founder @ hellas.ai - verifiable AI (catgrad & ???) - theorem provers (Metacat) - combinatorial syntax for compilers!
- Tweets 1,651
- Following 786
- Followers 726
- Likes 4,773
Pinned Tweet
cyclic open hypergraphs are tricky to draw.
Here's ONE SIMPLE TRICK to run your acyclic layout algorithm on cyclic diagrams!
GraphViz Users Hate This! (kidding i love graphviz)
blog.hellas.ai/blog/cyclic-h…
1
3
13
1,002
just saw a dude carrying a rifle with a bayonet affixed riding the back of a scooter
China bros what
2
87
You may not like it, but Edward Kmett is what peak performance looks like
THC now covers at least nominally all ~1500 or so of GHC 9.14.1’s primops, handles GHC’s full traced callstack and async exceptions, and can load and run real applications using cabal: e.g. hscolour, happy, alex, doctest, etc.
I’m very close on running pandoc (a minor wibble about mixing native and managed libraries and lack of rtsopt pass thru kept me from declaring victory) and can just barely load GHC itself as a library with a small source modification (due to limitations around signal handling and the JVM), so I have a path to a REPL.
Need to fix up disparities between the bytecode and AST backends to fully declare async exceptions resolved and both as supported.
And I still want native-image and/or crac support for pre-canning executables.
github.com/ekmett/thc
1
4
423
Teaching 3yo to read
Found @TeachMonsters and it's going great
Insane value for money
1
2
179
statusfailed retweeted
I am running a formal methods hackathon on November 1st, with a bunch of my friends. Contestants will compete to build normie software -- video games, AI agents, travel booking tools, etc. -- but formally verified. We are going to measure to what degree these tools are usable by SWEs with no prior FM background and no FM-specific education. This is a super exciting opportunity to see where the ergonomic gaps are that we, the FM community, need to fill in order to take advantage of our moment in the zeitgeist. If you want to get involved, please DM me! In particular, I think I've got location/food/tokens/photography all covered, but I need some sponsors for the prizes. If you make a cool gizmo or product we can give away and you're down to donate one, please do so!
18
30
249
18,618
based and proofs-are-not-programs pilled
Perhaps this is just a skill issue, but I’ve seen a lot of good programmers get CH-pilled and then waste a lot of time trying to think of the types of their working programs as rich propositions or trying to encode complex proofs about their programs in the types.
3
7
1,733
Prediction: within ~1 year, formal methods will produce a capability spike in AI-assisted coding comparable in magnitude to the jump from GPT-2 to Astra.
10
11
152
38,869
startup idea: generate startup ideas from old Schmidhuber papers
Congratulations to @typesafeai on Jev! In April 1991, when compute was a million times more expensive, I published the first Jev-like neural decision system in Munich [AC91, Sec. 2]. Back then we called this “adaptive confidence.”
[AC91] J. Schmidhuber. Adaptive confidence and adaptive curiosity. Technical Report FKI-149-91, TU Munich, 30 April 1991.
people.idsia.ch/~juergen/FKI…
4
283
finally, Minecraft is solved, we can all stop playing it!
Jev + Astra beats the Ender Dragon in Minecraft in 8 minutes 43 seconds! ⏱️
Cost less than $1 ($0.01 Jev, $0.96 Astra)
I open sourced the code and explain the harness setup below. This type of movement is only possible with Jev's near instant decisionmaking, and some continually learning skills from Astra.
Community note
This run used Peaceful difficulty and a pre-selected common speedrunning seed where humans finish in under 2 minutes, unlike the prior viral Astra attempt with hostile mobs like creepers over 100+ hours.
github.com/rmalde/minecra…
x.com/rronak_/status…
x.com/valsai/status/…
thegamer.com/minecraft-best…
3
538
"Most peasants do not regard leisure as legitimate" is the most American-protestant-work-ethic shit I ever read
Makes me optimistic for the future of us/china relations 😁
1
179
95% of my work is now just thinking
4% is writing down my thinking in a paper
1% is making sure codex understands the paper
2
7
225
