Why We’ve Tried to Replace Data Analytics Developers Every Decade Since 1974

Since 1974, new tools have promised to replace the data developer. From SQL to Self-Service BI, and now Generative AI, the dream is always the same: let business users get answers automatically. Yet, we are still here. Why? Because the bottleneck in analytics is not typing speed; it is thinking speed.

A data platform is a model of a complex business. Pointing an AI agent directly at raw data tables doesn’t bring instant insights—it brings confident hallucinations. An AI cannot simply guess your unique business logic. But this time, something fundamental has changed.

This talk explores the reality of AI-augmented data delivery. Instead of replacing us, AI upgrades our jobs. By letting LLMs handle the repetitive work such as writing boilerplate SQL, creating tests, and drafting documentation, data developers can finally focus on what actually matters: system architecture, governance, and business context.