When software and law are the same thing
summary
An experiment encodes statutes as executable Python, stores them versioned in git, publishes them GPL and serves them hosted as an API with machine-checkable regression tests and linters. The implementation models rolling time windows, complex family and residency rules, and edge cases like leap years and time zones. The deployed code finds a long‑standing arithmetic bug in a rates benefit (overpayments), is reusable across jurisdictions, and is already consumed or adapted by New South Wales and Singapore.
AI-written summary · about
chapters (36)
- 0:00Intro
- 0:40Brenda Wallace
- 1:03New Zealand flag
- 1:17Legal advice
- 1:30Turning rules into code
- 1:56Laws invented by humans
- 2:35Have you ever been asked
- 2:55Expert present
- 3:28Code is rules
- 4:03NZ legislation examples
- 4:55What legislation is code
- 5:53New Zealand legislation
- 7:48Singapore
- 8:13New Zealand
- 9:04Code base
- 9:44Payroll software
- 10:29How laws are made
- 12:04Spreadsheets
- 12:42Programming Languages
- 13:12Legislation as code
- 15:24Why Python
- 17:35The EOR Act
- 18:47Python
- 19:36JavaScript
- 20:02New South Wales
- 22:26Student allowance
- 22:57Citizenship
- 23:15When are you eligible
- 25:05Hero coders
- 25:16DevOps for law
- 25:52Point of time
- 26:30Calculator
- 27:45Variables
- 28:15Ledger
- 28:27Pull requests
- 28:39Reusability
related talks
28mExperience report: implementing a real-world, medium-sized program derived fr...
47mCodelaw
50mSoftware Is Details
09mLaw.Gov: America's Operating System, Open Source
14mOpen-source it right (and avoid calling your IP lawyer)!
24mLicense compliance with AI assisted coding
27mTalking tech to legal: part 2
32mFreeing software with governance, for public administrations
