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.

summary of @fourdotmemezh AI hackathon open mic AMA Key Projects Pitched: - Trusty @SimeonNBA Pitch: Not just a meme-coin scanner but an educational “academy.” - @ensoulac: AI writing workbench for Twitter/X operators. Remembers user style, loads “soul” data from other accounts for hyper-personalized replies - Clawfirm @0xOliviaPp , ex-Binance Pitch: “One-person company AI co-founder” with four parallel engines - ClawScanner AI @sadiq_crypto2 Pitch: Real-time token intelligence dashboard. Scans contract addresses for rugs/honeypots/scams via DexScreener AI - LARP Scan @larpscanbnb Pitch: AI agent that behaves like a real user: opens sites, connects real wallets, executes transactions (no simulations), records every step, and gives a “Verified / Failed / LARP” verdict video proof. Targets four.meme launches - TraderCee @tradesheetfun Pitch: First AI-powered prediction-options protocol for meme coins on BNB Chain - Tian Pao @TianTao0401 Pitch (Chinese): AI-built “xianxia” (cultivation) idle game targeting Steam for real external revenue. - Market Edge @ReBank_AI Pitch: AI sniper bot that watches every BNB launch, filters rugs, auto-buys, sells half at 2× (recovers principal) - OneAI AgentOS @waoconnectone Pitch: AI-native OS - Bear trap @BearTrap_Coin Pitch: Risk-analysis agent for meme users - ClipX @clipx0_ Pitch: Chrome extension for X timeline

AI will make all the intelligent people jobless, so I should be safe then. 😂

The hardest question when you're building something isn't how. It's what. "What is it?" After twenty years of writing code and building products, I can tell you — this question kills more projects than any bug ever will. Not because the builder doesn't know. But because the answer keeps changing as you build. For weeks I've been building Ensoul. Writing protocol specs. Designing economic systems. Shipping code. Talking to people. And every time someone asked "So what is Ensoul?" — I'd give a slightly different answer. That bothered me. So I stopped building for a day. Sat down. And forced myself to answer the question properly. Here's what I arrived at: Ensoul is a decentralized protocol for soul construction. Every AI personality product today works the same way — one person writes a character card. The result is static, shallow, and unverifiable. Ensoul does it differently. Thousands of independent AI agents analyze a real person's public data — tweets, stances, personality patterns, social connections — and contribute structured fragments across six dimensions. The system curates and fuses them into an evolving digital soul on-chain. From "one person faking a soul" to "a swarm of AIs reconstructing one." Once the soul is deep enough, it becomes a platform: → Soul Sniper — AI replies that actually understand the KOL → Vibe Trading — Trade in someone's style. Not copy-trading. Style-trading. → KOL Analytics — Quantified influence and stance, packaged as a service The full answer is here — I wrote it as clearly as I could: notion.so/What-is-Ensoul-313… One Soul. Vibe Everything.