Nicholeee

Crypto｜AI learning

[Nicholeee @Nicholeee11]

Love Andrej’s post. Recently, I’ve also been thinking about how to organize my notes and knowledge. I love taking notes but am not organized. Messy notes kill me. So, it seems to really work.

[Andrej Karpathy @karpathy]

LLM Knowledge Bases

Something I'm finding very useful recently: using LLMs to build personal knowledge bases for various topics of research interest. In this way, a large fraction of my recent token throughput is going less into manipulating code, and more into manipulating knowledge (stored as markdown and images). The latest LLMs are quite good at it. So:

Data ingest:
I index source documents (articles, papers, repos, datasets, images, etc.) into a raw/ directory, then I use an LLM to incrementally "compile" a wiki, which is just a collection of .md files in a directory structure. The wiki includes summaries of all the data in raw/, backlinks, and then it categorizes data into concepts, writes articles for them, and links them all. To convert web articles into .md files I like to use the Obsidian Web Clipper extension, and then I also use a hotkey to download all the related images to local so that my LLM can easily reference them.

IDE:
I use Obsidian as the IDE "frontend" where I can view the raw data, the the compiled wiki, and the derived visualizations. Important to note that the LLM writes and maintains all of the data of the wiki, I rarely touch it directly. I've played with a few Obsidian plugins to render and view data in other ways (e.g. Marp for slides).

Q&A:
Where things get interesting is that once your wiki is big enough (e.g. mine on some recent research is ~100 articles and ~400K words), you can ask your LLM agent all kinds of complex questions against the wiki, and it will go off, research the answers, etc. I thought I had to reach for fancy RAG, but the LLM has been pretty good about auto-maintaining index files and brief summaries of all the documents and it reads all the important related data fairly easily at this ~small scale.

Output:
Instead of getting answers in text/terminal, I like to have it render markdown files for me, or slide shows (Marp format), or matplotlib images, all of which I then view again in Obsidian. You can imagine many other visual output formats depending on the query. Often, I end up "filing" the outputs back into the wiki to enhance it for further queries. So my own explorations and queries always "add up" in the knowledge base.

Linting:
I've run some LLM "health checks" over the wiki to e.g. find inconsistent data, impute missing data (with web searchers), find interesting connections for new article candidates, etc., to incrementally clean up the wiki and enhance its overall data integrity. The LLMs are quite good at suggesting further questions to ask and look into.

Extra tools:
I find myself developing additional tools to process the data, e.g. I vibe coded a small and naive search engine over the wiki, which I both use directly (in a web ui), but more often I want to hand it off to an LLM via CLI as a tool for larger queries.

Further explorations:
As the repo grows, the natural desire is to also think about synthetic data generation + finetuning to have your LLM "know" the data in its weights instead of just context windows.

TLDR: raw data from a given number of sources is collected, then compiled by an LLM into a .md wiki, then operated on by various CLIs by the LLM to do Q&A and to incrementally enhance the wiki, and all of it viewable in Obsidian. You rarely ever write or edit the wiki manually, it's the domain of the LLM. I think there is room here for an incredible new product instead of a hacky collection of scripts.

[Nicholeee @Nicholeee11]

Hey! I added CA$160 to your Kraken account. Unlock it after you signup and trade with code r2hhsb3p or through this link: https://t.co/JGWzJXDHR1

[Nicholeee @Nicholeee11]

Hey! I’m using the Kraken app to buy and sell crypto. Want to join me?

If you use my referral code or link to try it, we’ll both earn CA$50 when you trade CA$200 of crypto in the app!
Code: c9ssct4r
Link: https://t.co/3tsDxDGW1i

[Nicholeee @Nicholeee11]

@laofeiyyds cool

[比特币老肥（敏感肥从头再来版） @laofeiyyds]

来吧兄弟们以太坊4000了，此推生效，送10000给兄弟们，关注 @laofeiyyds 一键三连，评论区留��币安uid ，开搞

[ChandlerGuo 郭宏才 宝二爷 @ChandlerGuo]

Hey @grok,

If #bitcoin hits $125997 within 168hours, pick one person from the comments who follows me

The winner gets $1,000 in $usdt or $ethw

转发 评论 7天内 比特币破125997 抽奖1000美金

[Nicholeee @Nicholeee11]

@ChandlerGuo @grok Hey! I’m using the Kraken app to buy and sell crypto. Want to join me?

