
In a shocking research paper, AI engineers including OpenAI's Jakub Pachocki, Geoffrey Hinton and Yoshua Bengio warn that self-recursive improvement (RSI) could lead to an "intelligence explosion", where AI improves on its own without human help, becoming a strategic priority for the sector.
AI-generated summary
Self-recursive improvement (RSI) refers to when the AI will be able to improve its own abilities without human help, being able to create an improved version of itself which will then design its successor, and so on.
This is no longer a science fiction scenario. In a shocking research text published overnight from Monday to Tuesday, the engineers who shaped artificial intelligence as we know it today warned against what presents itself as the new priority of the sector: the “explosion of intelligence”. A consequence of “self-recursive improvement (RSI)”, this concept refers to the moment when AI will be fully capable of improving its capabilities on its own, without human help, being able to create an improved version of itself which can then design its successor, and so on.
OpenAI’s chief scientist, Jakub Pachocki, himself announced in early September that RSI was now a priority for his company, and the “only way to stay at the forefront of AI research in the years to come.” Today, Pachocki, but also AI pioneers Geoffrey Hinton (Nobel Prize in 2024), Yoshua Bengio (Turing Prize in 2018), Jack Clark…
AI outlook — possibilities, not facts
ROI will become a major strategic priority for AI companies in the coming years
Very likely · Within months

Donald Trump said Tuesday that leaders of big tech companies had reached a moral agreement to self-regulate artificial intelligence, during a meeting at the White House with figures including OpenAI's Greg Brockman, Anthropic's Dario Amodei, Elon Musk and Jensen Huang. He also announced that he would sign a document establishing the term "super intelligence" in place of "artificial intelligence."

At a White House lunch with AI executives, Donald Trump defended self-regulation as the preferred approach, saying the industry's bosses wanted to "do the right thing." He announced that he would sign a document establishing the terminology "super intelligence" in place of "artificial intelligence." Of the 31 guests, only two women were present: Susie Wiles and Lisa Su.

OpenAI on Tuesday launched GPT-6.1 Sol, an intermediate AI model almost as good as GPT-6 Astra but five times cheaper, after abandoning GPT-6.1 Astra deemed unsafe. This release comes at the developers conference, amid increased pressure on AI security, with calls to slow development and incidents related to unauthorized behavior of advanced models.

Anssi has published a report establishing that the thefts of personal data from the DGFiP this summer, claimed by the ZeroBytes group and affecting nearly 700,000 individuals and businesses, are due to the exploitation of weaknesses in the systems, and not to a sophisticated attack. The agency recommends multi-factor authentication and a full audit ordered by Prime Minister Sébastien Lecornu.

Summary of three news stories: revelations from a laboratory on AI attacks linked to OpenAI, warnings from Bill Gates on the regulation of AI, and comparisons between calls to demonstrate and the Yellow Vests.

Overview of the news: Meta's facial recognition glasses project worries, Anthropic mandates Accenture for the security of its models, and calls to demonstrate against the price of fuel emerge on social networks.