
“Pacing AI via Compute Reallocation” by Thomas Larsen, romeo
This is a linkpost for pacing-ai.com. I encourage folks to read it via the site (which has interactive graphs and expandables), but we’ve copied the text here as well.
We offer a concrete proposal for pacing the frontier.
1. Transparency: Verified Compute Allocation Disclosures (VCAD)
2. Reallocation: Shift compute away from capabilities R&D
The first component is Verified Compute Allocation Disclosures (VCAD): AI companies regularly disclose their compute usage, and an independent auditor verifies the disclosures. The second component is to reallocate compute away from capabilities R&D and toward other use cases, such as serving customers and safety research (non-capabilities R&D).
We roughly estimate that pacing via compute reallocation starting January 2027, with a 15% cap, would delay the intelligence explosion by a year. This is most likely insufficient to prevent AI takeover (or mitigate other problems created by RSI), and so this year should be used to implement a more durable policy regime such as Plan A or Plan S.
Nearly every approach to pacing, including ones built on safety and security requirements, strongly benefits from VCAD:
- Any pacing proposal that actually slows down AI development must do so either by reducing one of the inputs to [...]
---
Outline:
(03:07) VCAD Implementation Proposal
(03:29) How AI companies could estimate compute usage
(10:00) Baseline Auditing
(13:22) Attacks & Loopholes
(16:05) Robust Auditing
(17:24) Implementation pathways
(17:59) Why would companies voluntarily pace?
(20:09) Implementation over time
(20:14) 1. Adoption
(20:55) 2. Coverage
(21:33) 3. Disclosure granularity
(22:38) 4. Combination with safety and security requirements
(23:12) 5. Verification
(24:31) Frequently Asked Questions
(24:35) Why pace?
(26:26) How much would compute reallocation slow things down?
(29:24) What about China?
(32:45) Why not a different mechanism for pacing?
(37:19) Supplements
The original text contained 6 footnotes which were omitted from this narration.
---
First published:
November 18th, 2026
Source:
https://www.lesswrong.com/posts/dBEDh2WeZfhuiDnJ5/pacing-ai-via-compute-reallocation
---
Narrated by TYPE III AUDIO.
---
Images from the article:
Apple Podcasts and Spotify do not show images in the episode description. Try Pocket Casts, or another podcast app.
Altri episodi di "LessWrong (30+ Karma)"



Non perdere nemmeno un episodio di “LessWrong (30+ Karma)”. Iscriviti all'app gratuita GetPodcast.








