I came to this idea from a personal experience.
As a programmer, I recently spent a lot of time thinking about how to build a better resume and how to present my real abilities in a fair and accurate way.
During that process, I realized how difficult it is for a traditional resume to truly show what someone can actually do.
Most resumes are still based mainly on things like previous employers, job titles, degrees, certificates, and years of experience. But in practice, these things do not always represent real capability.
Someone may have several years of formal experience and still have limited practical ability.
At the same time, another person may have little formal work history, but have strong technical knowledge, creativity, problem-solving ability, and the ability to build real things.
I also realized this problem in my own experience.
A significant part of my learning, development work, debugging, problem-solving, and technical growth has happened while working with AI tools such as ChatGPT.
But a normal resume cannot really show that journey.
It cannot show:
- what I actually understand
- what I can build independently
- where I need AI assistance
- how effectively I can work with AI
- whether I can detect when AI is wrong
- whether I can modify and improve AI-generated work
- how my abilities have developed over time
That led me to this idea:
ChatGPT could create an optional, user-controlled professional profile that reflects a person’s real capabilities based on evidence collected over time.
I call it the:
ChatGPT Skill Passport
Instead of simply saying:
“JavaScript — Advanced”
the profile could show different dimensions of ability.
For example:
JavaScript
Knowledge Depth: 65%
Independent Problem Solving: 72%
AI-Assisted Execution: 90%
Debugging Ability: 78%
AI Output Verification: 82%
This would show something that traditional resumes do not show very well.
A person may not know every advanced detail of JavaScript, but may still be extremely capable of completing real-world JavaScript projects using AI.
At the same time, the system could distinguish between someone who simply copies AI output and someone who actually understands it, verifies it, corrects it, and can work with it effectively.
I think this distinction is becoming increasingly important.
The important question is no longer only:
“How much does this person know?”
It is also:
“What can this person actually accomplish, with and without AI?”
ChatGPT already interacts with users while they:
- learn new technologies
- build projects
- debug problems
- make technical decisions
- solve difficult problems
- ask increasingly advanced questions
- correct AI mistakes
- improve generated solutions
Over time, these interactions could provide meaningful evidence about a person’s actual capabilities.
The system should not assume that frequent AI usage automatically means high skill.
It should try to distinguish between situations where:
- the user solved the problem independently
- the user needed significant guidance
- AI generated most of the solution
- the user successfully identified an AI mistake
- the user modified or improved the generated work
- the user understood why the solution worked
- the same skill was demonstrated repeatedly
If there is not enough evidence to evaluate a skill, ChatGPT could offer optional practical assessments or adaptive challenges.
The profile could then become a living professional record that develops together with the user.
Instead of a resume that is updated once every few years, a person’s Skill Passport could gradually become more accurate as they learn, build, solve problems, and gain experience.
Users could choose to publish selected parts of their Skill Passport through a shareable profile, similar to how people currently share a LinkedIn or GitHub profile.
For example:
An employer, client, or collaborator could then see a concise view of verified skills without seeing any private conversations.
Privacy would have to be fundamental to the entire system.
Private conversations should remain private.
The user should control:
- whether a Skill Passport exists
- which activities are used as evidence
- which skills are included
- which parts are public
- who can access the public profile
The system should be private by default and explicitly user-controlled.
I also think this could become much more than a resume feature.
Over time, ChatGPT could become a trusted reference for a person’s professional development and demonstrated capabilities.
For many people, especially self-taught developers, freelancers, students, career changers, and talented people without traditional corporate experience, this could create a much fairer way to demonstrate what they are capable of.
There is also another interesting effect.
If a user knows that their real learning, problem-solving, and project work can gradually contribute to a meaningful professional profile, they have a stronger reason to learn, build, and work continuously with ChatGPT.
Their work with AI would not simply disappear after a conversation.
It could contribute to a long-term record of their professional growth.
I have many more ideas about how this system could work, including skill verification, assessments, confidence levels, project evidence, privacy controls, public profiles, and how AI-assisted work should be evaluated fairly.
I am genuinely interested in seeing an idea like this explored and developed.
If there is interest from the OpenAI team, I would be very happy to share more detailed thoughts and expand the concept further.
The core idea is simple:
Don’t only measure what someone knows.
Measure what they can actually accomplish, with and without AI.
Proposed by:
Hamed Esmaeli