AI ALIGNMENT FORUM
AF

Paul Bricman

Sequences

Hypothesis Subspace

Posts

Sorted by New

11Cataloguing Priors in Theory and Practice8mo

0

15(Structural) Stability of Coupled Optimizers8mo

0

11Interlude: But Who Optimizes The Optimizer?8mo

0

17Representational Tethers: Tying AI Latents To Human Ones9mo

0

7Ideological Inference Engines: Making Deontology Differentiable*9mo

0

14Oversight Leagues: The Training Game as a Feature9mo

0

17Benchmarking Proposals on Risk Scenarios9mo

0

10Steelmining via Analogy10mo

0

Wiki Contributions

Comments

Linda Linsefors's Shortform

Paul Bricman8mo1

You mean, in that you can simply prompt for a reasonable non-infinite performance and get said outcome?