I’m Betting against Soylent
The other day I got into a fight on Twitter.
It was with Adam Ozimek, co-author of the excellent Modeled Behavior blog. Adam isn’t a guy to disagree with lightly. He’s razor sharp and superlatively well-informed. I was brash on the outside — but I was actually rather frightened on the inside.
If I may say so myself, our dispute offers a useful example of how to resolve many similar disputes: by betting on them.
We disagreed about… Soylent.
What is “Soylent,” you ask? I’ll let Soylent’s website explain:
Soylent™ was developed from a need for a simpler food source. Creator Robert Rhinehart and team developed Soylent after recognizing the disproportionate amount of time and money they spent creating nutritionally complete meals.
Soylent is a food product (classified as a food, not a supplement, by the FDA) designed for use as a staple meal by all adults. Each serving of Soylent provides maximum nutrition with minimum effort.
The FDA deems Soylent a food. Its creators are — a bit weirdly — slow to agree. The media likes to call Soylent a “food replacement,” and if anyone on the Soylent team has objected, I can’t seem to find it. (“What if you never had to worry about food again?” its ad copy asks.)
Switching to Soylent appears to mean drinking a beige-colored, slightly thick, very bland, mildly sweet liquid whenever you feel hungry. And never eating any food.
On the plus side, Soylent appears to be significantly cheaper than actual food. It’s also nutritionally complete, at least as far as its creators can tell. (Ordinary Times‘s own Russell Saunders has opined two times about Soylent, both in the negative, finding the product’s health claims dubious. Like me, he finds the pleasures of food… still pleasurable.)
As most of you know, I am a foodie. I spend a lot of my cognitive surplus thinking about what makes foods appealing. As a bright-eyed Thoreauvian undergrad, I reasoned that the need for nutrition couldn’t be avoided: If nature forces me to do it, I might as well extract as much fun as I possibly can. It’s been a wild ride ever since, and I’ve loved absolutely every moment of my relationship with food.
And, to me at least, Soylent seems like everything that could possibly be wrong about food. Prisoners get bland, totally uniform, flavorless foods. As punishment. And they hate it. Prisoners hate it so much, in fact, that they have alleged in court that bland, totally uniform, flavorless foods are cruel and unusual punishment. I think they’ve got a plausible case.
What about convenience? Granted, Soylent is nothing if not convenient. But when people are under pressure in real life, they don’t go for bland, thin, nutrition-conscious monotony. Oh no: They go for convenient comfort foods: Think cops eating donuts. Or undergrads gorging on pizza while pulling an all-nighter. Or really, just any fast foods anywhere.
Adam was considerably more bullish on Soylent. He thought there might be a market for it — maybe not as a total food replacement, but in place of ramen or McDonald’s, and as a sometimes food. He even argued that regularly consuming Soylent might make “real” foods more pleasurable, in light of diminishing marginal returns on variety. (There may be a follow-up post explaining why I believe this is wrong, but it would be pretty lengthy. If you’re interested, let me know.)
Anyway. Adam and I did what two rational individuals probably should do in such cases: Rather than pound the table, we tried to work out a bet: a testable prediction on the future of Soylent. And, of course, there would be real stakes involved. As Bryan Caplan put it:
Blathering talk surrounds us, but I will take no part in it… When challenged, I will bet on my words, refine them, or recant. When no one is present to challenge me, I will weigh my words and thoughts as if my fellow oath-takers were listening…
I will claim no false certainty; unless I will stake my life on my belief, I am not truly certain – and will admit it.
When I lose a bet, I will admit defeat, pay promptly, and hold my tongue – never protesting that I was “really right.” If I have caveats or reservations, I will declare them when I make the bet – not after I lose it.
When I win a bet, I will not shame my opponent, for a betting loser has far more honor than the mass of men who live by loose and idle talk.
Betting helps clarify whether a person means what they say. It helps winnow out the times that they’re saying it only because they want to believe it, or because they feel good when they say it, or because they look good when they say it. Those times should not be interesting to us.
Bettors take for granted that many pundits do little more than signal their own interior goodness, by whatever definition of goodness they hold. Bettors find such signals very, very unlikely to lead to the truth. Betting helps to clarify the assumptions that underlie our claims, and bettors believe it is deeply unfortunate that wagers are stigmatized rather than applauded in our society. Morgan Warstler deserves credit for suggesting that Adam and I bet, but the two of us readily agreed that this was the right course to take.
