(What I wrote is screenshotted below.)

They capture the exact moment when a developing heart shifts from silence to its first beat. There is no “switch”: many cells gradually become active and, upon crossing a critical threshold, the entire tissue suddenly synchronizes.

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.

🚨HOLY CRAP!!! The Atlanta TSA line has now stretched to a stunning 153 minute-wait-time... AT 6AM IN THE MORNING!!!! The line is WRAPPING AROUND BAGGAGE CLAIM!!! THIS IS PURE INSANITY!!!!!!

Thank you to everyone who joined us for the very first @stripe Commons. We are excited to be building the future of open finance together with you.

When the intended decompressor is very lossy (like a human mind), overspecifying the representation with lots of synonyms and syntactic sugar seems prudent. When the intended decompressor is closer to perfectly lossless (as LLMs are rapidly becoming), it makes less sense. (5/12)

Last year, we invested in @Monumental_Labs. They use AI and robotics to reduce the cost of stone fabrication by 90%. Construction tech is our game and while I think what they are doing with art and sculpture fab is amazing, the bull case for ML imo was doing building scale construction with stone. Essentially, reviving a dormant material type for type I construction to compete directly with steel and concrete. Before investing, I had to answer a question that's been nagging me for years: exactly when and why did we stop building with natural materials like stone? Where did beauty in our built environment go to die? It's mostly always been an economic story.

A publicly listed Nigerian salt company just put out a public announcement begging its biggest shareholder to please call them back. Union Dicon Salt Plc, listed on the NGX, says its largest shareholder, a Brazilian firm called Aims Limited, owns 64 million shares which is 40% of the entire company. And they can’t reach them. At all. They’ve tried multiple times over a prolonged period and got nothing. So the Company Secretary literally published a notice on March 3rd asking Aims Limited to contact them via their office in Kirikiri, Lagos, or call his phone number or email him. A whole publicly listed company is out here posting “please call me” to the entity that owns 40% of it. Quick history. This company was founded in 1984 as a joint venture between Aims of Brazil and Nigeria’s Defence Industries Corporation. Aims was the technical partner with 40% ownership. The company processed and packaged raw salt until it was shut down temporarily in 1988 then merged with Union Salt Limited in 1991 to become Union Dicon Salt Plc. Now here’s what makes this matter. A 40% shareholder has serious influence over board appointments, strategic decisions and shareholder voting. If they can’t be reached, the company can’t hold proper AGMs, can’t get key approvals and governance basically stalls. Meanwhile the stock is up 141% year to date. Went from ₦6.90 to ₦16.60. Up 68% in the last four weeks alone. Somebody owns 40% of a company gaining 141% this year and can’t be found. Nigerian stock market will never bore you.

I love this metaphor from Terence Tao—widely considered the world’s greatest living mathematician—about one of the drawbacks of using AI to solve hard math problems. theatlantic.com/technology/2…