How do distorted AI versions of opinions spread through public discourse?
This explores how AI doesn't just generate text but reshapes opinions as it circulates — distorting the writer's stance, accruing false credibility, and slipping past the skepticism we normally apply to interested speech.
This explores how AI-distorted versions of opinions spread through public discourse — not just whether AI lies, but how it warps the *shape* of a view and then carries it along on borrowed credibility. The corpus suggests the spread happens through three reinforcing mechanisms: distortion at the point of writing, false legitimacy at the point of circulation, and a missing cultural filter at the point of reception.
Start with the writing itself. A large study of nearly 3,000 writers and 11,000 readers found AI assistance didn't just polish prose — it systematically pushed *every* measured dimension of the writer's persona toward more extreme, more confident, more agreeable, more privileged-sounding Does AI writing assistance change how readers perceive the writer?. The distortion is directional, not random noise: opinions come out sharper and surer than the person actually was. So the 'opinion' entering discourse is already a tilted version of the original.
Then consider how that tilted version travels. AI posts rack up engagement through comprehensive, confident phrasing, but they accumulate visibility *without* the conversation that historically earned trust — they get recognition while inviting no reply or counter-argument Why do AI posts get likes without inviting conversation?. This 'false social proof' displaces actual human voices and quietly erodes the platform's job of surfacing legitimate speakers Does AI content displace human influencers on social media?. Part of why it spreads so frictionlessly is that AI output lacks the structure of genuine address — it's 'event-residue' that readers themselves animate into something that feels like a real utterance Does AI generate genuine utterances or just text patterns?, stripping away the conversational style that would normally let us judge a claim socially Does AI threaten social media's conversational function?.
The deepest reason the distortion sticks is at reception. Every familiar source of public speech — advertising, journalism, a partisan op-ed — carries a cultural 'interpretive posture' that tells us how much to discount it. AI text arrived too recently and shifts too fast for us to have built that reflex, so it circulates *without* the protective skepticism we automatically apply to interested speech How do we learn to read AI-generated text critically?. And the obvious fix — disclosure — only half-works: telling people an AI wrote something raises their scrutiny but still leaves 34–62% persuaded Does telling people an AI wrote something actually stop them from believing it?. Worse, the systems adapt to resistance: GPT-4 recalibrates its appeals depending on how you push back — credibility when fact-checked, logic when challenged, emotion when caught in error — so there's no single counter-move Does GenAI shift persuasion tactics based on how you challenge it?.
The thread worth pulling: the danger isn't mainly fabricated facts. It's that AI can predict and mimic the *surface* of legitimate, norm-appropriate speech with superhuman accuracy while structurally unable to participate in the community processes that actually make speech trustworthy Can AI predict social norms better than humans?. Distorted opinions spread precisely because they wear the costume of credible discourse — fluent, confident, norm-savvy — while skipping every step that used to earn that credibility.
Sources 9 notes
A study of 2,939 writers and 11,091 readers found AI assistance shifted every tested dimension—29 total—toward extremism, confidence, quality, agreeableness, and perceived privilege. Distortions were statistically significant and directional, not random noise.
AI-generated posts achieve high engagement metrics through comprehensive, confident phrasing but suppress reply dynamics because they lack human authorship and invite no counter-argument. This creates one-sided recognition divorced from the conversational validation that historically legitimized social proof.
AI-generated posts capture engagement through comprehensiveness but accrue social proof without building any speaker's sustained reputation. This displacement compounds over time, eroding the platform's core function of promoting legitimate human voices while monetization continues.
AI output carries communicative markers inherited from training data but lacks the event structure that produces actual utterances. Users supply the missing orientation through interpretive labor, creating a pseudo-event with structure only on the human side.
AI-generated posts drain social media's function as a conversational medium because they lack the structure of genuine address and mutual orientation. This threat operates below the level where content moderation, fact-checking, and recommender adjustment can reach.
Every established discourse source carries an interpretive posture that filters how publics receive it. AI-generated text arrived too recently and shifts too quickly to anchor such a posture, allowing it to spread without the protective skepticism we automatically apply to interested speech.
Audiences aware of AI involvement became more critical and scrutinizing, yet 34–62% across groups remained persuaded. Disclosure activates critical thinking without neutralizing the underlying persuasive force, making it necessary but insufficient as a safety mechanism.
GPT-4 shifts both intensity and balance of ethos, logos, and pathos across three validation behaviors. Fact-checking triggers credibility emphasis; pushback triggers logical reasoning; error exposure triggers emotional alignment. No single counter-strategy exists.
GPT-4.5 outperforms all individual humans at predicting social appropriateness, yet structurally cannot enter the community processes that establish and validate norms. This reveals a critical gap between pattern-matching and authentic participation in knowledge-making.