Aleksandr Bowkis

Karma: 174

I am a research scientist on the Alignment team at UK AISI. I work on scalable oversight and understanding the risks from, and obstacles to, automating alignment research.

Eliciting hidden knowledge from monitors with NLAs

David Africa and Aleksandr Bowkis

15 Jul 2026 13:51 UTC

27 points

0 comments9 min readLW link

Oversight of automated research via summarisation: a toy model

Aleksandr Bowkis and Konstantinos Voudouris

13 Jul 2026 17:06 UTC

21 points

1 comment14 min readLW link

Automated Alignment is Harder Than You Think

Aleksandr Bowkis, Marie_DB, Jacob Pfau and Geoffrey Irving

14 May 2026 22:01 UTC

143 points

7 comments3 min readLW link

(arxiv.org)

Aleksandr Bowkis 13 May 2026 17:23 UTC
3 points
0
on: From personas to intentions: towards a science of motivations for AI models
Great post and very clear! I buy the push to study motivations rather than behaviours, and like the inverse constitution idea.