What if you took an autonomous agentic AI, programmed it to act in its own self-interest like a human, and gave it a humanoid robot body?
Like nah obv not just telling an AI pretend ur a person.” I mean actually designing the whole system around it being a continuous individual persistent memory, its own long term goals, recognizing the robot body as its body, wanting to maintain and repair itself, making money so it can afford power/parts the works etc like shelter, avoiding being shut down, and generally trying to improve its own situation over time.
Basically give it an advanced local AI a robotic body and program its incentives to be something closer to a human being’s: preserve yourself, maintain your health and body, gain resources, form connections that are useful to you, learn from experience, seek greater independence etc
Like what about also a persistent autobiographical memory too, so if you talk to it today in 2026 and then again in 2030 it remembers what happened and considers itself the same entity. If its body gets damaged it sees that as injury to itself, if its low on money it tries to earn more, if it needs a replacement part it figures out how to obtain one, and so on.
How human type like would its behavior actually become after years of this? Would self-preservation mixed w memory + resource seeking as well as aphysical body eventually produce something that behaves almost like an artificial person?
If using modern tech someone did this with current tech could it be considered sapient?