Interesting article, but MBTI is not just contested, the concensus since the 90s is that MBTI is pseudoscience [1][2][3][4][5], so it's nearly 40 years.
Would have been interesting to test against the Big five instead !
Even if pseudoscience, it seems from the article that results almost always came back the same. If all AI models claimed their spiritual astrological sign is Libra, I'd still find it interesting enough to check out.
So, I tried it too, ChatGPT came back INTJ. But Gemini Flash 3.1 claims ENFJ once, INTP another time... Maybe OP could share their prompt and we'd understand why the results were so consistent.
I’ve noticed that people in my field tend to have relatively similar personalities and relatively similar MB categories (close to INTJ). That combo is relatively rare within the broader population. Given this consistency I don’t think we can say it’s pseudoscience in a strong sense.
Quite interesting! Opus / GPT / Gemini are all very similar and then Grok is the weird one :P