I asked ChatGPT to suggest four completely mad ideas the UK’s Conservative Party could include in their manifesto.

I fully expect to see at least one of these announced before the week is out.

Reader suggestion to expand the national service idea

Instruments, agents, and AI: novel epistemic categories of reliability

Another cool AI + philosophy of science paper by Eduede. Its starting point is that scientists increasingly rely on AI. But is this reliance warranted?

Nancy Cartwright for the win: reliable knowledge is a collective outcome (of 'a tangle of science')
global.oup.com/academic/produ…

Gemini 1.5 Model Family: Technical Report updates now published

In the report we present the latest models of the Gemini family – Gemini 1.5 Pro and Gemini 1.5 Flash, two highly compute-efficient multimodal models capable of recalling and reasoning over fine-grained information

I've really liked METR approach to rigorously evaluate the plausibility of concrete AI threat models, and say in advance how to mitigate them. I'm excited that we at Google DeepMind have now made our contribution!

x.com/AllanDafoe/sta…

Proud to share one of the first projects I've worked on since joining Google DeepMind earlier this year: our Frontier Safety Framework. Let’s proactively assess the potential for future risks to arise from frontier models, and get ahead of them! deepmind.google/discover/blog/…

🔭 Very happy to share Google DeepMind's exploratory framework to ensure future powerful capabilities from frontier models are detected and mitigated. We're starting with an initial focus on Autonomy, Biosecurity, Cybersecurity, and Machine Learning R&D. 🚀storage.googleapis.com/deepmind-media…

Can Generative AI improve social science?
If individual artificial networks can help us understand brains, why couldn't artificial agents / groups of agents help us understand society?
Interesting review of opportunities, challenges and ways forward.
FYI Stian Westlake