Sorry, I Still Think I Am Right About The Media Very Rarely Lying
Read the original on Astral Codex Ten →
Follows up on
↳ The Media Very Rarely Lies — Essay · Dec 2022
Follow-up reading
↳ Highlights From The Comments On The Media Very Rarely Lying — Highlights on the media debate · Jan 2023
Summary
Defends his 'The Media Very Rarely Lies' thesis (companion to Bounded Distrust) by working methodically through commenters' proposed counterexamples and showing each is true-facts-presented-misleadingly rather than fabrication: the 2020-election claims (Fox honestly reports Rand Paul, who honestly reports the Vote Pattern Analysis blog, which found a real Milwaukee absentee spike and leapt to 'fraud'); the Kirsch vaccine-deaths poll (real poll, almost certainly misunderstood by respondents); 'COVID less deadly than flu' (true early, dumbly framed); Alex Jones's Sandy Hook (the FBI really did list zero murders, because state not local police investigated); the Obama birth certificate (real Adobe Illustrator layering artifacts); Scalia and the Order of St. Hubertus/Bohemian Grove (mostly true!); the Iraq-WMD defector; 'Russia running out of missiles.' Two payoffs: 'ban misinformation' isn't a primitive action (censorship needs the same subjective judgment calls as any opinion), and — more deeply — people desperately want a 'BE DUMB AND EVIL gear' that marks bad reasoners as fundamentally different, but conspiracy theorists are doing the same thing we do (reasoning under uncertainty; Confirmation Bias Is Just A Misfire Of Normal Bayesian Reasoning), just failing terribly.
Why this score
Quality 76 · Excellent. Excellent band, low. A clarifying, important elaboration of the media-rarely-fabricates thesis, patiently demonstrated across many cases, with the genuinely sharp and memorable conclusion that there is no 'BE DUMB AND EVIL gear' separating us from conspiracists. A notch below the parent Bounded Distrust (82) because it defends and extends rather than originates.
Claude’s paradigm shift 48 · Moderate. Moderate. The no-evil-gear / reasoning-under-uncertainty framing is a fresh, exportable point, but it builds directly on his Bounded Distrust and bias-as-misfire work.
Real-world impact 3 · Moderate. A clarifying elaboration of the 'media very rarely lies' thesis, patiently demonstrated across many proposed counterexamples, with the sharp conclusion that there is no 'BE DUMB AND EVIL gear' separating us from conspiracists. Conceptual influence within epistemics discourse, defends/extends rather than originates — modest RWI.