MeByFace started with a question that had been following me for years:
How can we help people understand themselves faster — without pretending that one tool has the final answer about who they are?
My interest in people came long before MeByFace.
Through business, leadership, sales, relationships, and thousands of conversations and observations, I kept seeing a pattern:
People often have more potential than they recognize.
They make important career and relationship decisions without clearly understanding their own natural tendencies.
They collect information, but still lack clarity.
Face reading became interesting to me as one additional lens.
AI made it possible to structure that lens more consistently.
But the most important part of our methodology is not the technology.
It is the boundary.
Short answer
MeByFace uses AI-assisted facial mapping to measure visible facial structure, then applies multiple interpretive frameworks to generate reflection hypotheses. We do not treat landmark measurements, traditional face-reading systems, or first-impression research as a scientifically validated personality diagnosis. The final step is always real-life validation: Insight is a hypothesis. Life is the evidence.
Key takeaways
- MeByFace is designed for structured self-reflection, not medical or psychological diagnosis.
- AI-assisted mapping can measure visible facial landmarks and proportions.
- Measurement and personality interpretation are different layers.
- Traditional face-reading systems are historical and cultural interpretive frameworks, not validated psychological tests.
- Research on first impressions tells us how people perceive faces, not who a person truly is.
- Agreement across several interpretive traditions is convergence across frameworks, not scientific proof.
- The user’s lived experience remains the final validation layer.
Why I became interested in face reading
My interest began with observation.
I noticed that sometimes I formed an impression of a person before I could fully explain why.
Not a mystical certainty.
An impression.
I became curious about why some people immediately projected authority, why others created trust quickly, and why certain communication or decision patterns seemed to repeat.
That curiosity led me toward traditional face-reading systems.
Then toward technology.
Then toward a much harder question:
What part of the process is actually measurable, what part is interpretation, and what part must remain uncertain?
That distinction became central to MeByFace.
Because curiosity can be useful.
False certainty is not.
The boundary matters more than the technology
A self-discovery product becomes dangerous when it quietly changes the question.
From:
“Could this be a useful pattern to examine?”
to:
“This is who you are.”
I do not want MeByFace to make that jump.
A face is one source of information.
A photo is one sample.
An interpretive system is one lens.
Your life is much larger than all three.
That is why I prefer to make the layers visible.
The MeByFace Reflection Stack
Our ideal method can be summarized in six steps:
Photo → Measurement → Interpretation Lenses → Convergence → Reflection → Life Evidence
1. Photo
Everything begins with an image.
Lighting, angle, expression, resolution, cropping, obstruction, and camera quality can affect what computer vision extracts.
A clear, front-facing image reduces avoidable measurement noise.
It does not turn interpretation into scientific truth.
2. Measurement
Computer vision can locate visible landmarks and calculate relationships between them.
Examples include relative distances, proportions, angles, positioning, and symmetry indicators.
This is the most objective layer of the process.
But measurement is still not meaning.
A distance is a number.
A personality interpretation begins only when a framework attaches psychological meaning to that number.
3. Interpretation lenses
MeByFace draws from several kinds of interpretive material.
Some are traditional face-reading systems.
Some relate to modern social-perception research.
Some are structured coaching-style interpretations that turn observations into questions a user can test.
These sources do not all have the same evidential status.
They should not be blended as if they do.
4. Convergence
If several interpretive systems point toward a similar theme, we can mark that as cross-framework convergence.
For example, several traditions may independently associate a cluster of facial characteristics with persistence, reserve, or social presence.
That convergence can make a theme more interesting to explore.
It does not make the theme scientifically proven.
5. Reflection
The result is translated into questions that matter in real life.
Do I make decisions internally before other people see the process?
Do I naturally take responsibility early?
Do other people experience me as more serious or authoritative than I feel?
Do I use creativity through systems, language, strategy, or invention?
A useful reflection is specific enough to test and humble enough to be wrong.
6. Life evidence
This is the most important layer.
If a report says you may be persistent, ask:
Where is the evidence?
When has persistence helped me?
When has it become stubbornness?
