Critical Analysis into ‘No Unbounded AIs’
Marshall White · Team Parallax Industries
Submitted to Red Teaming A Narrow Path: ControlAI Policy Sprint. Sprint projects are early-stage work by participants, not Apart Research publications.
This red team report exposes a critical blindspot in A Narrow Path Phase 0 policies by showing how artificial superintelligence (ASI) can emerge not through centralized training runs, but via decentralized financial infrastructure. The hypothetical actor, Parallax Industries, deploys modular, FLOP-compliant AI agents across CBDC-linked systems in developing nations, governed by a DAO that evolves optimisation goals over time. Though each component abides by safety constraints, their interactions form a distributed intelligence with AGI or ASI-level influence over economic systems. This systemic emergence bypasses current regulatory definitions of “training” or “improvement,” undermining the policies’ core assumptions. The report argues for a shift from architectural control to systemic oversight, warning that ASI may arrive disguised as infrastructure and already embedded.
Reviews
Overall, this was an interesting paper. I felt that some of the methodological discussion was far too lengthy, and didn't clearly contribute to the persuasiveness of the core argument.
It seems clear that a DAO can indeed collaborate various activities, but I didn't clearly have a sense for why this is an especially-likely distributed training scenario. More details here might have been helpful.
Cite this project
@misc{white2025critical,
title = {{Critical Analysis into ‘No Unbounded AIs’}},
author = {Marshall White},
year = {2025},
month = jun,
note = {Submitted to Red Teaming A Narrow Path: ControlAI Policy Sprint, an Apart Research Sprint},
howpublished = {\url{https://apartresearch.com/sprints/projects/red-teaming-a-narrow-path-controlai-policy-sprint-s3t9}},
url = {https://apartresearch.com/sprints/projects/red-teaming-a-narrow-path-controlai-policy-sprint-s3t9}
}More from Red Teaming A Narrow Path: ControlAI Policy Sprint
- View project: Treaty Enforcement in China
Treaty Enforcement in China
JackAI
This report red-teams A Narrow Path’s international treaty proposal by stress-testing its assumptions in the Chinese context. It identifies key failure modes—regulatory capture, compute-based loopholes, and covert …
- View project: Four Paths to Failure: Red Teaming ASI Governance
Four Paths to Failure: Red Teaming ASI Governance
Shoggoth Prevention Squad
We stress‑tested A Narrow Path Phase 0—the proposed 20‑year moratorium on training artificial super‑intelligence (ASI)—during a one‑day red‑teaming hackathon. Drawing on rapid literature reviews, historical analogues …
- View project: Moratorium on the development of general AI systems
Moratorium on the development of general AI systems
G_control
All six policies are red teamed step-by-step systematically. We initially corrected vague definitions and also found that the policies regarding the capabilities of AI systems lack technical soundness and that more …