Teams at OpenAI are working on Strawberry, based on a recent internal document seen by Reuters in May.

The exact date of the document is unknown, but it outlines how OpenAI plans to use Strawberry for research. The plan is still in progress, and it's unclear when Strawberry will be publicly available. Even within OpenAI, how Strawberry works is a closely guarded secret.

The document describes Strawberry as a project that will enable OpenAI's AI to generate answers and plan ahead to navigate the internet autonomously and reliably for "deep research."

This level of reasoning has been difficult for AI models to achieve so far, according to interviews with over a dozen AI researchers.

When asked about Strawberry, an OpenAI spokesperson said: "We want our AI models to see and understand the world more like we do. Continuous research into new AI capabilities is common in the industry, with a shared belief that these systems will improve reasoning over time."

According to Bloomberg, OpenAI demonstrated a research project with new human-like reasoning skills.

An OpenAI spokesperson confirmed the meeting but did not provide details. Reuters could not confirm if the project shown were Strawberry.

OpenAI hopes this innovation will significantly enhance its AI models' reasoning abilities. Strawberry involves a unique way of processing an AI model after it has been pre-trained on large datasets. Researchers interviewed by Reuters say that reasoning is crucial for AI to achieve human or super-human-level intelligence.