Sara Holoubek (left), CEO of Luminary Labs, moderates a panel about AI with Aman Bhandari, executive director, data science and partnerships, Merck; Stephanie Bova, head of Europe and Canada and head of global operations for Takeda’s digital accelerator; and Michael Li, founder and CEO of the Data Incubator. Photo credit: Erica Berger

The rise in popularity of artificial intelligence in healthcare has been met with with both cheers and pilot programs, but pharmaceutical executives caution against over-promising about the technology’s potential

In recent years, drugmakers have looked at mobile apps as the end-all solution to their marketing challenges. “It’s the vehicle to do something,” said Stephanie Bova, head of Europe and Canada and head of global operations for Takeda’s digital accelerator, at the MM&M Transforming Healthcare conference on Monday in New York City. “Your app is not going to be the solution.”

See also: Pharma faces its ‘oh, shit’ moment

The same holds for AI, Bova added. Marketers need to ask what job needs to be completed and what problems need to be solved, and only then decide whether AI, machine-learning, or another new technology is the right fit. “We disregard the basics because of the shiny new toy,” she noted.

In Japan, Takeda uses an AI concierge service for its field force, designed to help sales reps who often market between 10 and 12 products. That’s a far cry from the the once-vaunted promise that AI will eventually replace doctors.

The other challenge for drugmakers lies in evolving their traditional corporate model into one that is data-driven and supported by data scientists. Michael Li, founder and CEO of the Data Incubator, said the same thing about data scientists that healthcare agency leaders often say about millennial-age talent: They want to work for companies that do good work. “Emphasize the core mission of the company,” he suggested.

See also: Pharma adopts data-science culture in move toward AI

Still, AI remains a tough sell, mostly because the role of technology in corporate healthcare is in its early days. Aman Bhandari, executive director of data science and partnerships for Merck, referred to AI first as a “black box” and second as the “Tower of Babel.”

Merck, like Takeda, is also working with AI, most recently alongside Amazon on a developer program that aims to improve care for people who have been diagnosed with Type 2 diabetes.