Algorithms have a vastly increased presence in financial markets, and they are also increasingly the understructure of other industries. In the past, financial traders saw algorithms as their “tools” and trading was “high human touch.” But algorithms have evolved: The first generation of algorithms learned from humans, the second from theories – this is when algorithms moved beyond imitating experienced traders. The third generation is about to become reflexive – algorithms now learn to use data and “read” news, and make their own trading decisions. This talk examines this development and asks what a posthuman financial market in which human traders and algorithms are both “actors” and “players” may look like, what the differences are between the two types of subjectivities involved, and what engagements and attractions between them develop. It draws an ongoing fieldwork on the trading floors of big banks which algorithms co-inhabit.