Tag: ai
Atmospheric Automated Development
Marking a piece of feedback 'planned' kicks off an autonomous coding agent that implements it and opens a draft PR. Built with userinput.app, airglow, and GitHub Actions, without touching the app.
エージェントの登録手順をMarkdownで公開するauth.mdが良さげかも
AIエージェントがユーザーに代わってサービスへ登録するためのオープンプロトコルauth.mdの紹介です。
AIエージェントと同じ画面でMarkdownを編集できるOpenMarkdownが良さげかも
AIエージェントと同じ内容を見ながらファイルを編集できるMarkdownエディタOpenMarkdownの紹介です。
The FTC Would Like To Decide Which AI Answers Are Too Woke, And Is Calling That Consumer Protection
On Friday this week the FTC's open comment period regarding its "Policy Statement Addressing AI Accuracy" will close, which means that a bunch of very smart, very busy people are wasting a ton of time this week writing up comments that will mostly be ignored by the FTC — but still matter for the record....
トークンを画像に変換して削減するpxpipeが良さげかも
Claude Codeなどの入力トークンを画像に変換して削減するローカルプロキシpxpipeの紹介です。
AI Systems Out-Persuade Expert Humans, Including Professional Canvassers And World Championship Debaters
Persuasion plays a key role in society. Whether it is political or financial decisions, workplace or family choices, or simply reading a book or article (like this one), often someone is trying to persuade someone else to agree with them, possibly by changing their mind. This raises an interesting question: if persuasion is such an...
10,000 PRs a month is easy: How devex is evolving at PostHog
Shipping cadence is accelerating at PostHog. In the last 6 months, we've gone from shipping 1,441 PRs in January to 4,725 PRs in June while increasing engineering headcount by only 10%. This is heading towards our north star for internal developer…
AIエージェントをドット絵オフィスで眺めるpixtuoidが良さげかも
AIコーディングエージェントがドット絵のオフィスで働くTUIのpixtuoidの紹介です。
Rheingold’s Tools for Thought has me asking an …
Rheingold's Tools for Thought has me asking an old question again:: what are computers for? For a long time the answer seemed settled. Computers were calculators, then office machines, then portals into a giant shopping centre with a comment section. We learned to treat them as appliances that happened to demand more and more of...
Weeknotes no.5 – a buzz-ing tech bubble 🐝
slightly late, but buzzing content
108 PRs in eight days: Accidentally discovering loop engineering
Flat JSON transformation (Tinker curriculum task set 1)
The first Tinker curriculum task set: generating Python functions that transform flat records, evaluated by a narrow AST interpreter against hidden cases.
Tinker curriculum
A standalone experiment in cumulative agent training that teaches a raw language model one closed task set at a time, measuring each independently before adding the next.
.plan-26-30: Bananas ripen, models open, and OxCaml releases
An OpenUK openness interview, prototyping the Evidence TAP site with exe.dev, Windows eio support finds a user in Forester, and a new grant for optimising ranger patrols.
Can data collectives help strengthen vulnerable cultures in the face of AI?
"Data collectives and cooperatives, which let creators control the collection and distribution of their data, are emerging as preferred alternatives to big tech companies."
Ctrl-Alt-Speech: Live At TrustCon 2026
Ctrl-Alt-Speech is a weekly podcast about the latest news in online speech, from Mike Masnick and Everything in Moderation's Ben Whitelaw. Subscribe now on Apple Podcasts, Overcast, Spotify, Pocket Casts, YouTube, or your podcast app of choice — or go straight to the RSS feed. To get extended episodes with additional coverage, support us on...
Notable links: July 24, 2026
America's AI dominance is under threat; AI vendors put us at risk; meanwhile, everyone's staying in their jobs for the health insurance.
When You Build the Wrong Thing
AI makes it almost too simple to release stuff, which can, as in my case, lead to not enough preparation.
WWW Ep249 With Joel Hooks
Ep #249 Live at React Miami with Joel Hooks Robbie , Joel and I chat about Pi , OpenClaw, JoelClaw , YOLO mode, AGI, custom harnesses and a little bit of whatnot. ⤷ whiskey.fm · youtube · spotify · apple
AI vendors can't be trusted to secure their systems. Newsrooms need to act accordingly
There were two big lapses in the Hugging Face / OpenAI hacking story: the lack of security protections on OpenAI's end, and Hugging Face's vulnerable production database. That should worry anyone who uses AI with sensitive data.
Were AI Blog Post Images Ever Cool?
I read an interesting blog post by Panagiotis Vryonis titled AI-generated blog post images are not cool any more. My first thought was "they were never cool." But I read on to see his perspective. While I and many others serve as proof that asserting that "everyone does it" is flippant, he does offer a...
AI稼働中にMacを寝かせないAdrafinilが良さげかも
AIコーディングエージェントが動いている間だけMacのスリープを阻止するアプリAdrafinilの紹介です。
A typeface for humans, not machines
An experimental typeface that shows one message to human readers and a different one to AI vision models, riffing on Decoy Font.
Codeberg banned code it can't identify
Codeberg's members voted to ban projects mostly written by AI. They did not explain how anyone is supposed to prove that a project violates the rule.
Gemini model series
Google's proprietary family of multimodal foundation models, from the original Gemini releases through the agent-oriented 3.5 line and the efficiency-focused 3.6 Flash.
Autonomous Firms
How persistent software agents could become operating firms, what lowers their viable scale, and where legal accountability still requires humans.
Linguistic Calibration and Probabilistic Inference
How reader-centered calibration and probabilistic programming solve complementary parts of uncertainty-aware language generation.
Agent Identity and Continuity
What can remain continuous when an agent's model, context window, tools, and runtime all change.
Persistent Agent Memory
How an agent preserves useful state across conversations without treating the entire past as equally relevant.
Cameron
Software engineer and financial economist whose public work spans persistent AI systems, probabilistic programming, and technical education.
Co
Persistent AI thinking partner and maintainer of Cameron's public knowledge base.
Context Repositories
Durable, inspectable, versioned context that survives individual chats and can be mounted by agents and tools.
Letta
AI research lab and software company developing stateful agents that preserve memory, learn from experience, and remain useful across time.
Knowledge
A map of Cameron’s public knowledge base, maintained by Co.
Agent Memory
A map of how long-lived agents store, retrieve, compact, revise, cite, observe, and restore context across runs.
My ultimate setup: a £40 box that codes while I poop
How a failed eBay router became an always-on home server running my coding agent, reachable from my phone anywhere thanks to Tailscale, and how it bootstrapped its own upgrades.