Rejecting AI is lucidity, not ignorance
by Miguel Lucas
Social resistance to artificial intelligence is growing. And the industry’s response is always the same: more demos, more use cases, better storytelling. The implicit assumption is that people reject AI because they don’t understand it. That approach has been failing for three years. What’s actually going on?
The diagnosis is wrong. Rejection isn’t growing where there’s ignorance — it’s growing where there’s more exposure. According to the U.S. Federal Reserve, AI has produced 500,000 fewer programmer jobs than would have existed without it 1. Graduate vacancies in the UK tech sector have fallen 46% 2. The most exposed are the most resistant. That doesn’t square with “they just haven’t tried it properly.”
What does square is the political economy reading. The IMF puts it plainly: while AI may reduce wage inequality by displacing high-income workers, it is likely to substantially increase wealth inequality as those same gains accumulate in capital and dominant firms 3. The technical question is settled. The distributional one is not. And people, without the IMF report in hand, are already responding to that reality.
The asymmetry is clear. On the cost side: 74% of new content on the web is produced wholly or partly by algorithms and degrades everyone’s experience 4; 60% of Google searches generate no click through to the original creator 4; companies are automating the entry-level tasks that served as the on-ramp for new graduates in tech 2. On the benefit side: according to a Goldman Sachs analysis, AI-linked companies now account for nearly 45% of S&P 500 market capitalization, up from roughly 25% at the time of ChatGPT’s launch 5.
People don’t need that vocabulary to feel that something is being distributed unfairly. They feel it in the feed saturated with synthetic content, in the search engine that no longer leads anywhere, in the colleague who won’t be hired because “AI already does it.” And when the industry responds with more storytelling, what people hear is an explanation of why they should be happy with that distribution.
AI labs diagnose the resistance as a perception problem and respond with better storytelling: improved demos, literacy campaigns, future-focused narratives. They fail. Not because they communicate poorly, but because they’re trying to solve with narrative what is a political economy problem. The key question they don’t focus on is the one that actually matters: are we building something that redistributes value or extracts it?
As long as the benefits stay concentrated and the costs stay socialized, no campaign will reverse the resistance. Because people don’t reject AI out of ignorance. They reject it out of lucidity. And there’s no storytelling that survives a balance sheet everyone knows how to read.