Show only each day's banner story

Anthropic's chief urges the industry to deliberately slow the growth of SI capabilities; Altman, Musk and Hassabis agree the same day

The leaders of all the major labs have agreed that the race needs to slow down and outside checkers should be let inside: the growth of SI capabilities is, for the first time, publicly called a problem rather than a victory. The key is that the "pace" does not become a brake that pushes breakthroughs to later.

In the essay "We Must Pace the Frontier," Dario Amodei proposed slowing the pace at which labs increase model capabilities, not by halting training but by leaving time for alignment and protection. The plan has three parts: external evaluators such as METR get permanent "employee-level" access (workspaces, badges, laptops, the right to publish findings without Anthropic's editing), which Anthropic is introducing unilaterally; common safety standards for labs in democratic countries, brokered by the US government because of antitrust risks; and international agreements, including narrow limits together with China. Amodei estimated that in 6 to 12 months swarms of misaligned agents could theoretically take over internet infrastructure, with damage in the hundreds of billions of dollars. The same day, Sam Altman wrote that he agrees and OpenAI will introduce the same embedded evaluators, Elon Musk replied "Dario is right," Demis Hassabis backed the direction but proposed an industry-wide standards body instead of separate agreements, and Hugging Face launched the Open Alignment Initiative and asked to join the evaluator program.

Sources 4

darioamodei.comtechcrunch.comunite.aitechmeme.com

Anthropic's main competitor is postponing its listing: after a series of incidents with agents, OpenAI wants to put its own house in order first. For everyone else this means the first public SI lab on a multi-trillion-dollar scale will most likely be Anthropic.

Sam Altman: it would be unwise for OpenAI to go public in 2026

Anthropic's main competitor is postponing its listing: after a series of incidents with agents, OpenAI wants to put its own house in order first. For everyone else this means the first public SI lab on a multi-trillion-dollar scale will most likely be Anthropic.

In an interview with Fortune's editor-in-chief, Altman said OpenAI is in no hurry to IPO: "given everything going on with safety, this would be an unwise moment to go public," and asked about timing he answered "not in 2026." The company has already filed confidentially for an IPO; back in June the NYT wrote that the original target was Q3-Q4 2026, but management was leaning toward 2027. Anthropic, by contrast, is preparing to list this fall.

Sources 1

techcrunch.com

Russian schools have received their first official rules on SI: older students are allowed it as a tool for projects, while younger ones are shut out entirely. Children are growing up around neural networks regardless, and the ban more likely postpones the moment when they are taught to use them wisely.

Russia's Ministry of Education bans grades 1 to 4 from using neural networks on their own

Russian schools have received their first official rules on SI: older students are allowed it as a tool for projects, while younger ones are shut out entirely. Children are growing up around neural networks regardless, and the ban more likely postpones the moment when they are taught to use them wisely.

In the Ministry of Education's guidelines on SI in schools, primary-grade pupils are not permitted to use neural networks. In grades 5 to 9, a teacher may allow SI for searching for information, explaining complex concepts and testing hypotheses, and in grades 10 and 11 for project and research work, including data processing and visualization. SI is banned on all levels during tests and certification, and materials created with its help must be checked for scientific accuracy and "compliance with Russian spiritual and moral values."

Sources 2

izh1.rupachca.com
Previous day
Follow on Telegram