#47: Industrial unpredictability
The dream food of the future.
I’m Vaughn Tan; this is another of my weekly attempts to make sense of the state of not-knowing. The first issue explains what the newsletter is about; you can see all the issues here.
My first book, The Uncertainty Mindset is a behind-the-scenes look at cutting-edge high-end cuisine … and what it can teach us about designing organizations to be more adaptable and innovative. You can get it here. If you like it, help me out by leaving a review somewhere. Book events are coming soon—tell me what you’d like to see and sign up for notifications here.
Yesterday, I got interviewed (on the phone, of course) by the author of a new book about the future of food that’s been twice delayed by the pandemic. Her last question was: “What future food has not yet been dreamed up?” For me, the answer is: industrial food produced on a large scale that is variable and unpredictable but in ways that surprise and delight.
There’s a lot to unpack in that statement.
The advantages of industrial food—food produced at large scale instead of small scale—are similar to the advantages of industrial anything: it seems more efficient to produce and is more consistent in output. Sameness and predictability come with scaling up production because diversity, exceptionality and unpredictability are not naturally consistent with large scale.
To see why this is the case, consider this pair of scenarios for growing ten thousand tons of corn:
Scenario 1 (relatively easier): All the corn sown is of the same variety, and it is all sown at the same time on a large expanse of land of the same soil composition, elevation, aspect, and climate. The sameness of the corn, the land, and the climate makes it more likely that all the corn sowed will germinate at the same time, go through its various developmental stages at the same rate, and be harvestable together. It is relatively easier to grow so much corn in this scenario because the processes required across the whole production can be managed in basically the same way.
Scenario 2 (relatively harder): The land is identical to Scenario 1 (uniform soil composition, elevation, aspect, climate). However ten different types of corn with the same yield are grown (vs just one type of corn). Each type of corn has a different ideal sowing and harvesting time. To produce the same amount of corn, the farmer will have to sow seed at ten different times (vs just one) and harvest at ten different times (vs just one).
In a world of limited resources—which include thought and attention—increasing the scale of production practically mandates increasing the sameness of what is produced and how it is produced.
Increasing the scale of production by increasing sameness also practically mandates increasing the predictability of what is produced. In Scenario 2, if the ten types of corn planted are each differently susceptible to pests and diseases or respond differently to weather patterns, unpredictable events (such as unexpected pests, disease, or weather) might affect some of the corn types but aren’t likely to affect all ten types of corn. In contrast, planting just one type of corn as in Scenario 1 means that the entire crop might be wiped out by unpredictable events which affect that specific type of corn.
Low diversity (= sameness) thus reduces a system’s resilience to uncertainty. This drives the system to invest in predictability, in the specific sense of designing itself around minimizing perceived impacts of unexpected variation in conditions.
So the single type of corn chosen in Scenario 1 will be one that the grower believes will be least susceptible to uncertainty. This is another way of saying that the type of corn chosen will be the one the grower believes to be most suitable to being protected from uncertainty by her farming decisions. She might not consider types of corn that are exceptionally tender and sweet if they are known to be exceptionally attractive to pests or diseases (as they likely are). She might not consider types of corn that ripen later and have more developed flavor because inclement late season weather might mess up the harvest (as it probably will). And so on.
Industrial food is virtuous in continually becoming more uniform and consistent with customer expectation because it becomes more same-y and predictable the larger the scale of production. These days, industrial food is very large scale indeed, and so what you expect from industrial food is usually what you get. This is why Starbucks, McDonalds, and Coca Cola are valuable brands.
But just because sameness and predictability have some value doesn’t mean that variability and unpredictability are necessarily bad.
Things can be different and unpredictable in ways that shock and displease (e.g., I wasn’t expecting to have to teach online for the next year because of coronavirus) but also in ways that surprise and delight (e.g., this new sculpture by Robert Irwin is unlike any previous piece of art by him and yet is thrillingly recognizable as his work).
We seem to have mostly forgotten this, but the tradeoffs we make to achieve predictability in industrial food make this especially clear. In an episode of How I Met Your Mother that I otherwise only vaguely recall, Marshall takes a bite of something and, with an expression of pleasure so great it qualifies as an excess of emotion, says to the camera: “I didn’t know food could taste this good.”
Industrial food, produced for scale and thus benefiting from and afflicted with sameness and predictability, never has the possibility of surprising and delighting the consumer in this way. Artisanal food, which is not produced for scale and thus needs neither sameness nor predictability, has at least the potential to surprise and delight.
However, creating surprising and delightfully unpredictable food is difficult. It can sometimes be the result of happy accident but is more often the product of native taste and skill combined with consistent attention to detail over time:
For [making wine] an apprenticeship, not scholarship, is required, which the Sculptor served during the years we spent in the vineyards above Carrara. Everyone agrees that the magnificent wine he makes is better than the neighbours’. There are many reasons for this, one is that nothing is put in it, medicaments I mean. An artist who has taken risks all his life accepts the risk of his wine “going off” and only takes the more care of his barrels. The agriculturalist spoils his wine by “making sure.” (from Patience Gray’s Honey from a Weed—one of the best cookbooks of the 20th century.
Taste and skill guided by experience are prerequisites for producing food that navigates uncertainty successfully. This is another way of saying that only those who embody the fallibility and subjectivity of being human (and who accept uncertainty in outcomes) are capable of producing creating things that can surprise and delight. This is why it is so rare to find food—really, any product or service—that has this quality.
So the Grail of future food is industrial food, produced at scale, which nonetheless embodies the kind of uncertainty that generates surprise and delight—and which until now has been rare. It isn’t even conceivable at the moment because it will probably require creating what amounts to generalised artificial intelligence, something which can develop an idea of subjective taste, acquire skill and experience, then use all three in the creation of something that necessarily cannot be fully defined in advance.
Bonuses:
Tomorrow (September 24), Bjorn Lee and I will be talking about innovation through uncertainty, thanks to Lam Yishan at Project Feed. You can RSVP for the session here.
Laetitia Vitaud wrote a very generous review of The Uncertainty Mindset (the book) …
Which is also mentioned in European Straits (subscribers only) …
In Tom Critchlow’s post on independent economics …
And in another issue of Sid Jha’s Sunday Snapshots.
Also, a team at UPenn’s Graduate School of Education has independently developed and released Planning for Uncertainty: An Educator’s Guide to Navigating the COVID-19 Era. This guide embodies many of the principles in The Uncertainty Mindset and applies them to the education context; we’re hoping to collaborate on some projects next.
Photos: Until Issue #52, I’ll be posting a few photos each week selected from the thousands I took during fieldwork for the book.
An intact, raw lamb brain curing in juniper wood infusion at Amaja (2011).
(Very) early days in developing a menu for a new restaurant at ThinkFoodGroup (2011).
Injecting uncertainty by design into the conference program for MAD5 (2016).
By the way: This newsletter is hard to categorize and probably not for everyone. But if you know people who might enjoy it, please send it to them.
You can find me on the web at www.vaughntan.org, on Twitter @vaughn_tan, on Instagram @vaughn.tan, or by email at <uncertaintymindset@vaughntan.org>. You can also find out more about my book atwww.uncertaintymindset.org.