The lie got retweeted 70% more. Your fast brain hit share before the slow one read it.
June 19, 2026 · Daniel Kahneman, Thinking, Fast and Slow~5 min read
There's a half-second you never notice. Your thumb is hovering over share, a headline has just landed in your chest — you won't believe what they did — and something in you has already decided. Not after weighing it. Before. By the time the careful, skeptical part of your mind clears its throat to ask "wait, is this even true?", the post is gone, flung out to two hundred people. And here's the uncomfortable part: if that post was false, it just traveled faster than the truth ever could. Not because you're gullible. Because of how a human mind is built.
The 70% nobody wants to hear
Researchers at MIT did something unusually thorough. They tracked about 126,000 stories spreading across Twitter over more than a decade, sorted into true and false by independent fact-checkers, and watched how each one moved. The headline finding, published in Science in 2018: false news was roughly 70% more likely to be retweeted than the truth. It didn't just win on volume — it spread farther, faster, and deeper into the network, reaching more people in less time. The truth, plodding and dutiful, rarely reached more than a thousand. The lies routinely reached tens of thousands. Same platform, same people, wildly different speeds.
Two minds, and only one of them is fast
To understand why, borrow Daniel Kahneman's two characters — the framework that won him a Nobel Prize. System 2 is the slow one: deliberate, effortful, the part that checks sources and notices "this smells off." System 1 is the fast one: automatic, always-on, running on feeling and gut. And System 1 has a tic Kahneman called the affect heuristic — when something hits you emotionally, the feeling itself becomes the answer. You don't reason your way to "share this." A jolt of surprise or disgust does the deciding, and the reason gets invented afterward, if at all. The slow mind could step in. It usually doesn't bother. So the share fires on emotion, and the fact-check arrives, if it arrives, far too late.
An MIT study of ~126,000 stories on Twitter (Vosoughi, Roy & Aral, Science 2018) found false news was about 70% more likely to be reshared than the truth, and spread farther, faster, and deeper — driven by novelty and the emotions it stirred: surprise and disgust. In Kahneman's terms, the fast, feeling System 1 hits 'share' before the slow, fact-checking System 2 wakes up. But mind the guardrail (How to Think Straight About Psychology): the 70% is a reshare-odds figure from one dataset, not proof people are stupid — correlation is not causation, and one platform's numbers may not be the whole web.
Why lies are built to feel shareable
Now the cruel design flaw. What makes System 1 light up? Novelty and strong feeling. And a lie has an unfair advantage on both: it isn't bound by what actually happened, so it can be engineered to be maximally surprising, maximally outrageous. The MIT team measured exactly this — false stories scored higher on novelty, and the replies they drew were soaked in surprise and disgust, while true stories drew calmer feelings like sadness and trust. Think of it as a translation: the truth is a documentary, accurate and a little slow; a viral lie is a thriller, written for the gut. Your fast brain doesn't read documentaries. It can't put down a thriller.
And the more you see it, the truer it feels
There's a second gear that makes it worse, and Kahneman named it too: the availability cascade. The more often you encounter a claim, the more available it is in memory — and your fast brain quietly mistakes "easy to recall" for "likely to be true." So virality becomes a self-feeding loop: a surprising lie gets shared, which means more people see it, which makes it feel more true, which gets it shared again. A 2024 study in the Journal of Personality and Social Psychology showed something sharper still — when people simply saw that a post was going viral, they perceived the threat in it as more real and reacted with more outrage, even for a made-up trend the researchers invented from scratch. The crowd's attention had become evidence. It isn't.
Surprise is not the same as truth.
A lie can be engineered to feel astonishing; the truth can only be what happened. Your fast brain rewards the first and shrugs at the second.
What this means for your thumb tonight — and one honest caveat
You can't disable System 1; it's faster than you and you need it. But you can insert one deliberate beat before the share. Three checks, ten seconds: First, name the feeling — if a post made you feel a hot rush of surprise or disgust, treat that not as proof but as a flag, because that's exactly the emotional signature lies are tuned to trigger. Second, notice the virality itself — "everyone's sharing this" is not a fact-check; the crowd is not a source. Third, ask the one question your fast brain skipped: do I actually know this is true, or does it just feel true? And the honest caveat, the one a good scientist insists on: that 70% is a reshare statistic from one dataset on one platform — it shows lies and virality travel together, not that you are stupid, and correlation is not causation. The point isn't that people are fools. It's that we're fast. The fix is to be, for ten seconds, slow.
Framework from Daniel Kahneman's Thinking, Fast and Slow (卡尼曼《思考,快与慢》) — System 1 / System 2, the affect heuristic, and the availability cascade. Rigor guardrail in the spirit of Keith Stanovich's How to Think Straight About Psychology: correlation is not causation, and a single dataset is not the whole world. The "70% more likely to be retweeted" figure is from Vosoughi, Roy & Aral, "The spread of true and false news online," Science (2018), studying ~126,000 stories. The virality-and-outrage finding is from Puryear, Vandello & Gray, "Moral Panics on Social Media Are Fueled by Signals of Virality," Journal of Personality and Social Psychology (2024). Popular-science reading, not professional advice.