Larks

karma rank #9 · 16,846 karma · 9 posts sampled · profile ↗

Top cause areas

longtermism xrisk 6.2 +1.8σ
ai safety 5.0 +2.5σ
ai governance 4.8 +3.5σ
ea meta 4.7 -0.4σ
Show all 15 cause areas
effective giving 4.2 +1.1σ
prioritization 3.1 +0.1σ
rationalist meta 2.6 +0.9σ
forecasting epistemics 2.3 +0.3σ
nuclear great power 1.7 +1.7σ
moral circle values 0.8 -0.4σ
biosecurity 0.7 +0.3σ
global health 0.4 -0.6σ
climate 0.4 +1.3σ
animal welfare 0.0 -0.7σ
mental health 0.0 -0.4σ

Distinguishing dimensions z-score vs top-30 cohort

▲ cause_ai_governance +3.5σ
▲ stance_ai_doom_axis +3.0σ
▲ cause_ai_safety +2.5σ
▼ style_personal_narrative -1.2σ
▼ tone_earnestness -1.1σ

Style quantitative / empirical / philosophical / speculative / operational / contrarian

quantitative 4.0 -0.0σ
empirical 4.6 +0.6σ
philosophical 4.2 +0.4σ
speculative 5.3 +1.1σ
operational howto 5.1 +0.5σ
critical contrarian 5.3 +1.2σ
Show all 9 style dims
personal narrative 1.6 -1.2σ
synthesizing 6.3 +0.9σ
question asking 2.3 -0.4σ

Stance signature top 3 stance dims furthest from neutral (5)

longtermist framing 7.8 +1.5σ
inside ea 7.4 -0.5σ
rationalist coded 7.1 +1.8σ
Show all 9 stance dims
ea org coded 6.8 +0.0σ
ai doom axis 6.6 +3.0σ
optimism ea institutions 5.2 -0.9σ
suffering focused 5.1 -0.1σ
animal moral status 5.0 -0.7σ
welfarist vs rights 5.0 -0.7σ

Tone

combativeness 2.9 +1.8σ
calibration 6.4 +0.2σ
jargon 6.1 +1.3σ
humility 6.2 +0.6σ

Quality

rigor 6.6 +0.3σ
originality 5.9 +0.4σ
clarity 8.0 -0.3σ

Neighbors in dimension space

Nearest: William_MacAskill

Foil: Ben_West🔸

Most characteristic post

Rolling Thresholds for AGI Scaling Regulation

baseScore: 60

This is a plan for how ASI could be relatively safely developed. Abstract: A plan that puts all frontier model companies on a unified schedule of model training, evaluation and approval, with regulatory compliance promoted through market access. This aims to combine (most of) the economic benefits of unrestricted competition but with more safety, (most of) the time-to-think benefits of AI pauses but with better compliance incentives, and (most of) the central oversight of a M…

Highest-karma post in sample

2021 AI Alignment Literature Review and Charity Comparison

baseScore: 176