5 Takeaways on Putting Generative AI To Work in Food Retail

The FMI Midwinter Executive Conference keynote session "Putting Generative AI to Work for Food Retail" brought industry leaders together to discuss the transformative impact of artificial intelligence in the grocery retail industry. Sponsored by Invafresh, Doug Baker, vice president, industry relations at FMI, led a conversation with Jose Luis-Gomes, managing director retail & consumer, at Google Cloud, addressing the current state of AI applications, future possibilities, and the challenges and responsibilities associated with AI adoption. 

The following are five takeaways from the session:

Purposeful AI Implementation: Joe Smirlies, senior vice president, product management at Invafresh, kicked off the session by emphasizing the importance of harnessing AI's power for purposeful change. The focus should be on delivering meaningful transformation that benefits businesses and society as a whole.

"While it really is the epitome of a transformative technology, the real question is, how do we harness its power?" asked Smirlies. “More importantly, how do we harness this technology to deliver purposeful change and use it in a pragmatic way that helps businesses and, more broadly, society, be better?"

Challenges and Responsibilities in AI Adoption: Luis-Gomes emphasized the need for a rigorous framework to evaluate AI initiatives, ensuring ethical considerations, and mitigating risks.

"Every single initiative within Google that has AI as a component goes through a very rigorous framework,” said Luis-Gomes. “We actually call it Lemonade With Google Cloud, and that means that any initiative, whether it's an internal Google initiative or whether it's something our customer has asked us to do, goes through this framework and goes through an evaluation of an independent committee.”

Regulation and Fast-Paced Innovation: Baker and Luis-Gomes touched on the multitude of ongoing discussions about AI regulation globally. As Baker noted, “this train is going extremely quickly.” Gomes acknowledged the increasing talks about regulation and the need for a balance between continued innovation and addressing potential risks.

"I think that regulation is going to catch up if it isn't behind,” said Luis-Gomes. “I just think that if you think about where we were a year ago and where the technology has moved right now, we're all in a little bit of a catch-up mode."

The Grocery Industry's Progress in AI Adoption: Luis-Gomes expressed optimism about the grocery industry's adoption of AI, noting its data richness as a foundation for applying AI at scale. He acknowledged the challenges, such as the need for engineering talent, but highlighted the industry's quick adoption of technology, likely because the grocery industry is less regulated than healthcare or financial services for example.

"Compared to many others, we actually have a phenomenal foundation to really apply AI at scale, so I'm really bullish about [the grocery industry’s] ability to do that.”

Data's Role in AI: Baker asked about the need for clean, accurate, unified data to effectively power AI. Luis-Gomes acknowledged the importance of good data but also highlighted AI's ability to create data and presented an example of using AI to enhance product attributes. He noted the company worked with one retailer with over a hundred million products, and the company built out 1.2 billion attributes for them, taking the average attribute for every product from about 1.5 to around 12. This enabled the retailer to drive much better personalization, both in search and recommendation. 

“Good data is super important,” said Luis-Gomes. “The great news is, and I'm probably going to use this catchphrase, is that AI can now also create the data for you. And that's a new paradigm.”

This was originally published on RIS News, a CGT sister publication.