Could AI attribution address consciousness, concentrated power, alignment and job loss?

A user argues that if people could identify and control whose pre-training data shapes each output token, problems involving AI consciousness, concentrated power, value alignment and job loss would largely go away. They propose paying workers when AI uses their ideas, while noting that some unionization might be needed to help workers whose ideas overlap.