Like some kinds of infinity, it is difficult to compare two lives, but it is clear that multiple lives are worth more than smaller numbers. This life-worth calculus is featured in many a philosophical discourse but is perhaps best summed up in the closing scenes of Star Trek II: The Wrath of Khan when Spock explains
The needs of the many outweigh the needs of the few
However much we like to think that human lives cannot possibly be compared to, say, a restaurant dinner, we are nothing if not an inventive species. The concept of money allows us to assign a numeric value to anything conceivable from rocks in the ground to complicated machines that move people around the country to even being able to bet on future outcomes like how many tornadoes there will be in 2016. Insurance companies and loss adjusters routinely value human life in this way (to calculate payouts for industrial accidents, or figure out loss of earnings in discrimination cases).
So – we can put a dollar value on a human life – call it an expected economic contribution.
Why not do the same for a robot? If you create a robotic version of me capable of contributing to the economy in just as efficient a way as the real me, then – strictly economically – are not our values the same?
Which brings us on to cost. All other things being equal, the reason I can continue to be gainfully employed is not that I have a higher value but simply that the robot has a higher cost.