@gerardsans
@NandoDF Thatβs a common misconception, ironically one sometimes reinforced by companies like Anthropic when they boast about β100% AI-generated code.β In reality, anyone who actually works with these systems knows the model is just a tool. The real work is in the steering: knowing the direction, shaping the prompts, and the constant oversight required to keep the output on track. That takes skill, competence, and domain knowledge. Without that, the tool is mostly useless, something even their own documentation implicitly acknowledges. https://t.co/Ut4hpvTU3C