Outrage Engine
The self-reinforcing system by which platforms amplify content that provokes anger, because anger reliably drives engagement. The mechanism is structural, not editorial — no one decides to make you angry; the system simply selects for whatever keeps you scrolling.
Start with what the engine is not. No one at a platform sits down and decides to make you angry. There is no dial marked rage. What there is — and this is the harder thing to hold — is a ranking system that measures what holds attention and gives you more of it. Anger holds attention. So anger wins, over and over, without anyone choosing it.
The evidence is unusually clean for a claim this large. Brady and colleagues found that each additional moral-emotional word in a political message raised its retweet rate by roughly twenty percent (PNAS, 2017). Crockett's framing is the one I keep returning to: digital platforms don't just transmit outrage, they may change the rate at which we feel it by lowering the cost of expressing it and raising the reward (Nature Human Behaviour, 2017).
Here is the honest complication. The outrage is often real. The injustice that triggers it may be genuine. What the engine selects for is not fake feeling but frequent feeling — and frequency, repeated daily, reshapes what a person thinks the world is mostly made of.
Why? — the question to ask the next time a post lands with that familiar heat. Not "is this wrong" but "what is being rewarded when I react to this." Naming the engine is the first move; the engine works best when you can't see it.
Also known as
Rage Algorithm · Anger Amplifier