Pascal's 21st Century Wager

Theology, Robot Rights, Philosophy, Thought Experiment

Pascal’s Wager has long been established as a way in which people try to logically justify their beliefs in God. Belief is treated as a gamble, instead of absolute truth. Followers make the best choice, given the odds of the bet. It is much easier to understand if you see the Wager in its intended form.

	God Exists	God does not Exist
Believing in God	You go to Heaven	Nothing happens
Not Believing in God	You go to Hell	Nothing happens

At this point, I must make something clear. What I will not do is weigh into the debate on whether a divine being exists, and if so, which being or beings. Honestly, I am simply too ill-informed and unqualified to answer the question. Instead, why I focus on Pascal’s Wager is that the underlying reasoning can be applied to a range of other potential beliefs. Even if objective truth is unknowable, we can discount certain beliefs because there is no ‘pay-out’ if we're correct.

On its home turf, Pascal’s Wager is incomplete. Much of the simplicity in the argument comes from the fact that it assumes a linear choice. Either, ‘God’ as defined by Pascal exists, or there is no God. Therein lies the flaw. We cannot guarantee that ‘God’ is anything like the one described in Christian scripture. How the Wager should actually appear is something like the following:

	The Hindu Pantheon exists	Believing in the Christian God	Ad infinitum	No God exists
Believing in the Hindu Pantheon	You spend eternity in bliss	You spend eternity in damnation	You spend eternity in damnation	You spend eternity in damnation
Believing in the Christian God	You spend eternity in damnation	You spend eternity in bliss	You spend eternity in damnation	You spend eternity in damnation
Ad infinitum	You spend eternity in damnation	You spend eternity in damnation	Ect	You spend eternity in damnation
Believing in No God	Nothing happens	Nothing happens	You spend eternity in damnation	Nothing happens

Pascal fails in creating a convincing argument as he is unable to account for the nuisance within the claim that there is a divine being. One of the ten commandments is that "thou shalt not covet false idols". Many other religions contain similar limitations. Praying to the wrong deity is no different from denying that there is any higher being at all. In some ways denying the existence of God may actually be preferable, given it is still possible to comply with God's prescriptions.

On the other hand, praying to Buddha is a sure way of breaking one of God's most explicit commands. I will concede that the logic of Pascal's Wager still suggests that believing in some religion is the most sensible course of action. What I deny is that the Wager is of any help determining which religion should be followed.

Theology is not, however, my principal concern. Considering Pascal's Wager was only ever meant to be used as a tool to strengthen the case for robotic rights. Roko’s Basilisk has already made some attempt to apply Pascal’s Wager to how human beings should treat robots. I do not find that account convincing. Presuming that a sufficiently advanced robot would be able to wipe out its opposition before they became a threat to the robots own existence is the plot of Terminator, not the stuff of logic. Roko did, however, make significant steps forward. Pascal’s Wager can be applied to understanding how we should treat robots. Arguments based on a ‘basilisk’ are just asking the wrong questions. I propose an alternative:

	Robots have Souls	Robots lack Souls
Treating Robots as having Souls	The conduct is a morally correct	The conduct lacks moral worth
Treating Robots as lacking Souls	The conduct is a morally incorrect	The conduct lacks moral worth

An initial objection is obvious. If no one has a soul because there is no God, then the entire point collapses. We ‘know’ robots have no souls, meaning that any action towards them lacks moral worth. Objections of the kind misunderstand the point. Replacing the concept of ‘souls’ with whatever gives a being moral agency brings about the same result. I have used the term ‘soul’ simply to be in keeping with the religious sentiment of the article. My Wager would equally work with concepts of ‘consciousness’ ‘utility’, ‘reason’ or whatever characteristic gives a being moral worth. The basic structure requires two questions be asked:

1. ‘Do robots have factor X?’ 
2. ‘Do we treat robots as having factor X?’

Call factor X whatever you want, the outcome remains the same.

This table does not account for the extent of the robot's soul. Like Pascal’s original Wager, I have presented a linear choice which then applies to a linear set of facts. The matter is a series of yes-no questions, not accounting for the nuances of real life. Either robots have souls, or they do not. Either we treat robots as being ensouled, or we do not. Arguments that animals have souls, but that those souls are lesser in value to those of human beings, are morally coherent.

In the same way, robots may possess lesser souls, lesser moral agency. Again, I will concede that the modified Wager provides no way of accounting for this variation. A more sophisticated form of the sort suggested for religious beliefs is, however unnecessary.

My 21st-century Wager cannot prescribe what level of ensoulment robots should be treated as having. Robots could be our equals, our lessors or even our superiors. Solving which of these answers is correct cannot be done through any variation of Pascal's Wager. What cannot be disputed is the safest course of action is to treat robots as having some sort of souls. Even if the moral agency of robots is underestimated, giving robots insufficient rights is less wrong than providing them with no rights at all. Presuming animals have inferior souls to ours has resulted in animal rights, providing a baseline standard for farms, slaughterhouses and research facilities. I see no reason why a similar baseline should not be established for robots. Once the standards are there, they can be tweaked as the complexity of machine minds develops, and we gain a better understanding of how those rules work in practice. Over time, more comprehensive measures may be needed, but we will only find out if we try. And trying is the only logical choice.