OpenAI has previewed its much-anticipated “Strawberry” model, officially named “OpenAI o1-preview.” Designed to tackle complex tasks in fields like physics, chemistry, and biology, this new model shows significant advancements in reasoning, positioning itself as a potential breakthrough in artificial intelligence.
Unlike its predecessor, GPT-4o, the o1-preview model is designed to ” think through questions before responding. OpenAI claims the model performs similarly to PhD students on challenging academic benchmarks. However, it’s still in an early stage of development, with some essential features like browsing and file uploading yet to be included.
This release comes amid growing excitement around artificial general intelligence (AGI). However, OpenAI’s CEO, Sam Altman, clarified that the model is “still flawed, still limited,” and does not yet represent AGI. With updates and improvements expected, the company has also teased a lighter version, “o1-mini,” for free users shortly.
Additionally, the model incorporates a safety-first approach, performing well in jailbreaking tests and reducing AI hallucinations. Thanks to its ability to evaluate multiple solutions before delivering the best answer, OpenAI suggests the model could be a valuable tool in healthcare, scientific research, and complex development workflows.
The model is now available to ChatGPT Plus and Team users, with more access expected soon. As AI continues to evolve, OpenAI’s “Strawberry” model might be a significant step toward a future where machines can reason on par with humans.
