In the cartoon show "Futurama," we can see robots with AI spending money, thereby contributing to the economy. In a general answer, if we had AI sophisticated enough to garner "Human Rights" and receive wages for their work, how might this affect the economy, assuming 1 Robot/50 Humans.
There are two ways to think about this question, one straightforward and one more interesting. To address the former, we can think of the “robots” as simply additional humans who add to the economy in the same way as a normal person: they save and consume to satisfy their preferences according to their utility function. Instead of gas and tacos, they may consume gear oil and ball joints. In this manner, the goods produced and consumed in the economy will change, so the composition is the only real change. If the dispersion is 1 robot/50 humans, the effect will likely be negligible in the grand scheme of things.
In a slightly different way, we can look at algorithmic trading as a proxy for what happens when AI begins to participate in the economy. This answer is admittedly a bit of a stretch, but it highlights a key point about AI (and technology in general): we have barely scratched the surface of its existence and cannot confidently project its role and interaction with society as we understand it today. The relevance of algorithmic trading is that the AI seeks to predict what a human or collection of them will do and then perform in the same manner, albeit more efficiently. A more tangible example is AIs trained to compete in chess.
The punchline is this: AI is more effective at most tasks than humans but generally less robust to periods of high volatility because they only know what they’ve seen before. In that sense, AI is often a complement to humans and rarely a substitute—barring sentience. One of the difficulties of sentient AI is that we can quickly degenerate to a Terminator-esque society where wages don’t matter, or the complete opposite happens and R2-D2 is the next president of the homeowners’ association.