[flagged]
LLM Wiki – example of an "idea file"
31–40 of 109 posts
Re: LLM Wiki – example of an "idea file"
#32I don't see why this wouldn't just lead to model collapse: https://www.nature.com/articles/s41586-024-07566-y If you've spent any time using LLMs to write documentation you'll see this for yourself: the compounding will just be rewriting valid information with less terse information. I find it concerning Karpathy doesn't see this. But I'm not surprised, because AI maximalists seem to find it really difficult to be...…
Edit for context: the sibling comment from karpathy is gone after being flagged to oblivion. Not sure if he deleted it or if it was just removed based on the number of flags? He had copy-pasted a few snarky responses from Claude and essentially said “Claude has this to say to you:” followed by a super long run on paragraph of slop. ———— Wow, I respect karpathy so much and have learned a ton from him. But WTF is the s…
Lol at that.
It's weird how some people cover the whole range of putting out some really good stuff and other times the complete opposite.
Feels as if they were two different people ... or three, or four.
Re: LLM Wiki – example of an "idea file"
#33He really wants to shine, but how is this different than claude memory or skills? When I encounter something it had difficulty doing, or consistently start off with incorrect assumptions, I solve for it and tell it to remember this. If it goes on a long trial and error loop to accomplish something, once it works I tell it to create a skill.
Re: LLM Wiki – example of an "idea file"
#34This is just RAG. Yes, it's not using a vector database - but it's building an index file of semantic connections, it's constructing hierarchical semantic structures in the filesystem to aid retrieval .. this is RAG. On a sidenote, I've been building an AI powered knowledge base (yes, it uses RAG) that has wiki synthesis and similar ideas, take a look at https://github.com/kenforthewin/atomic
> This is just RAG.
More to the point, this is how LLM assistants like GitHub Copilot use their custom instructions file, aka copilot-instructions.md
https://docs.github.com/en/copilot/how-tos/configure-custom-...
Re: LLM Wiki – example of an "idea file"
#35He really wants to shine, but how is this different than claude memory or skills? When I encounter something it had difficulty doing, or consistently start off with incorrect assumptions, I solve for it and tell it to remember this. If it goes on a long trial and error loop to accomplish something, once it works I tell it to create a skill.
> He really wants to shine, but how is this different than claude memory or skills?
It isn't different. This just tries to reinvent the wheel that all mainstream coding assistants have been providing for over a year.
Even ChatGPT rolled out chat memory in their free tier.
Re: LLM Wiki – example of an "idea file"
#36I don't think this is taking it as far as it can go. Everything should live in the repo. Code and docs yes. But also the planning files, epics, work items, architectural documentation and decisions. Here is a small example of my Linux system doc: https://github.com/gchamon/archie/tree/main/docs And you don't need to reinvent the wheel. Code docs can like either right next to it in the readme or in docs/ if it's too b…
> Everything should live in the repo. Code and docs yes. But also the planning files, epics, work items, architectural documentation and decisions.
You just described spec-driven development.
Re: LLM Wiki – example of an "idea file"
#37I don't see why this wouldn't just lead to model collapse: https://www.nature.com/articles/s41586-024-07566-y If you've spent any time using LLMs to write documentation you'll see this for yourself: the compounding will just be rewriting valid information with less terse information. I find it concerning Karpathy doesn't see this. But I'm not surprised, because AI maximalists seem to find it really difficult to be...…
Here in 2026, many forms of training LLMs on (well-chosen) outputs of themselves, or other LLMs, have delivered gigantic wins. So 2024 & earlier fears of 'model collapse' will lead your intuition astray about what's productive.
It is unlikely you are accurately perceiving some limitation that Karpathy does not.
Re: LLM Wiki – example of an "idea file"
#38Happy to see it gets attention. The friction shows up once you mix docs with structured things like work items or ADRs. Flat markdown doesn't query well and gets inconsistent. You can read TASKS.md fine. The agent can't ask "show me open tasks blocking this epic" without scanning prose or maintaining a parallel index.
The AGENTS.md approach papers over this by teaching the LLM the folder conventions. Works until the data gets complex but gets worse after many iterations.
Both are needed: files that open in any editor, and a structured interface the agent can actually query. Been building toward that with Binder (github.com/mpazik/binder), a local knowledge platform. Data lives in a structured DB but renders to plain markdown with bi-directional sync. LSP gives editors autocomplete and validation. Agents and scripts get the same data through CLI or MCP.
Re: LLM Wiki – example of an "idea file"
#39This thing already exists for multiple years - see https://deepwiki.com/ (99% it is autonomous, but it can be manually structured - see https://docs.devin.ai/work-with-devin/deepwiki#steering-deep...). There were also multiple attempts to replicate it with local LLMs.
The problem is that it is still a slop: not only it adds a lot of noise ("architecture" diagrams based on some cherry-picked filenames, incomplete datatables, hyperfocusing on strange things), it also hallucinates, adding factually incorrect information (while direct questions to LLM shows correct information).
Re: LLM Wiki – example of an "idea file"
#40I don't see why this wouldn't just lead to model collapse: https://www.nature.com/articles/s41586-024-07566-y If you've spent any time using LLMs to write documentation you'll see this for yourself: the compounding will just be rewriting valid information with less terse information. I find it concerning Karpathy doesn't see this. But I'm not surprised, because AI maximalists seem to find it really difficult to be...…
Edit for context: the sibling comment from karpathy is gone after being flagged to oblivion. Not sure if he deleted it or if it was just removed based on the number of flags? He had copy-pasted a few snarky responses from Claude and essentially said “Claude has this to say to you:” followed by a super long run on paragraph of slop. ———— Wow, I respect karpathy so much and have learned a ton from him. But WTF is the s…
It's not sad. He's a person like you and me. devnullbrain's comment is snarky. He invoked model collapse which has nothing to do with the topic of a wiki/kb, he wrote that karpathy is not normal, and then seemed to imply that the idea was useless. I'd be pretty in my feels and the fact that he wrote it and deleted it seems like a +1 normal guy thing.