Programming The Law: Best Practices
summary
Translating statutes into executable code is harder than arithmetic because of ambiguity and non-monotonic legal reasoning; dozens of government systems contain thousands to millions of lines (one payroll is ~20,000 lines in SQL) and documented failures cost hundreds of millions and left soldiers unpaid and ~10,000 US stimulus recipients missed payments. The solution path includes rigorous lawyer-reviewed unit tests, isolating a legal core, using declarative tools (Drools, OpenFisca, or the Catalan language that compiles to JS/C/R), formal proofs and open source reuse.
AI-written summary · about
chapters (11)
- 0:00Start
- 2:20Example from the Us Tax Code
- 10:00Louvre Failure
- 14:35Technical Best Practices
- 17:15Do's and Don'ts
- 30:30Organizational Best Practices
- 32:34Legal Framework for Automated Decision Making
- 32:51Evolution and Maintenance of Your Code
- 36:12What Is Formal Methods
- 39:02Conclusion
- 39:57Is the Law a Turing Complete Language
related talks
49mCode as Law
19mA General Library of Legal Components
01h 11mAlgorithm Design and Law: Potential and Challenges
01h 06mData and Society
59mPrinciples of Programming Language Translators
01h 02mProgramming Language Research for Technical and Social Good: What PL Can Do for Good?
09mLaw as Science, Science as Law
34mLegal Barriers To Interpretable Machine Learning