Will code remain a relevant user interface for end-user programming with genera...
summary
Argues the generative shift hypothesis: generative AI widens what end-user programming can do but does not eliminate code. Cites Copilot-style tools and experiments that show needs for prompt rewriting to make prompts match generated code, and for debugging and trust mechanisms because models hallucinate. Predicts a possible universal substrate that routes tasks to formulas, VBA, or native code, and calls for reproducible prompt methodology and reevaluation of attention-investment models.
AI-written summary · about
chapters (8)
- 0:00Introduction
- 1:42End-user programming
- 5:14Generative AI impact
- 8:25Code still matters
- 18:21Code versus prompting
- 21:36Research implications
- 31:52Conclusion
- 32:17Q&A
related talks
32mComputer Programming Is Dead; Long Live AI-First Programming
21mRevolutionizing software creation: Harnessing the power of GenAI
01h 32mCan AI replace programmers?
30mThe AI Revolution in Coding
20mWill AI Replace Coding or Improve It?
33mAI assisted software delivery
54mSoon replaced by AI? An overview of code generation
24mDoes working software still matter in the age of generative AI?