Comment by yazaddaruvala | Hacker News Reader

Comment by yazaddaruvala | Hacker News Reader

Parent Full thread

Note: I'm not the parent but like them I like, and consume Soylent and I agree that once a day is probably a good rule of thumb. I'll try and explain the parent's perspective in my words.

Soylent is deterministic. In-fact Soylent's goal is to be deterministic and as such will only become more deterministic as time progresses. Any two Soylent's will be the same.

Traditional food on the other hand, especially through variety, is by its nature non-deterministic. Example: You happen to be a patron of an icecream parlor. Today you buy/eat mango icecream, unknown to the parlor the mangoes happened to be from the east coast of Mexico. Tomorrow you buy/eat "the same" mango icecream, but these mangoes happened to be sourced from the west coast of Mexico. Was it the same? This example, doesn't even account for the variety involved with "traditional food", which only increases the stochastic nature of its nutritional properties.

Having defined that (hopefully we agree on the axioms), we can continue on to discuss the nature of problems where "we don't know [what] we need". Until we know what we need, the optimal strategy to satisfy an unknown goal is through non-determinism (again, hopefully we agree).

Based on the above:

I'll try to make a more succinct argument:

Given that Soylent contains everything we know we need.

And we know we don't know everything we need.

Then only consuming Soylent guarantees we do not satisfy all of our needs.

We must then consume things we do not know, to increase the probability of satisfying all our needs.

P.S. I am all for scientific experiments to better determine the unknowns and turn them into knowns. However, as long as there are any unknowns, the above points still hold.

I don't see why the larger variation in traditional diets guarantees a higher chance of providing some nutrients that we need but that we don't know we need.

Sorry, I guess I'm not able to explain it all that well.

I'll try one last time with an analogy.

Lets play a game: I will think of two numbers (range: 1..10) and you're going to guess which numbers they are.

I'm going to do you a big favor and guarantee you: I'm going to always pick 7. I will also guarantee you, the other number will never be 7.

To reiterate the givens: My numbers will be in the domains: [7] AND [1,2,3,4,5,6,8,9]

You get to pick two numbers. Would you pick 7 AND 7?

I hope not. 7 AND 7 by definition will always lose (probability of winning: 0%). Introducing variety for the second number will clearly increase your probability of winning to 1/8 = 12.5%.

Granted, the probability of winning is still low, and 12.5% assumes a random distribution, but its higher than 0%.

Now, similarly, lets a play a game: I am going to think of all of the nutrition that you need and you're allowed to guess two meals that will encompass all of that nutrition.

I will do you a favor and guarantee you: Soylent will get all of the food that we know we need (if you pick Soylent as one of the meals you will only need a tiny set of additional nutrition to win). I will also then guarantee you, we need things that we don't know, therefore since Soylent only has the things we know, we need at least one nutritional "thing" not in Soylent.

To reiterate the givens: Meals will be in the domains: [Soylent] AND [...everything except Soylent]

You get to chose two meals. Would you pick Soylent AND Soylent?

I hope not. Soylent AND Soylent by definition (in this game) will always lose (probability of winning: 0%). Introducing variety for the second meal will clearly increase your probability of winning. The increase in probability is not calculable, however it is non-zero.

Granted, the probability of winning is still very low, however, variety provides a non-zero probability while only-Soylent provides 0%. Clearly then, variety is the better option over only-Soylent.

Does that at all help? Am I just being trolled?

Either way, this was fun. glhf, and thanks for reading.

The problem with your analogy is that a traditional diet might be missing nutrients we know that we need (most people don't track very closely) and might also contain things that are actively harmful which Soylent does not contain. Lots of people eat traditional diets and have lots of nutrition problems.

To make the issue with your analogy clearer, note that you could invent any somewhat edible product and make the claim that adding it to your diet makes your diet safer, because of the possibility that it contains nutrients we need but which we don't know we need.

It's a simple matter of inability to prove a negative.

Reply on news.ycombinator.com