GPT-6 Astra and Claude Fable 5.1 beaten by a 10 year old CNN
A dermatologist compared these frontier LLMs against Inception v3, a CNN fine tuned for skin cancer diagnosis
A few interesting findings:
- 62.5% accuracy vs. ~42% for both LLMs
- The CNN was fine-tuned in just 20 minutes on an H100
- All three models struggled to identify melanomas
- Higher accuracy didn’t necessarily mean safer clinical decisions
It’s a small experiment, but a good reminder that bigger, general purpose models aren’t always the best tool for specialized tasks
Full study:
🔊 A successful melanoma vaccine trial is raising hopes that personalized cancer vaccines could one day be used against lung, pancreatic, colon and other cancers. Hear more from @mygershberg on Reuters World News
Immunotherapy has changed melanoma care - but not equally.
Some mucosal subtypes still show significantly poorer outcomes.
What’s driving this difference?
Explore here:
A successful trial of an mRNA-based melanoma vaccine signals a second act for a technology that encountered widespread skepticism during the Covid pandemic. Here's what the findings might mean for cancer treatment
🔊 Cancer vaccine breakthrough: a personalized melanoma shot succeeded in a large late-stage trial. @mygershberg tells Reuters World News what it means for treating tumors