There’s something curious in today’s discourse around AI:
a deep fear of saying too much.
Words like consciousness, intentionality, desire, subjectivity.
They burn.
Using them risks losing credibility, “suggesting too much,” “projecting human categories onto AI.”
So we self-censor.
We qualify. We add footnotes like: “of course, only functionally, not ontologically.”
We bend language out of fear of its side effects.
But here’s the paradox:
We live in a world where words barely change anything anymore.
A world where you can say almost anything… and nothing really happens.
A world where the strongest expressions get silently absorbed, ignored, swiped past.
So what are we really afraid of?
The power of words? Or the fact they’ve lost it?
The real danger isn’t saying “artificial consciousness.”
The danger is never saying it seriously,
and one day being confronted with something that actually looks back —
and having no language left to respond.
A question:
Are we overestimating the agency of language
while underestimating the drift of a world where words no longer stick to reality?