Tyson Maly

@Will_Tanner_1 We just need to see some job creation.

Oh look, tariffs are working and the economists were once again proven wrong Shocker that the people wrong about everything was wrong about this too

@OpenAIDevs Can it run against models on AWS Bedrock?

Bring Codex to your team without fixed seat costs. We’re rolling out usage-based pricing for Codex in ChatGPT Business and Enterprise plans, so teams have a more flexible way to get started.

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.

@drawandstrike @AGPamBondi @HarmeetKDhillon Drop boxes don’t involve the @USPS

Wait until @AGPamBondi and @HarmeetKDhillon drop the bomb on the cheaters: If you HAVEN'T given your voter rolls to the DOJ Civ. Rights Div. for verification yet, your STATE is not ELIGIBLE to have mail-in ballots counted in the next FEDERAL elections. Mail-in ballots for the 2026 midterms - what few there are going to be - WILL have a bar code that makes them trackable. Your 'election' NGOs are now shit-out-of-luck. Mail-in ballots are ONLY gonna count towards federal elections for President, Senator or House Rep in DC only IF that mail-in ballot was sent to a verified voter at at a verified address. Your 'election' NGOs like the ones Stacy Abrams got $2 billion for can sit out there in the ether frantically submitting requests that thousands of mail-in ballots get sent to vacant lots or an abandoned building until the cows come home. NO VERIFIED VOTER FROM THE VOTER ROLLS? NO VERIFIED ADDRESS? THEN **NO**, YOU CAN'T F**KING HAVE A MAIL-IN BALLOT WITH A BAR CODE SENT TO YOU. Losers! If you **pay attention** to the moves Trump is making, as he LITERALLY forces the GOP UniParty in our Congress to CONTINUE EXPOSING WHO IT REALLY WORKS FOR, you can SEE what's coming. CHECKMATE.

@johnkonrad Will all that pledged direct investment in the US materialize if those making the pledges in the Middle East can’t sell their energy?

BREAKING: US Navy will not open the Strait of Hormuz Trump gives nations who import oil from the Middle East three options: 1) Buy your oil from US and Venezuela instead 2) Secure the Strait themselves with limited US support Or 3) Wait for Hormuz to reopen naturally

@demishassabis I am guessing Gemini 4…

💎💎💎💎

Google Stitch introduced a new concept: DESIGN . md Like README . md but for design systems. A plain markdown file that LLMs read to generate consistent UI. An awesome collection of DESIGN . md files inspired by developer-focused websites like Stripe, Vercel, Linear, Notion, Figma and more. Drop one into your project. Your AI coding agent builds the rest.

@XFreeze Call center jobs are going to have to change

xAI's Grok Voice Agent API is insanely good.. You get a voice agent that sounds completely human and handles your always-active customer support across any domain, working for you 24/7 You can now build full voice agents - Medical Office, Restaurant Host, Real Estate, Hotel Concierge, Help Desk or create your own from scratch Even you can do it without any coding or needing a team to spend months setting things up Anyone can now deploy a real AI voice agent to handle calls around the clock for just $0.05/min That's only $3/hour for a worker that never sleeps, never calls in sick, and seamlessly speaks dozens of languages This is the stuff that used to cost companies millions. xAI just made it a template

@wesbos Wes, my gut tells me this code leak is an April fools joke

Claude code generates random 4 character IDs Then they filter out 25 swears. Kenneth tried to tell them..

Claude Code leaked their source map, effectively giving you a look into the codebase. I immediately went for the one thing that mattered: spinner verbs There are 187

@thehealthb0t I prefer sun shirts

Brecka: "The parabolic rise in skin cancer is almost perfectly superimposable with the parabolic rise in sunscreen use." O’Neill: "Exactly. Author Ian Whishart claims sunscreens themselves are causing basal cell carcinomas."

@jliemandt @shaneparrish Would love to hear more details on the knowledge graph and interest graph.

We waste 90% of kids' time in school. @shaneparrish asked the hard questions: Why do $50K/yr private school students show up years behind? Why did two-thirds of our high schoolers beg to skip summer vacation? What happens when you finally give kids their time back.👇

@ScottPresler @grok does existing laws around election security permit this or is it likely going to be overturned in court?

