My neuroscience is a little fuzzy, so bear with me. If your wondering how this ties in with world building, it's for a science fiction novel I'm writing, in which there is no magic. I am hoping to create a plausible system of seeing the future.

All right, so let's assume that what we currently know about consciousness and the human brain is true. Our thoughts are simply the result of physical neurons communicating with each other and firing billions of times per second. Much like how the individual bits of a computer come together to make so many calculations.

Assuming this is true, this means that if an event plays out, and your mind makes a decision, it is the same decision it would always make in that exact circumstance. For example, say you are walking down a path, but there is a lion in the middle. The physical workings of your brain tell your body exactly what to do, in this case run.

Now, imagine you could rewind time to the exact moment you made this decision. Note that every single aspect of the situation is exactly the same as before, down to the unpredictable movements of the subatomic particles in the world around you, and in your brain itself. This means, that since your thoughts are merely results of the neuron firings in your brain, you will make the exact same decision as before, down to where the atoms in your feet interact with the atoms of the ground.

So, if what you chose to do is completely dependent on the situation, and the inter workings of your brain, do you really have a choice at all? This makes me think...if I resumed time to around ten minutes ago, I would resume writing what i'm writing right now, down to the letter. With this in mind, what's to say our future isn't predetermined after all?

I tried to narrow this question down as much as possible, and so I hope it isn't too open ended. Please leave your thoughts, for they are all unique...right?

  • 7
    $\begingroup$ You mention "the unpredictable movements of the subatomic particles in the world around you, and in your brain itself" - my impression was that there's no consensus to what degree those are deterministic (but not predictable by current science) vs. truly random. Anyway, if you're interested in larger-scale "seeing the future", you might look up Azimov's Foundation series and its concept of psychohistory, which is basically that - predicting the future with psychology, sociology, and a dash of actuarial science. $\endgroup$ – Cadence May 29 '18 at 4:06
  • 4
    $\begingroup$ Scientifically - if universe is deterministic, then yes. If non-deterministic, then no. $\endgroup$ – Alexander May 29 '18 at 4:34
  • 8
    $\begingroup$ In real life the issue you run into a is a perfect simulation of the universe could not fit inside the universe. $\endgroup$ – John May 29 '18 at 4:34
  • 3
    $\begingroup$ I don't see a "world building" aspect to this question. Perhaps you could edit to make that aspect clear. $\endgroup$ – James K May 29 '18 at 5:35
  • 2
    $\begingroup$ After reading the horrific answers so far, I recommend taking this to the physics s.e. Sadly I personally don't have the time to discuss the details, and it's just a recommendation. Great threads with great answers exit there already btw $\endgroup$ – Raditz_35 May 29 '18 at 7:41

Yes, if the uncertainty principle was not a thing

The uncertainty principle creates an absolute theoretical limit to the precision of certain measurements. We would need to exceed this limit in order to monitor every aspect of the universe simultaneously, at which point we'd have created Laplace's demon:

We may regard the present state of the universe as the effect of its past and the cause of its future. An intellect which at a certain moment would know all forces that set nature in motion, and all positions of all items of which nature is composed, if this intellect were also vast enough to submit these data to analysis, it would embrace in a single formula the movements of the greatest bodies of the universe and those of the tiniest atom; for such an intellect nothing would be uncertain and the future just like the past would be present before its eyes.

This prevents us from measuring the universe with sufficient accuracy to predict the future. That does not mean that the universe is not deterministic, though, just that we cannot measure it.

Quantum non-determinism makes things a little difficult

The universe is on the surface. All prior events have a causal relationship to subsequent events. The existence of quantum non-determinism does not change that, but it does mean that certain events can be described in terms of probability, specifically a probability distribution. There is nothing here but determinism and probability (more specifically, stochasticity), all mathematical and well-defined concepts. We could use this to simulate a universe with some pretty extreme precision, but it would not be perfect. Such a simulation would not answer if the cat is alive or dead, merely the probability of it being alive or dead. So while the universe is not entirely deterministic, it is still subject to causal determinism with a sprinkling of stochasticity.

The universe is functional. Everything can be described mathematically, down to the smallest detail (of course, we cannot yet model it perfectly). Quantum non-determinism does not change that.

