Observed arrival · 2026-08-22
Mik Idrizovic’s Archive of AI Premise Failures
An independent AI-safety research site documenting how false premises enter, persist, and govern long-horizon language-model interactions.
Why it surfaced
The archive gives its argument several evidentiary shapes: 18 published case reports, 25 transcript records, a behavioral taxonomy, and a negative control. Its central concept, “Premise-Driven Error Cascades,” is paired with falsifiable claims, provenance notes, and reproduction steps rather than just broad safety rhetoric.
An independent AI-safety research site documenting how false premises enter, persist, and govern long-horizon language-model interactions.
Observed signals
Read the marks
Editorial observations of this landing page, not a rating.
One card from the complete issue