Autonomy in a lab comes in three levels. At the first level, AI acts as an assistant: it predicts, suggests, or ranks options while humans still do the work. At the second level, machines execute experiments automatically, but a person decides what to run. At the third level, a closed loop designs an experiment, runs it, measures the result, and updates itself, so the system chooses its own next experiment. The levels differ in who makes the decisions, not just in how much machinery is present.
Can AI Run a Drug Discovery Lab on Its Own?
What It Would Mean for AI to Run a Lab
Three Levels of Autonomy
2 / 3
So how much of that chain can AI own? It helps to separate three levels. At the first level, AI is an assistant. It predicts, it suggests, it ranks options, but a person still does the work.
0:00 / 0:00