Jacob Steinhardt

Article

Jacob Steinhardt is a recurring person in the Astral Codex Ten archive, appearing 6 times across 6 issues between November 01, 2021 and March 12, 2024. The archive places it in contexts such as “Jacob Steinhardt: “Earlier this year, my research group commissioned 6 questions for professional forecasters”; “Jacob Steinhardt’s claims here that a recent article committed “journalistic malpractice””; “some people I know in tech endorse Jacob Steinhardt’s claims”. It most often appears alongside Metaculus, Polymarket, Manifold.

Metadata

  • Category: People
  • Mention count: 6
  • Issue count: 6
  • First seen: November 01, 2021
  • Last seen: March 12, 2024

Appears In

Source Context

Recovered passages from the original issue text. When the raw archive preserved outbound links inside the source passage, they are listed directly under the quote.

November 01, 2021 · Original source
— Jacob Steinhardt: “Earlier this year, my research group commissioned 6 questions for professional forecasters to predict about AI.” Updates And Lessons From AI Forecasting.
July 01, 2022 · Original source
35: More in “NYT being bad” news - some people I know in tech endorse Jacob Steinhardt’s claims here that a recent article committed “journalistic malpractice” and “platform[ed] a bully” .
July 29, 2022 · Original source
21: Jacob Steinhardt on one year of AI forecasting: “While forecasters underpredicted progress on capabilities, they overpredicted progress on robustness. So while capabilities are advancing quickly, safety properties may be behind schedule.”
August 16, 2022 · Original source
5: Jacob Steinhardt describes the results of his AI forecasting contest last year. Short version: AI is progressing faster than forecasters expected, safety is going slower. Uh oh.
August 28, 2023 · Original source
4: Jacob Steinhardt reviews the first two years of AI forecasting:
March 12, 2024 · Original source
The first team is Halawi et al at Berkeley (also including Jacob Steinhardt, featured here before). They cite previous work on out-of-the-box AIs like GPT-4 or Claude. When these enter forecasting tournaments, they might beat some especially unskilled participants, but they lag behind the easiest aggregation method: “the wisdom of crowds”, ie a simple average of all forecasts. The wisdom of crowds is hard to beat - in my tournament, it scored at the 95th percentile.