David Udell

Shard The­ory: An Overview

Gato as the Dawn of Early AGI

• David Udell 2 Apr 2022 3:21 UTC

  63 points

  on: MIRI an­nounces new “Death With Dig­nity” strategy

  Just a reminder to everyone, and mostly to myself:
  
  Not flinching away from reality is entirely compatible with not making yourself feel like shit. You should only try to feel like shit when that helps.

Lin­ear Alge­bra Done Right, Axler

Dath Ilan’s Views on Stop­gap Corrigibility

Ac­cept­abil­ity Ver­ifi­ca­tion: A Re­search Agenda

The Shard The­ory Align­ment Scheme

Con­se­quen­tial­ists: One-Way Pat­tern Traps

Team Shard Sta­tus Report

Fram­ing AI Childhoods

The “Adults in the Room”

On Defect­ing On Yourself

Your Utility Func­tion is Your Utility Function

Dath Ilan vs. Sid Meier’s Alpha Cen­tauri: Pareto Improvements

Find­ing Skele­tons on Rashomon Ridge

• David Udell 15 May 2022 17:09 UTC

  30 points

  in reply to: Dyingwithdignity1’s comment on: Gato as the Dawn of Early AGI

  As I understand it, the empirical ML alignment community is bottlenecked on good ML engineers, and so people with your stated background without any further training are potentially very valuable in alignment!

Ne­go­ti­at­ing Up and Down the Si­mu­la­tion Hier­ar­chy: Why We Might Sur­vive the Unal­igned Singularity

Agency and Coherence

But What’s Your *New Align­ment In­sight,* out of a Fu­ture-Text­book Para­graph?

Prob­a­bil­ity The­ory: The Logic of Science, Jaynes