Lawrencestreet.ai is a philosophy of AI project exploring artificial agency and its relation to answerability and verifiable governance.

This work assumes that:

A technology powerful enough to amplify intelligence can become a native intellectual medium across domains.

AI is applicable to knowledge and design work in general.

Artificial agents must be designed to function normatively before they can be trusted.

Available models are not yet designed to function normatively to an acceptable standard.

Coding agents can write software based on philosophical principles.

As the cost of code trends down, technical expression of those principles decentralizes.

Philosophically principled symbolic systems can be specified.

Artificial agents can construct and empirically test the application of those systems.

Accordingly, LLM tools are instrumental to every facet of the project.

Three essays establish and defend the sociotechnical framework for discursive mindedness.

The LARRI open protocol and reference applications implement operational ideas drawn from the framework, scoped for use cases in trust-sensitive domains like healthcare and law.

Explainer videos (~5 minutes) are available for the essays, protocol and apps.

The rest of the site is a knowledge graph of public work influential or relevant to the project.

Please contact me nwillfreeman@lawrencestreet.ai with any comments or interest.

Thanks for visiting.