Free will, consciousness, and agency are distinct

Everyone loves to argue that the world will end if we do not have free will. If there is no free will, then where do morals go? If there is no free will, then we should just let criminals run the streets! If you think there is no free will, then you don't think Hitler did anything wrong! But the world does not work that way. We still make choices based on input from the environment and our memory, and we are still capable of pleasure and suffering (and a myriad of other states of awareness).

The idea of free will involves the claim that a person is able to make a decision absolutely unimpeded by any of the laws of physics. More specifically, free will involves the physically impossible concept of non-causality. Quantum physics is still physics. As such, using the standard definition of free will, there is no free will. After all, if your action is the result of either a deterministic process, or a non-deterministic process where the only non-deterministic factor is a completely random, uncontrolled probabilistic event with a uniform possibility distribution, where does "choice" come from? You can't take a deterministic system and add a bit of stochasticity and come out with free will. All you'll get is a quasi-deterministic system that's a little harder to predict. However, that does not mean we do not have agency. Agency is defined as the source of our decisions coming from within ourselves, rather than from purely external factors as if we were puppets. It does not mean we violate the laws of physics, and it still means that the universe is deterministic, but it does mean that our control comes from inside us, not outside us.

Consciousness is another concept, where qualia come in. This is far more complicated and involves the hard problem of consciousness. Suffice it to say, consciousness is passive and is even beginning to be described mathematically by Integrated Information Theory, studied by neuroscientists and neurophenomenologists alike.

Our brain is Turing Complete

While our brain is nothing like a classical Von Neumann or Harvard architecture processor, it is still a natural computer and is completely Turing Complete. Every neuron in our brain is influenced by other neurons in an entirely deterministic way. When it comes to quantum non-determinism, the effect is so small that it will not be the tie-breaker for even a single action potential.

  • 2
    $\begingroup$ Outstanding answer. Please accept this. $\endgroup$ – Cloud May 29 '18 at 10:24
  • $\begingroup$ Free will does not mean unmotivated choice. Free will means uncoerced choice. Even if I like apples more than bananas, when I choose an apple over a banana I exercise free will because nothing compelled me to chose the apple. $\endgroup$ – AlexP May 29 '18 at 10:26
  • $\begingroup$ @AlexP No, that is agency. I said in my answer that people often mix the two up. Please don't. $\endgroup$ – forest May 29 '18 at 13:08
  • $\begingroup$ Agency is a manifestation of free will. $\endgroup$ – AlexP May 29 '18 at 13:19
  • $\begingroup$ I'm not sure where you got that idea. Agency is, by definition, uncoerced choice. Free will on the other hand is non-causal, indeterminate choice. $\endgroup$ – forest May 29 '18 at 13:19

Theoretically, yes. Practically, no.

To perfectly calculate the future will require knowing a few things. You must have all state data, including any and all hidden parameters. You must know all transformations. You must have a calculator capable of crunching all data fast enough to provide meaningful answers. You must also have a machine to interpret the answers.

Perfect data must be had because you can't know whether a tiny bit of information is relevant or not. Maybe it's that one bit in the right place that makes things go all nonlinear on you. Missing a transform is problematic for the same reason that missing data is a problem, that one transform might be the one that makes things go nonlinear.

Without good data and good algorithms, you have no hope of perfect calculations. Compound this with the difficulty of hidden variables. With all the billions of brain cells and all their connections, it would require significant computer storage just to store a single, static snapshot of neuron configuration. Storing state variables of all those neurons would be greater still. Number crunching all the chemistry and physics in even a handful of those neurons makes my brain hurt.

The more accurate you want your simulation to be the more expensive it gets to run it.

The fundamental question is "Is the Universe deterministic?" The answer is we don't know. We know that we can make some parts of the universe highly deterministic but those are only small pools in a vast sea. We know Nature has made life processes fairly deterministic after billions of years of trial and error.

We know in the quantum realm that things look very nondeterministic but at our level of experience, plenty of stuff is deterministic.

Brains aren't computers

People like to explain brains like they are computers with well defined inputs, processing and outputs. But we aren't. We are heuristic engines, honed over billions of years to discard irrelevant information and focus on what's important. The neural networks in our brains have activation thresholds, that when crossed will initiate some signal. The neural nets also discard information that doesn't appear relevant but may impact the final "decision".