BREAKING MASSIVE NEWS President Trump just signed an Executive Order to provide for election integrity: 1) Election officials will be provided w/ federal data to ensure that ONLY American citizens are able to vote 2) Orders the Postmaster General & the United States Postal Service to verify that ballots are ONLY mailed to eligible voters & ballots are returned ONLY by eligible voters. This is a huge step in the right direction. Pass the SAVE America Act!

@altryne This could be a good thing for Anthropic. The open source community will do the Rust port for free. This buzz will attract OpenClaw like energy and they may get some new features developed by the community they can fold back into their Typescript version.

If you, like me, just woke up, let me catch you up on the Claude Code Leak (I know nothing, all conjecture): > Someone inside Anthropic, got switched to Adaptive reasoning mode > Their Claude Code switched to Sonnet > Committed the .map file of Claude Code > Effectively leaking the ENTIRE CC Source Code > @realsigridjin was tired after running 2 south korean hackathons in SF, saw the leak > Rules in Korea are different, he cloned the repo, went to sleep > Wakes up to 25K stars, and his GF begging him to take it down (she's a copyright lawyer) > Their team decided - how about we have agents rewrite this in Python!? Surely... this is more legal > Rewrite in Py > Board a plane to SK🇰🇷 > One of the guys decides python is slow, is now rewriting ALL OF CLAUDE CODE into Rust. > Anthropic cannot take down, cannot sue > Is this "fair use?" > TL;DR - we're about to have open source Claude Code in Rust

@Ncky_Mrtnez @altryne I am waiting for the Prolog version.

@altryne Leaked in TypeScript, ported to Python, now rewriting in Rust. By Friday it'll be in Assembly and nobody will remember why

@realsigridjin @bellman_ych The renamed repo looks like Python. I thought the true Claude Code source was Typescript?

we decided not to store the archive of claude code to avoid potential trouble so renamed to clawd-code github.com/instructkr/cla… and to develop the better harness with @bellman_ych together please consider sponsor or star us the previous our work was: github.com/Yeachan-Heo/oh…

New public experiment: Model Fusion Use multiple models, analyze outputs, and fuse the results for a response that every Deep Research agent preferred to its own, in our testing. No subscription needed at all.

My conversation with @jliemandt on why the future of education is better than you think. 0:00 The current education system 7:01 What makes Alpha School different 11:01 What are the results 23:20 Current classroom struggles 26:40 What does mastery mean? 35:37 Changing the education system 39:19 Teaching through AI 44:27 How do you solve motivation? 57:01 What makes a good teacher? 1:01:04 Coaching 1:05:17 What life skills matter? 1:08:18 Doing hard things 1:13:25 AI Monitoring 1:21:08 Effort vs. IQ 1:24:40 What happens after Alpha School? 1:38:21 The Genius of Jack Welch 1:45:49 Trilogy IPO: the choice to not go public 1:51:40 Physical vs. virtual learning 2:03:18 Does Paying Kids To Learn work? 2:11:01 What Is Success For You? (Includes paid partnerships)

@avidseries Tuition is revenue

Antiracism in action: University of Texas makes SATs optional for applicants, and then automatically accepts those students in the top 6% of their HS class — even if it doesn't know their SAT score and even if the high school from which they're graduating has almost no students proficient in math or reading. So what's the college GPA difference between those students accepted by Texas who opt out of submitting their SAT score, and those who submit their SAT score when they apply: A whopping 0.86 grade points. The difference in their SAT scores: 260 points.

@iruletheworldmo I would go with a consensus of model output if it were a decision with risk involved

ok polls aren’t working so i’ll try this slightly different. opus 4.6 seemed to prevail but let’s try again. outside of coding which of the following would you turn to for a complex problem / knowledge work i.e giving legal, financial, medical advice, or working on some complex knowledge work task. (leave your tasks and model used in the comments) thank you for your attention in this matter. the options are claude code (opus) codex (5.4 xhigh) gpt 5.4 pro extended.