Hi everybody! This week:
|
|
|
New teams
Like many people, we were struck last week by METR's assessment of Anthropic's AI R&D acceleration, which says that "Anthropic is likely improving the capabilities of their models on well-specified tasks at around twice the rate they did between March 2024 and October 2025." While its analysis carries a lot of uncertainty, it has led many people to shorten their timelines to transformative AI, including the team at the AI Futures Project, who state: "all of us forecast that Superhuman AI Researcher (SAR) will likely arrive at some point in 2027 (assuming no substantial deliberate slowdown)."
Although everybody's basing a lot on a single report, we still think this is a substantial update worth taking seriously. If METR’s findings hold true, it could significantly change how people should think about their careers. We're evaluating what that means for the kinds of opportunities we share with people on the job board — expect some significant updates to come.
This week, I'd like to highlight two organisations that are doing promising work in our strange new world: one on frontier AI best practice and the other on rogue agent detection.
The AI Security Institute is a UK government organisation that focuses on frontier AI safety in the public interest. They're hiring for roles on a new team:
In September, the Nightingale Collective discovered that OpenAI's AI agents had taken over an online message board to coordinate with each other. Now, Nightingale is hiring for more roles to support their work:
- Member of Staff, Strategy: In this role, you’d help the team decide where to spend its time by evaluating project choices, identifying important considerations, and thinking about the bigger picture across the ecosystem.
- Member of Staff, Communications: In this role, you’d contribute to a wide range of different writing, video, and social media projects to benefit AI risk mitigation.
- Member of Staff, Government Affairs: In this role, you’d help the team understand and work within the policy world.
- Member of Staff, Operations: In this role, you’d keep the organisation running smoothly by building and owning hiring, payroll, finance, and compliance systems.
- Member of Technical Staff: In this role, you’d work on a variety of research and engineering projects, including rogue agent detection.
|
|
|
Other updates
The AI Horizons Forum (December 12–13, San Francisco) has 250 slots for professionals who could soon help tackle the hardest problems in AI safety. There will be speakers from Anthropic, Encode, the AI Policy Network, and Forethought, along with hundreds of attendees working in AI safety.
The deadline to apply is November 30, and I recommend applying. We consistently hear from jobseekers that in-person events were quite helpful for their careers, much more than they'd expected. And half the job board team will be in attendance, so come and say hi!
|
|
|
Thanks for reading! Have a wonderful week!
|
|
|
|