Maxime Riché

Con­crete re­search ideas on AI personas

Con­di­tion­al­iza­tion Con­founds Inoc­u­la­tion Prompt­ing Results

A Case for Model Per­sona Research

Longter­mist Im­pli­ca­tions of the Ex­is­tence Neu­tral­ity Hypothesis

The Con­ver­gent Path to the Stars

Other Civ­i­liza­tions Would Re­cover 84+% of Our Cos­mic Re­sources—A Challenge to Ex­tinc­tion Risk Prioritization

For­mal­iz­ing Space-Far­ing Civ­i­liza­tions Sat­u­ra­tion con­cepts and metrics

De­ci­sion-Rele­vance of wor­lds and ADT im­ple­men­ta­tions

Space-Far­ing Civ­i­liza­tion den­sity es­ti­mates and mod­els—Review

Longter­mist im­pli­ca­tions of aliens Space-Far­ing Civ­i­liza­tions—Introduction

Think­ing About Propen­sity Evaluations

A Tax­on­omy Of AI Sys­tem Evaluations

[Question] What are the strate­gic im­pli­ca­tions if aliens and Earth civ­i­liza­tions pro­duce similar util­ities?

Maxime Riché′s Shortform

Scal­ing of AI train­ing runs will slow down af­ter GPT-5

An illus­tra­tive model of back­fire risks from paus­ing AI research

Ex­pec­ta­tions for Gem­ini: hope­fully not a big deal

[Question] How should OpenAI com­mu­ni­cate about the com­mer­cial perfor­mances of the GPT-3 API?

[Question] Can we ex­pect more value from AI al­ign­ment than from an ASI with the goal of run­ning al­ter­nate tra­jec­to­ries of our uni­verse?