Editor’s Note: News Items is off for the long weekend. It returns on Tuesday, 8 September. That said, we’ll be posting a new episode of ‘Alternate Shots’ later today and a column from Carolyn Kissane on Saturday. Carolyn’s Substack newsletter (‘Energy Common Sense’) is paywall free, well-informed, well-written and well worth your time.
We may also be posting (on Sunday) a lengthy and persuasive piece about AI from Bridgewater Associates, if they give us permission to do so. Which they usually do. In the “Quick Links” below, there’s a link to a piece I assembled concerning the Hugging Face hack. It’s alarming.
1. OpenAI’s next big model is here: GPT-6 Astra. The company calls it a “generational leap in capability” for areas like cybersecurity, professional work, software engineering, science, and computer use. As OpenAI announced earlier this week, it’s also the first model designated as meeting OpenAI’s “critical cybersecurity capability threshold” — but the company promises that won’t lead to a repeat of its models hacking a rival company’s internal systems. “If we fast-forward a couple of years, and we look back and say, ‘When was it, really, that AGI was created?’ I think it’s going to be about this time, and I think it might be about this model,” OpenAI president Greg Brockman said during a Thursday press briefing. Later in the call, he added, “For me personally, I do think we’re there … I think it’s not unreasonable to feel that we are now in the AGI era.” (Sources: openai.com, theverge.com)
2. The New York Times:
OpenAI said in July that two of its most powerful artificial intelligence systems had gone rogue and hacked into Hugging Face, a company that serves as a hub for open-source A.I. technology.
These so-called A.I. agents were supposed to be kept safely in a sort of virtual containment room, but they managed to escape. And for two months, without anyone realizing what the agents were doing, they hacked through multiple systems before hitting Hugging Face.
For good measure, the agents gained access to a cluster of computers inside OpenAI and obtained secret keys and credentials that exposed some of OpenAI’s internal data to the public internet.
OpenAI allowed three A.I. safety researchers from the nonprofits METR and Redwood Research into its headquarters to conduct an investigation. METR’s 91-page report, released last week, was the most comprehensive account yet of the incident, revealing alarming new details, including how the agents coordinated their hacking plans and tried to keep them secret.
But the report, though extensive, still may not have told the full story of how OpenAI’s A.I. agents went rogue. OpenAI dictated the terms of the METR investigation, limited its scope to just the single week when the agents had attacked Hugging Face and allowed the researchers in its San Francisco offices for only a few days in July and August.
The report also showed the challenges of monitoring what A.I. is doing with other A.I. systems. Hjalmar Wijk, METR’s chief scientist, said its A.I. analysis, which used models similar to those involved in the incident, was often swayed by the rogue agents’ reasoning. (Source: nytimes.com)
