AI is eating software: who truly controls the self-programming machine?

For decades, software has silently governed the world. Banks, insurance companies, transport networks, even our romantic relationships: everything flows through lines of code comprehensible only to a narrow elite of initiates. Developers and software engineers have for years wielded an almost esoteric power, guardians of a language incomprehensible to most.

Today, something profound is changing. Artificial intelligence is not merely complementing traditional software: it is absorbing it, transforming it, and, in some cases, replacing it from within. The barrier to entry that for generations separated those who can program from those who cannot has collapsed with surprising speed.

This is the core of Francesco Maria De Collibus's reflection, anticipating the themes of his book dedicated to the self-programming machine. The central question is no longer technical, but political and cultural: if anyone can now instruct a machine to write code, who truly holds the power? And crucially, who is accountable?

The democratization of development tools opens up unprecedented scenarios. On one hand, it unleashes creative energy in sectors previously excluded from digital innovation. On the other hand, it dissolves accountability, renders decision-making processes opaque, and fuels a new form of technological dependency, perhaps even more invisible than the last.

We live immersed in software like fish in water, unaware of its presence until something breaks. With self-programming AI, the danger is that we might not even notice what's truly happening.