Why do AI people think that everything needs to be automated? Why do they think that people even want to automate it?

A viral critique from a developer is sparking debate by questioning the AI industry's fundamental drive to automate everything, especially programming. The author, Petr Bena, pushes back against bold claims from AI enthusiasts that 'programming is solved' and that AI agents will liberate humans from the 'struggle' of writing code. He argues this perspective is detached from the reality that many developers find deep satisfaction in the craft of coding, viewing it not as a struggle but as a well-paying, enjoyable, and meaningful job.

The post draws a direct parallel to predictions made over 15 years ago, like those from Elon Musk, that human driving would become obsolete and potentially banned. The author, a manual transmission driver and motorcyclist, found such visions dystopian, celebrating an activity he loves. He sees the same pattern repeating with AI agents: a top-down assumption that automation is an unqualified good, ignoring human agency, pleasure in skilled work, and the value of the process itself.

This critique taps into a growing undercurrent of skepticism within the tech community. As companies like OpenAI, Anthropic, and Google push multi-modal models and agentic frameworks capable of writing and executing code, the narrative often focuses solely on efficiency gains. This post challenges that narrative's human cost, suggesting the goal should be augmentation—tools that assist developers—not presumptuous liberation from a profession many cherish. The implication is that for AI to be truly adopted, it must respect and integrate with human preferences, not seek to erase them.