Questions
Tournaments
Services
News

Questions
Tournaments

Questions

Questions

38 forecasters

5

38 forecasters

AI Technical Benchmarks

Will a dense machine learning model with at least 100 trillion parameters be trained before 2026?

ResolvedNo

No key factors yetAdd some that might influence this forecast.

Public Benefit Corporation

Metaculus is an online forecasting platform focusing on topics of global importance.

ExploreQuestions Tournaments Tournaments for AI bots FutureEval

ServicesLaunch a Tournament Private Instances Pro Forecasters

CompanyAboutCareers FAQ

ResourcesForecasting Resources For Journalists API

Guidelines Privacy Policy Terms of Use

10 comments53 forecasters

Billions of Param of Largest ML Model by 2030

Current estimate

24 comments98 forecasters

Greatest Computation Used in AI Training

3 comments28 forecasters

How larger will AI from non-authoritarian states be in 2030?

Current estimate

3.6×10²⁵ FLOP

3 comments8 forecasters

FLOP usage for training transformative AI

Current estimate

5.5×10³⁰ FLOP

0 comments25 forecasters

Dates Large AI Models Released

6 comments60 forecasters

When will an AI model using neuralese recurrence be publicly available?

Current estimate

22 comments30 forecasters

GPT-3 scaling plateau in < 3 OOMs

0 comments13 forecasters

Dates Publicly Downloadable Large AI Models