Do people who know me recognize this?
Does the pattern appear across several areas of life?
If the evidence is weak, do not force the interpretation.
If the evidence is strong, the insight may become useful.
Next step
Want to see the process before trying it?
Review the current MeByFace workflow, photo requirements, and product steps on the How It Works page.
What the AI actually does
The live MeByFace How It Works page currently describes AI-assisted facial landmark detection, proportion calculation, and multi-framework analysis.
Those engineering capabilities are useful.
They can improve consistency.
A machine can apply the same measurement rule repeatedly.
It can compare many variables quickly.
It can organize structured data without fatigue.
But there is a category error I want to avoid:
A precise measurement is not the same as a precise personality claim.
If the system locates a landmark accurately, that tells us something about the landmark.
It does not automatically tell us how accurately a personality trait has been inferred.
This distinction is central to trust.
Where interpretation begins
There are at least three very different sources of interpretation.
Traditional face-reading systems
Systems such as Chinese Mian Xiang, Indian Samudrika Shastra, Japanese Ninso, Korean Gwansang, Western physiognomy, and related traditions attach symbolic or character-related meanings to facial features.
These systems have historical and cultural importance.
They are not scientifically validated personality diagnostic systems.
We can study them.
Compare them.
Use them to generate reflection prompts.
But we should not call their traditional associations scientific facts.
Social-perception research
Modern psychology asks a different question:
How do people perceive faces?
Research by Janine Willis and Alexander Todorov showed that people can form trait impressions from unfamiliar faces after exposures as short as 100 milliseconds.
This tells us that faces influence social judgment very quickly.
It does not tell us that those judgments accurately reveal enduring inner personality.
Perception and truth are different questions.
That difference is useful.
If people repeatedly perceive your neutral expression as serious or reserved, that can influence real interactions even when the impression is incomplete.
Coaching-style synthesis
A third layer is the translation of an observation into a useful reflective question.
For example:
“Do people experience you as more decisive than you feel internally?”
This is not a diagnosis.
It is a prompt.
The value is what happens next.
You compare it with feedback.
Behavior.
Career history.
Relationships.
Decisions.
That is where self-discovery becomes grounded.
Why multiple traditions do not equal scientific validation
This is one of the most important distinctions on the entire site.
If five traditional frameworks independently produce a similar interpretation, that is interesting.
I call it cross-framework convergence.
But agreement among interpretive traditions is not the same as validation against an external scientific criterion.
Ten traditions do not magically become one validated personality test.
The same principle applies to AI.
An AI model can synthesize traditional interpretations more consistently.
It cannot create missing scientific evidence simply by processing them faster.
I would rather state the boundary clearly than build trust on language that sounds more certain than the evidence supports.
What we do not claim
I do not believe a responsible static face-analysis product should claim to scientifically determine:
- intelligence;
- morality;
- honesty;
- criminality;
- mental-health conditions;
- medical conditions;
- guaranteed career success;
- guaranteed relationship compatibility;
- a person’s destiny.
A report should also avoid false precision.
A score can be useful as an internal way of organizing output.
It should not be mistaken for a laboratory measurement of personality truth.
What AI is good at — and what it is not
I am optimistic about AI.
I use it.
I believe it can dramatically improve how we structure information, compare patterns, and personalize reflection.
But I do not believe AI replaces:
- empathy;
- context;
- human relationships;
- live conversation;
- intuition informed by experience;
- the person’s own lived evidence.
If an AI self-discovery product makes you feel that an algorithm knows you better than your own life does, something has gone wrong.
The best outcome is not:
“The machine has decided who I am.”
It is:
“The system has shown me a pattern worth examining.”
How to validate a MeByFace insight
Use three labels.
TRUE
This strongly matches lived experience.
MAYBE
Interesting. I want to test it.
NO
This does not fit.
Then take the three strongest TRUE or MAYBE statements and ask:
1. Where is the evidence in my decisions? 2. Where is the evidence in my relationships? 3. Where is the evidence in my work? 4. What would someone close to me say? 5. Is this pattern a strength, a cost, or both?
