DeepSeek to make its open-source models more open, potentially pushing Meta, Baidu, and OpenAI to do the same
DeepSeek, the free and open-source AI company that’s pushed some of its competitors to make their models free, is now potentially pushing them to be even more open, Bloomberg reports.
The Chinese startup plans to make its code repositories available to all developers and researchers, allowing anyone to download and build or improve upon it. It’s an unusual step beyond the typical understanding of open-source in that it’s sharing even more of its core technology with the outside world.
“DeepSeek says it intends to go further by publicizing the underlying code, the data used to create it, and the way it develops and manages that code,” Bloomberg wrote. “By making its coding secrets freely available, DeepSeek is helping to ensure wider adoption of its technology.” This may also nudge its competitors into doing the same.
Its competitor Meta is already open-sourced, but to a lesser extent. After DeepSeek’s splash onto the scene last month, competitor Baidu announced it was also transitioning to an open-source model. OpenAI CEO Sam Altman has since said his company, which was once open-source but no longer, is on the “wrong side of history.”
“DeepSeek says it intends to go further by publicizing the underlying code, the data used to create it, and the way it develops and manages that code,” Bloomberg wrote. “By making its coding secrets freely available, DeepSeek is helping to ensure wider adoption of its technology.” This may also nudge its competitors into doing the same.
Its competitor Meta is already open-sourced, but to a lesser extent. After DeepSeek’s splash onto the scene last month, competitor Baidu announced it was also transitioning to an open-source model. OpenAI CEO Sam Altman has since said his company, which was once open-source but no longer, is on the “wrong side of history.”
