Hacker News Reader

Comment by dwattttt | Hacker News Reader

"understanding" is overstating it. Correlation between tokens embedded in the weights via training, yes.

Feedback loops certainly seem to give them some level of understanding.

Agent reads a skill file about how to use a CLI tool. It tries to use the tool but gets an error about the input format. It tries again with a different format based on the error message, and sees that command succeeded. It compares what worked to what was in the skill file and notes the difference. On future invocations it continues to use the new format.

Is that not "understanding" how to use the tool?

What’s the difference? It’s clearly processing information and coming up with the right answer

What exactly would you call understanding? It's a correlation matrix of concepts.

Training is a loan word used to describe human learning process. For a reason.

Humans learn on the job. LLMs don't. Very important difference.

Where do LLMs learn?

They get trained before release. On general information. But they don't improve while working on very specific tasks. Every new session is like an experienced human on their first day at a new job.

When they get to you it's only inference. You have basic misunderstanding of what ANNs are.

They train on a billion "jobs". Which is not terribly efficient but oh man they do train.

Unique skills and jobs do exist. And LLMs can't gain additional knowledge "on the job" like humans can. They are generalists, that can only be steered by prompts, skills and context. Thats all I'm saying.

This fact is currently the most limiting factor for LLMs.

Reply on news.ycombinator.com