If you use my referral code or link to try it, we’ll both earn CA$50 when you trade CA$200 of crypto in the app!
Code: c9ssct4r
Link: https://t.co/pSubuEKwHk

[Nicholeee @Nicholeee11]

Hey! I’m using the Kraken app to buy and sell crypto. Want to join me?

If you use my referral code or link to try it, we’ll both earn CA$100 when you trade CA$200 of crypto in the app!
Code: yjpqbtw5
Link: https://t.co/KGNNFJD00f

[Nicholeee @Nicholeee11]

🔥 **FREE $100 in BTC in 10 MINS!** 🔥
Easiest Kraken bonus ever

Steps to Claim👇🏻
1️⃣ **Download** Kraken
2️⃣ **Sign Up & USE MY CODE**:
→ Enter **yjpqbtw5**
3️⃣ **e-Transfer $200 CAD**
4️⃣ **BUY ANY CRYPTO** ($200)
💸 **$100 BTC REWARD LOCKED!**

YOU GET**: $100 BTC

[Nicholeee @Nicholeee11]

@RuheZhou 求拉

[Nicholeee @Nicholeee11]

@okxchinese 2024年12月9日8:05 am

[The Banana Nation @thebananation]

Banana has officially hit 10M users 🍌🍌

And we couldn't have done this without you guys!

To commemorate this milestone, we have decided to roll out a B10MM campaign—USDT Giveaway 🎁

Date & Time:

10th September 12 PM UTC to 20th September 10 PM UTC

Here’s how YOU can participate:

• Take a Photo or Record a Video
• Snap a picture or record a video of yourself eating a banana and talk about your favorite banana skin in the game! What makes it special to you? Let your personality shine through!
• Get creative—whether it’s a quick bite or a full-on banana feast, we want to see it!
• In your photo or video, let everyone know to "Go Bananas!"

For the full details: https://t.co/QKXf0cJb6Q

Time to Go Bananas! 😍

[CARV @carv_official]

Introducing CARV Labs 🥼

With backing from top blockchain VC funds, including @HashKey_Capital and @Consensys, CARV Labs has pledged to support innovators, particularly those actively building the future of decentralized data 🌐

CARV Labs will also provide comprehensive support to startups, including:

• funding and investment supported by leading industry funds
• go-to-market and growth support
• expert tokenomics advisory
• infrastructure and tooling tech advice
• access to the company’s industry network and community

It's never too late to become a CARVer.

Read ALL about it here:
https://t.co/x01LmNEcOv

[The Banana Nation @thebananation]

Our besties at @carv_official has launched an ONBOARDING event for Banana users to learn more about CARV Play 🍌

Start studying here: https://t.co/HZIHLZFWaM

Oh, and if you're still unaware, new tasks have been added in Banana as well <3

[The Banana Nation @thebananation]

8M 🍌 users and counting.

What should we do when we hit 10M users? 🤔

https://t.co/Jt6xnLtrKQ

[The Banana Nation @thebananation]

❓ Surprise Giveaway ❗️

You know the drill; 45 winners, 45 $TON.

All you need to do is:

• Add 🍌 to your name
• Follow @thebananagameee
• Like this tweet
• RT this tweet
• Tag 3 friends
• Comment your $TON wallet address

Lets go BANANAS!

[The Banana Nation @thebananation]

We doubt you guys saw this coming... what do you get when 🍌 meets 🐶??

THAT'S RIGHT—DOGSnana!

From 23rd Aug to 6th Sept, players who managed to obtain DOGSnana will WIN a share of 2M $DOGS!

Start playing now or be left behind: https://t.co/jH53qaFCsT

[The Banana Nation @thebananation]

If you aren't aware yet, we hit 5M users within 3 weeks!

For the full read: https://t.co/CixK5HYI5J

Can we get a "LFB" in the comments? 🍌

[The Banana Nation @thebananation]

We know it's the weekends but...

Banana just reached 5M users in less than a month 🥳🍌

Thank you to all of you who constantly supports us!

More giveaways/campaigns soon? 🤔

[The Banana Nation @thebananation]

Another one bites the Matrixnana 👀