In a candid Reddit AMA, OpenAI CEO Sam Altman admitted that the company is losing its lead in the AI race to Chinese companies like DeepSeek, which OpenAI alleges might have stolen its intellectual property. Altman also expressed his personal view that OpenAI needs to rethink its open-sourcing strategy, a sentiment not shared by everyone at the company.

Altman's comments come at a time when OpenAI is trying to strengthen its relationship with Washington and pursue an ambitious data center project, while reportedly preparing for one of the largest financing rounds in history. The company has been criticized for its proprietary, closed-source development approach, which Altman believes might need to change.

Kevin Weil, OpenAI's chief product officer, added that the company is considering open-sourcing older models that are no longer state-of-the-art. This move could potentially level the playing field and allow other companies to build upon OpenAI's research. However, it's unclear how this would impact OpenAI's competitive advantage.

DeepSeek's progress has also pushed OpenAI to reconsider how its models, like the o3-mini model released today, show their "thought process." Currently, OpenAI's models conceal their reasoning to prevent competitors from scraping training data for their own models. Altman and Weil hinted that they might reveal more about their models' thought processes in the future, but emphasized the need to balance transparency with competitive concerns.

In other news, Altman and Weil dispelled rumors that ChatGPT, OpenAI's popular chatbot app, would increase in price. Altman expressed his desire to make ChatGPT "cheaper" over time, if feasible. This is notable, given that OpenAI is reportedly losing money on its priciest ChatGPT plan, ChatGPT Pro, which costs $200 per month.

Weil also discussed the importance of compute power in developing better AI models, which is driving OpenAI's massive data center project, Stargate. Serving a growing user base is also fueling compute demand within OpenAI, according to Weil.

When asked about recursive self-improvement enabled by powerful models, Altman said he thinks a "fast takeoff" is more plausible than he once believed. This has sparked concerns about the potential risks of advanced AI systems. However, Weil reassured users that OpenAI's models, self-improving or not, would not be used to develop destructive weapons, citing the company's partnership with the U.S. government and its trust in the scientists involved.

The OpenAI team also answered several technical questions, including the release timeline for its next reasoning model, o3, and its next flagship "non-reasoning" model, GPT-5. Weil hinted that a successor to DALL-E 3, the company's image-generating model, is in the works, promising that it would be "worth the wait."

Overall, OpenAI's Reddit AMA provided a rare glimpse into the company's thoughts on its position in the AI race, its open-sourcing strategy, and its future plans. As the AI landscape continues to evolve, it remains to be seen how OpenAI will adapt and respond to the challenges ahead.