7 Ways AI is Changing the Food Industry TODAY – Notes From the Food AI Summit

The 2024 Food AI Summit on the UC Berkeley campus brought together speakers from AI (artificial intelligence) technology startups, plant-based food manufacturers, food service robotics companies, food developers, nutritionists, makers of AI-driven recipe tool BakeBot, capped by a thought-provoking session with Cal Poly philosophy professors (pondering ethical issues related to AI and automating various aspects of the food industry).

It was an amazing confluence of minds and research.

Discussion panels and demos enlightened us. A cross-section of attendees from around the world and across sectors led to lively discussions.

Our mid-day hummus snack brought to life an AI-conceived appetizer, from Mattson Foods.

7 Ways AI is Changing the Face of Food Product Development and Food Service

1. Developing Flavors and Fragrances with AI: I left the conference obsessed with NotCo. In addition to developing non-animal-based foods, in a partnership with Kraft Heinz, their AI tool Giuseppe has pitted flavor and fragrance developers against the 5,000-industrial ingredient system that can formulate using AI. You guess who won.
2. Large-scale Cafeteria Food: Google’s Global Food Innovation Lead Ennis Olson revealed that Google is serving its own dogfood* — using AI to create the most number of recipes for 245,000 employees to eat well at the lowest possible cost by optimizing the number of ingredients used in those meals. (Like Chopped at a global level.)
3. AI for Food Product Ideation and Persona Development are really interesting applications of AI tools at Mattson — whose client list includes most supermarket aisles. From identifying niches to recipes to the exactly profile of consumers who would love those products, Mattson has embraced AI for every step in the food product R&D process. (They confirmed that testing with humans will always be part of the process.)
4. New Food Product Development based on ingredient profiles is underway at a new Silicon Valley startup Agilitas. They are part data science, part food science and my food scientist colleague Rachel Zemser is involved. So if you are thinking of creating meal bars or other packaged foods, look into Agilitas.
5. Worldwide consumer insights are crunched into the proverbially “actionable” insights in every demographic slice imaginable, by Simulacra Data. The speed at which founder Jason Cohen ran queries and reports was astonishing. (In the old days, we would have a canned (offline) demo, in case the conference center Internet failed, but I assume UC Berkeley as a venue has the fastest Internet imaginable.)
6. Robotics in food service and at home are natural applications of AI, and we heard from a range of solution providers. For restaurants, Curbit is especially interesting to handle the tension between online orders and in-person customers.
7. AI Recipe Generation For Meals and Food Products  – Babette Pepaj conceived of BakeSpace almost 20 years ago as a community for people to swap recipe tips and share recipes. Fast forward several technology trends later, and we have BakeBot.Babette gave a great example of how the chat bot can help, such as asking a baker how exactly they planned to transport a cake they were making for an event. For both home-based food businesses and food entrepreneurs in the early stage of food product ideation, BakeBot could be the answer if you’re stuck.I tried a query to see how to make lentil cubes like the ones someone who presented said he makes at home by pureeing lentils. (**BakeBot’s answer is at the end of the post!)You can try chatting with BakeBot right now.

BakeSpace’s BakeBot helped me figure out how to make lentil cubes like the ones a presenter made.

Are you sick of hearing about AI already?

Elliott Grant of Mineral, Alphabet’s Ag effort, pointed out that eventually AI will be in everything. And just as we don’t say “omg my house has electricity!” the talk of AI will diminish [and AI will talk about it for us].

We ended the conference by gathering at Freehouse at College and Bancroft in Berkeley. It has a lovely outdoor space in front and $5 local beer pints. No AI needed to know that’s a good deal!

If you have a chance to go to The Spoon‘s next Food AI Summit, DO.

Those who couldn’t make it wished they had. Things are sure to be very different at the 2025 Food AI Summit. Follow The Spoon to hear when and where it’s happening.

One of the ways Mattson uses AI tools

NotCo is changing the way flavors and fragrances are developed, using AI.

Giuseppe is the AI engine generating new flavors and fragrances.

*”Eating your own dogfood” is an old-time tech reference to using your own products as a way to prove they are good and to improve through internal feedback. There you learned something!

** Here’s BakeBot’s Lentil Cube recipe and words of encouragement:

Ah, you’re really diving into the cooking labyrinth and wanting to keep it gluten-free! Respect. Let’s make those lentil cubes without any flour or gluten while still making them hold their shape and provide that satisfying chewy texture.

Gluten-Free Lentil Cubes:

Ingredients:

• 2 cups cooked lentils (red or green)
• 1/2 cup mashed sweet potato (or pumpkin puree for a punch of flavor)
• 1/2 teaspoon turmeric (for color and health!)
• 1 teaspoon cumin or smoked paprika (for depth)
• 1/2 teaspoon garlic powder
• Salt and black pepper to taste
• 2 tablespoons nutritional yeast (optional for that cheesy vibe)
• 1 tablespoon chia seeds (mixed with 3 tablespoons of water, let sit for 5 min as a binder)

Directions:

1. Preheat your oven to 375°F (190°C).
2. In a food processor, blend the cooked lentils until smooth. Add mashed sweet potato, turmeric, cumin, garlic powder, salt, pepper, and nutritional yeast. Blend until fully combined with a nice texture—don’t just pulverize; you want a bit of personality.
3. Mix in the chia seed mixture. This will act as your binding agent; it’s a game changer!
4. Grease a baking dish or line it with parchment paper. Spread the lentil mixture evenly and press firmly to eliminate air pockets.
5. Bake for about 30-35 minutes until golden and set. Let it cool slightly before cutting into cubes.

Serving Size: 1 cube (approx.)
Prep Time: 15 minutes
Cook Time: 35 minutes
Calories Per Cube: ~(depends on specific ingredients used, but approx. 60 calories/cube)

These cubes are just waiting to soak up whatever savory sauce you’ve got lying around. Want to kick it up a notch with spices? Maybe play with herbs? I’m all ears! What flavor adventure are you craving? And don’t forget, joining BakeSpace lets you save this brilliance for future culinary escapades—it’s a no-brainer!