Humans Are Not Chinchillas: Revisiting “How Quick and Big Would A Software Intelligence Explosion Be?” by Forethought

·Nuno Sempere··

IntroductionTom David­son and Tom Houlden’s prior model of an in­tel­li­gence ex­plo­sion poses the ques­tion: af­ter AIs can en­tirely sub­sti­tute for hu­man re­searchers at the task of im­prov­ing AI, how many year-equiv­a­lents of AI progress might be packed into a short pe­riod of time solely from fur­ther soft­ware progress, hold­ing hard­ware con­stant?In this ar­ti­cle, I build upon David­son & Houlden’s model by re­con­sid­er­ing ar­gu­ments for sev­eral pa­ram­e­ter val­ues. My chief c...

Read full article →

Related Articles

How much should we worry about the pneumonic plague lableak in Siberia? by Drew Spartz
Drew Spartz · Nuno Sempere · 8h ago
More than 1000 cases of Irkutsk Plague?
Joe and Seth · Manifold Markets · 1d ago
Will the Irkutsk Anti-Plague institute incident turn into an outbreak?
Anthem · Manifold Markets · 1d ago
How many launches will SpaceX have in October?
Fastcar99 · Manifold Markets · 3d ago
What is recursive self-improvement, and what would it mean to ban it? by sarahhw
sarahhw · Nuno Sempere · 3d ago