I believe that given sufficient resources, scientific method could accurately predict the future.
The resources are not there because humans are machines of such incredible complexity that a theorem or a program sophisticated enough to account for all relevant determining factors is presently beyond even imagining.
But this is a practical limitation, not a theoretical one.
I do believe that humans are machines, and that behaviour is basically mechanics, and that free will is, in a strictly literal sense, illusory.
Resistance to such an idea, especially when so baldly expressed, comes from all fronts, from commonsensical empiricism, from wounded political thinkers, and from those who fear correctly that the general adoption of such a belief would be interpreted as a licence to wallow in apathy and pessimism.
I sympathise greatly with the latter position: but I still feel that whether or not it is generally adopted, and if so whether or not its general adoption will have the kind of effects we envisage, is already determined.
But that doesn’t mean that anything need change. Our general perception of phenomena has no effect on the phenomena themselves, which go their happy way utterly unchanged by however we choose to understand or misunderstand them.
So obviously if the universe is determined then it has always been determined, and if the free will model ever worked it will still work.
The notion of free will, though false, is still a perfectly adequate one so far as it goes, as a model for explaining our interrelation with other objects and existents in a multi-dimensional universe.
When Newtonian mechanics was first shown to not be universally infallible it didn’t mean it is was suddenly going to start letting you down in the areas where it had always held good.
So too with free will: it's an illusion, but if it always was then there is no need to abandon it up to the point where it still works.