He’s saying that because humans can see things in the future (such as a coming flood due to an oncoming weather pattern) and then take preventative action, we are exercising free will in the most meaningful way possible (harm avoidance)–even if we are actually living in a deterministic world.
I would say that what we get here is the illusion of free will, but not free will itself. But it’s true that it’s practically valid, since it feels exactly like what we would expect free will to feel like, but we’re still just responding to a physical, causal world; it’s just so complex and advanced that it yields free-will-like results.
Remember, factory manufacturing lines can detect problems and make changes to production flows and such as well. They’re sensing their environment and responding to it in order to avoid harm in much the same way that humans do. It’s not free will; it’s sufficiently advanced programming so as to present the illusion of choice.
It’s also interesting that he goes on to plainly state that any sort of quantum trickery does nothing to give us more freedom–which is a view I’ve always held as well. The quantum bits really lend nothing to the free will argument; they only serve to potentially sink a rigidly deterministic view of the universe. But again, that extreme is not in any way necessary for the argument against free will. ::