LLMs "increasingly… role of being transformers for back-end symbolic data stores (knowledge graphs)… become a natural language bridge for querying the stores, and once the results are produced, they also become a similar bridge for converting the results of such queries into natural language output”

Other talks by-
✅Andy Neale “Open Legislative Data Services”, from NZ Parliamentary Counsel's Office
✅Sarah Sutherland, with interesting take on limits of #RulesAsCode
✅AustLII on their experiments in using RAG on their huge body of legal data
✅Lexum & CanLII, incl on AI-assisted search on CanLII

Our @mattwadd.bsky.social presented “Visualising legislative logic” (slides osf.io/uk2vy/files/...) - demo of interactive logic mapping tool (github.com/smucclaw/l4-... & jl4.legalese.com?id=8831700c-...) using L4 language for coding law. More at crlp-jerseyldo.github.io/work/an-ide-...

Just spotted that that link was to the map of a simpler version of the “farmer must feed an animal”, which doesn't have the condition that the farmer is not exempt. If you want to try a map with that negative condition added in as an extra twist, go to jl4.legalese.com?id=8831700c-...

The demo example in the video is at jl4.legalese.com?id=8b65fc3b-..., where you can play with it (or make your own if you're that way inclined).
Some more explanation in our latest Substack post substack.com/home/post/p-... and on LinkedIn www.linkedin.com/posts/matthe...