LessWrong (30+ Karma) podcast

“Delaying Internal Deployments Until After X Days of Safety Testing” by annawang

0:00
14:45
15 Sekunden vorwärts
15 Sekunden vorwärts

This post reflects my own takes, and not that of my employer's.

This is a simple idea: before a new model can be deployed internally, you must first test it for a minimum, fixed period.

You can adjust various axes of this proposal:

  1. Duration of the wait: in calendar time, or until you fulfill some condition.
  2. What it blocks: external deployment, internal deployment, or the training of future checkpoints.
  3. When the fixed period starts: from the first checkpoint, the final checkpoint, or any checkpoint that is “sufficiently similar / representative” to the one that is ultimately deployed.

One concrete target that I like is: “No checkpoint may be used internally for purposes beyond safety testing, until at least X calendar days after its hash is filed with an auditor.”

This is one specific implementation of internal deployment lags, an idea that the AI Futures Project (AIFP) briefly discusses in How to pace the US frontier as “AI R&D can only be conducted or assisted by AIs that were trained at least 9 months ago”.

What does this look like when implemented?

Let's run through a scenario to pin down the details of this proposal. The places where [...]

---

Outline:

(01:23) What does this look like when implemented?

(04:13) Open questions

(10:48) Could this be used as a speed dial?

(12:31) What are the benefits and drawbacks of this proposal?

---

First published:
November 11th, 2026

Source:
https://www.lesswrong.com/posts/XNLhpWgzhF8GYZ4Er/delaying-internal-deployments-until-after-x-days-of-safety

---

Narrated by TYPE III AUDIO.

Weitere Episoden von „LessWrong (30+ Karma)“