- Senior AI leaders from Dell Technologies recently participated in a virtual panel discussion about AI and customer experience.
- While voice assistants and autonomous technology can dominate conversations around AI, decision-making powered by machine intelligence is the vast middle ground that is realistic for the majority of companies.
Companies new to implementing AI should consider focusing their adoption efforts on machine intelligence that can be applied to automating data-driven decision-making, according to a recent panel discussion of technologists about how AI is evolving, sponsored by Dell Technologies.
“If you have a business process — sales, support, services, engineering, anything where there are decisions made based on data — the careful introduction of machine intelligence that changes the precision of those decisions can improve the productivity of your outcomes,” said John Roese, chief technology officer for Dell Technologies. “Honestly, a five percent improvement on Dell’s supply-chain predictive capabilities could be hundreds of millions, if not billions of dollars, of impact.”
Roese estimated that machine-driven decision-making is a massive middle ground where many businesses will play, while two other common AI implementations — human-like features such as voice assistants and full autonomy as seen in self-driving vehicles — will continue to be more rarefied.
The adoption gaps Roese describes were borne out by a special report on AI trends Business Insider published last year. The report cited research showing only 15% of small businesses had implemented AI-powered chatbots while mid-sized companies showed 39% adoption rates, and large companies came in at 67%.
In addition to Roese, the panel discussion “The Real-Time Evolution of AI: Servicing Customers in an Accelerated World,” featured Doug Schmitt, president of services for Dell Technologies and Michael Shepherd, a distinguished engineer in AI research, also from Dell’s services division.
Schmitt said that companies of all sizes can scan their customer-facing systems for opportunities to implement machine intelligence for tasks that are predictive, proactive, and prescriptive.
He defined “predictive” as technology that identifies problems before they occur. A “proactive” application corrects an issue with little to no involvement from the customer, perhaps before they’ve even become aware a potential problem existed.
“‘Prescriptive’ is really leveraging data so we can take action with a customer to improve the environment or the technology and solutions that they’re using,” Schmitt said.
He gave a technology manufacturer’s customer insights about laptop hard drives as an example.
“The predictive side of this would indicate, ‘Hey, we think there’s an issue.’ And so we identify that issue early on, before your hard drive crashes. The proactive side is alerting you the way that you want to be alerted, whether it be email, text, or call, that we want to come out and replace that hard drive. And then the prescriptive is letting you know you’re running out of hard-drive space.”
Describing what he sees as the next wave of predictive customer-support AI, Shepherd alluded to Snapchat-like augmented-reality enhancements appearing on customer’s smartphone screens as they point their cameras toward equipment that needs repair. Only instead of seeing comical eyes or ears augmented the camera view, the user can see a detailed diagram walking them step by step through a repair.
Shepherd said he hopes functions like those can move away from the concept of AI meaning “artificial” intelligence and think of the “A” standing for “augmented.”
“‘Artificial’ conjures up artificial grass, or artificial sweeteners,” he said. “The reality is, AI is not artificial at all. And as we augment the human’s ability to do menial tasks, it actually allows humans to interact with other humans much more.”
The AI panel discussion was part of the new series “The Technologies Keeping Humanity in Business.”