Fable 5 and Karpathy's knowledge base trick are basically a cheat code. Someone just dropped a full walkthrough on building the most powerful second brain ever Andrej Karpathy shared a simple idea: use an LLM to build personal knowledge bases. Index your sources, let the model organize them, and... use Obsidian as the front end. Someone just built the full thing with Claude Code and Fable 5, and walked through every step You drop your sources into a folder (YouTube transcripts, PDFs, articles, meeting notes) and Claude Code reads them, splits them into linked wiki pages, and builds a living second brain that connects everything. It grows on its own as you add more. Where Fable 5 pulls ahead: it found a connection buried across two sources you'd miss reading them separately. And asked to turn the messy web of links into something a beginner could understand, it produced clean, simple HTML in one shot. The same task took a full day on Opus 4.8 and came out worse. The best part: it's all just markdown files with routing. So it's not locked to Claude Code. Plug it into Codex or any agent you want. Give your AI a real knowledge base, and it stops guessing and starts knowing. Bookmark this.show more

The index drifting when sources contradict or update is the part nobody shows in the demo. Without explicit contradiction detection and re-index triggers, the second brain quietly becomes a liability.

interesting point 👀

the obsidian second brain pipeline has been 'almost there' for 3 years this might actually be the one that doesnt fall apart after week two 🔥

Fable 5 is already the coolest product to date

I think it will be like this for a long time

I was very surprised when I heard this

I had a similar thing

banger article

insane gem. stealing this playbook

no problem bro

AI tools become genuine assets with real skill

Whoa bro it changes everything

Now I'm waiting for Andrej Karpathy to announce Fable 6

time will tell, heh :)

Thanks for such a useful and cool tip 🤝

I'm glad you liked it

This Chinese guy surprised me a lot

It's sooo good! I'm using it to better my knowledge about ecommerce

i adore fable 5

The blind spot in most second brains is video. What you learn on YouTube stays locked in audio your model can't index. Pull the transcript first: YouTube's own transcript panel for one video, or yt-dlp if you code. Then feed it in like any other source.

Second brain on autopilot neat

Love the energy, and the pattern genuinely works — but there’s a second half to this story most threads skip. Worth 2 minutes before you commit your whole vault:

thanks for new banger bro

second brain is the best thing there can be

I watching all video this guy

Nice. This combo sounds game-changing—totally agree that indexing sources + LLMs makes a strong second brain. Let’s see it in action.

A second brain that grows as you add sources sounds great

Indexing is just the start. The real challenge is retrieval efficiency in dynamic datasets.

个人知识库接上LLM后，价值不在存得多，而在能把分散材料串起来。只要源文件干净、结构稳定，后面扩展到不同代理会很顺。

legit

that's the real unlock, not some magic trick or 'cheat code', but collapsing the time between research and application. when you can build on top of existing knowledge like this, that's when the real progress happens.

Sounds like a tidy way to make chaos feel organized. Still waiting to see how often it holds up outside the demo.

The LLM-indexing-into-Obsidian idea is genuinely something Karpathy has discussed, and it's a real working pattern. But "Opus 4.8 took a full day and came out worse" is a specific comparison with no benchmarks or examples shown — reads more like promotion for a particular model than a reproducible test.

Impressive, the ability to auto‑link across sources truly elevates a second brain. Shows how LLMs can move from guessing to genuine knowledge

No. Simply put. This is grifter influencer guru BS. You can’t outsource graph linkage for effective traversal without defined logical models etc. this is amateur 101. Stop posting this BS.

gonna throw this to my ai agent starting to set up obsidian today!