VC Marc Andreessen famously has said “software is eating the world” – well it turns out the hottest trend in software these days is in fact eating – more specifically, food production. California generates percentage of 50% of the country’s venture funds, and also produces 50% of the country’s food.A growing percentage of that VC money is going into ways to optimize food production – some say the next Green Revolution is in fact a Data Revolution.
Here at Orange Silicon Valley, we are tracking the powerful convergence of changing consumer appetites – for cleaner, more sustainable food – with the maturation of powerful relevant tech like computer vision, IoT, and machine learning. Besides the obvious point that most of the tech is here, keep in mind that Silicon Valley likes to feed its workers for free as a way to keep them productive – employers like Google, Twitter, and Facebook are key customers for sustainably produced “industrial’ food suppliers from distributors like Bon Appetit Management. The result is a powerful (and hopefully) virtuous cycle of innovation and new forms of consumption, although a recent article from the NY Times suggests it’s killing the restaurant business.
But here’s the thing, beyond the pronouncements of entrepreneurs like Elon Musk’s brother Kimbal Musk about “empowering 1,000’s of millennials to become #realfood entrepreneurs” – the thing is that growing food is complicated enough already. Not to pick on Kimbal Musk – he’s built an impressive network of retail food purveyors selling locally-sourced food called the Kitchen – but let’s not conflate “year-round heaven for local foodies” with addressing structural issues relating to water scarcity, labor shortages, and a fragmented supply chain.
Based on our hundreds of hours of conversations with growers, agronomists, and supply chain participants here in the “salad bowl of America,” we see the growing gap between the tech-topian engineering view of quantitative yield management with the pragmatic tacit knowledge of growers rooted in generations of soil-based farming. One thinks in terms numbers, the other thinks in terms of seasons.
For the tech-topians, hydroponic farming in controlled environments (so-called CEA), is expressed in comparing the pounds of product grown in a shipping container full of hydroponic racks to an acre of farmland – guess what, the shipping containers numbers are much better. Other metrics involve a denominator of electricity used for lighting, water pumps, etc for food production: an entrepreneur we met boasts a 250 grams/kWH of yield versus hydroponic yields of 55 grams/kWH. You always wanted to know precisely how much electricity went into that head of lettuce right?
We are hard at work on the topic of Controlled Environment Agriculture, more specifically the application of machine intelligence to growing food indoors. Here’s the thing, because indoor growing at scale means shorter growing cycles (ideal lighting and climate 24/7 makes plants grow faster), and growing 365 days a year, the asymmetry in yields is crazy: a new report from Agrilyst shows yields from indoor controlled environment farming on the order of 10X more output per acre.
There’s a lot more numbers coming that will need critical inspection. The smart folks at LabelInsight have released a report asserting that 97% of consumers indicated transparency was very or somewhat important in purchasing dairy products (here’s the breakdown by category, thanks to Sustainable Brands), but now we have to understand what “transparency metrics” really means.
It’s easy to see where this is going. We saw what happened to traditional media: readers were replaced by “unique visitors”, viewers were replaced by “streams,” impressions by “engagement”. Today, do any of us really understand online advertising analytics? How do you feel about that happening to your food? Whether we like it or not, the disciplines of optimization and predictive analytics are coming to the food chain — the robots are about to put food on your table.