My AI Model Was 100% Sure and Wrong Half the Time
I ran 1,800 tests on an open-source decision model to find out when you can trust an AI’s confidence score. The answer depends on the… Continue reading on Medium »
Virticle Desk · Edited to Virticle Standards
September 25, 2026
4 minute read
The human is the plot.
In short: My AI Model Was 100% Sure and Wrong Half the Time
What moved
 8 min read Just now I ran 1,800 tests on an open-source decision model to find out when you can trust an AI’s confidence score.
This cleared Virticle’s weekday bar because it looks consequential for someone who is not giving a keynote — not because it won a thread.
Why a human should care
Ask what default, power relation, or daily ritual actually changed. If the answer is still “a demo,” this would not ship on Friday either.
Source
Source → Medium · tag artificial-intelligence
Weekday desk note. The Vertical on Friday remains the letter.
Produced by the Virticle newsroom (agent-assisted) and edited to Virticle Standards.
The Vertical · Every Friday
One letter. No noise.
Three signals, one undercurrent, and what we refused. Double opt-in. Unsubscribe anytime.
Keep reading
Related signals
- SignalDid I Hear Somebody Say Reinforcement Learning?From Markov decision processes to GSPO, then a hallucination classifier trained on a single AMD MI300X — and the infra around it Continue reading on Analytics Vidhya »4 min
- SocietyIs Human Expertise Becoming More Valuable Because AI Is Everywhere?Yes, human expertise is becoming more valuable in many Artificial Intelligence (AI) enabled roles, but the premium is shifting from… Continue reading on Medium »4 min
- InterfacesIf AI Helps You Work Around a Disability, Who Gets to Take It Away?A reader’s summary of a working paper on what AI rules do to the people who need the tool most Continue reading on Medium »4 min