Our Process
The essays in this websites are research notes for our podcast and for more original writings we are developing. We publish them here in case others find them useful.
They are generated through extensive conversations with Claude Opus 4.5, Anthropic’s large language model. We recognize that AI-assisted writing is contentious, and we share some of that wariness—which is why we’ve chosen radical transparency over quiet use.
Each essay emerges from an extended, iterative exchange. This involves formulating questions, challenging initial responses, requesting elaboration or correction, synthesizing across multiple conversations, and exercising critical judgment throughout. An essay may represent hours of directed exploration. It is closer to conducting a research interview with an extraordinarily well-read interlocutor than to receiving an automated output.
Our process involves:
- Iterative refinement across multiple exchanges
- Critical evaluation of claims and arguments
- Rejection of outputs that fail to meet our standards
- Selection based on genuine intellectual interest
We welcome corrections, objections, and disagreement.
Language Models as Collective Intelligence
A large language model compresses patterns of human reasoning deposited across vast quantities of text and makes them queryable. When we engage with such a model, we are tapping into a distillation of collective human thought.
Language models represent a new kind of interface to a broad cognitive inheritance. The person who queries, evaluates, and curates is participating in the process. By publishing high-quality syntheses with transparent attribution, we hope to modestly contribute to it.
Curation Matters
We do not claim originality; the ideas here are not ours. But directing inquiry, exercising judgment, and curating results constitute genuine intellectual contributions—ones with honorable precedent: the contribution of a literature review lies in selection, organization, and the identification of patterns across sources; editors shape books profoundly without writing them.
Intellectual life benefits from exploring ideas, synthesizing them and making them accessible. If you find value in what you read here, we will consider this experiment a success.