OpenAI plans to release a new ‘open’ AI language model in the coming months


Openai says it intends to release the first “open” language model GPT-2 “In the coming months.”

According to this a Contact form The company was published on Monday’s website. Openai’s uniform “developers, researchers and [members of] Would you like to see a broader community in an open model of “to fill” Openai? “Includes questions like?” and “What open models did you use in the past?”

“To collect entries and make this model as useful as possible, we are pleased to cooperate with researchers and a wider community,” he said. “If you want to join the opinion session with the Openai Team, let us know [in the form] below. “

Openai plans to collect opinions and hold developer to collect the model’s demo prototypes in the future. The first developer will follow the sessions in San Francisco, Europe and Asia-Pacific.

Openai faces the growing pressure of rivals such as China AI Lab DeepSeek, who accepts an “open” approach to the “open” approach to starting the models. Unlike the Openai strategy, these “open” competitors are experimenting and in some cases, they present to the AI ​​community for trade.

Proven to have a wildly successful strategy for some outfits. Meta, which invests a lot of open AI models to the Llama family, said he was Llama in March before More than 1 billion loaded. Meanwhile, DeepSeek, gathered quickly to a large user base in the world and drew the attention of local investors.

In Last Reddit Q & AOpenai CEO SAM Altman said that when Openai came to open its technology, he said he thinks that history is on the wrong side.

“[I personally think we need to] Understand a different open source strategy, “Altman said. Everyone in Openai does not share this picture and is not our current highest priority […] We will produce better models [going forward]But we will protect our leadership less than we do in previous years. “

Altman Adani has expanded in open model plans X in X on Monday afternoon xOpenai’s Open Open Open Model Openai’s O3-Mini’s lines will be “justificial” opportunities.

“[B]Leave the Effe, we will appreciate this model [to] Our preparation frame, as we want for any other model, “Altman said.”[A]ND, given that we know this model will be sent after this, we will do extra work […] [W]I am pleased to see what the producers are using and using large companies and governments that a model prefers to work themselves. “

Wall Street Journnaly Journalist Keach Hagey’s part-time partition claim This Altman approached Openai leaders with model security reviews before short folder In November 2023.





Source link

Leave a Reply

Your email address will not be published. Required fields are marked *