How Gen AI Will Accelerate Food Industry Innovation

The dismal statistics around new product rollouts are well known in the food industry. New product failure rates are as high as 90% by some calculations. But what if product developers could dramatically increase the odds of success—by using generative AI (Gen AI)?

Unlike traditional AI (artificial intelligence), which relies on human-programmed rules to produce output based on structured data, Gen AI doesn’t just follow rules. It learns from the data it is trained on, finds patterns, and provides novel insights—often from datasets beyond the ability of the human mind to comprehend. It can supply answers to questions we didn’t even know we had.

Instead of a food technologist working with, say, 20 ingredients to choose from to develop a new product, we can give a Gen AI 20,000 ingredients and tell it to optimize not just for flavor, but also for ingredient sustainability, nutrition, cost, supply chain—the whole package. This should vastly improve the strike rate for successful new product launches.

In the past, large food companies with the most consumer data were the winners of the new product game. They could mine their data for insights that were valid for an extended time period, sometimes several years. It was considered acceptable to develop a new product from ideation to release in up to 18 months using the now outdated Stage-Gate process. Today’s consumer is far more dynamic, and six months is now an eternity for developing and releasing a new product.

The consequence is that, as time passes, the legacy data advantage that large CPG companies enjoyed will shrink to almost nothing because only the last six months or less of consumer data will be relevant or useful.

Another Gen AI–powered advance will be the digitization of flavor. This will progress to the point where it can be modeled accurately for any geographic, demographic, or psychographic cohort. A signal of this impending future is the “cyborg” Koniku computer chip. This cyborg chip comprises genetically engineered mouse neurons on a silicon chip and is capable of smell and taste.

Imagine this connected to a Gen AI platform that can profile the smell and taste of any food product and match this to any consumer profile. No more focus groups. No more huge consumer taste tests. Much faster product development cycles.

In this future, food companies of any size will simply be able to ask their Gen AI to obtain a product formulation optimized for numerous variables from ingredient companies, and seconds later they will receive multiple competitive quotes. They’ll then pay via their own Gen AI and have a product that is—if not 100%—then 90% or more ready to go. How much will this cost? Whole genome sequencing has gone from $100 million in 2001 to potentially $100 in 2025. Why should a product formulation cost any more than this?

As I discussed 12 months ago in Food Technology, consumer food personalization is the future. Key to Gen AI’s role in food or any other space is data. For consumers, the data will be provided by sensors. Currently, most sensors can only capture a limited range of molecules like glucose, lactate, and ketones. In the future, we’ll be able to analyze a much wider range of molecules using everything from biowearables to sensors integrated into clothing, and from tattoos to semi-permanent and permanent implants.

Eventually personalization will take the form of Gen AI–powered digital twins like virtual physiological twins Harvey and Harvetta. These male and female digital twins model 80,000 biochemical reactions, 26 organs, six blood types, and the microbiome to monitor how food affects us—all fed by data from our sensors. This will mean true personalization of our health and nutrition. We’ll each be surrounded by a personal “AI Shell” powered by sensors and quantum computers interacting inwards with our health and nutrition sensors and outwards with the world around us.

All of this barely scratches the surface of how Gen AI will accelerate food innovation. It’s simply the most influential technology of the last 20 years in food—or any other industry.

With the product development and consumer playing fields leveled, will small, agile companies catering to consumers’ personalized needs cause the disintegration of large global food companies? Only time will tell, but one thing is clear: the first signals of this industry and consumer future are already here.ft

The opinions expressed in Dialogue are those of the author.