OpenAI was on the 'wrong side of history' with open source - Altman
Kyiv • UNN
Sam Altman acknowledged that the Chinese chatbot DeepSeek has weakened OpenAI's leadership in the AI industry. The company plans to revise its policy on open source and disclosure of information about the “thought process” of models.
OpenAI CEO Sam Altman has admitted that Chinese chatbot DeepSeek has weakened OpenAI's leadership in the AI industry and said that in his opinion, OpenAI was "on the wrong side of history" with regard to open source of its technologies, UNN reports citing TechCrunch.
Details
Although OpenAI has opened source code for models in the past, the company has generally preferred a proprietary, closed development approach, the publication notes.
"[I personally think we need to] come up with a different open source strategy," Altman said at the Reddit AMA event on Friday. - Not everyone at OpenAI shares this view, and it's not our current highest priority either... We will create better models [in the future], but we will have less leadership than in previous years.
Kevin Weil, OpenAI's chief product officer, said that OpenAI is considering opening up the source code of older models. "We'll definitely think about doing more of that," he said, without going into details.
In addition to prompting OpenAI to reconsider its philosophy on releases, Altman said DeepSeek pushed the company to potentially disclose more information about how its so-called reasoning models, such as the released o3-mini model, reveal their "thought process." Currently, OpenAI's models hide this, a strategy designed to prevent competitors from obtaining data to train their own models. In contrast, DeepSeek's reasoning model, R1, shows its entire chain of thought.
"We are working on showing much more than we are showing today - [showing the model's thought process] is coming very, very soon," Weil added. - "It's not certain yet - showing the entire chain of reasoning leads to competitive distillation, but we also know that people (at least power users) want it, so we'll find the right way to balance that.
Altman and Weil also tried to dispel rumors that ChatGPT, the chatbot platform through which OpenAI runs many of its models, would become more expensive in the future. Altman said he would like to make ChatGPT "cheaper" over time if possible.
Earlier, Altman said that OpenAI was losing money on its most expensive ChatGPT plan, ChatGPT Pro, which costs $200 per month.
