Hacker Neue

Preferences
nobunaga parent
ITT: People who think LLMs are AGI and can produce output that the LLM has come up with out of thin air or by doing research. Go speak with someone who is actually an expert in this field how LLMs work and why the training data is so important. Im amazed that people in the CS industry seem to talk like they know everything about a tech after using it but never even writing a line of code for an LLM. Our indsutry is doomed with people like this.
usef-
This isn't about being AGI or not, and it's not "out of thin air".

Modern implementations of LLMs can "do research" by performing searches (whose results are fed into the context), or in many code editors/plugins, the editor will index the project codebase/docs and feed relevant parts into the context.

My guess is they either were using the LLM from a code editor, or one of the many LLMs that do web searches automatically (ie. all of the popular ones).

They are answering non-stackoverflow questions every day, already.

nobunaga OP
Yeah, doing web searches could be called research but thats not what we are talking bout. Read the parent of the parent. Its about being able to answer questions thats not in its training data. People are talking about LLMs making scientific discoveries that humans haven't. A ridiculous take. Its not possible and with the current state of tech never will be. I know what LLMs are trained on. Thats not the topic of conversation.
oezi
A large part of research is just about creatively re-arranging symbolic information and LLMs are great at this kind of research. For example discovering relevant protein sequences.
semiquaver
> Its about being able to answer questions thats not in its training data.

This happens all the time via RAG. The model “knows” certain things via its weights, but it can also inject much more concrete post-training data into its context window via RAG (e.g. web searches for documentation), from which it can usefully answer questions about information that may be “not in its training data”.

planb
I think the time has come to not mean LLMs when talking about AI. An agent with web access can do so much more and hallucinates way less than "just" the model. We should start seeing the model as a building block of an AI system.
raincole
> LLM has come up with out of thin air

People don't think that. Especially not the commentor you replied to. You're human-hallucinating.

People think LLM are trained on raw documents and code besides StackOverflow. Which is very likely true.

nobunaga OP
Read the parent of the parent. Its about being able to answer questions thats not in its training data. People are talking about LLMs making scientific discoveries that humans havent. A ridiculous take. Its not possible and with the current state of tech never will be. I know what LLMs are trained on. Thats not the topic of conversation.

This item has no comments currently.