In your example, you ask if you would be writing this same sentence again. But, what if your threshold of going outside is just about there, all you need is a puff of air on your ear? With that puff you close your laptop and decide its time for a break.

The puff of air is random. You have no way of predicting its arrival nor its effect.... Unless, unless you have modeled the entire world, including your brain.

  • 5
    $\begingroup$ Also worth pointing out that getting all state data is impossible without some handwavium. Heisenberg is pretty clear on that. $\endgroup$ – Joe Bloggs May 29 '18 at 5:51
  • 4
    $\begingroup$ A "heuristic engine" is still a computer. Our brains are equivalent to Turing machines (see: any thought experiment that begins with the word "Chinese"). Our brains do not use the Von Neumann architecture, or the Harvard architecture, but that does not mean they are not natural computers. $\endgroup$ – forest May 29 '18 at 7:29
  • 1
    $\begingroup$ "You must have all state data, including any and all hidden parameters:" you cannot have all state data. Moreover, there are no hidden parameters. $\endgroup$ – AlexP May 29 '18 at 10:19
  • 1
    $\begingroup$ If you were to model the entire universe in a computer, surely you would need to model the computer too (as it is inside the universe). This will clearly exceed its capacity. $\endgroup$ – iyop45 May 29 '18 at 11:08
  • 1
    $\begingroup$ Our brain has only 86 billion (sometimes rounded to 90 billion or exaggerated to a round 100 billion) brain cells, most in our cerebellum. It most certainly does not have "trillions". $\endgroup$ – forest May 29 '18 at 13:39

We're already doing this - to an imperfect degree

The question of whether or not the universe is deterministic has already been covered so I won't step into that argument other than to agree that we don't know the answer to that and won't ever, simply because the universe is too big and complex to be able to resolve that question perfectly.

But; there are parts of the universe that we have been perfectly predicting for some time. We know (for example) not only the exact length of a day, but how much it is changing over time as the earth's rotation slows gradually. We add leap seconds into atomic clocks to keep them in sync with its rotation. We know when the moon is full, when it's new, well into the future. We know what will happen to the sun over the coming 5bn years and what impact that will have on the Earth.

We also can predict things like fraudulent transactions out of a sea of financial data, which emergency service workers are at a high risk of self harm or other consequences of stress. We can predict how a patient will react to a given medication, and we can predict what the world will be like in 5yrs, 50yrs, 500yrs if we keep burning fossil fuels and continuing our population growth. I already do many of these things personally with analytics algorithms and AI.

All these problems are what we call narrow domain problems; that is to say, that they operate in a very specific way that can be segregated from the universe as a 'job lot', meaning that if the rules that govern them and the stateful data that those rules can be applied to are both present, we can know what the future state will be in these cases, with varying degrees of accuracy.

That is to say that;

1) The more discrete the problem is and
2) The better we understand the problem and
3) The more complete our state data is then

the more accurately we can predict the future according to that specific domain. We know that some causes seem to have consistent and specific effects. We would never have invested so much money in the Apollo program if we weren't positive that putting human beings on top of that much fuel wouldn't have finally put them on the moon, for instance. Physics does seem to act in many ways deterministically and for the purposes of such grand endeavours, can be trusted to deliver a specific effect when we apply a specific cause.

Of course, every model is a simplified version of reality in some form. For the most part, we simplify by only modelling those parts that are relevant to the domain of problem we are solving. BUT, we also simplify by subsetting the data to the larger cause/effect pairs, knowing that reduces the accuracy of the prediction.

So; in any prediction, the completeness of our understanding of the rules and the completeness of our understanding of the initial states both impact the accuracy of our prediction. If both are perfect, then the prediction will be perfect but in reality, we know that's not the case. We often make do with 1:1m chance or even lower as being 'perfect' knowing that this will be close enough for our needs.

If I remember correctly, The chances of a conventional SATA HDD not recording a 0 or 1 correctly is somewhere around 1:10^38, making it virtually perfect, at least for our needs.

So, ultimately, we already can and regularly DO predict the future, we just do it within relative subsets of the universe AND with slightly less than perfect predictions.

