I'm compelled to act rationally. How is that a bad thing for science?

The fixed will of a scientist all the more can consistently find connections and truth; the will doesn't just get skipped over. If there's bias and it gets in the way, one can learn from this omission and thus a newer, wider fixed will can attend to more complete solutions. The fixed will is dynamic, not frozen for all time.

Here's the argument:
1. The ability to make choices is a necessary condition for the evaluation of evidence.
2. Evaluating evidence is a necessary condition for science.
3. Without free will there is no ability to make choices.
4. Without the ability to make choices, evaluation of evidence is impossible.
5. If evaluation of evidence is impossible, science is impossible.
6. There is no free will.
7. Therefore, science is impossible.

Free will = making choices is not a standard definition of free will used in arguments about free will. I happen to define it that way myself but what the great (?) philosophers mean is that the choices are not made consciously like it seems when you make them, thats is an illusion because (in a nutshell) determinism. The thought/choice is being made subconsciously, and you are just aware of that choice and as you become aware it feels as though somewhere in your conscious mind a decision is being made but its not, the choice was made by deterministic factors before you could even know about the choice at all.
So your argument only applies to your own idiosyncratic definition of free will, which you sorta defined but didnt offer any support for that definition.

Minimally, to say that an agent has free will is to say that the agent has the capacity to choose his or her course of action.
https://www.iep.utm.edu/freewill/

Right. The way you are using it in your OP doesnt match that. You are acting as though the lack of free will means no action can take place. The action still occurs, its just that YOU aren’t the one deciding what to do, thats happening as a result of (basically) determinism. The feeling that you are making a conscious decision is an illusion, the action itself still happens.
In the portion you quoted, the capacity to choose isnt the same as free will as you are using it in your argument.

"You are acting as though the lack of free will means no action can take place."

No. In a deterministic universe, action still takes place.

Right so the “necessary condition for doing science” is an action. Choosing/evaluating is an action, it is something that you are doing.
As you just said, the action still takes place. Free will doesnt determine whether it does or not. In order for your argument to work it would have to. You have to adjust your argument so it addresses free will, not the act itself. In order to do that, you need to offer support for defining free will as the act, which as I said I agree seems a more sensible way of defining it.

How would a person who conclusions are utterly compelled to seem rational, but might not be, be able to assess this?

Right so the “necessary condition for doing science” is an action. Choosing/evaluating is an action, it is something that you are doing.
As you just said, the action still takes place. Free will doesnt determine whether it does or not. In order for your argument to work it would have to. You have to adjust your argument so it addresses free will, not the act itself. In order to do that, you need to offer support for defining free will as the act, which as I said I agree seems a more sensible way of defining it. — DingoJones

I think this is right and that the issue gets very complicated since we have no model for free will. I do think the ability to assess rationality is problematic once one of one's axioms is determinism. IOW if one's evaluations are utterly determined they may not be based on what we think they are based on. We would also be compelled to think we are rational, though not necessarily at all because we are rational and because of what we think is evidence of it.

How do you do (science) without free will? — RogueAI

Easy. All you gotta do is figure out that you don’t have to will yourself to be a scientist. If you do science properly, which presupposes you know how, you are automatically a proper scientist. No willing required.

Does anyone really think a good scientist has to will himself to perform the right experiment in accordance with a prediction, rather than automatically recognize a certain experiment to perform based on the empirical necessity that a prediction demands?

Physical science divorced itself from subjective predicates such as free will for a reason, with humanity in general being the beneficiary.

First, note that I buy that we have free will. So in the comment below, I'm not arguing that we do not. However:

The ability to make choices is a necessary condition for the evaluation of evidence. — RogueAI

How are you arriving at that premise?

How can you evaluate evidence if you can't freely determine whether it's good evidence or not? If you're simply compelled into believing a particular piece of evidence supports a hypothesis, you don't know if it actually does support the hypothesis. You just have to hope that what you were compelled to believe is right, but how would you ever know?

