# How could someone calculate all potential outcomes of an action?

In my universe, some individuals possess the capacity to assess all outcomes for an action. As you can imagine, this is very useful.

The only way I can think of a plausible way to make that happen is to have my individual calculate probabilities extremely quickly and react to the one which is the most likely to occur.

What would be the least magical way of making that happen, and still make that capacity quick enough to reach a decision in less than half a second?

For the purpose of the question, the tech level is high science fiction, but I want all "strange" capacities to be grounded in something at least marginally plausible.

Edit thanks to @Alexander : I am searching for something a single individual can know. If you have a great answer including 'omniscient' knowledge, you can still try, but this is not the outcome I'm searching for.

• Welcome to Worldbuiding.SE, Morgan. I edited your question to fix some grammatical issues and to make it clearer, including the title. If I changed anything in a way where it doesn't mean what you meant, please go ahead and edit it. I added the [mathematics] tag because of the issue of calculating probabilities, but I'm not sure about it. I thought about adding [artificial-intelligence] but I didn't think that's what you were after. Feb 19 '19 at 15:38
• Recommended reading: Wikipedia's article for Laplace's Demon, especially the "arguments against" section. There are multiple reasons why it's impossible to 1) Know all the relevant details for a prediction and 2) calculate a reliable, long term prediction even if all the details where known (see also butterfly effect for how small details can have very significant effect on the outcome). (continued) Feb 19 '19 at 16:01
• (continued) I'm not saying you can not use this in a sci-fi context, but it may help to understand why this is considered impossible from a scientific perspective. Also, some of the terminology ("Quantum mechanical indeterminacy", "Laplace's Demon" etc. etc.) might be useful for a more scientifically sounding handwave ;) Feb 19 '19 at 16:07
• "all outcomes for an action" is terribly vague. How far into the future? Think of the butterfly effect. A tiny action by a very powerful person could change the course of world history for thousands of years. Does your savant know all possible future human histories? Feb 20 '19 at 10:26
• In the Mistborn novels by Brandon Sanderson some characters are able to use Atium(spoilers) to gain almost exactly the ability you describe. Granted the explanation there is essentially 'because magic', but this might still be a good ressource/inspiration for you. Feb 20 '19 at 13:14

# They don't calculate, they perceive

Rather than calculating all possible outcomes, which can be problematic*, consider this instead:

Your gifted individuals can somehow see/sense into the future - but not just "our real future", but many possible ones (maybe even every possible one).

You can use the "Many worlds interpretation" of quantum mechanics to handwave this ability. That interpretation treats any action as actually having all possible outcomes - each in its own different "parallel world". An individual capable of visioning these many worlds can navigate to the most suitable one by choosing his actions accordingly.

This can also help you set various limitations to this super-power: maybe they can't see that far or maybe reviewing the options requires time and concentration, forcing them to sometime act based on less than perfect and complete information. Finally, when you use the "sensing" concept, you can block or interfere with it e.g. power doesn't work next to handwavium wave generators or is foggy/blinding in highly diversifying circumstances (such as a casino, a hailstorm or when being fired at by a machine-gun).

*: As mentioned in my comments to the question, Laplace's Demon - essentially the theoretical notion of calculating every outcome for an action - has multiple arguments refuting it. There are multiple reasons why it's impossible to:

1. Know all the relevant details for a prediction
2. Calculate a reliable, long term prediction even if all the details where known

You'll also get some strange paradoxes if there are several such gifted human calculators - they'll need to model the other gifted humans modeling the others modeling the others. That's a bit like asking what is the "final" reflection in two mirrors reflecting each other.

Using the "sensing" concept rather than the "calculation" one does not have these limitations.

• The paradox of multiple "gifted" humans is very interesting. It could open a world of things to try, and help flesh out the world ! Feb 20 '19 at 7:59
• The method of perceiving the future is also used by in Mistborn novels by Brandon Sanderson. Some characters there are able to use Atium(spoilers) to shortly see into the future - if they encounter each other the abilities essentially cancel each other out. (see also my comment under the question). Feb 20 '19 at 13:16

The brain of these individuals is able to carry out some sort of quantum annealing

Quantum annealing (QA) is a metaheuristic for finding the global minimum of a given objective function over a given set of candidate solutions (candidate states), by a process using quantum fluctuations. Quantum annealing is used mainly for problems where the search space is discrete (combinatorial optimization problems) with many local minima

Think of the process of finding the lowest place in a large landscape. You can walk the landscape, measuring the height of each spot, then finding the minimum. That would take a lot of time.

