Depth Profile · 6 min read · Personality Science

Big Five vs MBTI: Which Personality Framework Works Best for AI Prompts?

If you've ever tried to personalize ChatGPT, Claude, or Gemini by telling it your personality type, you've probably used one of three frameworks: Myers-Briggs (MBTI), the Enneagram, or the Big Five.

They all claim to describe who you are. But when it comes to feeding personality data to an AI, they are not remotely equal. One of them gives AI models dramatically better signal — and it's probably not the one you've used before.

The Three Contenders

MBTI (Myers-Briggs Type Indicator)

You know this one. Four letters: INTJ, ENFP, ISTP. It's the most popular personality framework in the world — roughly 2 million people take the official test every year, and "What's your type?" has become a dating app icebreaker.

How it works: You're sorted into one of 16 types based on four binary dimensions: Introversion vs. Extraversion, Sensing vs. Intuition, Thinking vs. Feeling, Judging vs. Perceiving.

The problem for AI: MBTI is binary. You're either an I or an E, a T or an F. There's no spectrum, no gradation. If you score 51% Thinking and 49% Feeling, you get the exact same label as someone who scores 99% Thinking. That's a massive loss of information.

For AI prompts, "I'm an INTJ" tells the model you're probably analytical and independent — but it doesn't tell it how analytical, how independent, or how you compare to other INTJs (who can be wildly different from each other).

There's also the reliability problem. Research shows that nearly 50% of people get a different MBTI type when they retake the test within five weeks. A personality framework that can't even agree with itself isn't giving AI stable data to work with.

Enneagram

Nine types, each with a core fear and core desire. Type 3 (The Achiever), Type 5 (The Investigator), Type 8 (The Challenger), and so on.

The problem for AI: The Enneagram is motivational, not behavioral. It describes why you do things, not how you do them. An AI model generating a response needs behavioral data — how do you want information structured? How direct should feedback be? How much detail do you want?

Telling ChatGPT "I'm a Type 5" helps it understand you value knowledge and privacy. But it doesn't tell it whether you prefer bullet points or paragraphs, whether you handle criticism well, or whether you want creative suggestions or conventional ones.

Big Five (OCEAN Model)

Five dimensions, each measured on a continuous spectrum from 0 to 100:

• Openness to Experience — curiosity, creativity, preference for novelty
• Conscientiousness — organization, discipline, attention to detail
• Extraversion — sociability, assertiveness, positive emotionality
• Agreeableness — cooperation, empathy, conflict avoidance
• Neuroticism — emotional volatility, anxiety, stress reactivity

Why this works for AI: Every dimension maps directly to how you want an AI to communicate with you.

Why Big Five Wins for AI Personalization

Here's the core insight: AI models need quantitative, behavioral data — not categorical labels.

When you give an AI your Big Five scores, each dimension translates directly into response calibration:

MBTI can't do this because it's binary. Enneagram can't do this because it describes motivation, not behavioral preferences. Big Five does exactly this because it was designed to measure observable personality traits on continuous scales.

The Spectrum Advantage

Consider two people who both score as "Introverted" on MBTI:

• Person A: Big Five Extraversion score of 12/100 — deeply introverted, finds most social interaction draining, strongly prefers written communication
• Person B: Big Five Extraversion score of 42/100 — mildly introverted, enjoys small groups, can do presentations when needed

MBTI gives both of them the same label: "I." Big Five captures the 30-point difference that fundamentally changes how an AI should interact with them.

The Research Backing

This isn't just opinion. The Big Five model is the most validated personality framework in psychological research. It emerged from decades of factor analysis — researchers analyzed thousands of personality-describing words across multiple languages and consistently found the same five dimensions.

• Test-retest reliability: 85-90% consistency vs. MBTI's roughly 50%
• Cross-cultural validity: Replicated in over 50 countries and dozens of languages
• Predictive power: Big Five scores predict job performance, relationship satisfaction, and health outcomes better than any competing framework
• Continuous measurement: Captures nuance that categorical systems miss

How to Use Big Five Scores With Your AI

The practical application is straightforward:

• Take a Big Five assessment. Depth Profile offers a free assessment battery that includes Big Five scoring along with nine other personality dimensions.
• Get your scores. You'll receive a percentage score for each of the five dimensions, plus a narrative explanation of what your specific combination means.
• Feed it to your AI. Paste your profile into ChatGPT's custom instructions, Claude's system prompt, or Gemini's context. The AI immediately calibrates its communication style to your personality.
• Experience the difference. The AI stops being generically helpful and starts being specifically helpful — for your brain, your communication style, your preferences.