OpenAI aims to release a new AI model “Strawberry” in fall, which is expected to revolutionize the field of artificial intelligence with advanced capabilities. This new product is expected to significantly surpass the capabilities of current AI models, according to a report by Seeking Alpha, citing a paywalled article from The Information.
Strawberry is anticipated to handle tasks that have so far been beyond the reach of existing AI technology. The new model is reported to solve math problems it has never encountered before, perform high-level tasks like crafting market strategies, and tackle complex word puzzles. It is also designed to conduct “deep research,” making it a versatile tool for various applications.
The development of Strawberry marks a significant step in enhancing AI’s reasoning capabilities. While the model aims to push the boundaries of what AI can achieve, specific details about its functionality and release timeline still need to be made clear.
AI’s Potential to Revolutionize Industries
Ghazenfer Mansoor, founder and CEO of Technology Rivers, highlighted the transformative potential of AI advancements. “OpenAI’s pursuit of human-level reasoning isn’t just a technological marvel; it’s about pushing boundaries and sparking new possibilities across sectors,” he said in an interview with PYMNTS. Mansoor noted that AI could drastically improve supply chain management, provide precise market trend forecasts, and personalize customer experiences on a large scale.
As OpenAI aims to release a new AI model Strawberry in fall, the tech industry is preparing for a potential leap in machine learning and reasoning. OpenAI has released new features and products over the past few months. In July, the company introduced an advanced voice feature for its ChatGPT platform, utilizing its GPT-4o model. This feature provides hyper-realistic audio responses and enables real-time, seamless conversations, even allowing users to interrupt mid-sentence.
Additionally, OpenAI launched SearchGPT on July 25. This new tool, currently available to a limited test group, aims to provide more concise and relevant search results compared to traditional search engines, delivering summarized answers with direct source links instead of a list of links.
Another notable release in July was GPT-4o mini, a smaller and more affordable version of OpenAI’s AI model. Announced on July 19, the GPT-4o mini replaces the previous GPT-3.5 Turbo model and has demonstrated superior performance in academic benchmarks across textual and multimodal reasoning. It also supports a wide range of languages and has shown improved performance in function calling and long-context tasks.
“Strawberry” and Its Advanced Capabilities
The new model, internally referred to as “Strawberry,” is said to solve math problems it has never seen before and handle complex tasks like developing marketing strategies and solving intricate word puzzles. In a notable demonstration, Strawberry solved the New York Times word puzzle “Connections,” showcasing its advanced problem-solving abilities.
Strawberry is also expected to achieve over 90% on the MATH benchmark, a collection of championship-level math problems. For comparison, the GPT-4 model scored 53% on this test, and the GPT-4o model scored 76.6%. As of July, GPT-4o held the highest MATH benchmark score among currently available AI models, indicating that Strawberry could place OpenAI well ahead of its competitors if it delivers on its promises. Given that OpenAI aims to release a new AI model Strawberry in fall, there is growing interest in how this model will handle advanced tasks such as market strategy development.
The development of Strawberry has not been without controversy. Initially known internally as Q (pronounced Q-Star), the model was central to a brief period of turmoil at OpenAI last year. This period included the temporary ouster of CEO Sam Altman, who returned to the company shortly afterward. Concerns about a major AI breakthrough involving Q were reportedly a factor in this upheaval. A group of OpenAI researchers had raised potential risks associated with advanced AI, specifically highlighting Q in a letter to the board before Altman’s departure.