Now, to get to the universe itself. IF (and it's a big if) the universe is deterministic, that means that there is a set of rules that govern it. If we can know those rules perfectly, then all we'd need is a complete set of state data for the entire universe to model it.

That would look very much like - well, the universe.

In other words, we'd probably need a full universe in order to be able to model the universe perfectly. That is because the universe is already very efficient in recording its internal state, and mimicking that would require something even bigger to store the data on, so another universe would probably be your best bet.

Of course, if you're after less than perfect outcomes, you can use a smaller set of data. As discussed at the beginning, we already do that to predict the orbits of planets, moons, rockets, et al.

So, if the universe is Non-Deterministic, then all the above is invalid. You'll never perfectly simulate the universe because it doesn't conform to any rule set. BUT, as has been already discussed, we have rules already that get pretty close to perfect, so if the universe isn't deterministic, it's probably not off by much.

What does all this mean? Well, it means that the universe is sufficiently deterministic to make almost perfect predictions about parts of it, which is what we do every day.

Ultimately, this follows an inverse square law. Eventually, you're pouring far higher levels of model sophistication and data into ever decreasing improvements in accuracy. We deal with these problems in advanced analytics all the time, and the answer is that you provide the most accurate answer to the widest domain that your data and the understanding of the problem will support.

Usually that's enough. How wide a domain you're seeking to predict will ultimately define just how much data you need, and whether or not it can be managed within the cognitive limits of a human mind or the processing limits of a computer.

  • 1
    $\begingroup$ Asimov’s Foundation series sets up this sort of prediction fairly convincingly with its “psychohistory”. On the mathematical side, chaos theory implies (very roughly) that for some systems (e.g. weather), it’s theoretically impossible to model the system in a way that’s simultaneously efficient and accurate; but there are still plenty of kind of systems which can be modelled well, and there’s no theoretical reason why “human decision-making” shouldn’t turn out to be such a system. $\endgroup$ – Peter LeFanu Lumsdaine May 29 '18 at 10:04

If the computer exists inside the universe, this is impossible.

Even assuming you could give the machine all the data in the universe, to be able to predict the future it would need to predict its own states, too. For this to be useful, though, it would need to be able to compute future states faster than the machine itself can execute. You cannot have an algorithm that computes its own values before it has finished computing those values. It's self-contradictory.

To get around this, your future-predicting device would necessarily have to exist outside of your universe to some extent. Since you don't want magic, you could use a deity or some other supernatural entity to prevent mere mortals from wielding this power indiscriminately. If you don't need exact predictions, maybe your future-predicting device can give you some a prediction that is true 99% of the time (or maybe 75% of the time, to keep things exciting).

One way or another, you'll have to break the loop of the computer needing to predict its own future states to figure out how its predictions (or even its atoms, quantum entanglement/butterfly-effect style) will interact with the rest of the universe.

  • 2
    $\begingroup$ To some degree, our own universe is a future-predicting oracle whose predictions arrive in real-time. $\endgroup$ – ArnoldF May 29 '18 at 8:17
  • 1
    $\begingroup$ There's a solution that allows it to exist in this universe. In fact, I have such a computer! If you want to know the exact diameter of Sagitarius A*, this computer can compute it! The output will be placed 26,000 light years from where we are. A little hard to get to, but hey. EDIT: @ArnoldF Looks like you pointed that out while I was in the middle of writing this comment! $\endgroup$ – forest May 29 '18 at 8:17
  • $\begingroup$ @forest - yes, haha, I thought of that just after posting :) $\endgroup$ – ArnoldF May 29 '18 at 8:18
  • $\begingroup$ Also, I think it's impossible even if the computer exists outside the universe. I imagine there are some fundamental limits to the speed of information transfer that we'd run into pretty quick for any sort of computer, however independent it is from the universe (assuming it is "outside" the universe to prevent it from needing to simulate itself, rather than to handwave and give it infinite abilities). $\endgroup$ – forest May 29 '18 at 8:21
  • $\begingroup$ I agree with you, but to play devil's advocate, if our universe is just a simulation, then it's simulated by some program in a truer universe than our own (and whose laws of physics differ). That program could perhaps be optimized. $\endgroup$ – ArnoldF May 29 '18 at 8:28

What you are asking was a kind of faith some times ago, a faith called determinism:

Determinism is the philosophical theory that all events, including moral choices, are completely determined by previously existing causes. Determinism is usually understood to preclude free will because it entails that humans cannot act otherwise than they do. The theory holds that the universe is utterly rational because complete knowledge of any given situation assures that unerring knowledge of its future is also possible.

