As a data scientist at one of those firms that likes to use data to sell you things, I can attest that predicting human behavior is indeed very possible and profitable. However, as the very next paragraph states, very tiny errors can accumulate to useless predictions: and the data quality at these large firms is, while in some places spookily accurate, more often than not wrong or at least slightly off. So the predictions are useful, but it’s not too hard to cover your tracks enough to make it a lot harder for them to target you.
But even if we had perfect data about you and infinite computing power, our models wouldn’t be able to predict everything you would do _unless_ the models have as much freedom as the human brain.
That is, in order for such an almanac to be able to predict everything a human might be able to do, it must first have the power to enumerate all the actions that a human might be able to do. If the almanac doesn’t know about eating, for example, it doesn’t matter how much it knows about the rest of the universe: it will never be able to predict the behavior of eating.
And so very quickly you end up with needing to build a model that is _at least_ as complicated as the human mind. It has to be more complicated, of course, if you want it to be able to do more than predict a single human’s behavior.
But, practically, big firms don’t need or care to know everything you are going to do. They just want to know how to get your money (usually by encouraging you to buy things). And for that, models can be quite simple (by comparison to the brain) indeed.
All that being said, to answer the question, yes it is possible to create an almanac that predicts everything a person will do. The universe (the whole thing) is doing it right now. That is, there’s this thing called the past. Now you might say that it’s not fair to call that a prediction since we already know what will happen, and you’d be right, but prediction models are trained and validated on data where the events/outcomes are known. And when it comes to validation: the model doesn’t know that the events are already known. So as far as the “almanac” is concerned, one could hypothetically use the whole of the past as the training data (or even the model itself). The question then is, for events in the future, does such a model underfit (not have enough predictive power) or overfit (predict the past so well that it sucks at predicting the future b/c it is just regurgitating the past). Or, is the arrow of time in this universe just an illusion to those of us stuck in it, therefore the whole universe is actually a constant when viewed from the “outside”, and therefore the universe itself is as good as a mode as you’ll ever need? In any case, none of that gets us closer to something useful that you or I could use.
As for the philosophical idea of free will, I have to say that we like to wrap up free will in the physics of things, but to be perfectly honest, I don’t think that physics holds the answer. We can still use logic to reason about it, but I don’t think science can truly answer the question. Whether the universe is governed by deterministic or non-deterministic rules, at the end of the day, your will is not separate from those rules. Sure, if you believe your “self” is separate from the universe then by all means believe however you wish, but that’s beyond the realm of what we can really discuss in a logical fashion. So why does that matter? Consider that you truly have free will. If you do, then you must be able to decide to do anything you want, whenever you want (your body might limit you, but it’s the deciding in the mind that’s important for this discussion).
Now, that’s all well and good, that you can decide what you want, but _why_ did you decide it? Say you decide to eat a bar of chocolate and that is evidence for your free will. Okay, so why did you decide to eat a bar of chocolate?
If it is because you like chocolate, then, if you truly have free will, you must be able to decide to _not_ like chocolate. You could argue that this is unfair, as your preference is decided by your body, not your mind. But if that is true, then the question has to be asked: how much of your will is governed by parameters internal to your mind, and how much of it is governed by your body? Your mind is, of course, part of your body, so the line cannot be drawn.
Put more abstractly: let’s say you want to change something (it doesn’t matter what that something is). In effect, you are changing your will (you go from a state of accepting something as it is to a state where you are willing yourself to change that something). What motivates that change in state?
If it is your will that motivates the change in state, then what is motivating your will to motivate the change in state? And now we can go in this spiral forever.
And that brings us back to the core of the issue: your mind is governed by the rules of how this universe works. It doesn’t matter that the rules are deterministic or non-deterministic: the rules were not our choice, and neither was our genetics, and we, at best, have only limited control over our nurture (in that whole nature vs nurture debate). So no matter what choices you make, you are still making them as a function of the inner workings of your mind, which are governed by the rules of this universe.
Let’s say you could go back before you were born and change your personality with the objective of arriving at a personality in which you have total free will despite whatever rules of physics are in play (so this is a universe-agnostic thought experiment, so long as said universe is logical).
Well here’s the problem: how would you make that choice of what to change? And more importantly, who would be making that choice? “Well”, you say, “it’s me who is making the change!”
But it’s you as you are today (or sometime in the past, or sometime in the future, it doesn’t matter)! It’s a personality you did not choose that is making the choice! So no matter what change in your own personality you decide to make, that decision is ultimately a function of your current personality. Even if you decide to throw dice and the universe happens to be gracious enough to be nondeterministic, the very decision to throw dice is a result of your current personality. And if you leave the decision up to someone else then you end up right back where you started: with a set of genetics or place in history that you didn’t pick. We can change the objective (or get rid of it altogether) and we are still left with the same problem.
TL;DR: you may have free will, on the surface. But you don’t have control over your will itself, you can’t will the will. Supposing you could, what would you do with it? And who would be doing what with it? It’s like your tongue trying to taste itself, or your eyes trying to see themself, or the universe trying to contain itself. And this is why the underlying physics doesn’t really matter for the problem of free will. Nondeterminism might make it harder, even impossible, to make predictions about what people will do. But that doesn’t mean the people are suddenly able to command the rules of physics within their own minds to be able to make any kind of decision they want. And even if it did, how would such a person make the decisions to change the rules themselves?