Quick one that matters if you ever worry about AI getting too expensive. A new model called GLM-5.2 was just released for free. Anyone can use it, and companies can even run it on their own computers at no license cost. The interesting part: on certain coding tasks it scored higher than GPT-5.5, one of the big paid models, while costing roughly one sixth as much to run. In plain terms, the free and open side of AI is catching up fast to the expensive paid side. That is good news for normal people. More competition usually means better tools and lower prices for everyone. You do not need to switch anything today. The takeaway is simpler: you are not stuck paying premium prices to get strong AI. The cheap and free options are getting genuinely good. Have you tried any free AI tools that surprised you with how good they were? Tell us which one below.

Here is a headline almost nobody saw coming. SpaceX, the rocket company, has agreed to buy an AI coding tool called Cursor for 60 billion dollars. Cursor is a program that writes and fixes computer code for you. You type what you want in plain English and it builds it. A lot of developers swear by it. So why would a rocket company want it? The short version: every big company now wants AI building things in house instead of renting it from someone else. Owning the tool means owning the speed. This is one of the largest deals ever in the AI world, and it landed just days after SpaceX went public on the stock market. The bigger picture for the rest of us: the everyday tools people use are becoming so valuable that the biggest companies on earth will pay a fortune to own them. What is the one AI tool you would be sad to lose if it vanished tomorrow? Drop it in the comments.

Last week a major AI model got pulled offline by a government order, just days after it launched. If that made you wonder what happens to your work when the company behind your tools can flip a switch, you are asking the right question. Here is the other side of that story. There is a whole class of AI called open weight models. The company releases the actual model file to the public, so anyone can download it and run it on their own computer. Once it is out there, no single company or government can quietly take it back. This week one of those open models, called Kimi, scored higher than one of the big paid flagship models on a test of doing real work with other tools. So this is not a case of free but worse. On that test, the free downloadable one won. What this means for you: the tools you rely on do not all have to live on someone else's server. Keeping one option that runs on your own machine is starting to look less like a hobby and more like common sense. Would you ever want a private AI that runs entirely on your own computer, or do you prefer the convenience of the big online ones? Comment below.

You can run a real AI on the computer you already own, for free. No subscription, and nothing you type ever leaves your machine. It is more approachable than it sounds. You install one free app, download a model once, and from then on you have a private AI that works offline with no monthly fee. Good for anything you would rather not put into someone else's website. I wrote up the full step by step: the exact app to use, and which model to start with depending on how powerful your computer is. The PDF is attached to this post, grab it below. Give it a read, then tell me in the comments: what is the first thing you would ask an AI that you knew was completely private?

Welcome to the AI Untapped Workshop. This is the room behind the newsletter and the YouTube channel. The whole point: cut through the AI noise, faster, together. Here's what lives here: → One new tool + prompt every Sunday. Posted in the feed. Copy, paste, run it. → The Field Notes: five short PDFs covering local AI setup, prompt patterns, the skip list, the maker stack, and the Sunday routine. Find them in the Classroom tab (top nav) or scroll the feed today, where each one is posted separately so you can ask questions in the comments. → An open thread where you drop the AI tools you're testing and what's actually working. → Behind-the-scenes context I don't put on YouTube. What I'm asking from you: → Post what you're testing. Wins and failures both. The room learns faster when we're honest. → No pitching. No "I built an agency, DM me." This is a workshop, not a marketplace. → If a prompt works, drop the result in the comments. If it breaks, drop the error. I'm here daily. Drop a hello below and tell me one thing you're trying to figure out with AI right now. I read every reply. — Dusty