In college I took a class on what it means to be human. We spent weeks eliminating the obvious answers: intelligence, tool use, language. Now AI is working on all three.
Не знаю, очевидно ли вам, как и автору, почему intelligence, tool use и language не делают человека человеком, но мне было не очевидно.
Вот что claude пишет про все три:
сначала мой TL;DR
Интеллект — вороны, осьминоги и шимпанзе проходят когнитивные тесты, иногда обыгрывая людей.
Инструменты — шимпанзе ловят термитов веточками, вороны гнут проволоку в крючки, осьминоги таскают кокосы как укрытие.
Язык — обезьян учат жестам, у китов и дельфинов сложная вокализация, луговые собачки кодируют в криках тип, размер и цвет хищника.
а ниже полный ответ на английском
These three were the classical answers to “what makes humans unique,” each dominant in different eras of Western thought. Aristotle called man the “rational animal” (intelligence). Benjamin Franklin called him the “tool-making animal.” And language has been the go-to answer from Descartes through 20th-century linguistics — Chomsky built a career on the idea that grammar is uniquely human.
They got eliminated because animal research kept knocking them down, one by one:
- Intelligence. Once you try to define it rigorously, other animals keep clearing the bar. Crows solve multi-step puzzles. Octopuses pass cognitive tests. Chimps beat humans at certain working-memory tasks. Whatever “intelligence” means, it’s a spectrum, not a human monopoly.
- Tool use. This was the cleanest answer for a long time — until Jane Goodall watched chimps strip leaves off twigs to fish termites in 1960. Since then: crows bending wire into hooks, otters using rocks, octopuses carrying coconut shells as portable shelter. Tool use turned out to be everywhere once people looked.
- Language. This is the most contested one. Apes have been taught sign language (Koko, Washoe). Dolphins and whales have complex vocalizations. Prairie dogs have calls that distinguish predator type, size, and color. Critics argue none of this is true language — no recursion, no grammar, no displacement (talking about things not present). But the line keeps getting blurrier, and “language” as a clean human/animal divide doesn’t hold.
So by the time the author took that class, these were the exhausted answers — the ones every intro course dispatches in the first weeks to clear ground for harder questions.
And here’s why it matters for the essay’s argument: the author is setting up a parallel. Just as animal research eroded those three pillars of human uniqueness, AI is now eroding them again — from the other direction. Models are intelligent (in some sense), use tools, and handle language. So the question “what’s left that’s distinctly human?” gets sharper, not duller, with AI. The class’s answer — coordination around shared fictions — becomes the author’s wedge into taste.
It’s a clever move: the same answers that failed against animals are failing against machines. Which suggests we were looking in the wrong place all along.