Simple learned response

So the next strategy is to allow for a more complex CS that will be able to process the signals as they're received, and will perhaps allow for at least learned responses. But this, just as such, is difficult, since every new signal (and these are constant) requires the CS to have to recalculate or recalibrate all of the other signals impinging upon it, together with the rapidly shifting internal state of the organism, in order to decide upon a single action. The problem is that the CS is still, in a sense, wired directly into environmental stimuli, and lacks a "wholistic" view that would provide a basis for better decision-making. It's clearly more sophisticated than the simple table look-up, but still a control system with limited flexibility.