If you instead flood the entire landscape and slowly remove the water, you will be able to find the lowest spot by observing where the water remains the longer. In this way you won't need to walk the entire landscape point by point, thus the process will be faster.

• Nit-pick: the Wikipedia quote does states that QA is "used mainly for problems where the search space is discrete" - real-life situations won't be easily modeled into discrete spaces... That said, anything using the term "quantum" is automatically high-quality handwavium to explain sci-fi non-magical powers... Feb 19 '19 at 15:11

Instinct

Your character doesn't manually process every possible combination of events. They simply get a feeling about whats the best way to react. Their body naturally processes all the signals and spits out the best result using instinct to guide your character.

A famous example would be Spider mans Spidey Sense. You don't need to think or calculate it. Your body does its thing and you just need to follow it.

• Which is pretty much how most human decisions are made, no matter how much we try to convince ourselves we are being logical. Note that we get better at this with more and more experience - so it may be that the people with the near superpower are unusual in that they are both extremely long-lived and can keep learning throughout their long lives. Feb 20 '19 at 9:35

A flavor of autistic savants.

I am autistic myself and have a savant skill called "pattern recognition"

Your people could have a form of pattern recognition enhanced beyond that of a savant, so that your individuals would be able to analyze patterns and predict outcomes.

A limitation you could place on this would be that the person doing the predicting would need knowledge of similar situations. The more knowledge, the better the skill at prediction.

Your specially trained analysts would need to have in depth studies of past events in their particular discipline, be it finance, military strategy, or what have you.

Computer scientists like to break machines down into finite state machines, so they know all the possible states of a current machine. By doing so, they know exactly what the machine is doing, what it can do, how it will work, where it can move onto, everything. Basically, someone who saw that diagram could know everything about that machine.

You could say these people have an uncanny ability to see all the states and the states leading on from it, they have a great ability of foresight to see and analyze all the outcomes of one particular variation. That is for their current situation.

Chess algorithms do a similar thing but assign a value to each outcome and decide which is the best path to follow through with. They constantly check after each move to see if there is a new 'better' path to occur based on what their opponent played.

Basically, quick processing speed and the ability to adapt will allow people to basically 'consider' all outcomes and decide which to choose.

We all already do this in fact, we're just not incredibly effective at it, however even in the modern day, some people are wired up differently and can do stuff like this to an extreme level of efficiency. Bobby Fischer was a world chess champion and I believe was Autistic. He could do incredible things at the time and analyze so many outcomes in a few seconds. However, that came with practice.

You could tie in a natural ability for certain humans to have this feat and the ability to quickly learn, allowing them to become a master of anything with half the effort. Hence they could become incredible at analyzing things in that relevant field in a blink of an eye. Sherlock Holmes is another example if you're looking for people who can do what you suggest.

Regardless of the mechanic you use to explain a character's prescience, you will likely need to account for any actions that require or inspire the (re)actions of others. Once other individuals become involved, it is impossible to know with certainty how they will react for they bring their own set of perceptions and motivations to the interaction. If the characters are well-known to each other one might be able to infer what the other will do in a particular situation. Overall, I think the Pattern Recognizing savant is the best way to go. It gives you lots of wiggle room for developing the circumstances in which it works and how it is developed.

The only way I can think of a plausible way to make that happen is to have my individual calculate probabilities extremely quickly and react to the one which is the most likely to occur.

But how?

By thinking like a programmer! If you need to find a number in a billon, this will take a lot... except if all the numbers are SORTED and then you can use binary search. Consider how fast google answer for a question. In short, the keys are:

• Sorted data, when possible
• Specific data structures that allow to answer quickly for search (for example a HashMap provide a "instant" (O(1)) solution for search for a single value in the key position.
• Use of the appropriate algorithm for the above, like binary search.

So, in first place, you DON'T want to "calculate all outcomes for an action". Nope, you first step is to cut down, aggressively, the search space. Then you need to build a "data structure" with things for search.

A example for real use in humans: A memory palace.

After build a data structure, you apply an algorithm on it. Take in account you can optimize for "speed" or "for space", ie: If need to get a answer as fast as possible, to risk to have bad results (that is what google do). However, if your human can have some time, it can optimize for space, and increase the range of possibilities (aka: Data and accuracy) to yield a more accurate results.

Also, take in consideration that experience & training is already a real-world way to manifest this on humans, "calculate posibilites" is part of what brain do. Plus, heuristics, like for example a Fermi estimate.