The world says no: 63 % believe AI will not surpass us, against 38 % who think it will. Not a single undecided voter: on an inherently speculative question, everyone takes a side.
Men reject the idea fairly clearly. Women split exactly in two, making them the group most open to the possibility of being surpassed.
Age separates little. The 18-24s follow and reinforce the global majority, while the over-65s divide into two equal halves. It is one of the rare rounds where the generation most exposed to these tools is also the most sceptical.
The debate sits almost entirely on the refusal side, and it avoids the science-fiction register. The main argument is about nature, not performance: AI does not perceive, does not live and does not understand the world as a human does. It processes information and produces answers from models, which is very different from conscious experience.
The thread adds a reversal that travelled well: today it is rather AI that depends on us. We feed it constantly with our data, our knowledge and our usage; we nourish it as much as we consume it. One reply pushes further: if we feel we need it so badly, it is mostly because we have grown used to producing more and faster, in a system where competitiveness comes before quality.
The « yes » camp does not try to contradict any of this. Its single contribution sets a far more modest limit than the vote suggests: AI can surpass us, but not in every domain. The disagreement is therefore less about the machine's power than about what « surpassing us » means.