci Gusher, principal for data, analytics and artificial intelligence at KPMG, talked about the emerging trends in enterprise AI adoption.
KPMG is helping its customers harness the
power of AI through something it calls the connected enterprise. AI Business
sat down with Gusher to find out more.
“What this is really about is how do you
connect the back office to the middle office to the front office and bring
artificial intelligence through all three layers of an organization,” she said.
“A lot of what we've seen in artificial
intelligence in the enterprise over the last few years has really been very
customer-centric. And that's not a bad thing. But we've gotten a little bit
stuck on the customer.
“We need to take a step back and look at
how does that customer experience, and the AI we're driving in the customer
space, how does it connect the middle office? How does it connect to the back
office, and how do we connect it all together? And it's a message and an
approach that's really resonating with a lot of our clients, particularly those
that are that are aiming to get both revenue and cost savings benefits out of
Gusher also mentioned two common misconceptions
about enterprise AI: “One the amount of time and effort it takes to train AI. I
think there's often a misconception that I'm going to turn the machine on, and
it's just going to learn, and suddenly I'm going to have these amazing insights
- but it really does take a lot of work to train AI to be effective.
“I think the second misconception is that
AI is all about replacing humans, because it really is about augmenting humans.
One of the things that I talk a lot about with my clients is [the idea that]
the best AI is the AI that is closely connected and complemented by the human
involvement. If you don't have close human involvement from the very beginning,
that it's really AI for AI sake, versus something that's really going to drive