Public arguments are not primarily contests between the two visible disputants but performances meant to persuade a third, silent audience who compares competing cases. Large language models can manufacture plausible-sounding positions, but because they lack adversarial testing and social judgment, their arguments risk filling the public sphere with untethered rhetoric that looks persuasive but hasn't survived scrutiny.
— If true, this shifts how we should regulate, design, and use AI argument tools: focus less on policing content and more on preserving adversarial testing, provenance, and cues that signal which claims have been meaningfully contested.
Bo Seo
2026.09.11
85% relevant
Bo Seo’s advice emphasizes persuading onlookers and structuring arguments for people not directly engaged in the clash (e.g., clear signposting, framing, and audience‑facing clarity), which directly matches the claim that competitive debating’s function is to convince the silent audience rather than to score logical purity against an opponent.
Arnold Kling
2026.09.07
85% relevant
The article uses the Bundy–Morgenthau televised exchange to show how a debater's public performance (Bundy) persuaded audiences and deferred doubts even though the critic (Morgenthau) was substantively right; this directly illustrates the idea that debates function primarily to persuade passive publics rather than to discover truth.
Jesse Singal
2026.03.26
63% relevant
Singal's call for definitional clarity is implicitly addressed to the broader audience that consumes debates (not just participants): by making terms explicit, debates become informative for listeners rather than performative displays, aligning with the notion that debates should serve the silent audience.
Jerusalem Demsas
2026.03.15
100% relevant
Joe Weisenthal’s claim that LLMs can construct coherent arguments for any position and the author’s judge/debate analogy showing that the real target of public argument is the silent audience.