A decade-old CNN beats frontier LLMs at diagnosing skin cancer

A 10-year-old CNN outperforms frontier LLMs for skin cancer diagnosis

A dermatologist tested GPT-6 Astra and Claude Fable 5.1 on 200 biopsy-proven skin lesions, alongside Inception v3, a 2015 CNN fine-tuned for about twenty minutes. The CNN named the correct diagnosis 62% of the time; both LLMs managed 42%. The LLMs also missed melanomas and discharged dangerous lesions, while the CNN's errors leaned toward caution. The network is small enough to run in a phone browser via WebGPU.