If algorithms optimize culture, then to what purpose should they be optimized when it comes to the state? In this chapter, I lay out three different potential approaches for optimizing algorithms towards the goals of the state: a ‘republic of virtue’ approach, a ‘preservation of the self ‘ approach and a ‘preservation of rights’ approach. I then move onto talking about how, in the United States, the overwhelming emphasis of tech-activists is for algorithms to be optimized towards the preservation of rights. I briefly discuss the differences between US and European approaches and explain why the preservation of rights approach can be incomplete.

A ‘republic of virtue’ optimization

Civic Republican political theory posits that the state has an affirmative role to play in promoting civic virtue (Sandel 1988). If we can agree on what civic virtue is, we could optimize algorithms to promote good citizenship. Adherents of a classical view would build on the Aristotelian concept of ‘telos’ which suggests that everything is built with a natural purpose. A teleological view suggests that embedded within all humans is a natural inclination towards certain virtues, such as preserving life, educating the young, and preferring order – much like a watch is designed to tell time. If a watch fails to fulfil its natural purpose, it is considered a watch in need of repair. Similarly, if humans fail to fulfil their telos, they are not achieving their moral purpose.

This idea implies that humans can be evaluated based on their alignment with this purpose. Correspondingly, algorithms could be trained to encourage virtuous behaviour. Feeds might be populated with examples of people engaging in such behaviour or show people punished for being unvirtuous.