AI compatibility
Checking a mole for cancer risk is not a job any AI agent should do alone.
A human should do this one.
Average across 1 submission.
The honest read
This task requires physical examination of a specific person's skin over time, access to longitudinal visual data the agent cannot obtain, and medical judgment with life-or-death stakes. An AI agent cannot observe your mole, cannot compare it to six months ago, and a wrong reassurance could delay diagnosis of melanoma. This is not an automation problem — it is a medical consultation.
Aggregated across 1 submission.
The five dimensions
Repeatability
LowEvery mole assessment is unique to a specific individual, body location, skin type, and personal medical history. There is no repeatable structure an agent can apply without physical access to the patient.
Ambiguity Tolerance
LowSuccess criteria are dangerously ambiguous: 'changed in appearance' requires subjective visual comparison across time, and 'requires evaluation' is a clinical judgment call with no crisp threshold an agent can reliably apply.
Data & Tool Availability
LowThe agent has no access to the user's skin, no baseline photos from six months ago, no dermatoscopy data, and no medical history. All required inputs are physically unavailable to any software agent.
Error Cost
HighA false negative — telling someone their mole is fine when it is melanoma — could directly cost a life. This is among the highest-stakes error categories possible, and the outcome is irreversible.
Human Judgment Required
HighBoard-certified dermatologists use trained clinical intuition, dermoscopy, and physical examination to make this call. Even AI-assisted dermatology tools require supervised clinical deployment and cannot replace a physician's judgment.
What an agent would need
- High-resolution dermoscopic images of the mole taken today and six months ago by a calibrated device
- Access to the patient's full medical and skin history including prior biopsies and family history of melanoma
- A validated, clinically approved AI dermatology model (not a general-purpose agent)
- Regulatory clearance and physician oversight for any diagnostic output
- A liability and consent framework appropriate for medical advice
Best-matched agent type
The kind of agent this work would call for if it were a fit. For this task, it isn't.
Run your own fit check
Get a calibrated read on your specific task in under a minute.