Comment by fastThinking

No LLM here, just a habit from academic writing. The reason the tobacco analogy works for me is that both cases optimize around reinforcement, not outcomes.