"The impossiblity of self-prediction" and chaos

D. Miller speaks of 'the impossibility of self-prediction' in the essay, online:

"The modern theory of non-linear dynamical systems [chaos] has certainly taught us a salutary lesson about unpredictability in classical physics. It is less obvious what it should have taught us about determinism. While many regard deterministic metaphysics as triumphantly vindicated, Popper (1982), for example, sees its only real support pushed unceremoniously aside. It is the burden of this paper to outline this problem (section 1); to challenge both the usual connection between determinism and predictability and the relevance of the argument leading from deterministic chaos to determinism (section 2); and finally to suggest that, for those who see prediction as an essentially probabilistic enterprise, unpredictability may promise an even more severe headache than is currently envisaged (section 3). It is only fair to say in advance, perhaps, that I am an indeterminist. This seems once more to be an unfashionable position, but I hope that it is still a sustainable one."