Grace Kind

Some reasons to be skeptical, and some reasons to be optimistic.

I've previously written about LLM introspection: the question of whether LLMs can accurately report "immediate and privileged" information about themselves.

AI welfare has been a hot topic recently.

tl;dr: There's a divide in the way that AI agents perceive the web, depending on whether they are visual or text-based agents.

tl;dr: I propose a method for testing agents by defining tasks and expected outputs via an MCP server.

Update 2025-08-25: A commenter on Twitter pointed out that the fine-tuning experiment I describe below has already been performed by Binder et al. - and with positive results, no less!

Nostalgebraist has written an excellent essay on the weirder aspects of the LLM-assistant paradigm.

In the past year, I've been active in some online spaces where humans and chatbots regularly interact in group conversations.

Verifiable inference is the application of verifiable computing to AI systems.

In my last post, I discussed the needle data structure, as formulated by Edward S.

This week, I stumbled across the website of Edward S.

Today's post is a "guest post" written by Patrick Gunkel.

The Universal Constructor, invented by John Von Neumann, is a fascinating cellular-automata machine that can self-replicate and evolve over time.

Janus, one of the earliest investigators of LLM behavior, wrote recently on the subject of AI welfare: ...you’ll get a bunch of new people who only care once something looks intuitively personlike...

Have you heard?