The project’s goal is to enable AI to generate responses to queries and perform thorough internet research independently by planning and navigating, a capability not achieved by previous AI models
read more
OpenAI plans to release its reasoning-focused artificial intelligence, named “Strawberry”, as part of its ChatGPT service in the next two weeks, The Information reported on Tuesday (September 10), citing two people who have tested out the model.
What distinguishes Strawberry from other conversational AI is its ability to “think” before responding, rather than immediately answering a query.
Although it will be a part of ChatGPT, Strawberry is still set to be a standalone offering, according to the report. However, exactly how Strawberry will be offered remains unclear.
One plausible way to extend the model to users may be to include it in the dropdown menu of AI models that customers can pick from to power ChatGPT.
A report by Reuters in July had said that Microsoft-backed OpenAI was working on a novel approach to its AI models in a project code-named Strawberry.
The project’s goal is to enable AI to generate responses to queries and perform thorough internet research independently by planning and navigating, a capability not achieved by previous AI models.
Strawberry reportedly involves a specialised way of processing an AI model after it has been pre-trained on very large datasets. OpenAI hopes the innovation will improve its AI models’ reasoning capabilities dramatically Reuters had quoted a source as saying.
When asked about Strawberry at that time, an OpenAI company spokesperson had said, “We want our AI models to see and understand the world more like we do. Continuous research into new AI capabilities is a common practice in the industry, with a shared belief that these systems will improve in reasoning over time.”
Researchers are of the view that reasoning is critical to AI achieving human or super-human-level intelligence. Reasoning is important because large language models often fall short on common sense problems whose solutions seem intuitive to people, like recognising logical fallacies and playing tic-tac-toe.
With inputs from Reuters