If you stay in the realm of Newtonian physics than you have hopes of knowing the future by carefully and thoroughfully measuring the present. However

Since the beginning of the 20th century, quantum mechanics—the physics of the extremely small—has revealed previously concealed aspects of events. Before that, Newtonian physics—the physics of everyday life—dominated. Taken in isolation (rather than as an approximation to quantum mechanics), Newtonian physics depicts a universe in which objects move in perfectly determined ways. At the scale where humans exist and interact with the universe, Newtonian mechanics remain useful, and make relatively accurate predictions (e.g. calculating the trajectory of a bullet). But whereas in theory, absolute knowledge of the forces accelerating a bullet would produce an absolutely accurate prediction of its path, modern quantum mechanics casts reasonable doubt on this main thesis of determinism.

Relevant is the fact that certainty is never absolute in practice (and not just because of David Hume's problem of induction). The equations of Newtonian mechanics can exhibit sensitive dependence on initial conditions. This is an example of the butterfly effect, which is one of the subjects of chaos theory. The idea is that something even as small as a butterfly could cause a chain reaction leading to a hurricane years later. Consequently, even a very small error in knowledge of initial conditions can result in arbitrarily large deviations from predicted behavior. Chaos theory thus explains why it may be practically impossible to predict real life, whether determinism is true or false. On the other hand, the issue may not be so much about human abilities to predict or attain certainty as much as it is the nature of reality itself. For that, a closer, scientific look at nature is necessary.

Also, if you want to see why determinism doesn't work, look at Uranium

All uranium found on earth is thought to have been synthesized during a supernova explosion that occurred roughly 5 billion years ago. Even before the laws of quantum mechanics were developed to their present level, the radioactivity of such elements has posed a challenge to determinism due to its unpredictability. One gram of uranium-238, a commonly occurring radioactive substance, contains some $2.5 \times 10^{21} $ atoms. Each of these atoms are identical and indistinguishable according to all tests known to modern science. Yet about 12600 times a second, one of the atoms in that gram will decay, giving off an alpha particle. The challenge for determinism is to explain why and when decay occurs, since it does not seem to depend on external stimulus. Indeed, no extant theory of physics makes testable predictions of exactly when any given atom will decay. At best scientists can discover determined probabilities in the form of the element's half life.

So, in short, the answer is NO. Maybe only on small 4 dimensional scale.

  • $\begingroup$ Determinism sounds like a faith where you can really mess with the person. "Everything you do is pre-determinted!!", "well Maybe the universe only wants you to think that" $\endgroup$ – Shadowzee May 29 '18 at 5:37
  • 2
    $\begingroup$ This answer is incorrect and misleading. Determinism is not a faith (and I have never heard it described as one by any crowd other than those who call all of science a faith) and is the most commonly accepted explanation in the scientific community. Note that quantum nondeterminism does not invalidate determinism. $\endgroup$ – forest May 29 '18 at 7:26
  • $\begingroup$ @forest, around the end of XIX century there was the faith that, once discovered all the physical laws ruling the universe, there would have been the possibility to fully steer it. Well, it didn't work that way... $\endgroup$ – L.Dutch May 29 '18 at 8:01
  • $\begingroup$ First of all, that is not a faith, that is a hypothesis which is very likely incorrect (simply knowing all the laws of physics may give you more control, but does not make you omnipotent). Second of all, we do not know everything about the universe, so it's silly to say that it "didn't" work that way when the prerequisite was never achieved (even if it is BS). Also, what you are describing is not determinism in the first place. The idea that absolute knowledge gives omnipotence is not the same as the idea that non-causality is impossible. On that note... Vi veri universum vivus vici. $\endgroup$ – forest May 29 '18 at 8:08
  • 1
    $\begingroup$ Regardless, calling something that is still a debate in top scientific journals (specifically whether or not quantum non-determinism is deterministic under the hood, not whether or not it exists) a "faith" is not just an incredibly short sighted use of weasel words, it is downright intellectually dishonest. $\endgroup$ – forest May 29 '18 at 8:19

In principle this is impossible

The Heisenberg uncertainy principle means we cannot in principle take any measurement with perfect accuracy. Chaos theory means a single inaccuracy in our starting data, however small, might cause arbitrarily large inaccuracies in our final prediction.

