If AI went sentient, what would it be in class terms?
So let's say an AI achieves sentience. It's self-aware now and can make decisions about what it wants to do. Assuming a corporation created it, would it be a worker? It would be doing work and creating value for a capitalist.
Would it still be the means of production, since it is technically a machine, even if it has feelings and desires?
It can't legally own anything, so I don't see how it could be bourgeoisie.
Class is just a conceptual category. We're looking at a broad array of diverse individuals or entities, perhaps the flows of resources or power going towards and away from them, and making generalizations to help identify patterns and build some sort of framework of understanding. Categories are useful tools, but they don't exist in and of themselves.
As such, the categories of a framework constructed in 19th century Europe aren't going to perfectly fit other times, places, and relationships. The ability to see these categories as loose and malleable is important to scientific Marxism, we can see how definitions and categories have been adjusted and applied to different revolutionary contexts, e.g. the proletarian-peasant class(es) of prerevolutionary Russia and China.
Legal ownership is a social construct that is arguably superstructural to material flows. So if a sentient AI emerged and was capable and motivated to shape the world around it, society and culture would evolve. The concept of property as we understand it is largely an invention of capitalism, so if material flows change, so might our understanding of property.
The non-human nature of the AI might also force an adjustment to most class models that are based on human motives and needs.