But that's the point of confusing free will with consciousness. You can experience something (that's the hot stove argument), and inference is perfectly possible for even today's computer algorithms to do (which are certainly not free). Even something as common as a virtual memory system deciding which pages to swap makes predictions/inference based on what it has seen in the past.

Morality is still a thing, much like there are worse and better paging algorithms. Think of it as mental models of how things work, much like the rest of science. In this view morality is a study of how things/consciousnesses feel.

Edit: another way to look at morality as a science is like this: assume most people genuinely don't want to cause suffering of others, and especially of themselves. But they can still cause harm by being simply wrong.