Conjecture - LessWrong Tag Page

Summary

This is the LessWrong tag page aggregating posts related to Conjecture, an AI safety research organization focused on cognitive emulation (CogEm) and interpretability approaches to alignment. Conjecture's work explores building AI systems whose reasoning processes are understandable and human-like rather than opaque.

Key Points

• •Conjecture is an AI safety organization pursuing cognitive emulation (CogEm) as a core alignment strategy
• •CogEm aims to build AI systems that reason in ways humans can inspect and understand
• •The tag aggregates LessWrong posts by and about Conjecture researchers and their alignment approaches
• •Interpretability is central to Conjecture's research agenda, seeking transparency in AI cognition
• •Conjecture has been an active contributor to the LessWrong alignment research community