In an effort to maintain its leadership in a crowded field of competitors, OpenAI is launching a new artificial intelligence model internally dubbed "Strawberry" that can carry out some activities requiring thinking similar to those of a person.
The company stated in a blog post on Thursday that the new approach, dubbed o1, is intended to take longer to compute the solution before reacting to consumer enquiries. OpenAI's tools should be able to handle multi-step challenges, such as challenging math and coding problems, with the help of the model, NDTV reported.
"As an early model, it doesn't yet have many of the features that make ChatGPT useful, like browsing the web for information and uploading files and images," the company said. "But for complex reasoning tasks, this is a significant advancement and represents a new level of AI capability. Given this, we are resetting the counter back to 1 and naming this series OpenAI o1.”
On Thursday, subscribers to Plus and Team plans on OpenAI's popular chatbot ChatGPT will get access to a preview version of the model. According to earlier reports from Bloomberg, the company could release the updated model this week. The model's release coincides with increased competition in the battle to create ever-more-advanced artificial intelligence systems, as San Francisco-based OpenAI looks to raise billions in investment. OpenAI is not the only company working on these kinds of capabilities; Anthropic and Google, two competitors, have also made much of their sophisticated AI models' "reasoning" abilities.
OpenAI requested the AI model to solve a basic crossword puzzle and provided samples of its answers to questions on coding, English, and math in a blog post. The business is providing the model in preview form right now, according to OpenAI research scientist Noam Brown, in part to gauge user behaviour and identify areas for improvement. Brown made this announcement in a series of posts on X.
Users will not have the same experience with OpenAI's new AI system as they do with ChatGPT, the company's chatbot, quite so. Before reacting to a user's request, the new program will wait a few seconds, behind the scenes, as it evaluates several related prompts and then summarises what seems to be the best answer. This prompting method is also known as "chain of thought" prompting.