Animal welfare #EA is motivated by virtue signalling but pretends to base itself in some objective measure of total +/- qualia.

It is usually better to ground something in those factors that actually cause it.

The objective qualia basis for animal welfare EA will not work. https://twitter.com/RokoMijicUK/status/1281898049707610114

In the familiar domain of farm animals and pets, the basis of virtue signalling and qualia maximisation agree, so it seems like they're the same thing.

A man who tortures pets or farm animals is a genuine threat to other humans, and this is true even if pets are p-zombies.

But in the domain of insects, small fish, etc, there is no real signalling implication; people who swat flies are not more likely to be psychos.

However actual theories of qualia maximisation struggle to replicate this result.

Worse still is wild animal suffering.

For wild animal suffering, the "qualia maximisation" basis for animal EA and the "virtue signalling" basis point in opposite directions! This is why wild animal suffering is a "difficult" topic; it's a place where actual motives clash with claimed justifications.

@petrlorg is the ultimate problem for qualia-based animal EA. Whatever your objective, qualia-based theory of animal suffering is, there's a disembodied RL algorithm that's computationally identical but has 0 virtue signalling value.

Actually making a way to compare and measure qualia across extremely different minds is probably impossible. It's like division by 0. I would like to be able to prove this, but that's a serious academic project. The stuff in these tweets is stuff I'm relatively certain of.