Hacker News

points by whycombilater 3 days ago | 0 comments

LLMs are just regurgitating an output according to a set of rules generated from harvesting all human output.

Language to humans is a symbolic mapping of labels to reality based on actual experience.

"Language" to LLMs is just a string of tokens to manipulate based on relationships harvested from recorded human experience.

It doesn't mean anything other than 1s and 0s - it's not a language because the LLM isn't communicating with you, just as the sum of humanity isn't communicating with you.

It's just a really good "choose your own adventure" book.

[–] hardbass 3 days ago

...do you think molecules and their interactions are also not bounded physical systems? What part of a human being do you think isn't a turing computer? Please release your groundbreaking physics paper proving this.

[–] whycombilater 2 days ago

Yes they are.

Humans aren't magical.

They also aren't a glorified library. They are authors, they generate the books in the library.

Cross reference a load of books and you will get useful relational info, but it's still a library index in a library.

[–] hardbass 2 days ago

How do you know an llm cannot create new content. Humans create what they do out of existing information and molecular and neural interactions, so do llms except with electrical interactions. I have had llms produce novel content all the time.

[–] whycombilater 1 day ago

You're confusing hidden relational information extracted from training data for "new content".

[–] hardbass 18 hours ago

How do you think humans generate new content? Do you think they get new information from thin air?

[–] whycombilater 15 hours ago

From their own subjective experience.

[–] hardbass 15 hours ago

Yes and? Still not seeing whats the special part you are seeing that's different here. LLM's have both the massive data they are trained on to draw from and working memory. I am not sure what this subjective experience you think is other than data and signals.