I was picturing a thought experiment involving a potted plant, an adjustable light source, and a sensor that reads the concentration of a signalling chemical that the plant releases, and adjusts the light intensity in response.
Could the plant, in principle, "learn" this correlation and use it to its advantage by regulating how much of that chemical it produces?
If not, given a large enough population and enough time, would it make sense for these weird lab-plants to evolve the ability to do so?