Like you, I want to avoid this collapse, but I also want to stress maybe a bit more forcefully the ontological or metaphysical import of the second claim.
That’s because the intellectual active powers of man (both practical and theoretical) are what it is that Earth’s past potency is relative to, on the one hand, and the specific actualizations of those potencies in the shape of our form of life is what it is that we now actively sustain, develop, revise, etc. The universe, we might say, does give rise to us (or creates us) but it doesn’t do so independently of us, since we aren’t separate from it, and we merely seem to be separate from (and hence rigidly constrained by) its “past pre-biotic state” due maybe to something like Kim’s fallacious Causal Exclusion Argument (or van Inwagen’s Consequence Argument) that make it look like our own past constrains us from outside of ourselves, as it were.
So far so good, but I also wonder if the foregoing is sufficient to forestall a view that I don’t want to suggest. There is a sort of restricted social constructivism (advocated by John Searle) according to which there are socially constructed items such as money, institutions, wars, promises, etc. This is supposed to be contrasted with things like rocks, planets, and electrons. I of course want to endorse a more radical form of constructivism according to which this contrast is drawn between two sorts of ontological domains that both are constructed in the sense that what it is that counts as an exemplar of either still is dependent on the conceptual nets (and scientific practices, including quantum theory, etc.) that we cast. So, there is the danger of giving the impression that there already were “things” making up the pre-biotic Earth that are being individuated differently when caught under our nets than they are “in themselves,” and the fish example, where we grant that since fish, qua autopoietic entities, individuate themselves, encourages this picture.
Likewise, if we say that what it is that makes up the boundaries of a mountain (e.g., where it is that the mountain ends and the valley begins) are social conventions regarding something like an underlying rocky terrain that is allegedly not socially constructed, we invite the same misunderstanding. What I’d rather want to argue is that we are like the fish in some respects (since we are animals) and unlike it in others (since we are navigators of the space of reasons, which is, in my view, just the natural/social world individuated in a particular way different from the way a fish individuates it into the affordances of its Umwelt) but there is no pre-biotic world that we can make sense of as being other than a determinate set of potentialities for the actualization of our Umwelt or, equally intelligibly to us, for a fish’s Umwelt.