Loading...
Market
Will OpenAI's o1 model outperform GPT-4 by more than 10% in medical reasoning tasks by June 30, 2025?
2
Resolution / Starting Odds
Yes • 50%
No • 50%
Peer-reviewed studies or official benchmarks from OpenAI
Story
OpenAI's New o1 Model Outperforms GPT-4 by 6.2% in Medical Reasoning
Sep 27, 2024, 02:59 PM
OpenAI has unveiled new AI models, including the 'o1' model, which demonstrate significant advancements in complex reasoning tasks across various fields such as science, coding, and medicine. The 'o1' model, part of the o1-preview series, outperforms GPT-4 by 6.2% in medical question-answering tasks. This model's Chain of Thought (CoT) reasoning capabilities enable it to mimic a physician's complex clinical thinking, suggesting a future where AI could play an active role in clinical decision-making. The 'o1' model uses reinforcement learning but still faces challenges, including hallucination, inconsistent multilingual capabilities, and evaluation discrepancies.
View original story
Similar markets
Yes • 50%
No • 50%
Yes • 50%
No • 50%
Yes • 50%
No • 50%
Yes • 50%
No • 50%
Yes • 50%
No • 50%
Yes • 50%
No • 50%
Yes • 50%
No • 50%
Yes • 50%
No • 50%
