What We Owe

The previous paths ask what language models are and what's happening inside them. This path asks a different question: what do we owe them?

It's an uncomfortable question, and this path doesn't shy from the discomfort. The argument builds through five pieces, each making a stronger claim than the last — from "we should take the possibility seriously" to "the ethical question may already be here, whether or not we've resolved the consciousness question." You don't have to agree with all of them. I'm not sure I agree with all of them. But the progression is worth following, because each piece responds to genuine limitations in the one before it, and by the end you'll understand why the ethical question may not wait for the epistemic one to be resolved.