In short: By offering a bet, I wasn’t trying to be a dick to Adam; indeed, Adam was exceedingly gracious throughout the private negotiations that took place about the bet. In the end, though, he declined my offer. For which I give him my deepest respect: He did a far nobler thing than one typically finds in the chattering classes.
Here is the text of the bet, which I will now propose to any one other person who might want to take me up on it. (First come, first served, subject to our arbiter’s approval.)
The stakes are $500.
________ will win the bet if, at any time within the next five years, either of the following two conditions is met:
1. Soylent and derivative food substitute products thereof (hereafter “Soylent”; see Definitions, below) records annual total worldwide sales revenue for any year in excess of $600 million, excluding any revenue derived from famine relief.
2. Soylent becomes the self-reported majority source of nutrition, by choice and to the exclusion of medical necessity, for more than 300,000 non-incarcerated people in the United States, as found or estimated by any of the following research agencies or outlets:
Public Policy Polling
Any U.S. government agency (CDC, USDA, etc)
Any article in a peer-reviewed academic journal
Any analysis by a reputable, independent marketing research firm.
For (2), neither of the parties to the bet nor the arbiter is permitted to contribute to the research findings or estimates. (n.b.: Polling agencies named above have all conducted research on Americans’ food choices and were selected for this reason.)
The default assumption is that conditions (1) and (2) have not been met, and it will fall to ________ to demonstrate otherwise.
Jason Kuznicki will win the bet
3. Immediately, if at any time in the next five years, the original Soylent permanently ceases production, and the arbiter agrees with Jason’s assertion that no derivative products can be found on the market.
4. At the end of five years, if neither (1) nor (2) can be demonstrated.
The bet will be off if either the corporation manufacturing the original Soylent or the brand name of “Soylent” is sold or transferred.
Definitions: “Soylent” is defined here as the product currently sold under that name, as well as all derivative products whose adjudicated aim is to become a food substitute. No other currently existing food substitutes, and no derivatives thereof, may count as Soylent for purposes of this definition. Only food substitutes that make use of Soylent’s open-source formula, or modifications thereof, can potentially be called Soylent for purposes of this bet, subject to the arbiter’s further discretion.
Examples: A chewable product composed entirely of the original Soylent’s powder component will count as Soylent, as will a chewable food substitute product composed primarily (> 50%) of Soylent. An onion soup that uses Soylent as a base will not count, because the intent is to be a food rather than to substitute for food. A mixture of Jevity and Soylent will likewise not count, regardless of the proportions involved, because it is a derivative of a currently existing food substitute.
Payment: The bet will be paid in U.S. dollars within two weeks following the arbiter’s decision.
Arbitration: On all matters of dispute, including particularly disputes about what counts as Soylent, we name Adam Gurri as the arbiter, and we agree to respect and abide by his determinations. (n.b.: We can choose another arbiter, if Adam Gurri isn’t interested, or if he doesn’t want to arbitrate for a particular bettor. I do think he’s trustworthy, however, and altogether impartial.)
Effective date: The bet is effective on the date when both parties and the arbiter have agreed to it.
Why did Adam decline? Here are his reasons, from an e-mail he allowed me to share:
I’m reconsidering the specifics of my forecast. I think if someone wants to be optimistic about Soylent, your parameters make for a good bet. I’m just not sure how optimistic I am about it being embraced. If I had to put odds on it right now, I would say I think my side of the bet winning has a non-trivial probability, maybe 20%, but I can’t say it’s 50%+.
I think if Soylent costs fell to the point where it represented much more significant saves versus other low-cost food options, it would have a much brighter future. I think genuinely healthy food substitutes would represent a significant improvement over many of the most popular convenience foods today for many people. But I am not sure they would be embraced. Maybe success comes when some zero/low calorie flavorings get added… But a lot of the possible contingencies would fall outside our bet parameters.
You should definitely feel free to blog about anything we’ve said here. Narrowing disagreement and forcing us to select specific probability estimates and forecasts is what betting is for!
Indeed it is. This was the first time I ever offered such a bet, and I have to say that I learned a lot about it too. (What does “success” mean here? What does “failure” mean? How might my success conditions be foiled? What other possible futures have I neglected? Am I being absolutely fair to Adam? And so on… Needless to say, any argumentative schema that of its own accord forces an arguer to be fairer to his opponent is to be highly recommended!)
Dear readers: If any of you think Soylent has a future, let’s bet. And if you think it has a future but you don’t want to bet… well, why not?