
Professor of Planetary Computing @Cambridge_CL
Cambridge, England
Joined April 2008
- Tweets13K
- Following1.3K
- Followers6.3K
- Likes19.4K
Pinned Tweet
Made the cover of @CACMmag with "A Decade of Docker Containers", recapping much systems work! Docker grew so fast in those early days that we never got a chance to write an academic paper about it, so this writeup has been a long time coming: cacm.acm.org/research/a-deca….
A heady mix of Go, OCaml, Swift, Linux, nested kernels, ancient network protocols like SLIRP, all mixed up into a container cauldron...
The article also comes with cool art and a video interview that was a lot of fun to film around @pembroke1347 and my overgrown office. I left some musings on my blog too anil.recoil.org/notes/cacm-d… with @mugofsoup and @justincormack
Anil Madhavapeddy retweeted
Historical moment indeed: formal methods are now verifying real-world binaries at scale! But the specs are single-entry (not API), proofs are huge, and reuse is largely copy-and-adapt. Compositional reasoning and modular abstractions are how the FM community can make this scale.
New on our Frontier Red Team blog: using Claude to formally verify software at large scale.
Formal verification can prove that code is free of certain bugs, but historically it has only been feasible for small codebases (or required a lot of expert time).
With AI models, however, this is beginning to change. In collaboration with @theoremlabs, we used Claude to prove security properties in over one hundred real-world open-source libraries over the span of a few weeks.
Read our research in full here: anthropic.com/research/forma…
Anil Madhavapeddy retweeted
What if your workspace was a live wiki, a graph with everything extractable, linkable; and that works across whatever hosts you do your work on? And what if it was optimized for reading?
Someone at lunch commented to me that they'd heard on the BBC today that an AI hallucination drug called PHP has been doing the rounds...
Upgrading from OpenBSD 7.9 to 8.0 is just too easy now. I miss having to print out the FAQ, grab the install ramdisks, merge /etc manually, and pray to the BSD gods for driver rain
The @trailofbits+OCaml "patch the planet" week is ongoing, and the lesson we've learnt is to ponder really, really, really hard about exposing more C interfaces from our favourite programming languages ever again
Anil Madhavapeddy retweeted
10 Downing Street is recruiting a Director of AI in the No10 Policy Unit, where I work. They will advise the Prime Minister on AI.
This is an important role for the UK!
.plan-2026-w42: Lectures start with unexpected questions as always, and a sampling of some of the meetings I'm lucky enough to take: from the Japanese secretariat, to hosting samosa socials at Pembroke College on responsible AI, to undisclosed locations in London to hear about AI safety from... well, read on for yourself: anil.recoil.org/notes/2026w4…
Anil Madhavapeddy retweeted
One of the best things to happen to Indian academia is this human dynamo moving back to India and making awesome things happen there. IIT-M is lucky to have him and smart to promote him.
something that trips up students every single year is why OCaml (and Scala and Standard ML and GHC) have list tail functions that error on an empty list, instead of just returning an empty list
And me! I'm deploying so many throwaway services on @exedev and it's got an auth layer so it can be selectively shared
Welcome to the 2026 Cambridge 1A CST students for their first Foundations of Computer Science lecture! They’ve started a journey in the most rapidly paced eras in our field I’ve ever seen. May the functional programming gods spark the same enduring joy in CS as it did for me :-)
2026w41: a week into full term, and our Computer Science Part II projects are so much fun: everything from 'debugging Bach' with a musical OCaml DSL to figuring out how to teach AI to stop predicting that elephants swim in the oceans off the Gabonese coast anil.recoil.org/notes/2026w4…
You gotta feel (*) for Google, being scooped by OpenAI and Anthropic via phat tweet proof drops while they patiently waited for the Science peer review process (presumably for months).
(*) Like a mild twinge in my neck, I'm not crying or anything
Another big AI math drop. This time from Google DeepMind
Google DeepMind’s AlphaProof Nexus AI solved mathematical questions that had resisted progress for decades.
The system autonomously resolved 9 of 353 open Erdős problems, including 2 that had remained unsolved for 56 years, and also proved 44👀 of 492 open conjectures from the Online Encyclopedia of Integer Sequences.
It also resolved an open question in algebraic geometry and improved a known bound in min-max optimization.
The key: pairing an LLM’s ability to generate proofs with Lean’s formal verification. The AI proposes a proof, the checker tests every logical step, and failed attempts produce feedback for another try. Proofs must pass the checker to be accepted.
Researchers also tested a simpler approach. Repeatedly generating candidate proofs with an LLM and checking them in Lean, alongside their more sophisticated search framework.
Finished up the Cambridge 1A Foundations of Computer Science course notes, now as a forester with toplevels. cl.cam.ac.uk/teaching/2627/F… This'll may be the last year this course (which has been going for thirty years) will run in its current form, but I need to meditate to figure out its new shape.
For those curious, our course materials are public in Cambridge. Here's the 1998 version of the Foundations course and the 1A CST Tripos. Robin Milner used to give the intro lectures! It was Standard ML back then, OCaml since 2019. cl.cam.ac.uk/teaching/1998/I…
Anil Madhavapeddy retweeted
🗞️ unikernels were hard. key word: were.
ghuntley.com/unikernels
Justin Cormack, who worked on MirageOS back in the day, called me at the pub to talk about why unikernels are coming back. The friction that made them hard is gone; the operating system is design debt, and if you care about security, you should seriously consider them.
This hardware looks awesome but for god's sake please secure your software layer after reading our paper on NERVE attacks against BCI devices arxiv.org/abs/2609.08971
we've raised $50M from Khosla Ventures, Accel, Initialized, Kevin Weil, DST Global
to build the world's most wearable BCI cap at Sabi
read more here:
forbes.com/sites/rashishriva…

















