OpenAI reportedly plans to release Strawberry in the next two weeks, as new reports build anticipation for the model’s capabilities.
The latest intel about the AI model’s release date comes from The Information, which cites two people who have tested the model and shared further details. OpenAI did not respond to request for comment by the time of publication.
Strawberry is reportedly a generative AI model that’s capable of advanced reasoning beyond the back-and-forth responses that the current models provide. Based on earlier reports from Reuters, Strawberry can perform multi-step tasks based on a single query and “plan ahead enough to navigate the internet autonomously and reliably to perform what OpenAI terms ‘deep research.'” This would put the model one step further towards OpenAI’s goal of developing autonomous AI agents.
It’s unclear whether Strawberry is related to SearchGPT, OpenAI’s search engine that is currently a prototype only available to a small group of users. But according to The Information, Strawberry is reportedly a standalone model that can also be used with ChatGPT, possibly through a drop-down option within the app.
OpenAI CEO Sam Altman teased Strawberry in August with a wink-wink image of a strawberry plant, further adding to buzz about the mysterious new model. OpenAI’s Dev Day is October 1 in San Francisco, which is when OpenAI was expected to announce something about Strawberry. But now it looks like we’ll hear about it sooner than that.
The recent outlook on generative AI tools was somewhat tempered recently by talks of an AI bubble, fueled by waning investor interest, consumer backlash, and apathy towards its general usefulness.
But if OpenAI can successfully tackling the tricky “reasoning” problem as reports of Strawberry suggest, it might just reignite the AI hype machine. If the rumors about investing the AI company are true, Apple, Nvidia, and Microsoft certainly believe in OpenAI’s promise.
0 Comments