Loading...
Market
What score will OpenAI's Strawberry model achieve on the MATH dataset by end of 2024?
3
Resolution / Starting Odds
Less than 90% • 25%
90%-94.99% • 25%
95%-99.99% • 25%
100% • 25%
Official announcements or publications by OpenAI
Story
OpenAI's 'Strawberry' AI Model Achieves 90% on MATH Dataset, Aiming for Superintelligence by 2026
Jul 15, 2024, 04:52 PM
OpenAI is reportedly working on a new AI model code-named 'Strawberry', previously known as Q*. The project aims to enhance AI capabilities in reasoning, advanced planning, and superintelligence. According to sources, OpenAI has tested an AI internally that scored over 90% on a MATH dataset, a benchmark of championship math problems. The model is designed to enable the AI to not just generate answers but to plan ahead and navigate the internet autonomously for deep research. Sam Altman, CEO of OpenAI, remarked that AI is currently the dumbest it will ever be, indicating significant future advancements. The project is expected to be fully realized by 2026.
View original story
Similar markets
Yes • 50%
No • 50%
Yes • 50%
No • 50%
DeepSeek-R1-Lite-Preview • 25%
OpenAI's o1-preview • 25%
Google DeepMind's model • 25%
Other • 25%
Significantly better • 25%
Slightly better • 25%
On par • 25%
Worse • 25%
OpenAI's model • 25%
Deepseek-Math (7B) • 25%
Gemini 1.5 Pro (May) • 25%
GPT-4o • 25%
Yes • 50%
No • 50%
Yes • 50%
No • 50%
Yes • 50%
No • 50%
Yes • 50%
No • 50%
Yes • 50%
No • 50%
