Loading...
Market
What will be OpenAI o1 model's performance on benchmark tasks by end of 2024?
2
Resolution / Starting Odds
Top 10% • 25%
Top 1% • 25%
Top 5% • 25%
Below Top 10% • 25%
Official benchmark results published by OpenAI or independent evaluators
Story
OpenAI Releases o1 and o1-mini AI Models with Advanced Reasoning and Fact-Checking Capabilities
Sep 12, 2024, 05:15 PM
OpenAI has officially released its new AI model, OpenAI o1, internally known as Strawberry, after several months of development. This model is designed to enhance reasoning capabilities and solve complex tasks in fields such as mathematics, science, and coding. The o1 model series includes a smaller, cost-efficient version called o1-mini and a free tier version of ChatGPT. Both models are available to ChatGPT Plus and Team users, with o1-preview and o1-mini being selectable in the model picker. The new models are reported to perform at PhD-level accuracy on benchmark tasks in physics, chemistry, and biology, and can reason through problems similarly to human thinking. Additionally, the o1 model can fact-check itself, marking a significant milestone in AI development.
View original story
Similar markets
Yes • 50%
No • 50%
Reinforcement learning • 25%
Search-based reasoning • 25%
Thinking before answering • 25%
Other • 25%
Yes • 50%
No • 50%
Yes • 50%
No • 50%
Yes • 50%
No • 50%
Improvement in medical reasoning • 25%
Improvement in coding tasks • 25%
Improvement in scientific reasoning • 25%
Other • 25%
None • 25%
1 to 2 • 25%
3 to 4 • 25%
5 or more • 25%
Yes • 50%
No • 50%
Less than 80% • 25%
80% to 85% • 25%
85% to 90% • 25%
Over 90% • 25%
Yes • 50%
No • 50%
