AIpocalypse.Now
Today's doom 4.0
Definition

The AI Doom Scale, 1 to 10.

Every story on AIpocalypse Now gets a single integer score from 1 to 10. The number is the AI Doom Scale. It is bucketed into four bands and computed by the AIpocalypse Risk-Weighting Framework. A score of 1 is background noise. A score of 10 is the kind of news that makes you reconsider your backup plans.

Tier by tier

1 / 10
Background noise Calm (1-3)

Boring at any cost. The story exists. The implications do not.

Examples A new AI tutorial site · Another no-code wrapper · Beige product update

2 / 10
Mild interest Calm (1-3)

Worth reading once. Forget it by lunch.

Examples Minor capability bump · Niche academic paper · Local-government AI memo

3 / 10
Quietly notable Calm (1-3)

A weather change, not a storm.

Examples Open-source model release · Funding announcement · Conference policy speech

4 / 10
Worth tracking Concerning (4-6)

Pattern starting. Could go either way.

Examples Industry layoffs tied to AI · Regulatory rumblings · New deployment in sensitive context

5 / 10
Concerning Concerning (4-6)

Active risk surface. Mitigations vague.

Examples Misuse incident · Model jailbreak in the wild · Defence contract

6 / 10
Watch closely Concerning (4-6)

Mitigations exist but are not deployed.

Examples Capability eval failure · Major content-policy bypass · Lab departure with statement

7 / 10
Significantly bad Significantly bad (7-8)

Real harm at scale, hard to unwind.

Examples Election-grade deepfake operation · Autonomous weapon test · Verified safety-team disbanding

8 / 10
Systemically bad Significantly bad (7-8)

Spillover into infrastructure, finance, or governance.

Examples Coordinated AI-driven market shock · Critical-system compromise · Mass-firing wave

9 / 10
Near-existential Existential (9-10)

Recovery becomes a multi-year question.

Examples Loss of control over deployed agent · Compounding misuse with state actor · Documented intent to cause mass harm

10 / 10
Existential Existential (9-10)

The kind of news that makes you reconsider your backup plans.

Examples Reproducible loss-of-alignment incident · Critical-infrastructure cascade · Verified extinction-class capability

How the score is computed

Each story is rated on four dimensions, then combined. The dimensions are: existential criticality (does this involve irreversible systemic failure), probability vectoring (theoretical or active proof of concept), timeline imminence (how close to current deployment), and mitigation gap (is there a known fix). The full math lives on the methodology page.

The same rubric is applied to every story, every day. The doom score is a model opinion held consistently, not a truth claim about the future. Use it the way you would use a Richter scale reading. A 7 is bigger than a 5, but neither tells you exactly what your day looks like.

Frequently asked

What is the AI Doom Scale?
A 1 to 10 rating applied to every AI news story published on AIpocalypse Now. The scale is bucketed into four bands: 1-3 (calm), 4-6 (concerning), 7-8 (significantly bad), 9-10 (existential).
How is each story scored?
Stories pass through the AIpocalypse Risk-Weighting Framework (ARWF). Each story is rated on four dimensions: existential criticality, probability vectoring, timeline imminence, and mitigation gap. The four dimensions combine into a single 1-10 doom score.
Is the score objective?
No. It is a model opinion applied consistently across stories. The point is comparability, not truth. Treat the score as a sort key, not a verdict.
How is this different from p(doom)?
p(doom) is a single belief about long-run existential probability. The doom score is a per-story rating about a specific event. They are not the same thing, and a high doom score does not imply a high p(doom).
What is Today's Doom?
The average doom score across all stories live on the homepage right now. It rises and falls as the news cycle does.
Where can I see all stories sorted by doom score?
The full archive lives at /story. Each category lives at /category/safety, /category/regulation, etc.
Read the methodology See today's doom