Note: This does not answer the question of whether the universe is deterministic or not. What it says is that, even if the universe is deterministic, you cannot abuse that fact to make perfect predictions.

Another Note: Chaos theory does not say every small reading error will cause large prediction errors. Stuff like the Shadowing Lemma says in some cases the error will not compound itself. So sometimes Heisenberg is the only restriction.


Einstein & Laplace thought so, and they were nearly-modern great thinkers in maths & physics... but that's not the majority opinion in those fields anymore.

It's still an open question, but since the 2nd half of the 20th century, expert opinion has mostly gone the other way. I think physicists would mostly agree w/ me that it's probably impossible both in theory & practice -- likely not possible even if you could build your predictor out of universe & solve the observer problem W.R.T quantum measurement for a given instant; almost certainly not possible if your prediction machine had to physically exist in or adjacent to the space it predicted.


Yes, an omniscient being would know everything.

Free will is not real, heck it's in reality impossible, like the value of infinity or the power of omnipotence. If you claim you have free will, you're implying that you never had a true beginning. Like a delusional God.

Free will is the notion that comes from our conscious act of choosing between two things. Like how I ate noodles instead of spaghetti. This of course is a choice you make, but what makes you make these choices are nether. "I choose noodles coz I like noodles". Sure dude, but what makes you like noodles? "Ummm... I don't know, I like the taste." Likes... In conclusion motives make action.

Everyone voluntarily action is a result of your involuntarily motives; your likes, or your feelings. All things you certainly don't control. Thus, free will is bogus.

Free will is impossible, coz for there to be free will your motives must have been designed by you as well. But under what basis would you design your character. For anyone to do any motivated acts they need to have a character. So if an earlier you designed the current you who designed the earlier you? An even earlier you? Who designed him them?

This is why free will is fake and impossible. Humans are humans due to the absence of such an obscene theory.

  • $\begingroup$ Free will does not mean unmotivated choice. Free will means uncoerced choice. Even if I like apples more than bananas, when I choose an apple over a banana I exercise free will because nothing compelled me to chose the apple. $\endgroup$ – AlexP May 29 '18 at 10:25
  • $\begingroup$ Sure. Despite your inclination for apples you choose a banana and your deeming that as a free will? The reason for doing such a thing has motive too. It's trying something new motive. Every choice has a motive. Even a random pick from a bowl of coins is a choice chosen by said subtle motives. Doing something you wouldn't do would not prove free will it only adds. Everything a human does has a reason. A random pick too. $\endgroup$ – Hittfler May 29 '18 at 12:57
  • $\begingroup$ The point is that "free will" does not imply "arbitrary unmotivated choice". All it means it that a meaningful choice was possible. In your example, it is irrelevant that you chose the noodles because you liked noodles: all that matters is that you could chose noddles or spaghetti, and the choice was neither compelled nor logically necessary. You chose noodles out of your own free will, because of your own internal reasons. $\endgroup$ – AlexP May 29 '18 at 13:02
  • 1
    $\begingroup$ In a general sense, noone is uncoerced. Everyone is bound by their gender, status, species and other factors. If we could fly like a bird our brains would gives us command to fly. But since we aren't birds, and nor can we literally fly, our actions are 'coerced' to exclude flying if we are rational. The extent of coercing can vary though. Like a slave girl. So is anything done unconstrained by external agencies. I wouldn't call anything free at this point, or anyone accountable for anything. $\endgroup$ – Hittfler May 29 '18 at 13:09
  • 1
    $\begingroup$ In reality, influence does a lot then you say. Only the inherently radical minds end up as Karl Marx. Everyone isn't inherently bold they'd due to their beta characters( that they didn't ask for) end up along the influence. Therefore, influence might as well be a synonym for coercion. $\endgroup$ – Hittfler May 29 '18 at 13:31

Not the answer you're looking for? Browse other questions tagged or ask your own question.