The Fear of AI

Even as we fine-tune our AI strategy — in our workplaces, our organisations, our personal lives — that lurking fear persists. Let's break it down. What, exactly, are we scared of?

1. Agency

For all of known history, humans were the only things on earth that could decide, create, and understand. Everything else — fire, rivers, animals — could be acted upon, but nothing acted back with intent. Now there is something that acts, chooses words, seems to understand — and it isn't us. The long-cherished hierarchy is showing cracks.

We know the machines won't kill us. We're good enough at that ourselves. What unsettles us is the vertigo of no longer being the sole author of the world. We've stumbled onto an agency we can't negotiate with — one whose conscience we can't appeal to, because there is no conscience (something we'd always assumed we had) to appeal to. Power without a face is a uniquely human terror.

2. Meaning

Every tool before this replaced muscle — the ox, the loom, the engine — and we always had a story for why that was fine: the machine does the drudgery, the human does the soul. But this is the first tool that reaches into judgment, language, art — the very things we used as proof that we mattered. "Will the machine take my job?" is no longer just an economic question. If a machine can write the poem, make the music, hold the conversation — and do it all more efficiently than us — then what are we here for? That's the existential ache behind the practical anxiety.

3. The mirror

AI holds up a reflection and shows us that much of what we called "intelligence" and "creativity" is pattern recognition — replicable, mechanisable. It threatens our specialness, not our safety. And that is harder to bear than any Terminator fantasy, because you can't build a wall against it.

The answer to this human anxiety lies in control. Who controls AI is the real question. A few unaccountable corporations holding the most powerful tool ever made is genuinely frightening. I have always argued that communities must re-appropriate technologies — and the means to control them. And that re-appropriation is possible only in open source. The question of open source has never been as relevant as it is in the age of AI.