Dennett has argued that free will is best ascribed to sufficiently smart deterministic robots in a deterministic universe.
These robots have been designed to look out for themselves and survive and even perhaps reproduce themselves.
If they do not ‘believe’ (behave as if) that they are part of the causal chain in that universe they are unlikely to survive.
Being smart they hold a poor model of their part of the universe and imperfect theories on causal chains.
Occasionally they learn new things about their vicinity and consider possible future situations that may come about.
Some of those situations may be more favorable to the robot than others.
The theory of causal chains allows the robot to choose an act that makes the favorable situation more likely.
It takes that action and increases its expected lifetime.
It has made a decision.
It looks to me like free will.
From another vantage point the robot is merely part of the deterministic universe; its actions can “in principle” be explained that was with no need for the free will concept.
The robots knowledge of the universe is necessarily imperfect and the above analysis is irrelevant.
Seeing as how ‘truth’ is something that we fellow humans, along with many prehuman ancestors evolved with to help us survive and thrive, any claim that our decisions are pointless to act upon is a bad meme.
A world model in the intellectual repertoire of a robot which omits the robots decisions in the causal chain of that world, will likely be fatal to that robot.
A viable robot, even in a deterministic world must have a world model which includes the robots decisions in the causal chain.
Humans need that too.