From COBOL to Copilots: Why “No-Code” Never Kills the Coder
summary
Claims that English‑like languages or no‑code will replace programmers get tested against concrete failures: Flow‑Matic and COBOL require experts, HyperCard built Myst but had to be fully ported, 4GLs and low‑code handle niches but fail at auth scaling and CI, and Copilot and LLM agents produce hallucinations, prompt‑injection and reproducibility gaps that demand supervision. The pattern repeats over ~70 years: tools lower friction for ideation and scaffolding but not for production reliability. Concrete takeaway: AI speeds scaffolding; engineers still finish, verify and secure.
AI-written summary · about
chapters (16)
- 0:00Introduction
- 1:11English-like languages myth
- 2:58COBOL’s birth with Flow-Matic
- 3:39COBOL’s promises and limits
- 5:03BASIC and programming democratization
- 6:48Fourth-gen languages niche
- 9:15HyperCard and visual programming
- 12:02Visual Basic and low-code
- 13:46Executable UML failure
- 14:08Low-code/no-code in 2010s
- 16:33AI coding assistants hype
- 18:17Cyclical tool promises
- 20:24AI in modern development
- 23:16Why AI won’t replace coders
- 24:20Working with AI tools
- 25:47Q&A
related talks
18mGrounded Copilot: How Programmers Interact with Code-Generating Models
47mCode Is Cheap. Software Isn’t.
16mBeyond the IDE: The Future of Developing Software With Agents
30mBeyond the Hype: Using AI Coding Agents Today
43mWill code remain a relevant user interface for end-user programming with genera...
20mWill AI Replace Me? A Philosophical Mind Dump.
01h 32mCan AI replace programmers?
45mProgrammer’s Apprentice Season 2: Future Directions in AI-assisted Coding