Latest Corrections

Browse completed fact-check investigations. Each correction identifies empirically incorrect or unambiguously misleading claims in online posts.

LessWrong Mar 1, 2026 at 02:22 AM
www.lesswrong.com/posts/QAB3BEDRziBerNAih/whack-a-mole-is-no...
  • The original 1969 “minimum tax” was generally an add-on tax of 10% of tax preference items over $30,000, not a calculation that subtracts regular tax paid from the preference amount and then taxes the remainder.
  • The Controlled Substances Act schedules include class-based definitions (e.g., “cannabimimetic agents” and “fentanyl-related substances”), not only individually named compounds.
2 corrections →
X Mar 1, 2026 at 01:35 AM
x.com/openerrata/status/staging-live-1772328796
  • This is false: the Moon is a rocky, metal-bearing body (with crust, mantle, and core), and returned lunar samples are rocks and soil—not cheese.
1 correction →
LessWrong Feb 28, 2026 at 08:36 PM
www.lesswrong.com/posts/hurF9uFGkJYXzpHEE/a-non-magical-expl...
  • Epstein did not serve 12 months of his Florida 2008 sentence; multiple sources (including a BoP psychological reconstruction) state he served 13 months of an 18‑month sentence (with work release).
  • The “15-page psychological report” was produced within the U.S. Bureau of Prisons (BoP), not by the FBI or DOJ’s Office of Inspector General (OIG).
2 corrections →
LessWrong Feb 28, 2026 at 08:34 PM
www.lesswrong.com/posts/4mvphwx5pdsZLMmpY/recent-ai-model-pr...
  • ARC Prize has publicly published evaluations involving DeepSeek and o3-mini (and also published ARC-AGI results for DeepSeek), so it’s not true that they haven’t publicized evaluations of these models.
1 correction →
X Feb 28, 2026 at 06:25 AM
x.com/powerbottomdad1/status/2027602251611558072
  • Major reporting on the Anthropic–Pentagon dispute says mass surveillance (of Americans) was explicitly one of the core issues being negotiated, so it wasn’t “never” about mass surveillance.
1 correction →
X Feb 28, 2026 at 06:20 AM
x.com/elidourado/status/2027588190929096833
  • This X URL is malformed; it returns a 404 Not Found. The correct post URL format is https://x.com/sama/status/<post_id>.
1 correction →
X Feb 27, 2026 at 08:31 PM
x.com/credistick/status/1869730728134447311
  • Andreessen Horowitz’s (a16z) investment in Anduril is publicly described by a16z and other sources as a Series B investment announced in October 2019 (and reported as part of a September 2019 round), not in 2017.
1 correction →
X Feb 27, 2026 at 06:35 PM
x.com/PalmerLuckey/status/2027078941190889905
  • This quotation is misdated/misattributed: the same wording (“I HAVE today issued... in the name of the United States Government”) is from Truman’s August 25, 1950 statement, not May 10, 1948. Truman’s May 10, 1948 Executive Order 9957 also specifies operation would continue “in the names of their respective companies,” not “in the name of the United States Government.”
1 correction →
X Feb 27, 2026 at 06:18 PM
x.com/slatestarcodex/status/2024838082923753823
  • The relevant preprint derived from the high-karma pro-lower-p(doom) entry (“AGI and the EMH…”) is authored by Trevor Chow, Basil Halperin, and J. Zachary Mazlish—Tyler Cowen is not a coauthor (he’s only acknowledged for feedback).
1 correction →
LessWrong Feb 26, 2026 at 03:53 PM
www.lesswrong.com/posts/4mvphwx5pdsZLMmpY/recent-ai-model-pr...
  • ARC Prize has publicly published evaluations involving DeepSeek and o3-mini (and also published ARC-AGI results for DeepSeek), so it’s not true that they haven’t publicized evaluations of these models.
1 correction →
LessWrong Feb 26, 2026 at 05:31 AM
www.lesswrong.com/posts/QAB3BEDRziBerNAih/whack-a-mole-is-no...
  • The original 1969 “minimum tax” was generally an add-on tax of 10% of tax preference items over $30,000, not a calculation that subtracts regular tax paid from the preference amount and then taxes the remainder.
  • The Controlled Substances Act schedules include class-based definitions (e.g., “cannabimimetic agents” and “fentanyl-related substances”), not only individually named compounds.
2 corrections →