This has been so inspiring to watch happen. vox.com/future-perfect…

SFF has launched a new funding round focused on demonstrating the feasibility and advancing the technical maturity of 'Flexible Hardware Enabled Governors' (flexHEGs). Applications close on Sep 15th. 🏁 More info: survivalandflourishing.fund/sff-2024-flexh… 🧵

thezvi.substack.com/p/danger-ai-sc…

Nearly half of all proposed housing units were attempted to be blocked by environmental lawsuits in California, CEQA Works

Excited to share a new blog post on Planned Obsolescence (first one in a while!) by my colleague Luca Righetti 🔸! Luca says dangerous capability tests need to get *way* harder: planned-obsolescence.org/dangerous-capa…

We should take the precautions described in this article. foreignaffairs.com/world/new-biow…

.Epoch AI just keeps putting out bangers. epochai.org/blog/can-ai-sc…

"I’m increasingly frustrated by the tendency to pit 'doomers' against AI 'optimists'...most optimists are not, fundamentally, AI optimists - they are superintelligence skeptics" slowboring.com/p/what-the-ai-…

YIMBYs have been taking victory laps this week on the back of some great speeches at the DNC. Open Philanthropy has been the movement's biggest funder for most of the past decade, and I’m super proud of the progress. But we’re still very early in this fight: 1/ x.com/JerusalemDemsa…

Here's an actual YIMBY triumph.

Worth observing that these people would be just as dead if it was p-hacking or other questionable-but-not-outright-fraudulent research practices that created the incorrect result

I love this Max Roser manifesto: ourworldindata.org/much-better-aw…

Great opportunity for AI and governance experts. AI office role, deadline on September 6. You need to be a EU citizen, but the role is in SF. …ternational-partnerships.ec.europa.eu/jobs/policy-of…

We are hiring! Google DeepMind's Frontier Safety and Governance team is dedicated to mitigating frontier AI risks; we work closely with technical safety, policy, responsibility, security, and GDM leadership. Please encourage great people to apply! 1/ boards.greenhouse.io/deepmind/jobs/…

Seems kinda important…

Commerce has released the interim rule for reporting on AI models (>1e26 ops). Quarterly reports on training activities, cybersecurity measures, model ownership, and red-team testing results. A great policy: minimal burden, gives gov't needed visibility—no more, no less. 1/

Fascinating article, worth reading. . .this paragraph jumped out at me!

🛜 "If-then" commitments are the new frontier in AI safety. This emerging framework aims to mitigate AI risks without needlessly stifling tech advances. Increased interest in the framework will accelerate its progress & maturity, writes Holden Karnofsky. carnegieendowment.org/research/2024/…

The global nature of AI risks makes it necessary to recognize AI safety as an international public good, and work towards coordinated governance of these risks. Statement made in Venice: idais.ai/idais-venice/ Associated The New York Times article: nytimes.com/2024/09/16/bus…

This seems bad