This Brookings working paper presents an analysis of how institutions could be set up to prepare for a scenario of increasingly smart autonomous machines which could replace human labor and lead to lower wages is presented. Three potential consequences of this situation are discussed, and recent predictions and objections to these consequences are evaluated. The optimal way to handle a situation where income from autonomous machines rises and the value of labor decreases is explored, with the conclusion that it may be best to phase out work, beginning with workers who have low labor productivity and job satisfaction. In contrast, welfare systems that require individuals with low labor productivity to work would not be helpful. If there are significant wage declines, new methods of distributing income, such as capital ownership or benefits, would need to be implemented to avoid mass misery. It is suggested that work could still be engaged in for its own sake if individuals find it rewarding, and that public policy may be necessary to encourage work if it leads to positive externalities or if individuals don't fully appreciate the benefits of work. However, it is proposed that in the long run, alternative methods of providing these benefits should be developed.