From: bencollver@tilde.pink   
      
   The Dream of Coding Without Coders: A History of a Persistent Promise   
   =====================================================================   
      
   by Marios Karagiannis   
   May 19, 2025   
      
   For as long as software has existed, there have been promises, often   
   grand, sometimes naive, that the need to "know how to code" would   
   soon vanish. The vision: ordinary people, business analysts, or even   
   executives designing powerful applications without writing a single   
   line of code. From the earliest days of computing to today's AI   
   revolution, this dream has been revived again and again. Yet, despite   
   billions in investments and waves of hype, the core of software   
   development, the logic, structure, and abstraction, remains   
   stubbornly human.   
      
   The 1960s: COBOL and the Business User   
   ======================================   
      
   In the 1960s, COBOL (Common Business Oriented Language) was created   
   to make programming accessible to business people. With its   
   English-like syntax, COBOL was supposed to bridge the gap between   
   domain experts and machine code. The dream was clear: managers and   
   analysts would write software themselves.   
      
   But COBOL, while more readable than assembly, still required   
   training, structure, and logical thinking. The dream didn't   
   materialize. COBOL coders,still in demand decades later, became their   
   own specialized workforce. Instead of removing the need for   
   programmers, COBOL expanded the profession.   
      
   The 1980s-90s: 4GLs and Visual Tools   
   ====================================   
      
   Fourth-Generation Languages (4GLs) promised another leap. Tools like   
   Fox Pro, Power Builder, and Oracle Forms let users "draw"   
   applications. Visual Basic allowed developers to build GUIs with   
   drag-and-drop components. At the time, these were seen as the end of   
   traditional coding.   
      
   But while these tools simplified UI creation and database binding,   
   complex business logic still required real coding. The abstraction   
   broke down quickly as projects grew. Power users emerged, but   
   professional developers remained essential.   
      
   The UML Era: Modeling as Programming?   
   =====================================   
      
   In the late 1990s and early 2000s, the Unified Modeling Language   
   (UML) was heralded as the new foundation for software development.   
   Why write code, the thinking went, when you could diagram it? With   
   Model-Driven Architecture (MDA), one could draw class and activity   
   diagrams and automatically generate applications from them.   
      
   Despite heavy support from enterprise vendors, this approach never   
   took off at scale. Software is not just structure; it's behavior, and   
   behavior is messy. Diagrams became too complex, brittle, and   
   incomplete to replace real code. UML found a niche in documentation   
   and architecture, but the coder was not dethroned.   
      
   The No-Code/Low-Code Renaissance   
   ================================   
      
   In the 2010s, a new generation of no-code and low-code platforms   
   emerged: Bubble, Out Systems, Mendix, and others. These platforms   
   boasted intuitive interfaces for building web apps, workflows, and   
   integrations. This time, the audience expanded to entrepreneurs and   
   startups.   
      
   While successful for prototyping, internal tools, or constrained   
   domains, these platforms hit a wall when it came to scalability,   
   customization, and maintainability. Developers were still needed to   
   extend functionality, ensure security, and keep performance in check.   
   Once again, the promise remained only partially fulfilled.   
      
   Now: AI Will Replace Coders?   
   ============================   
      
   The latest iteration of the promise centers around artificial   
   intelligence. Tools like GitHub Copilot, ChatGPT, and Claude can   
   write code, refactor it, explain it, and even suggest solutions.   
   Surely now, many claim, AI will finally eliminate the need to know   
   how to code.   
      
   But even AI doesn't remove the core challenge of software   
   development: understanding what needs to be built, translating that   
   into logical structure, and debugging edge cases. AI is a powerful   
   tool--perhaps the most powerful yet--but it is a copilot, not a   
   captain. It accelerates developers, it doesn't replace them. Just as   
   calculators didn't eliminate the need to understand math, AI won't   
   eliminate the need to understand code.   
      
   Why the Dream Won't Die--and Why It Won't Come True   
   ===================================================   
      
   The repeated promises share a common mistake: underestimating what   
   software development actually is. Coding is not just syntax; it's   
   problem-solving, system design, abstraction, trade-offs, and   
   communication. Each time we try to automate or abstract it away, we   
   rediscover how central human reasoning is to the process.   
      
   Software is not a commodity product. It's a living, changing   
   expression of intent. Until we can automate intent, and all the   
   ambiguity, creativity, and complexity it entails, there will always   
   be a place for coders.   
      
   From:    
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)