The company has announced O3-Mini, the first reasoning model that will be distributed free of charge to all users. We are not faced with an attempt to democratize the technology by the company, but to a survival strategy not without risk.
All the plans have changed after the launch of Deepseek, the Chinese IA model that made the Silicon Valley. Openai, a leading company in the artificial intelligence sector, also decided to review its strategy. As the company’s last announcement shows: “We are released a new artificial intelligence model for free”. Is called O3-Mini and the first reasoning model available for everyone.
As Openai explains: “For the first time we are bringing the ability to reason to our free users, an important step towards the expansion of accessibility toAdvanced artificial intelligence At the service of our mission, “said Openai. We are not faced with an attempt to democratize technology by the company, but to one survival strategy not without risk.
What does Deepsek have to do with it
With only 6 million dollars, Deepseek launched a free artificial intelligence model capable of challenge the main US tech companies Like Openai, Microsoft and Google. The development of cutting -edge artificial intelligence to a fraction of the cost compared to western rivals has raised new doubts about the hundreds of billions of dollars planned in the sector.
Deepseek is already disappeared: the Chinese artificial intelligence app can no longer be downloaded in Italy
Deepseek has shown that you can win by changing the rules of the game. The Chinese company has in fact decided to change the fundamental structure of the IA models and use its resources more effectivelywhich are limited due to Export blocks imposed by the United States.
Openai’s answer
The CEO of Openai, Sam Altman, reacted to the Deepseek earthquake with new promises. “We will commit ourselves to creating better models by engaging and accelerate the versions of the products“He explained. O3-Mini for free is part of the counter-offensive. It is no coincidence that the announcement came a few days after the release of Deepseek.
O3-Mini is free but there will be limits. Package users Pro di chatgptwhich costs $ 200 per month, will in fact have unlimited access to the IA model, while users of the cheaper Plus tariff and users who use the free service will have some limits.
Because the O3-Mini model could be risky
Deepseek opens a new front of the Cold technological war Between the United States and China. It is a race, and the apnea of the competition, risks pushing companies to release models that are not without risk. And in fact the O3 model was reported byInternational at the Safety Report. The main author of the study, Yoshua Bengio, He explained that the skills of the model have frightened it “can be risky”.
The performance of O3 in an abstract reasoning test have “have stunned us,” he explained Well. “In some tests, O3 has passed many human experts.” To quote Geoffrey Hinton, Nobel Prize for Baptized Padrine of Artificial Honestyle, technology is advancing faster than expected, and we do not have never had to do with something more intelligent than us.
