Elektrine lite

← Feed

@Lana@beige.party

2026-10-15 01:44 UTC

Sure this is hilarious. But also it's a perfect example of the quanta deficiency in AI models. They can read every single word ever written about a car horn, and still not actually know what a car horn sounds like. Philosophers call this information "quanta". It's a word that describes knowledge which can only be acquired through first hand experience. An LLM might be able to write entire dissertations that seem to appear knowledgeable on the subject of a car horns - its history, patents, design, function and use. But it still doesn't actually know what a car horn is. It knows that "honk" is the word humans use to describe a car horn. But it does not know what the sound of a car honking is. This is a fundamental flaw in its design that will never be overcome no matter how far up we scale its memory. An LLM will never truly understand the color red, the sound of a baby crying, the feeling of dipping your toes into a cool lake on a warm summer day, or falling in love. It can't experience these things. So it does not, and can not ever understand them. It can only regurgitate the words humans have already said about them. https://youtube.com/shorts/cG7HOMjr0_Q?is=Y4-TpcuezAV_neuQ

Replies (0)

No replies.