Intelligence may emerge partly through a system’s behaviour and partly through another system’s interpretation of it. What might that mean for how we understand AI?
One way of understanding intelligence is as an emergent phenomenon.
On this view, sufficiently complex systems produce intelligence as a higher-order property.
Brains and nervous systems interacting with the environment, and perhaps AI systems.
However, this assumes that the emergence occurs within the system itself.
It may instead be that complexity produces sufficiently rich behaviour for another system to represent it as intelligent.
If so, emergence may occur partly in the system and partly in the perceiver.
This may matter for how we think about artificial intelligence.
At some point, sufficiently rich behaviour stops looking like instruction-following and starts looking like intelligence.