Argument: Social and moral cognition can produce large internal rewards that dominate local costs.
Addiction: What about agents who self-pleasure to the detriment of their entire welfare integral?
Argument 1: The reward circuit considers the addictive behavior to be a positive investment, i.e. returns are higher than cost.
Argument 2: The welfare integral defines the optimal behavior, i.e. what the agent wants to do ideally. Agent behavior is an approximation to the policy. There are many bugs in agent behavior such as these pleasure traps. Low intelligence can also affect proximity of actual behavior to optimal behavior.
Argument: This is a disorder of the reward circuit itself, the theory assumes a functional reward circuit. Non functional reward circuits are NOT stable behavioral solutions under natural selection.
Argument: Pharmacological and neurological interventions can induce valence–motivation decoupling, but across evolutionary timescales selection favors architectures that keep them functionally coupled because coupling better promotes survival and reproduction. [1]
Conclusion
A minimal theory of volition needs only , , and time.
From this view, value is the shape of expected pleasure and pain across a life, and volition is the policy that optimizes that shape under constraints.
Everything else, including patience, sacrifice, and cooperation, is downstream strategy.
Claim: Volition is strictly a function of two variables: pleasure and pain .
Background: This article is a movement towards formalizing the definitional basis of volition in Extrapolated Volition / Coherent Extrapolated Volition.
Candidate Definition
For any sentient agent with a functioning motivational system, volition is the optimization of its own lifetime welfare integral:
Where is pleasure, is pain, is the agent's birth, and is its expected death.
What the agent wants to do, it's volition, is to not feel bad and to feel good for as many moments as possible throughout their expected life.
No additional terminal value is required to explain behavior at this level, apparent counterexamples seem to be explained through the time integral.
Candidate Counterexamples
Conclusion
A minimal theory of volition needs only , , and time.
From this view, value is the shape of expected pleasure and pain across a life, and volition is the policy that optimizes that shape under constraints.
Everything else, including patience, sacrifice, and cooperation, is downstream strategy.
It is NOT a coincidence that the extremes of pleasure and pain perfectly map to reproduction and the inverse of survival.