Martin Milani: “Bertrand Russell didn’t trust language to express truth. He built a new system—formal logic—to make thought precise. In Principia Mathematica, Russell didn’t try to say things clearly. He tried to prove them. Today’s AI skips that step. It treats language as knowledge. But LLMs don’t understand the world—they understand statistical shadows of words about the world.

If Russell were here, he’d remind us:
Words aren’t knowledge. Reason is.
Language isn’t intelligence. Math is.
LLMs might be how humans talk to machines—but they won’t be how machines think.
They’re not the mind. They’re the keyboard.” Link

Martin also recommends this article: “The Illusion of Thinking: Understanding the Strengths and Limitations of Reasoning Models via the Lens of Problem Complexity“