RSS

Three Laws

Karma: 4

Our homepage: https://​​threelaws.net
You can contact us at: roland@threelaws.net

About

• Three Laws is an agentic AI alignment research collective investigating how fundamental principles from biology and economics can inform safer, more aligned AI systems.

• Our work centres on homeostasis, multi-objective balancing, sustainability, and universal human values — drawing from nature’s time-tested strategies for maintaining equilibrium — to develop benchmarks that expose dangerous failure modes in current AI approaches.

• We also research frameworks that mitigate these risks. We believe that shifting AI design from “maximise forever” toward “maintain a healthy equilibrium” is a crucial and underexplored part of the alignment solution space.

Research Interests

Open-source LLMs ad­minister max­i­mum elec­tric shocks in a Mil­gram-like obe­di­ence experiment

7 Jul 2026 20:05 UTC
8 points
0 comments25 min readLW link
(arxiv.org)

Re­search agenda for train­ing al­igned AIs us­ing con­cave util­ity func­tions fol­low­ing the prin­ci­ples of home­osta­sis and diminish­ing returns

28 Dec 2025 21:53 UTC
14 points
0 comments8 min readLW link

Black-box in­ter­pretabil­ity method­ol­ogy blueprint: Prob­ing run­away op­ti­mi­sa­tion in LLMs

22 Jun 2025 18:16 UTC
17 points
0 comments7 min readLW link

Sys­tem­atic run­away-op­ti­miser-like LLM failure modes on Biolog­i­cally and Eco­nom­i­cally al­igned AI safety bench­marks for LLMs with sim­plified ob­ser­va­tion for­mat (BioBlue)

16 Mar 2025 23:23 UTC
45 points
8 comments16 min readLW link

Why mod­el­ling multi-ob­jec­tive home­osta­sis is es­sen­tial for AI al­ign­ment (and how it helps with AI safety as well). Subtleties and open challenges.

12 Jan 2025 3:37 UTC
48 points
7 comments15 min readLW link

Build­ing AI safety bench­mark en­vi­ron­ments on themes of uni­ver­sal hu­man values

3 Jan 2025 4:24 UTC
18 points
3 comments12 min readLW link
(docs.google.com)