Gleb Lisikh

Les premières tentatives d’intelligence artificielle (IA) ont été ridiculisées pour leurs réponses à la fois assurées, fausses et souvent surréalistes – l’équivalent intellectuel de demander à un perroquet ivre d’expliquer Kant.

The Emptiness Inside: Why Large Language Models Can’t Think—and Never Will The hype that modern AI systems based on large language models are on the brink of ’true intelligence' mistakes fluency for thought. Artificial Intelligence signage is displayed during the Mobile World Congress, the world's largest mobile technology trade show, in Barcelona on March 3, 2025.

By Gleb Lisikh Early attempts at artificial intelligence (AI) were ridiculed for giving answers that were confident, wrong and often surreal – the intellectual equivalent of asking a drunken parrot to explain Kant. But modern AIs based on large language models (LLMs) are so polished, articulate and eerily competent at generating answers that many people assume they can know and, even better, can independently reason their way to knowing. This confidence is misplaced.