That's 5D-educational chess.

#FuckGenAI #ChatGPT #GenAIsucksCamelDong

@Eatsbluecrayon Another demonstration would be to bring a chessboard to class, along with extra pieces, and have the class play chess with an AI of their choosing. They’ll get to watch with their own eyes as the AI fabricates positions and pieces.

This is especially useful because computers have been beating human opponents at chess for a long time now, so people know that chess is something computers can do. That an AI can’t indicates it is worse than its predecessors.

@WhiteCatTamer @Eatsbluecrayon that's cause it's not an AI, but a really fancy autocorrect. DeepBlue and the like were probability machines using complex computational algorithms and ChatGPT just strings words together that make some kind of grammatical sense, in which grammer =! logic. The mistake that people are falling for is the thought that language is mathematical. It is not. Language is culture, which cannot be summed up through math and numbers. Culture is the synthesis of human emotion and connection. No machine can replicate that. Ever.

@jadedtwin @WhiteCatTamer @Eatsbluecrayon

If you don’t believe everything can be expressed in numbers then consequently you must believe there is some “magic” in which numbers are meaningless.

That simply isn’t the case. From the count of neurons firing, to their relations and positions: every “emotion” can be described with numbers.

Magic or numbers. That’s the choice.

Thus: language is math (numbers encoded).

@Eatsbluecrayon @jadedtwin @WhiteCatTamer @altruios

One would have to define what ”expressed in numbers” and ”described with numbers” mean. Is Pi a number? And if it is, does the symbol express it, or is it an expression for the inability to describe that function as a number?

But the main thing is this: Language has no fixed meaning. A symbol (like a word) is not one thing, but many. And it shifts.

Math doesn’t.

@jadedtwin @WhiteCatTamer @Eatsbluecrayon @altruios

I’m no mathematician, but language would be something like the n-body problem. In a sense it can be expressed as numbers, but it can’t be solved.

@thelovebing @jadedtwin @WhiteCatTamer @Eatsbluecrayon
It can be expressed as numbers. As you say: we agree? The meaning of words is a vector list of numbers. Here’s a video on the subject:

https://youtu.be/iErmK_sJtag?si=HaIk2FhXQ9BHNKe4

Word Embeddings: Word2Vec

YouTube

@jadedtwin @altruios @WhiteCatTamer @Eatsbluecrayon

I don’t do Youtube. If you have a point I’m sure you can make it yourself.

@thelovebing @jadedtwin @WhiteCatTamer @Eatsbluecrayon the video explains how you use math on language to encode meaning. Examples like “king - man = queen” relational vector math (where each word is a vector {list of numbers}). Word2vec is the keyword to research more.

Those numbers are variable, depending on the relational web of the vocabulary of the language, of course…

@altruios @thelovebing @jadedtwin @WhiteCatTamer @Eatsbluecrayon Which has, as it should, a "limitations" section at the end... which only briefly covers the many limitations of such things.

"Context" is a massive one which changes from person to person, use to use let alone from culture to culture and then from language to language (while translation apps do a reasonable job, especially at the basics, they'll always fall down with e.g. more complex prose where the translation becomes opinion).

@jadedtwin @altruios @level98 @WhiteCatTamer @Eatsbluecrayon

Precisely. ”Queen”, for example, is a word that has (had) many meanings. ”Queen Kristina” of Sweden –again, an example– wasn’t a queen. She never married, and the word used to signify the spouse of a king. She was crowned a king, though. Which of course changed the meaning of those words, at least in Sweden, but not the same way then as it later did.

@level98 @altruios @WhiteCatTamer @jadedtwin @Eatsbluecrayon

It is enormously complex, chaotic even. And while every meaning of a word might be described in numbers (or a dictionary), those descriptions are not very exact (and that’s why translation is really hard, because it isn’t just about denotation).

@level98 @altruios @WhiteCatTamer @jadedtwin @Eatsbluecrayon Margritte explained this better than anyone, the proof thereof being that tech bros mostly don’t get the explanation.