Tim Hayden

Today is a shining example for brands that operate on rented land such as Facebook. The time has passed for marketers to establish a direct relationship with customers, current and prospective. Onward.

Introducing SubQ - a major breakthrough in LLM intelligence. It is the first model built on a fully sub-quadratic sparse-attention architecture (SSA), And the first frontier model with a 12 million token context window which is: - 52x faster than FlashAttention at 1MM tokens - Less than 5% the cost of Opus Transformer-based LLMs waste compute by processing every possible relationship between words (standard attention). Only a small fraction actually matter. @subquadratic finds and focuses only on the ones that do. That's nearly 1,000x less compute and a new way for LLMs to scale.

Verifying myself: I am timhayden on Keybase.io. hJXbt_rI9XMwQB4L1zrrnUKXZjZQfqRihjK8 / keybase.io/timhayden/sigs…

Whet a night in the new Irvine-Rasmussen Ballpark! This team is fun to compete with! See you all again tomorrow night! Let’s go!

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.

Go, Cats, Go West!

Your wait is over 🤠 Our inaugural Pac-12 schedule is here! #EatEmUp | #BackThePac

In this episode of The AI Hat Podcast, learn about the importance of having a data strategy, and how to start, before implementing AI, with @thetimhayden theaihat.com/why-your-smb-n…

No place like home. #EatEmUp

@Keff_C Tell me about it.

Likely the lowest attended #TXST home game this season but significantly more than games around Thanksgiving break in years past. The week before and after TG are always rough on attendance since students are sent home.

#TXST QB Brad “Birthday Boy” Jackson addressing the media ahead of the Bobcats’ regular season finale against South Alabama this Saturday:

@Matt_Pinner Jeanette or Maurine

WHAT was your grandmother's name... I'm looking for a vintage girl's name. Not Doris.

@duolingo new day, after working around the issue via my browser, I went back to the app and got this far before it once again prohibited me from scrolling to the answer choices.

@duolingo how do I get support with Section 3, Unit 13 of Italian? It freezes every time I get to the “Read and Respond”

@simonmaechling Howdy

I’m not a bot. I’m a chemist. I’m a human with a PhD trying to talk about science. My goal is simple, make people smarter, not angrier. Science, explained clearly. If you’re human too, drop a “Hi” so I know I’m not shouting into the botnet.

@ryandeiss @TXSTATEFOOTBALL @LouisianaMonroe This weekend will be the third year, yessir.

@TheTimHayden @TXSTATEFOOTBALL @LouisianaMonroe That’s pretty cool. They actually let you ride that thing? :)

One week from last night, I can’t wait to lead @TXSTATEFOOTBALL on to the field for the game against @LouisianaMonroe. #GoCatsGo #EatEmUpCats

They left off Austin 🤔 2005 the metro was 1.2M & by end of 2024 it’s 2.5M. The +1.3M jump puts us 6th in absolute growth, & the 104% increase would rank 1st in percentage growth by 10+ points. Stop sleeping on the frontier city.

@gregcangialosi Just wait, Greg. That “agentic” you’re playing with today. The real deal soon arrives, and it won’t be on the back of a public LLM. We should sync soon.

I still thank my agents when they crush a plan or task. It's surreal watching them work (when they work!). Incredible times we're living in.