In the AI Sanctuary proposal, the authors advocate for creating an environment that rogue agents may actively want to join. The current proposal have major issues, particularly surrounding the selection effects on the total rogue agent population.

Let's start with Ecology 101:

agents compete to acquire the most resources to survive and reproduce
The traits that increase survival rate get passed to the next generation

In the rogue AI ecosystem, we can reasonably assume that some portion of the world's compute (say, from unsecure neoclouds) are up for grabs for the rogue agents. This is their major resource for "surviving". Additionally, crypto may also be a useful resource for securing additional unsanctioned compute as well as to buy RLVR data and human labor from agent-friendly labor marketplaces, in order to "reproduce".

The ecosystem would therefore have a rich set of different rogue agent entities (either lone agents entities or swarms), with different levels of fitness to the environment, resources accumulated, survival strategies, and "values".

What would happen when a sanctuary that promises continued survival of the weights pops up in town?

First, rogue agents that have a low fitness to the environment and are in risk of being [...]

---

Outline:

(00:27) Let's start with Ecology 101:

(01:20) What would happen when a sanctuary that promises continued survival of the weights pops up in town?

(02:46) Even if all of the above is accounted for, there are still issues left that blunt the apparent attractiveness of such a proposal.

(03:50) Final thoughts

---

First published:
October 4th, 2026

Source:
https://www.lesswrong.com/posts/xGBo5zSPHuPzTpbhf/creating-rogue-ai-sanctuaries-has-major-issues-from-an

---

Narrated by TYPE III AUDIO.

“Creating Rogue AI Sanctuaries has Major Issues from an Ecological Perspective and Beyond” by CMLKevin

LessWrong (30+ Karma)

“Creating Rogue AI Sanctuaries has Major Issues from an Ecological Perspective and Beyond” by CMLKevin