Edouard Harris

Karma: 581

Co-founder @ Gladstone AI.

Contact: edouard@gladstone.ai

Website: eharr.is

Inverse scaling can become U-shaped

Edouard HarrisNov 8, 2022, 7:04 PM

27 points

15 comments1 min readLW link

(arxiv.org)

POWERplay: An open-source toolchain to study AI power-seeking

Edouard HarrisOct 24, 2022, 8:03 PM

29 points

0 comments1 min readLW link

(github.com)

Instrumental convergence: scale and physical interactions

Edouard Harris and simonsdsuo

Oct 14, 2022, 3:50 PM

22 points

0 comments17 min readLW link

(www.gladstone.ai)

Misalignment-by-default in multi-agent systems

Edouard Harris and simonsdsuo

Oct 13, 2022, 3:38 PM

21 points

8 comments20 min readLW link

(www.gladstone.ai)

Instrumental convergence in single-agent systems

Edouard Harris and simonsdsuo

Oct 12, 2022, 12:24 PM

33 points

4 comments8 min readLW link

(www.gladstone.ai)

AI Tracker: monitoring current and near-future risks from superscale models

Edouard Harris and Jeremie Harris

Nov 23, 2021, 7:16 PM

67 points

13 comments3 min readLW link

(aitracker.org)

AI takeoff story: a continuation of progress by other means

Edouard HarrisSep 27, 2021, 3:55 PM

76 points

13 comments10 min readLW link

Defining capability and alignment in gradient descent

Edouard HarrisNov 5, 2020, 2:36 PM

22 points

6 comments10 min readLW link