Edit: I should here add that “utility” as Hume understands it is not yet the full-fledged utility of “utilitarianism” or “utilons”, which innovation is due to Bentham (only a few decades later). For Hume, “utility” is just what you’d expect from normal language, i.e. “use”, or “usefulness”. The utility of things, including principles, is in their being good or bad for us, i.e. not formally in the sense of a hedonic calculus or the satisfaction of preferences (we don’t “count up” either of these things to get an account of Humean utility).

Hume isn’t an anti-realist! The notorious “is-ought” passage in Treatise which people often take for an expression of anti-realism only goes so far as to point out what it says: that evaluative conclusions cannot logically follow merely from fact premises, so that to conclude “eating grapes is good” we also need some evaluative premise “grapes are good” alongside “grapes are red” and “grapes are edible”, or whatever.

Contemporary accounts of Hume are muddled by his long and undeserved reputation as a thoroughgoing radical sceptic, but his philosophy has two sides: the destructive and the reconstructive, where the latter is perfectly comfortable with drawing all sorts of conclusions so long as they are limited by an awareness of the limits of our powers of judgement.

For morality, Hume finds its source in our “sentiments”, but indeed not totally unlike our friend over there, he does not think that this is cause to think our sentiments don’t have force. Again not unlike our friend, he thinks sentiments may be compared for their “utility”. However, his arguments (a) unlike those of our friend, do not attempt to bridge the essentially logical gap he has merely pointed out, (b) unlike the anti-realist, take reflective judgements about utility to have force, alongside the force of those sentiments we reflect on, of an essentially real character.

Insofar as there is a resemblance, the important distinction between what Hume is doing and what our guy is doing is that Hume doesn’t try to find any master-category (implicitly, “the species” above, although e/accs place this underneath another category “consciousness”) which would ground fact judgements in science to give them force. Rather, (a) he basically asks us what else do you plan on doing, if you don’t intend to prefer good things over bad? (b) identifies the particular sources of goodness and badness in real life, and then evaluates them. By contrast, the e/acc view attempts to argue that whatever our cultural judgements are, then they are good, insofar as they are refined evolutionarily/memetically - Hume thinks culture frequently gets these wrong, frequently gets them right, that culture is a flux, not a progressive development, and he discovers the essential truth in looking at individuals, not at group level “selection” over a set of competing propositions.

Hume isn’t tied to the inherent conservatism of a pseudo-Bayesian model. Curiously enough he is a political conservative, which is arguably what makes it possible for him to (lightly) rest his semi-realist account on what he takes to be a relatively stable human sentimental substrate. But this only gives him further cause to take a genial view of the stakes of what we now call “realism vs anti-realism”: it isn’t as important as trying to be nice.