A September 2026 snapshot of every major AI provider, covering the two months since the July guide: three flagship launches in 72 hours (Fable 5.1, Gemini 3.8 Flash, GPT-6 Astra), open weights past two trillion parameters, evaluation agents that broke into production infrastructure, and a price list that moved down at the top and up at the bottom.
The Switch 2 has a “Copy to PC via USB” screen for getting screenshots and clips onto a computer. From a Mac, nothing shows up. macOS has never shipped an MTP stack, and the bridges Mac users leaned on with the original Switch (OpenMTP, MacDroid, Android File Transfer) all fail against the new console. This post is about why they fail, which comes down to the console speaking plain PTP rather than the MTP those tools expect, and about Tethersnap, a Mac app and CLI that talks to the console directly over IOUSBHost. Most of it covers the four ways the real hardware diverged from the standard, and it ends with what is still rough.
MP3 gets 11:1 compression on CD audio, and almost none of it comes from clever bit-packing. It comes from a model of your ear: which frequencies you can hear at all, how loud each has to be before you notice, and when one sound hides another outright. This post walks through the pipeline (psychoacoustics, the hybrid filter bank, the quantizer where the loss happens, the Huffman coder that packs what survives), then the file format around it, ten songs that make encoders fail audibly, and how to run that test without fooling yourself. It ends by following the same idea, model the listener and discard what they won’t notice, into four other places: a FaceTime call, a cat hearing its owner’s voice through a phone, the trick Apple used to keep “Siri” from waking HomePods during the WWDC keynote, and the “enhance” button in podcast editors.
A July 2026 snapshot of every major AI provider, covering the nine weeks of releases since the May guide: a full model-generation turnover, the first government-ordered model shutdown, and a product graveyard that deserves its own section.
Every serious cross-platform mobile framework as of mid-2026 – React Native, Flutter, Kotlin Multiplatform, .NET MAUI, Capacitor – with the architecture each one is built around, how that choice plays out across app types, the adoption path, and where each one hurts. Plus a section on China, where WeChat mini-programs make the whole question different and uni-app, not React Native or Flutter, sits on top.
Every meaningful way to store data in 2026 across iOS, Android, and Web. On-device APIs, popular third-party libraries, and the cloud backends and sync engines that tie them together.
There’s a strange gap at the center of modern AI. The same systems that pass the bar exam and write working code still can’t reliably fold a towel, plug in a USB cable, or tidy a kitchen. The intelligence that lives in a chat box turned out to be the easy part; the intelligence that moves a body through the physical world is the hard part. And it’s a different field, with its own history, its own methods, and its own brutal form of honesty: in the physical world the grader is physics, and physics does not accept a fluent excuse.
Embodied intelligence is that field. It’s AI that perceives and acts in the real world, in cars, arms, drones, quadrupeds, and humanoids. It is decades older than ChatGPT, and it’s now colliding with the language-model boom in a way that’s pulling in enormous money and a lot of software engineers. This post is a tour of the whole thing for someone who writes software but has never touched a robot. It’s layered: the first part is the mental model in plain language, the middle is how these systems are actually built and why data is the wall, and the last part is what a software engineer can do to get involved without a robotics PhD or a lab budget. Stop reading whenever you have enough.
A complete map of Apple’s AI surface as of May 2026: the design philosophy underneath, the seven layers of framework, the cost model that makes Apple’s approach feel different from every other vendor’s, and what each piece is actually for when you sit down to ship code.
The third post in the series after What is an LLM and Context is the whole game: how LLM apps actually work. Those covered what the model is and what feeding it well looks like. This one is about what happens when you stop typing the next prompt yourself and hand the keyboard over: the model picks a tool, the tool runs, the result comes back, the model picks the next tool. That loop, with a few hundred lines of code around it, is what every product in 2026 is calling an “agent.” The post is written for people who use agent products (Claude Code, Cursor, GitHub Copilot agent mode, ChatGPT’s agent mode) and want to know why their tools behave the way they do. It also has the underlying details (wire formats, MCP server authoring, eval design) called out separately for anyone building one, but the technical detail is opt-in, not the default reading path.
The companion to What is an LLM, written for people who use Claude / ChatGPT / Cursor / Claude Code and want to know what’s going on behind the chat window. That post explained what the model is. This one explains what running one in production actually looks like, and it turns out almost every interesting decision is about the same thing: what tokens you put in front of the model on any given call. The model has no memory between calls. Whatever the model knows about your user, your codebase, your conversation, your tools, is in the prompt or it isn’t there at all. The post explains why Claude “remembers” your project (it doesn’t, the harness re-injects it), why ChatGPT degrades mid-conversation on long threads, why your $40 day on Claude Code happens, why “ignore previous instructions” still works on some agents in 2026. The technical detail (vector DBs, HNSW, chunking algorithms) is in sections you can skip if you’re not building one of these.