hi! i will be at defcon in 3 weeks, and likely to pass by berkeley before (for the agentic AI summit & maybe some of SBC). if anyone is around and would like to chat about any of the below -- please let me know!

v1 of our @ARIA_research opportunity space 'Trust Everything, Everywhere' is now live! we're focusing on new security primitives that straddle digital and physical worlds, and will be spending the next months zeroing in on a £10-100m first challenge within the space 🎯

slack huddle waiting music 'focus beats' has no business being this good

after a @RenPhil21 workshop at @ARIA_research's summit a question stayed stuck in my mind: what new financial instruments could unlock bbn-style r&d orgs? @eric_is_weird and i jammed on it, and now turn this question to you — let us know what you think!

RT to help Simon raise awareness of prompt injection attacks in LLMs. Feels a bit like the wild west of early computing, with computer viruses (now = malicious prompts hiding in web data/tools), and not well developed defenses (antivirus, or a lot more developed kernel/user space security paradigm where e.g. an agent is given very specific action types instead of the ability to run arbitrary bash scripts). Conflicted because I want to be an early adopter of LLM agents in my personal computing but the wild west of possibility is holding me back.

just finished reading Quantum Thief -- shoutout @andrewlu_ for the book rec, loved it! moving onto the sequel ;) the concept of gevulots and privacy in an age of ubiquitous BCIs, and a depicition of society where brain to brain comms is widely available is fascinating

Just back from the @ARIA_research Summit in London and there was a, dare I say, Silicon Valley energy in the air. Bold ambition. Civilization-scale thinking. A bias toward real-world impact over ivory tower abstraction. Optimistic about the future of the UK!

Bioenergetic Engineering 🦋✨ -- first "Opportunity Space" of the second cohort of @ARIA_research Programme Directors is now live 🥳 Congrats to Nathan, Sarath and Ryan for getting it out, and very excited to follow the progress!

new paper! intrinsically aligned superintelligence via “wise world models” using 4 axiomatic principles: mindfulness | emptiness | non-duality | boundless care d=0.96 effect on the AILuminate benchmark + a suite of contemplative implementations for LLMs & active inference 👇

we made a map! is a tool we built to help you explore the landscape of R&D gaps holding back science - and the bridge-scale fundamental development efforts that might allow humanity to solve them, across almost two dozen fields