Can artificial intelligence/robots possess thought and perception that would enable them to surpass humans, or can they not exceed the stage of mere thinking and preferring?

The problem lies in defining the meaning of thinking and choosing. If it means absolute independence in perception and preference, then this is negated for a machine. However, if it means using premises to reach conclusions and processing inputs, then a machine can do that. A calculator, for example, performs complex operations that exceed human capacity, but it does not think or choose; rather, it uses programmed algorithms. Therefore, even if a machine performs some functions of the human mind, it does not rival it or render it unnecessary. There is no harm in scientific research in this field, as a machine does not deviate from its true nature and does not exceed human limits.