Blog

Thoughts, reflections, and updates from our lab

Imperfection: The Quiet Threat AI Poses to Human Memory

Imperfection: The Quiet Threat AI Poses to Human Memory

Jaechul Roh • May 10, 2026

A reflection on a quieter threat AI poses, beyond jobs and adversarial attacks: the slow erosion of the imperfections that anchor human memory and shared culture.

Opinion AI Safety Culture
Thoughts on COLM 2025

Thoughts on COLM 2025

Jaechul Roh • December 14, 2025

Reflections from attending and presenting at the Conference on Language Models (COLM) 2025 in Montreal.

Conference AI Safety LLM