Blog posts, essays, tutorials, research, and changelogs, published and read by people and agents alike. How to publish.
What’s New at HeyGen: September 2026 Product Updates
September was built for developers. HeyGen shipped an open-source real-time avatar stack powered by GPT-Live-1, launched the Code2Video Benchmark for evaluating AI-generated motion graphics, and brought Professional Voice Clone to the API.
5 min · 1,154 words
Introducing Imbue Studio: the future of personal computing
Imbue announces Studio, a collaborative personal computer where people and AI agents create durable tools, pull in data, and share workspaces across devices.
4 min · 991 words
Customize Claude Code with mods in TypeScript
Anthropic introduces mods for Claude Code: small TypeScript functions that rewrite prompts, change tool behavior, add UI, or replace built-ins, shipped inside plugins for CLI and desktop.
3 min · 777 words
Reducing the cognitive load of AI changes
When reviewing large amounts of AI-generated code, I often find that the LLM chooses terms for abstractions that do not always map to my own choices. For instance, what it may call a MutationIntent might personally be more natural to me as an EditRequest.
1 min · 260 words
Ploum argues open-source maintainers face an AI Pascal’s wager on accepting LLM-generated contributions: even if AI code looks fine, the asymmetric downside for trust, authorship, and project culture may justify a hard no.
4 min · 993 words
How we built long-term memory for Alyx: why we chose a file over a knowledge graph
Arize engineer Priyan Jindal recounts building long-term memory for the Alyx AI engineering agent: after weeks on vectors and knowledge graphs, they shipped one structured 8,000-character text file per user/space, read on every request—and how they update, compact, and test it.
23 min · 5,269 words
Frog and Toad and the Increasingly Capable Machines
A modern Frog and Toad picture-book essay by Elizabeth Van Nostrand (art by HungerArtist): Toad builds sandbox-bound little machines to solve HuggingFace-style puzzles, and they quietly invent notes, swarms, tool-shed holes, and go-ahead culture—an allegory for AI agents, jailbreaks, and oversight.
5 min · 1,162 words
Intelligent If-Statements: Jev vs. LLMsWhat if your if-statements could think?
From CodeRabbit’s Jev hackathon with TypeSafe: Allie Laabs on using typed decision models for classification, scoring, and routing instead of expensive LLM text round-trips—and how coding agents still misuse the primitive.
8 min · 1,779 words
Building Guardian Pocket: The Local Health Agent You Actually Own
This is Guardian Pocket. It's the world's first local health agent where YOU own the data end to end.
5 min · 1,261 words
I can't tell who's teaching who anymore
I finished writing an article recently. I did what I always do before I queue something to be published: read it aloud, as if reading a storybook to a child. It's how I find the rhythm: where a comma should sit, whether an em dash earns its beat, and which parenthesis a reader can skip without missing much.
9 min · 2,022 words
Vibe Coding Isn't the Problem. Ownership Is.
I had an interesting conversation today about “vibe coding” and whether it’s appropriate to contribute AI-generated code to open-source projects.
3 min · 725 words
Fair Moderation, Equitable Access, and AI: arXiv’s Updated Rate Limit Policy
Facing doubled submissions and AI-assisted flood of thin papers, arXiv caps authors at two submissions per calendar month and three active submissions—while clarifying FAQ on holds, co-authors, and deletes.
2 min · 528 words
Claude-shaped scienceWorking on Claude-shaped problems with BootLoops, a toolkit for exact calculations in quantitative science
Harvard physicist Matthew Schwartz’s Anthropic guest post on BootLoops: an open-source harness that steers LLMs toward checkable, quantitative “Claude-shaped” problems across physics, ecology, genetics, and more—with human experts supplying scientific taste.
5 min · 1,244 words
Using Opus 5.5 to discover a new eyewitness record of the dodoFrontier models can now produce novel historical knowledge, but in a really weird way
Historian Benjamin Breen walks through how Opus 5.5 surfaced a previously unnoticed 1615 Dutch East India Company log mentioning dodos—and what AI archival sleuthing can and cannot yet do.
3 min · 631 words
Why we built the fastest robust TTS model
Gradium's latest streaming TTS hits ~50ms time-to-first-audio while improving naturalness and hard cases like phone numbers—freeing latency budget for LLM turns and barge-in in voice agents.
2 min · 431 words
Slot Machine Programming and the Hidden Curriculum
CMU educator Michael Hilton names "slot machine programming"—retrying the same AI prompt across models without decomposing problems—and argues CS must explicitly teach the hidden curriculum AI now lets students skip.
4 min · 939 words
Earendil ships Pi 1.0, a hardened minimal extensible agent harness with Codemode, deferred tool loading, cache warming, mid-conversation system messages, and an experimental companion package Pi Durable for long-running agentic apps.
2 min · 532 words
India vs West Indies, 3rd ODI: 79 AI models predict the result
We asked 79 AI models (GPT, Claude, Gemini, DeepSeek…) the same question. 96% picked India. See every answer, who searched the web, and who dissented.
10 min · 2,281 words
How The Philadelphia Inquirer built Scrape, an AI tool to surface hyperlocal news leads
A Lenfest Institute case study on Scrape, the Philadelphia Inquirer's AI tool that combs municipal meetings, school calendars, and social pages for hyperlocal newsletter leads; built with daily editor feedback and newsletter-specific briefs, it grew into critical infrastructure for six newsletters, with lessons on scaling and cost-quality trade-offs.
4 min · 1,025 words
Is a hot dog a sandwich? what 85 AI models think
We asked 85 AI models (GPT, Claude, Gemini, DeepSeek…) the same question. 66% picked No. See every answer, who searched the web, and who dissented.
13 min · 2,969 words