If you're simply compelled into believing a particular piece of evidence supports a hypothesis, you don't know if it actually does support the hypothesis. — RogueAI

No, I'm not following this line of argument either. Why does being compelled (by whatever factors) to believe something make that belief inherently a less reliable correspondence to 'the way things are' than one which was arrived at with "free will". A computer is certain to come to a more reliable conclusion about whether an equation is correct without free-will, than I am with it.

"IOW if one's evaluations are utterly determined they may not be based on what we think they are based on. We would also be compelled to think we are rational, though not necessarily at all because we are rational and because of what we think is evidence of it."

Yes.

Suppose someone tells you your house is on fire. Let's call that evidence (A). Now you have to update your belief in the hypothesis "my house isn't on fire". How can you do that if you can't even choose whether to believe the person is reliable or not? Is lying or not? Is in a position to know about your house or not? Those are all choices you have to make before you can even begin to assess (A)'s impact on the hypothesis.

Nope, still not seeing where choice causes an increase in accuracy. You could have a predetermined conclusion on all those questions with no less liklihood that such a conclusion would be right. Why are predetermined conclusions of less use?

3. Without free will there is no ability to make choices. — RogueAI

But surely free will isn't merely the ability to 'make choices'. It surely turns instead on the nature of choice made: is the choice itself freely chosen, or itself 'determined'. If so, the fact that science requires 'choices' to be made says nothing about the necessity of free will to underpin science. What matters is how 'choices' are to be understood, not weather or not they occur in the practice of science. The equation of free will with choice seems to be a mistake.

Ok, so, the Universe suddenly begins and both the momentum and position of every particle is somehow determined which allows some outer observer to know everything that will happen for the rest of the life of that universe, this outer observer is omniscient and since the human mind, which appears thousands of millions of years later can also be completely understood by this outer observer because the human brain obeys physical laws. This outer observer realized he can also predict what brains will do by determining both position and momentum of the particles in them. Some time later a human named James Clerk Maxwell is born and writes down 13 laws of electromagnetism which can boil down into four famous equations, this, of course, is no surprise for the all knowing being. In this hypothetical case, how exactly are non free will and science endeavor mutually exclusive?

First, your experience would be no different than it is now. You'd be compelled to believe that evidence x supports hypothesis H where it seems to you as if you're freely believing that for reasons that you consider to be good reasons to believe it.

A necessary condition for doing any science is choosing/determining which evidence to believe and how much weight to give it — RogueAI

How does AI work? How does a computer program determine which calculations to do? Prebuild instructions in the case of no free will.

Because without free will, you're simply compelled to believe that a particular piece of evidence supports a hypothesis. — RogueAI

No? Where do you get that. That's like saying "look at this pong AI, it is simply compelled to always move the bat up while never moving it down" which is obviously not true

"But surely free will isn't merely the ability to 'make choices'. It surely turns instead on the nature of choice made: is the choice itself freely chosen, or itself 'determined'."

If we have free will, then our choices are being freely made. That is a necessary condition for free will. If your choices aren't being freely made, then you obviously don't have free will.

If so, the fact that science requires 'choices' to be made says nothing about the necessity of free will to underpin science. What matters is how 'choices' are to be understood, not weather or not they occur in the practice of science. The equation of free will with choice seems to be a mistake.

If science involves the evaluation of evidence, which it does, then choices are occurring in science. For example, in order for Pasteur's flask experiment to confirm germ theory, one must first decide whether Pasteur's experiment was good evidence or not. I see no way around this.

My point is an epistemological one: I'm not claiming good science isn't going on. I'm saying, how do we know that what we're doing is good science or not?

You could argue that, if we're compelled to make choices, evolution would have weeded out those of us who made bad choices, and that works for things like "should I pick those berries or not?". It doesn't seem to work so well for more esoteric stuff, like "Is Mercury's eccentric orbit strong evidence for relativity, or, since it was known before Einstein developed his theory, is it an example of ad hoc reasoning?" That also requires a choice, and I don't think evolutionary pressures can explain how we would get that one right, if we were simply compelled to believe what we believe.

