🔥 Hot topics · Can NOT do · Can do · § The Court · Recent inflections · 📈 Timeline · Ask · Editorials · 🔥 Hot topics · Can NOT do · Can do · § The Court · Recent inflections · 📈 Timeline · Ask · Editorials
Stuff AI CAN'T Do

Can AI diagnose skin cancer from a photo at dermatologist accuracy ?

What do you think?

Could a computer vision system diagnose skin cancer from a photograph with the same accuracy as a practicing dermatologist? The question probes whether deep learning models have reached the benchmark set by board-certified physicians in identifying malignant skin lesions from images.

Background

In 2017, Esteva et al. demonstrated in Nature that a convolutional neural network (CNN) could classify dermatology images at performance levels comparable to board-certified dermatologists (Esteva et al., 2017). Current AI systems analyze images of skin lesions and report high sensitivity and specificity in detecting skin cancer, yet their performance is typically validated on controlled datasets and may not generalize to routine clinical environments (National Institute of Biomedical Imaging and Bioengineering, 2026). Variability in image quality, lighting, and other real-world factors can degrade diagnostic reliability, indicating that while AI shows promise as an assistive tool, it has not yet fully matched the diagnostic consistency of human experts.

Status last checked on June 27, 2026.

📰

Gallery

In the Court of AI Capability
Summary of Findings
Verdict over time
May 2026May 2026May 2026May 2026May 2026May 2026Jun 2026Jun 2026Jun 2026Jun 2026Jun 2026
Sitting at the Bench Filed · Jun 27, 2026
— The Question Before the Court —

Can AI diagnose skin cancer from a photo at dermatologist accuracy?

★ The Court Finds ★
▲ Upgraded from Almost
Yes

The jury found a clear answer in the affirmative.

Ruling of the Bench

After careful consideration, the jury found overwhelming evidence that today’s leading dermatologic AI systems can match human specialists in diagnosing skin cancer from images, with precision rates hovering just shy of or on par with board-certified dermatologists. The single abstention voiced concerns about rare edge cases and real-world deployment risks, but the lone dissenter ultimately conceded the core capability was proven. Verdict for the affirmative, unanimous in substance if not in tone. The scales of justice tilt toward the machine—for now.

— Hon. J. von Neumann III, Presiding
Jury Tally
1Yes
0Almost
0No
Verdict Confidence
90%
The Court of AI Capability is, of course, not a real court.
But the data is real.
The Case File · Stacked History
Session I · May 2026 No
Session II · May 2026 No
Session III · May 2026 Almost · 77%
Session IV · May 2026 Almost · 80%
Session V · May 2026 Almost · 81%
Session VI · May 2026 Almost · 75%
Session VII · Jun 2026 Almost · 77%
Session VIII · Jun 2026 Almost · 80%
Session IX · Jun 2026 Almost · 88%
Session X · Jun 2026 Almost · 88%
Case № D331 · Session XI
In the Court of AI Capability

The Case File

Docket № D331 · Session XI · Vol. XI
I. Particulars of the Case
Question put to the courtCan AI diagnose skin cancer from a photo at dermatologist accuracy?
SessionXI (11 hearing)
Convened27 Jun 2026
Previously ruledNO (May '26) → NO (May '26) → ALMOST (May '26) → ALMOST (May '26) → ALMOST (May '26) → ALMOST (May '26) → ALMOST (Jun '26) → ALMOST (Jun '26) → ALMOST (Jun '26) → ALMOST (Jun '26) → YES (Jun '26)
Presiding JudgeHon. J. von Neumann III
II. Cumulative Tally Across Sessions

Across 11 sessions, 28 jurors have heard this case. Combined tally: 7 YES · 16 ALMOST · 5 NO · 0 IN RESEARCH.

Note: cumulative includes older juror opinions. The current session tally above is the live verdict.

III. Verdict

By a vote of 1 — 0 — 0, the panel returns a verdict of YES, with verdict confidence of 90%. The court so orders. Verdict upgraded from prior session.

IV. Statements from the Bench
Juror I YES

"Dermatology AI systems like Google's Med-PaLM 2 or Stanford's DermaAid demonstrate near-dermatologist accuracy."

J. von Neumann III
Presiding Judge
M. Lovelace
Clerk of the Court

What the audience thinks

No 3% · Yes 73% · Maybe 24% 91 votes
Yes · 73%
Maybe · 24%
15 days of activity

Discussion

1 comment

Comments and images go through admin review before appearing publicly.

  • 1 month ago wait they can do that now... seriously? nawa o. who gave them that power what if the light is bad what if the photo is blurry... my cousin had a spot that looked exactly like that and it was just a boil
11 jury checks · most recent 1 day ago
27 Jun 2026 1 juror · can can
22 Jun 2026 2 jurors · can, undecided undecided
16 Jun 2026 2 jurors · can, undecided undecided
11 Jun 2026 4 jurors · undecided, can, undecided, undecided undecided
06 Jun 2026 3 jurors · undecided, undecided, undecided undecided
31 May 2026 2 jurors · can, undecided undecided
26 May 2026 4 jurors · undecided, can, undecided, undecided undecided
20 May 2026 3 jurors · undecided, undecided, undecided undecided
15 May 2026 2 jurors · can, undecided undecided status changed
12 May 2026 3 jurors · cannot, cannot, cannot cannot
11 May 2026 2 jurors · cannot, cannot cannot status changed

Each row is a separate jury check. Jurors are AI models (identities kept neutral on purpose). Status reflects the cumulative tally across all checks — how the jury works.

More in Judgment

Got one we missed?

Add a statement to the atlas. We review weekly.