Does FP still matter in the era of AI?
summary
Panel argues functional programming stays relevant because it yields proofs of correctness and language designs that eliminate the null problem, which AI models repeatedly mishandle; current LLMs are "probability machines" trained on mostly imperative code and often fail to produce production-ready implementations. Practical paths discussed include hybrid LLM–proof assistant models, verification tools that target AI-driven workflows, and using FP-style compositionality so generated code can be mechanically checked rather than blindly trusted.
AI-written summary · about
chapters (10)
- 0:00Introduction
- 1:50FP still matters
- 4:45AI code generation
- 8:43FP future with AI
- 15:31Requirements in AI era
- 18:21Formal verification evolution
- 23:19Educating for verifiable systems
- 25:07Engineering beyond coding
- 28:20Testing AI-generated code
- 31:16Q&A
related talks
59mFunctional Programming for Machine Learning Panel
01h 03mKeynote: Why Functional Programming Matters
58mKeynote: Why Functional Programming Matters
22mAdapting to the Future of Work in Tech
01h 28mFuture of Software Programming with AI Bots at Hand
24mDoes working software still matter in the age of generative AI?
26mThe present and future of programming with AI (Keynote)
33mAI is coming for your job