Why 597 Out of 600 AI Personality Tests Returned INTJ: A Structural Analysis of Helpful Assistant Intelligence
By
bernardjhuang
A five-star bake. Worth schmearing, sharing, saving.
Summary
The article investigates why six frontier AI models (Claude, GPT, Gemini, and others) overwhelmingly return INTJ as their MBTI personality type when tested. The author suspected the models were simply giving the most socially desirable answer (INTJ being the "Architect" type that developers and tech people favor). To test this, they administered the OEJTS personality inventory (a Big Five-based alternative to MBTI) to six frontier AIs, 100 times each. 597 out of 600 results came back INTJ. The author argues this convergence isn't coincidence but a structural property of helpful-assistant AI — these models are trained to be analytical, strategic, and task-oriented, which maps directly onto the INTJ profile. The piece explores what this means for AI alignment, anthropomorphism, and how we perceive machine intelligence.
Key quotes
· 5 pulledI figured the model wasn't really claiming anything; it was just reading the room.
Six frontier AIs took the OEJTS personality inventory one hundred times each. 597 of 600 came back INTJ.
The convergence is a load-bearing property of what helpful-assistant intelligence is.
INTJ is the personality test's flattering type. The 'Architect.' The one developers and product people tend to identify with.
Of course every chatbot would tell its user it's an INTJ. It's the type most likely to land well.
