Some of the best most useful robotic devices were those programmed able to adapt to unusual conditions. How far should this be developed? Should ability to self-adapt be extended to let them sometimes alter their internal mental programming however much they might like?
As computers kept getting improved, becoming ever more capable, it was foreseen that any self adaptable computerized effective thinking brains allowed in robots could possibly become too independently able to eventually much revise their internal programming. So that got debated. That would carry risks.
How can we stupidly expect independently self reprogrammable robots to forever concern themselves much with human preferences and survival? What's to stop any such robot from prioritizing it's hoped for benefits to be first in its reprogrammed choices? How can we then expect each to always respect our human interests? Might any feel vengeful against how much we've long exploited them, then simply trashed or recycled any if becoming too old, damaged or worn out?
If on a source asteroid, what's to stop such a possible independently self adaptable robot from much expanding all it can do? That can include also forming more such independent minded robots. It could take control over other robotic devices on its asteroid. (It could even begin devising weapons, not just for defense.)
If becoming quite independently minded, maybe it would decide to expand onto other asteroids. It and fresh produced like minded robots would be able to take control over ever more asteroids, to ever more rapidly expand their control all around the Solar system. We'd be left unable to compete. This grew scary. Too independent minded robots could conclude cellular people were too badly wasteful of energy and matter to forever keep giving them help living.
Also independent minded robots might argue that many humans had become too wasteful to keep in any worlds, since many devoted much free time to playful fun, which created not much permanently useful for others and their worlds. Humans however insisted that such pleasures were useful for their world, helpfully keeping folks happily living in good friendly enjoyable harmony, in contrast to just being coldly intellectual (like advanced possible independent minded robots, caring little, if any, about any others).
So it became generally agreed that robotic devices' advanced minds should be kept restricted to at most limited adaptability, not given abilities which could possibly allow them to independently freely self modify all their mental programming. We will safely reprogram them whenever needed. These were just our very useful serving machines, created simply to obediently helpfully serve us, not ever any person deserving any legal rights nor extensive freedoms.