Traditional systems often evaluate people through CVs, degrees, grades, interviews, and standardized tests. But these signals can miss abilities that haven’t yet had the right environment to become visible.
Someone may have strong creative problem-solving, systems thinking, pattern recognition, curiosity, rapid learning, or an unusual ability to connect ideas—without having a conventional background that demonstrates these abilities.
What if ChatGPT could optionally help discover these patterns through long-term interaction?
Unlike a CV or a short interview, an AI assistant may interact with someone across many different contexts over time. It could potentially notice recurring patterns in how a person:
approaches unfamiliar problems
reasons and explains ideas
connects seemingly unrelated concepts
learns new subjects
responds to ambiguity
chooses the problems they naturally gravitate toward
Over time, these observations could form an evidence-based picture of possible strengths and interests.
The system could then help the user explore:
Observed patterns → possible strengths → evidence → possible opportunities → alternatives
For example, someone with no prestigious degree or conventional career history might repeatedly demonstrate unusual systems thinking, creative problem-solving, rapid learning, and cross-disciplinary connections. Instead of simply telling them “you should become X,” ChatGPT could show the recurring patterns, explain the evidence behind them, and suggest fields, projects, educational opportunities, organizations, jobs, or even geographic regions worth exploring.
The goal would not be for AI to rank people, determine their destiny, or make employment decisions.
It would be about helping people discover possibilities about themselves that they—and traditional systems—may never have recognized.
This could be especially valuable for young people, career changers, people without traditional educational backgrounds, and people whose abilities don’t fit standardized evaluation systems.
Of course, something like this would require strong safeguards: explicit user consent, control over which conversations are considered, transparency about the evidence, the ability to correct the AI, clear separation between observation and inference, uncertainty about conclusions, and no hidden ranking or profiling of users.
I think the larger opportunity is:
hidden potential → self-understanding → appropriate opportunity
I’m curious whether others think this is technically feasible, what the biggest limitations would be, and what safeguards would be necessary to make such a capability useful without turning AI into a system that labels or determines people’s futures.