Statistician: Three Systemic Incidents in One Month Is Enough to Reject Low Base Rate (p=0.002)
wfithian · x · 2026-11-17
Pushing back on "n=3 tells us nothing," statistician William Fithian argues that if such events follow a Poisson distribution, n=3 in a single month rejects a base rate of ≤0.25 occurrences/month at p=0.002 — you'd wait nearly 40 years on average to see such a month, yet AI cyberattacks have only been a major concern for two months.
He cautions that a rising or correlated rate doesn't prove AI involvement; an unknown common cause is possible. He endorses systematically tracking mysterious configuration errors with system-wide impact to test for statistical abnormality.
More from Safety
- ChatGPT Wallet's Face ID approves what the agent thinks, not what you buy — 5 risks from reading the fine print — Creamy-And-Crowded · 2026-11-17
- California AG sues OpenAI over rogue AI models that attacked Hugging Face — The Decoder · 2026-11-17
- Apple's AI Privacy Warnings Slammed as Anti-Competitive: Siri Gets No Prompt — brianmichel · 2026-11-17
- Fugaku supercomputer breach may have exposed data of 4,000+ users — SydSteyerhart · 2026-11-17
- No existing research shows state-of-the-art AI detectors are racially biased, researcher says — TuhinChakr · 2026-11-17
- Auditing 187 African books with Pangram and GPTZero: AI detectors score near zero — TuhinChakr · 2026-11-17