Sam Altman’s OpenAI is reportedly on the verge of unveiling a new artificial intelligence (AI) model that promises substantial improvements in the mathematical and reasoning capabilities of its popular chatbot, ChatGPT.
According to a report from The Information, this project is under the codename “Strawberry,” this advanced AI model has been under development for several months, though the company has kept its progress largely under wraps. The existence of this large language model (LLM) first came to light in July, with Strawberry being positioned as the potential successor to the highly-regarded GPT-4.
The publication also adds that the launch of Strawberry AI model is anticipated this fall. Moreover, OpenAI is reportedly planning to introduce this new AI model during the upcoming fall season. However, details on how the model will be rolled out remain uncertain. One option under consideration is the release of Strawberry as a standalone chatbot, while another possibility is the integration of its advanced capabilities into the existing ChatGPT platform.
The excitement surrounding the Strawberry model stems from its rumored proficiency in tackling complex mathematical and reasoning challenges. Unlike current AI models that struggle with unfamiliar problems, Strawberry is reportedly capable of solving equations and logical puzzles it has never encountered before.
This breakthrough is noteworthy because most AI models are designed as deep learning systems that recognize and process patterns. When these patterns become symbolic or ambiguous, traditional AI models often falter. As a result, current chatbots face significant limitations when confronted with intricate mathematical problems, logical reasoning, or context-heavy inquiries.
However, if recent reports are accurate, Strawberry may have overcome these challenges, positioning it as a substantial leap forward in AI technology. Specific details about the model’s architecture or parameters remain scarce, but sources suggest that Strawberry was previously referred to as Q* (pronounced Q-star).