OpenAI recently upped the level of its AI game with a newly introduced internal model aptly named “Strawberry.” This latest work is the next step in achieving human-like reasoning abilities and should lead to useful AI agents that can perform more sophisticated tasks across a wide variety of domains like science, code,math.
OpenAI Launches New AI Model ‘o1‘ with Advanced Reasoning Skills
A New Level of AI Performance
Crucial to the development of this novel model is an increased amount of processing time on AI responses, so that it can tackle more complex problems. OpenAI has built two models that are specific for different purposes:
o1-preview:
A more complex model, with a larger parameter size (in line with the current state-of-the-art) capable of solving structurally sophisticated reasoning problems; it is also beneficial to our evaluation process given its extensive knowledge base.
o1-mini:
80% less expensive than o1-preview, this model is nevertheless quite competitive when it comes to coding activities.
With rate restrictions of 20 requests per minute (RPM), both models are presently undergoing beta testing. However, the o1-mini is a formidable competitor because to its superior speed and efficiency.
Enhanced Ability to Solve Problems by o1-mini
The o1 series is intended to perform more extensive computations prior to answering, especially for situations involving multiple steps. Because of this, it excels at solving complex problems, such as complicated mathematical calculations and coding riddles.
Even while this early version doesn’t have some of the services that ChatGPT users might be used to—like web browsing and file uploading—OpenAI emphasizes that these developments in reasoning herald a new age in artificial intelligence capabilities.
Competition and Availability of OpenAI o1-mini:
OpenAI’s preliminary version of the o1 model will be made available to subscribers of OpenAI Plus and Team through ChatGPT, beginning on Thursday. This announcement occurs during a period when OpenAI is in pursuit of more investment and encounters expanding rivalry from major players in the field such as Google and Anthropic, who are also enhancing their AI’s logical reasoning capabilities.
As OpenAI progresses in refining its models, the introduction of the o1 series represents a major milestone in the company’s goal to create more advanced artificial intelligence technologies.