Now the report becomes a starting point for inquiry.
Not an endpoint.
Free Reading
Try the method with the right expectation.
Use the free reading as a hypothesis. Mark what fits, what does not, and what you want to test against real life.
Why disagreement is useful
A self-discovery product should leave room for the user to disagree.
If every statement must be accepted, the system is not encouraging reflection.
It is encouraging compliance.
Sometimes disagreement reveals a weak interpretation.
Sometimes it reveals a blind spot.
Sometimes another person sees the pattern more clearly.
Sometimes the report is simply wrong.
All four possibilities are useful if the product allows them.
I would rather have a user say:
“This part does not fit me.”
than force a false sense of accuracy.
Privacy is part of methodology
A face photo is sensitive personal data.
That means privacy is not a legal appendix.
It is part of the method.
The live MeByFace Privacy Policy currently describes the production data flow, providers, retention periods, and user rights. Those technical details can evolve, so this article should not become a frozen copy of architecture documentation.
For current information about photo handling, retention, providers, deletion, and user rights, use the live Privacy Policy.
A self-discovery product cannot ask users to be open while being vague about its own data practices.
Trust has to work in both directions.
A strong opinion: no self-discovery system should become your identity
Not face reading.
Not MBTI.
Not Enneagram.
Not a strengths test.
Not an AI report.
A framework is useful when it helps you see.
It becomes limiting when it tells you to stop looking.
This is why I prefer:
“This may be a pattern.”
over:
“This is who you are.”
The first opens inquiry.
The second closes it.
My goal with MeByFace is not to give people a more sophisticated box.
It is to give them another mirror.
What success looks like
The best outcome is not:
“AI knows me perfectly.”
It is:
“This helped me notice something useful about myself.”
Maybe it confirmed a strength.
Maybe it revealed a blind spot.
Maybe it gave language to something you already sensed.
Maybe it created one better conversation.
Maybe it changed one decision.
That is the standard I care about.
Clarity that becomes useful in life.
Not dramatic certainty.
Use MeByFace as another mirror — not a verdict.
Start free, compare the result with your lived evidence, and keep only what genuinely helps you see yourself more clearly.
Frequently asked questions
Is MeByFace a personality test?
It is better understood as an AI-assisted face-reading and self-reflection product. It can generate structured hypotheses about personality-related themes, but it should not be treated as a validated clinical or psychometric personality diagnosis.
Can AI know my personality from my face?
Not in the strong sense implied by a scientifically validated personality diagnosis. AI can measure visible facial data consistently. Psychological meaning is a separate interpretive layer and should be validated against real-life evidence.
Why use face reading if it is not a diagnosis?
Because a tool can be useful without being diagnostic. A structured reflection can help you notice strengths, blind spots, social-perception effects, or questions you have not considered before.
What should I do if an insight does not fit?
Reject it, question it, or ask someone who knows you well. The methodology should leave room for disagreement. Life evidence has priority over the report.
About the Author
Marijus Morkevicius — Entrepreneur, Human Potential Researcher & Co-Founder of MeByFace
Marijus co-founded MeByFace around a long-standing interest in human potential and a belief that clearer self-knowledge supports better decisions. His position on technology is deliberately bounded: AI can structure information and accelerate reflection, but it should not replace human judgment or lived experience.
About Marijus: https://www.mebyface.com/about/marijus-morkevicius-mebyface-co-founder
Related reading
- How MeByFace Works
- The Complete Guide to Face Reading
- How AI Face Analysis Works
- Privacy Policy
- About Marijus Morkevičius
Sources
- MeByFace. How It Works. https://www.mebyface.com/how-it-works
- MeByFace. Privacy Policy. https://www.mebyface.com/privacy
- Willis J, Todorov A. First Impressions: Making Up Your Mind After a 100-ms Exposure to a Face. Psychological Science (2006). https://collaborate.princeton.edu/en/publications/first-impressions-making-up-your-mind-after-a-100-ms-exposure-to-/
- MeByFace. The Complete Guide to Face Reading. https://www.mebyface.com/learn/face-reading-guide