Altman itselfOpenai CEO, created Striking admission On Friday, his company “On the wrong side of history” regarding the artificial intelligence of Open Source, signaling a potential seismic change of strategy, because competition from China is intensifying, and efficient open models gain adhesion.
Candid confirmation took place during Reddit “Ask me about anything“Session, just a few days after the Chinese company AI Deepseek Global global markets with the R1 Open Source model, which claims that comparable performance with OpenAI systems for a fraction of costs.
“Yes, we talk [releasing model weights]” Altman wrote. “Personally, I think we were here on the wrong side of the story and we must come up with another Open Source strategy.” He noticed that not everyone at Openai share their view and this is not the company’s highest priority.
The statement is an amazing departure from An increasingly reserved approach of OPENAI In recent years, which attracted the criticism of some AI researchers and former allies, in particular Elon Muskwho sucks the company for allegedly betraying his original Open Source mission.
Altman itself on Deepseek: “We will keep less lead”
Altman’s comments appear among Market confusion triggered by the appearance of Deepseek. Claims of a Chinese company about building advanced AI models for justice $ 5.6 million training costs (Although the total development costs are Probably much higher) sent nvidia supplies violentDestruction of market value almost $ 600 billion-the greater one-day decline for every American company in history.
“We will produce better models, but we will maintain a smaller advantage than in previous years,” admitted Altman in the same AMA, directly dealing with the influence of Deepseek.
Altman himself admits that the closed Otnai strategy could also be wrong
The breakthrough of Deepseek, regardless of whether his specific claims prove accurate, emphasized the change in dynamics in the development of artificial intelligence. The company claims that it has only achieved its results GPU 2000 NVIDIA H800– APPLY lower than estimated over 10,000 tokens often implemented by the fundamental AI laboratories.
This approach suggests that algorithmic innovations and architectural optimization could also be more essential than the harsh computing power. The revelation not only threatens the OpenAI technical strategy, but its entire business model based on exclusive access to mass computing resources.
Open Source debate: Innovation vs. Security
However, the growth of Deepseek also intensified Fears of national security. The company stores user data on servers in continental China, where it might be subject to access to the government. Several American agencies have already passed to limit its use, and NASA has grow to be the latest blocking of the application, citing “concerns about security and privacy.”
The potential Openai turn point for Open Source would mean a return to its roots. The company was founded as a non-profit organization in 2015. With the mission of providing artificial general intelligence, it’s going to profit humanity. However, his transition to the “limited from the organization” and an increasingly closed approach caused criticism of Open Source supporters.
“The correct reading is:” Open Source models exceed the reserved “, wrote the fundamental scientist AI Meta Yann Lecun on LinkedInresponding to the appearance of Deepeks. “They got here up with latest ideas and built them in addition to the work of other people. Because their works are published and Open Source, everyone can profit from it. This is the strength of open research and open source. “
A brand new chapter of the development of AI
Although Altman’s comments suggest that a strategic change may occur, he emphasized that Open Source is not currently the highest priority of OPENAI. This hesitation reflects the complex reality, which is facing AI leaders: balancing innovation, security and commercialization in an increasingly multiply world of artificial intelligence.
The rates extend far beyond the bottom line of OPENAI. The company’s decision can transform the entire AI ecosystem. Key Open Sourcing models can speed up innovation and democratize access, but it might also complicate efforts to make sure the safety of the Openai mission.
The time of accepting the Altman, approaching the shock of the Deepseek market, and not earlier, suggests that Opeli can react to market forces, not to steer them. This reactive attitude is a striking reversal of a role for a company that has long been positioned as the northern star of AI.
When dust is based on the Debut Deepseek, one thing becomes clear: real interference is not only technology or market value – it is about questioning the assumption that strictly guarded AI models are the surest path to artificial general intelligence. In this light, the adoption of Altman can consist in being on the wrong side of history and more recognition that the story itself has modified the course.
