Optimizing AI Pilots and Employee Input

Use cases for generative artificial intelligence (AI) differ among organizations. To determine those best uses, many organizations engage in pilot programs. But turning pilot programs into successful implementations can be challenging.

It’s critical to set up the pilot or proof-of-concept program with the right KPIs so ensure its potential, Erika Pabo, chief operations and transformation officer at Humana, during a Reuters Events webinar earlier this month. It’s equally critical to set up pilot programs where it makes most sense — and to not saturate the business with them.

“With 400 use cases, how do you prioritize resource investment?” Pabo said during “Realize Your GenAI Ambitions From Proof of Concept to Scale.” She continued, “And how do you operationalize that many pilots and use cases at once?” While organizations cannot realistically conduct 400 pilots at once, they might be able to do 40 if they’re in different pieces of the business, she said.

According to EY’s How Can You Realize The Promise Of Transformational Technologies? report, “The majority of organizations currently investing in emerging technologies are still only at the proof-of-concept or pilot phase of deployment, regardless of technology. AI and automation are the most established, with limited or extensive deployments accounting for 45 percent and 35 percent of the organizations currently investing in these technologies.”

Nearly two-thirds (63 percent) of survey respondents said that it’s challenging to scale “successful emerging technology trials into pervasive organization deployment.” That’s an increase from 46 percent, according to the report.

According to EY research, 90 percent of organizations are utilizing AI technology, said moderator and panelist Samta Kapoor, partner and principal at EY. Generative AI is a significant component, she said, adding that 83 percent anticipate an increase in their gen AI budget. Additionally, studies show that employees are concerned for their jobs in the face of increased AI usage.

How can organizations implement AI without making customers anxious or employees worried about job displacement? What is the optimal time for adoption?

It’s “a challenging reality” that implementation of AI and other emerging technologies will replace work people are doing as well as the roles they have, Pabo said. Therefore, businesses should approach implementation with “thoughtful retraining” and redeployment of those resources to other parts of the business, she said.

It’s also important to ensure your customers are ready for AI implementation and to not move faster than their capabilities, Pabo said. Listen to feedback along the journey, she said. Have a customer-facing element in the testing model, where you learn whether the pilot is working from a business perspective as well as from user adoption user comfort perspective, she said.

“Transparent communication is key to lessen the anxiety that (AI adoption) might create,” said the third panelist, Leo Barella, chief technology officer at Takeda. Additionally, any AI budget should contain funds for training and change management, so employees can learn new capabilities and understand how AI is being used and augments their roles, he said.

Barella recommends, he said, “a feedback mechanism so employees to also share what first of all how the technology is improving their ways of working as well as the anxieties or fears that this technology might be generating.”

Regarding optimal timing, Pabo said she doesn’t think of AI adoption “any differently than I think about a rollout of any large-scale transformation. But teams and organizations can only absorb so much change at a time. You need thoughtful infrastructure around the rollout of change.”

She notes, “This is such a watershed, transformative moment in terms of how much a technology will change how we operate. I liken it to the advent of the Internet or the start of personal computing.”

It’s important that leaders as well as employees understand its implications and how it can aid them in their jobs.

(Image credit: Getty Images/Arkadiusz Wargula)