The use of AI for diagnosing skin cancer is being tested at a London hospital, claiming a 99% accuracy in identifying benign cases. However, there are concerns about the practicality and efficiency of this approach, especially given that only 5% of urgent referrals actually turn out to be cancerous. Critics argue that the emphasis should be on reducing the number of false negatives and improving initial assessments, as many urgent cases are incorrectly flagged. Some commenters believe that the AI hype is overstated, suggesting that similar technologies existed over a decade ago without the current labeling. There are also discussions about the potential for at-home diagnostics to ease hospital burdens, pointing out the need for a better filtering system before urgent referrals are made. Overall, there’s skepticism about whether AI is the right solution to the existing challenges in skin cancer diagnosis.