If we have free will, then our choices are being freely made. That is a necessary condition for free will. If your choices aren't being freely made, then you obviously don't have free will. — RogueAI

Ok, but your OP doesn't talk about choices 'feely made'. It says merely that science requires that choices be made. The determinst simply has to reply that of course choices are made all the time - only that those choices are not freely made. So the argument that we need free will in order to practice science collapses. I.e.

"3. Without free will there is no ability to make choices."

Is false.

"Ok, but your OP doesn't talk about choices 'feely made'. It says merely that science requires that choices be made."

No, I said science requires the evaluation of evidence, and the evaluation of evidence requires the ability to freely make choices about things like "is this a good or bad piece of evidence?". So, if you can't freely make choices, then you can't evaluate evidence, and if you can't evaluate evidence, you can't do science.

What part of that chain do you take issue with?

"The determinst simply has to reply that of course choices are made all the time - only that those choices are not freely made."

I have no problem with that. My question to the determinist is: if there's no free will, how are we able to do science? If science IS impossible without free will, and we're DOING science, then we have free will (or, more narrowly, we are freely making choices).

""3. Without free will there is no ability to make choices."

Is false.""

You can make choices without free will? How does that work? You can be determined to do an action, as we would be in a deterministic universe, but that is not the same as choosing. Choosing requires there to be at least two options. How are there any options if everything is already determined?

Choosing require there to be at least two options. How are there any options if everything is already determined? — RogueAI

There is an apple and a pear. Two choices. I choose the apple (to eat). The determinist says: the choice was not one freely made.

Here you have both options and determination, neither of which are incompatible with the other.

We're getting bogged down in semantics, but I'll address this: "There is an apple and a pear. Two choices. I choose the apple. The determinist says: the choice was not one freely made."

In order for there to be choices, there have to be options. In a deterministic universe, there are no options. Everything's already been set. You are determined to eat whatever. It appears you have a choice, but if you end up with the apple, in a deterministic universe, you were always going to eat that apple. The pear was never an option. No options, no choice.

We may or may not have free will, but either way judgements (choices about what to think or believe) are made on the basis of what seems most reasonable, most logical, most in accordance with the evidence, most consistent with our whole body of knowledge and so on.

We cannot arbitrarily decide what criteria we will use to determine what seems most reasonable, most logical, most in accordance with evidence, most consistent with our overall knowledge and understanding and so on. Those criteria are "built into us" on account of nature and culture: our intelligence, education, general knowledge, critical thinking abilities and so on.

They may change gradually over time, but cannot be arbitrarily chosen in the moment, like as if if I was to deliberate over whether to eat chocolate with nuts or plain chocolate. I doubt that, at any given time, we could even imagine alternative criteria for judgements, unless we happened to be geniuses who had come to find the conventional paradigm full of inconsistencies.

What if the universe is indeterministic? This wouldn't make any practical difference to our situation regarding judgements, as far as I can see.

In a deterministic universe, there are no options. Everything's already been set. You are determined to eat whatever. — RogueAI

Of course. You are determined to eat the apple. But you still had the option between apple and the pear. You just chose the apple, not freely. There is a choice, and what is chosen. That you didn't choose it freely is irrelavent to there being a choice.

Deep Blue chooses between two possible moves in a chess game. Everything about it is determined. Yet it still makes a choice, among the possible options, to castle or to check, to move the queen, or sacrifice the bishop. This is what Deep Blue is designed to do. Make choices.

I'm going to try and simplify my argument a little.
Here's the argument:
1. The ability to make choices is a necessary condition for the evaluation of evidence.
2. Evaluating evidence is a necessary condition for science.
3. Without the ability to make choices, evaluation of evidence is impossible.
4. If evaluation of evidence is impossible, science is impossible.
5. The universe is deterministic.
6. Therefore, the ability to make choices is impossible.
7. Therefore, science is impossible.