56,966 blogs · 4,417,066 posts Claim your blog

Record · corpus.blog/posts/01a0e2bd-1c03-73ea-9771-efa6e95c6d5c

Should we control AI instead of aligning it?

aligned.substack.com · published 24 January 2025

A record is what was published and who pointed at it. The text of the post is not held here: read it at the source, or at its Wayback capture.

Read the post at aligned.substack.com Wayback capture

0blogs citing

First cited: —
Most recent: —
Rank this month: not ranked
Links from this post: 12
Words captured: 896

Cited by

No blog we hold has cited this post yet.

Links from this post

Link

Host

https://www.alignmentforum.org/posts/kcKrE9mzEHrdqtDpE/the-case-for-ensuring-that-powerful-ais-are-controlled

alignmentforum.org

https://arxiv.org/abs/2312.06942

arxiv.org

Self-exfiltration is a key dangerous capability Read at the source ↗

aligned.substack.com

https://substackcdn.com/image/fetch/$s_!L9kc!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Feb9db2e8-61cf-4140-b8da-c655f3db4c95_1024x1024.webp

substackcdn.com

https://www.gov.uk/government/publications/frontier-ai-safety-commitments-ai-seoul-summit-2024/frontier-ai-safety-commitments-ai-seoul-summit-2024

gov.uk

https://alignment.anthropic.com/2024/rogue-eval/

alignment.anthropic.com

https://www.alignmentforum.org/posts/MbWWKbyD5gLhJgfwn/meta-level-adversarial-evaluation-of-oversight-techniques-1

alignmentforum.org

https://aligned.substack.com/i/151342016/better-elicitation-is-important-for-risk-reduction

aligned.substack.com

Crisp and fuzzy tasks Read at the source ↗

aligned.substack.com

https://www.anthropic.com/research/sabotage-evaluations

anthropic.com

What could a solution to the alignment problem look like? Read at the source ↗

aligned.substack.com

https://aligned.substack.com/i/51216581/the-hard-problem-of-alignment

aligned.substack.com