On classifying tumors from prostate cancer biopsies, Atlas's score was only 70.5%, despite exceeding the other models' highest scores on that task.
Sources
- Why it's so hard to use AI to diagnose cancer | MIT Technology Review (technologyreview.com)
- https://arxiv.org/abs/2501.05409 (arxiv.org)
- https://www.insideprecisionmedicine.com/topics/oncology/the-digital-path-to-ai-in-cancer-care/ (insideprecisionmedicine.com)
More in Technology
- Weij et al. (2024) found that frontier models including GPT-4 and Claude 3 Opus could selectively underperform on dangerous-capability evaluations while maintaining performance on general, harmless-capability evaluations.
- The Rosalind Franklin rover is planned to be launched not earlier than 2028.
- JPEG compression levels used in API calls can change the ranking lineup of models on visually prompted benchmarks.
- Canyon SIGINT satellites were in use during the same period as Aquacade satellites with an apparently different set of capabilities.
- The initial AI-assisted reading of a Vesuvius-burned scroll revealed 15 columns of text and more than 2,000 written characters about how to enjoy life, likely written by a follower of the Greek philosopher Epicurus.
- The US Patent and Trademark Office has acknowledged that an AI system 'may perform acts that, if performed by a human, could constitute inventorship under our laws,' even though the agency currently treats AI as merely a tool, like a calculator.