This essay argues that the emergence of LLMs offers not only a challenge to human creativity, but an opportunity to examine human cognition. Drawing on questions from art, cognitive science, and theories of grounding, it argues that human cognition is relational, emergent, environmentally scaffolded, and mediated by inherited models rather than directly grounded in an unmediated reality. This has consequences for how we understand AI-assisted creativity and, more fundamentally, the alignment problem: rather than aligning an ungrounded artificial system to humans assumed to be directly grounded in reality, we may need to understand how differently mediated cognitive systems can be coordinated within shared causal environments.