OpenAI is gearing up to release its next-generation artificial intelligence product, Strawberry, this fall. The new AI model aims to outdo the present models and solve problems and tasks that AI systems cannot solve.
According to Seeking Alpha, which cited a report from The Information, Strawberry will introduce groundbreaking features in AI problem-solving. The model is intended for dealing with new mathematical tasks, creating effective market strategies, and solving complicated word tasks. Furthermore, Strawberry is capable of carrying out complicated searches to conduct research activities; this improves the system’s information processing capability.
OpenAI Continues Push for Human-Like AI Reasoning
The announcement of Strawberry follows previous developments in AI technology by OpenAI. According to Ghazenfer Mansoor, the founder and CEO of Technology Rivers, such development can assist in transforming business ecosystems, including supply chain management, to deliver value to the client’s party.
“OpenAI’s pursuit of human-level reasoning isn’t just a technological marvel; it’s a narrative of pushing boundaries and sparking new possibilities in every sector.”
Ghazenfer Mansoor
In July, OpenAI made several important changes to its product offerings. A new feature called ‘voice’ was introduced for the ChatGPT. The company used the GPT-4o model for this. This new feature provides clear audio feedbacks and such feature makes it possible for a conversation with ChatGPT to be ongoing without having to stop.
SearchGPT was released by OpenAI on July 25 as a tool that is expected to provide shorter and better search results than conventional search engines. The prototype provides brief answers with links to sources, enhancing the effectiveness and recency of the information found.
Additionally, OpenAI launched a cheaper version of its AI model ChatGPT called GPT-4o mini on July 19. This model supersedes the previous GPT-3.5 Turbo while enhancing the model’s capabilities in textual understanding, multimodal, and long context understanding.
GPT-4o mini also supports multiple languages and improves the ability to call functions for developers who apply AI to different systems. The company stated, “GPT-4o mini surpasses GPT-3.5 Turbo and other small models on academic benchmarks across both textual intelligence and multimodal reasoning and supports the same range of languages as GPT-4o.”
OpenAI Forms Similar Partnerships
Condé Nast and OpenAI have recently agreed to collaborate with the aim of incorporating content from Condé Nast’s established media outlets namely Vogue, Wired, and The New Yorker into OpenAI’s products. This partnership includes the famous ChatGPT and the SearchGPT model, which gives the users an opportunity to read Condé Nast publications.
The details of the financial aspect of the agreement have not been released to the public. However, it follows similar arrangements that OpenAI has made with other top publishers such as Time magazine, Financial Times, Axel Springer, Le Monde, and Prisa Media in the recent past. These agreements enable OpenAI to have access to huge data sets for its training of language models and the fetching of up-to-date data.