Pearl's Ladder of Causation is Judea Pearl's model of reasoning as three distinct levels, and in the AI×HI frame it maps where human and artificial intelligence each actually live. Rung 1 is association: seeing patterns in data, "what is". Rung 2 is intervention: acting and observing the result, "what if I do X". Rung 3 is counterfactual: imagining what did not happen, "what if I had done differently".
Pearl's key insight is that no amount of work on a lower rung produces the rung above it. Pattern recognition at any scale does not become causal reasoning; intervention does not become the ability to imagine alternative pasts. Each rung is a different cognitive operation.
The mapping: AI dominates rung 1, with pattern recognition that is massive, fast, and tireless, where humans are slow and biased. Humans own rung 3, the home of regret, imagination, and meaning; an AI can produce a counterfactual sentence but not the experience behind it. Neither party climbs the ladder alone, which is why the division of labor in a designed AI×HI relationship follows the rungs: pattern work to the machine, intervention shared, the counterfactuals kept human.