Sam Altman Admits OpenAI Got Open Source All Wrong
To conclude a day filled with product announcements, researchers, engineers, and executives from OpenAI, including CEO Sam Altman, participated in a comprehensive Reddit AMA on Friday.
OpenAI is currently in a somewhat difficult position. It is contending with the belief that it is losing its edge in the AI race to Chinese entities like DeepSeek, which OpenAI alleges may have infringed on its intellectual property.
The company is working to strengthen its ties with Washington while also embarking on an ambitious data center initiative, alongside reportedly preparing for one of the largest funding rounds in history.

Altman acknowledged that DeepSeek has narrowed OpenAI’s lead in the AI sector, stating his belief that OpenAI has been “on the wrong side of history” regarding the open-sourcing of its technologies. Although OpenAI has previously made some models open source, it has typically preferred a proprietary, closed approach to development.
“[I personally believe we need to] develop a different strategy for open sourcing,” Altman commented. “Not everyone at OpenAI agrees with this perspective, and it is also not our top priority right now … We will create better models [in the future], but we will hold less of an advantage than we had in past years.”
In a subsequent response, Kevin Weil, OpenAI’s chief product officer, noted that the company is contemplating open sourcing older models that are no longer cutting-edge. “We’ll certainly consider doing more of this,” he stated, though he did not provide further specifics.
In addition to encouraging OpenAI to rethink its release strategy, Altman mentioned that DeepSeek has motivated the company to possibly disclose more information about the reasoning behind its so-called reasoning models, such as the o3-mini model introduced today, and how they demonstrate their “thought processes.”
At present, OpenAI’s models do not disclose their reasoning, a tactic intended to prevent competitors from utilizing their training data for their own models. Conversely, DeepSeek’s reasoning model, R1, displays its entire line of thought.
“We’re working on revealing significantly more than we currently do — [showing the model’s thought processes] will happen very soon,” Weil added. “It’s still to be decided on all — revealing the entire reasoning chain could lead to competitive disadvantages, but we recognize that people (especially power users) desire this, so we’ll seek the right way to strike a balance.”
Altman and Weil sought to quell speculation regarding potential future price increases for ChatGPT, the platform through which OpenAI launches many of its models. Altman expressed a desire to make ChatGPT “cheaper” over time, if possible.
Altman has previously stated that OpenAI is incurring losses on its most expensive ChatGPT plan, ChatGPT Pro, which costs $200 monthly.
In a related discussion, Weil mentioned that OpenAI continues to see indications that increased computing power results in “better” and more effective models. This is a significant factor driving initiatives like Stargate, OpenAI’s newly announced large-scale data center project, Weil explained. The demand for computation is also growing within OpenAI due to an expanding user base, he added.
When asked about the potential for recursive self-improvement that these powerful models might facilitate, Altman indicated that he now believes a “fast takeoff” is more likely than he previously thought. Recursive self-improvement refers to a process whereby an AI system could enhance its own intelligence and abilities without human intervention.
It’s notable, however, that Altman has a reputation for making exaggerated claims. Not long ago, he reassessed OpenAI’s ambitions for artificial general intelligence (AGI).
One Reddit user inquired if OpenAI’s models, whether self-enhancing or not, could be utilized in the creation of harmful weapons — particularly nuclear arms. This week, OpenAI announced a collaboration with the U.S. government to provide its models to the U.S. National Laboratories for nuclear defense research, in part.
Naijaeyes Report
Join Our Social Media Channels:
WhatsApp: NaijaEyes
Facebook: NaijaEyes
Twitter: NaijaEyes
Instagram: NaijaEyes
TikTok: NaijaEyes