Software history is riddled with promises that business folk would simply describe what they want and let the machine do the rest. COBOL and BASIC claimed managers would write their own programs. HyperTalk let anyone click-build a HyperCard stack. BPML and UML tried to generate systems straight from diagrams created by managers and experts in meetings. Constraint based rule engines argued that if you just captured every rule, which managers could just write down, then the code would materialise. Low-code platforms repackaged the dream for the web, and now LLM agentic assistants and “natural-language” IDEs are the latest contenders.

So why do we keep hearing the same pitch? Because messy requirements, tacit knowledge, and perpetual change leak through every abstraction. Instead of vanishing, the complexity just moves... While developers, adapt, debug, and keep the lights on. History shows each wave actually creates new niches for people who can straddle business intent and technical reality.

In this talk I’ll:

• Zip through six decades of attempts to eliminate developers, their bright ideas, brief wins, and spectacular failures.
• Unpack the recurring traps: expressive ceilings, accidental complexity, maintenance drag, governance headaches.
• Show how every cycle boosted demand for devs by making it easier to do business rather than development.
• Hand you dead simple heuristics for sniff testing the big promises, spotting where these tools can help instead of hinder, and how to team up with AI tools without handing over your craft.

Expect a breezy ride full of war stories, face-palm moments, and a few “hey, that actually worked” surprises.