The Blame That Never Runs Out

Here's the single most dangerous idea in modern discourse: that you can be morally responsible for violence committed by a stranger you've never met, based purely on the statistical likelihood that someone, somewhere, might act on words you said to millions of people.

This is "stochastic terrorism" — and it's a trap disguised as an insight.

The Seductive Logic

The argument sounds rigorous: if a pundit says "X is evil" to ten million listeners, and statistically some tiny fraction of unstable people will hear "kill X," then the pundit has essentially committed murder-by-proxy. No explicit call to violence needed. The math does the incriminating.

It borrows the language of actuarial science — the same logic insurers use to predict how many drivers will crash this year without predicting which ones. Applied to speech, it feels scientific. It isn't.

Why the Framework Collapses

1. It's unfalsifiable. Any sufficiently large audience produces some violent outlier eventually. This means literally any widely-heard opinion — political, religious, scientific — can be retroactively declared "stochastic terrorism" once someone unstable acts. The theory never fails to find a culprit; it just waits for tragedy, then reverse-engineers blame.

2. It has no principled stopping point. If rhetoric-plus-probability equals guilt, then climate activists are responsible for eco-terrorists, war critics are responsible for deserters-turned-shooters, and civil rights leaders are responsible for riots. The logic doesn't discriminate by ideology — only its application does. That selectivity is the real signal.

3. It confuses correlation with intent. Moral responsibility has always required some connection between what you meant and what happened. Stochastic terrorism severs that link entirely. You can be blamed for outcomes you actively condemned, simply because your audience was large enough.

A Better Mental Model: The Responsibility Gradient

Instead of binary "guilty/innocent," think in terms of a gradient:

• Direct incitement ("Go attack them now") → high responsibility.
• Dehumanizing rhetoric with explicit calls to action → moderate responsibility.
• General criticism, even harsh → responsibility rests with the actor, not the speaker.
• Popularity alone → no responsibility. Size of audience is not evidence of intent.

This gradient forces us to ask what was actually said, not how many people heard it.

Why This Matters Beyond Politics

The stochastic terrorism frame isn't really about preventing violence — it's a weapon for silencing disfavored speech while exempting favored speech from the same standard. Once you accept "large audience + eventual violence = guilt," you've built a censorship machine that runs on outrage, not evidence. It will always find a new target, because tragedy is inevitable and audiences are large.

The antidote isn't defending inflammatory speech — it's defending the principle that blame requires intent and proximate cause, not statistical odds.

The Real Insight

Ideas are dangerous. So is the idea that ideas are automatically dangerous enough to warrant suppression whenever misfortune follows them. The moment we let probability substitute for culpability, we've handed enormous power to whoever gets to define "harmful" — and history shows that power rarely stays neutral.

Takeaways

• Stochastic terrorism is unfalsifiable — it always finds someone to blame after the fact.
• Audience size ≠ intent. Responsibility requires a causal, not statistical, link.
• Use the Responsibility Gradient to separate incitement from ordinary speech.
• Watch who wields the frame — its selective application reveals its true purpose.

Sources & Further Reading

https://www.astralcodexten.com/p/against-stochastic-terrorism