The debate on the security in artificial intelligence returns to the center of the scene. Sam Altman, CEO of Openai, was one of the most anticipated protagonists in the last TED event, where publicly addressed the growing doubts about the future of AItheir impact on society and the responsibility that great technological ones have on their development.
His intervention was not exempt from tension or surprises. During a moderate talk by Chris Anderson, the theme of the safety of AI and its possible lack of regulation occupied a good part of the conversation.
Altman, in a measured tone, He acknowledged that he has not yet lived «terrifying moments» related to the development of smart systemsbut he did not hide his concern for what could come. «AI with agency capacity is the most interesting and consequent security problem we have faced,» he said.
More users in OpenAI and an unexpected revelation
One of the most commented moments occurred when The host revealed a figure that Altman did not expect to make publicand it is the number of weekly active users of the products of OpenAi would have reached 500 millionwhich represents approximately 10% of the world’s population.
«I said that in private,» was Altman’s rapid response, generating a mixture of laughs and surprise among the public.
Ted YouTube channel images
This data reveals the enormous reach of platforms such as Chatgpt, and at the same time, increases pressure on the company to act with transparency and responsibility. Altman said Openai is not hiding any conscious model or experimenting with artificial intelligence that is defendantAlthough he admitted that the risks have grown accelerated.
In the conversation there was also space for self -criticism. Altman distanced himself from one of his best known proposals, the creation of a federal body that supervises the development of large artificial intelligence models. In 2023, he had publicly requested this type of regulation. However, he now recognizes that his vision has changed.
«Since then I have learned more about how the government works. It is not the appropriate framework«He said, suggesting that this approach is not viable or effective in practice.
Despite that, Altman does not uncheck the need for regulation, but proposes more flexible and technical alternatives that respond to the rhythm of change. For many experts, this position reflects the tension between rapid innovation and institutional structures that fail to follow the step.
Open source privacy
The OpenAi CEO also addressed a topic that generates debate, the increasingly active role of artificial intelligence attendees in our lives. He affirmed that these systems are becoming extensions of peopleeven able to anticipate your needs.
Although the idea sounds practice, it also poses serious problems of privacy and personal autonomy, something that many critics are not willing to overlook.
On the other hand, Altman said Openai prepares to launch an open source system that could be «close to the technological border.» He acknowledged that his company has been slow to react to competitors As Deepseek, but said they are ready to recover ground.
Altman was willing to maintain an open dialogue on the limits and opportunities of artificial intelligence, but made it clear that there are no simple solutions. Transparency, in your opinion, is essential, but it is also to make accurate and responsible decisions in real time.
Know How we work in NoticiasVE.
Tags: Artificial intelligence