FeedCity

A small Mechanical Habit today:

Commenting Guidelines This website has an active comments section, carefully maintained and curated by the author of this website. When commenting on this website, please keep the following points in mind: You may include HTML or Markdown in your co...

Managed agents (cloud-hosted agent harnesses) are powerful, but locking yourself into a frontier lab's platform now is risky - here's why and what to do instead.

“11 AI agents” is meaningless as a phrase.

If I said “I have 11 spreadsheets” or “I have 11 browser tabs” to do my work, it means about the same thing.

— Boris Mann

Tags: ai-agents, ai, agent-definitions

Today's links Billionaire solipsism, dictator solipsism, AI, and the fascist paradigm: AGI works best in a K-hole. Hey look at this: Delights to delectate. Object permanence: Woz's remotes; Furbeowulf; Vinge on AR; Oligarch buys FSU; DNC x GOP megadonors; John Key v Panama Pa...

a polycausal theory

An independent benchmark of the ecosyste.ms Python fund

Tool: CSP Allow-list Experiment

An experiment that shows that you can load an app in a CSP-protected sandboxed iframe (see previous note) and have a custom fetch() that intercepts CSP errors and passes them up to the parent window... which can then prompt the user to add that domain to an allow-list and then refresh the page.

I built this one with GPT-5.5 xhigh running in the Codex desktop app.

Tags: content-security-policy, iframes, security

Release: datasette 1.0a29 New TokenRestrictions.abbreviated(datasette) utility method for creating "_r" dictionaries. #2695 Table headers and column options are now visible even if a table contains zero rows. #2701 Fixed bug with display of column actions dialog ...

Release: llm 0.32a2

A bunch of useful stuff in this LLM alpha, but the most important detail is this one:

Most reasoning-capable OpenAI models now use the /v1/responses endpoint instead of /v1/chat/completions. This enables interleaved reasoning across tool calls for GPT-5 class models. #1435

This means you can now see the summarized reasoning tokens when you run prompts against an OpenAI model, displayed in a different color to standard error. Use the -R or --hide-reasoning flags if you don't want to see that.

Tags: llm, projects, openai, generative-ai, annotated-release-notes, ai, llms

This year Elon Musk has started banging the drum about building AI datacenters in space. As the only person who owns a successful space company and a (moderately) successful AI company, this is a sensible way to boost his profile and net worth. Is it a sensible way to build ...

In short: I just published Open Link in Unloaded Tab, a little Firefox extension that adds “Open Link in Unloaded Tab” to the right-click context menu.

In Firefox, you can unload tabs to save system resources. But there’s no way to open a new tab in the unloaded state…until now! I built a very simple extension that adds a new option to do this. (It even has a cute icon which I paid ~$15 for.)

I’ve built one-off extensions before, but this is the first one I’ve submitted to the Firefox Add-ons directory.

Download the extension here or check out the source code.

The way I used to design UI was to sit with Figma, or a React prototype with no back end. Imagine every failure mode. Design beautiful flows for each. I slipped into this mistake while working on LanWhisper, a voice dictation app. I designed for the failure modes I could im...

Now, if your CEO has never heard the phrase Ralph Loop, oh man, you are less than 30 days away from your next promotion. I'm not even exaggerating. Walk into his office, close the door, and say, hey chief, been experimenting with something. It's called Ralph Loops. And I th...

The thing about 90% of TDMs [Technical Decision Makers] is that they're motivated primarily by NOT GETTING FIRED. These aren't people who browser Lobsters or push to GH on the weekend. These are people that work 9 to 5, get paid, go home, and NEVER THINK ABOUT WORK AGAIN. So to achieve all that, they follow secular trends supported by analysts and broad public sentiment. Oh, Gartner said that "AI strategy" is most important? McKinsey said "context" needs to be managed? Well, "Context Engine for AI Apps" is going to be defensible. Buy it.

— Mitchell Hashimoto, in a conversation about the design of the Redis homepage

Tags: marketing, mitchell-hashimoto, redis

Artificial intelligence platforms may be just as susceptible to social engineering as human beings, but they are proving remarkably good at finding security vulnerabilities in human-made computer code. That reality is on full display this month with some of the more widely-used software makers -- including Apple, Google, Microsoft, Mozilla and Oracle -- fixing near record volumes of security bugs, and/or quickening the tempo of their patch releases.

How curl's disclosure policy filtered an AI scanner's findings at source

Here’s Scott Jenson in his insightful piece “The Ma of a New Machine”: the chatbot interface [makes us] feel like deep cognitive work is happening. But the interface is fundamentally reactive. It spits complex text at you, you skim it quickly, and you immediately type a rea...

Last year I said I'd probably never recommend another Bambu Lab printer again.

I still use my P1S, but after Bambu Lab started pushing their always-connected cloud solution as the new default:

• I blocked the printer from the Internet via my OPNsense Firewall
• I stopped updating the firmware
• I locked the printer into Developer mode
• I deleted Bambu Studio and started using OrcaSlicer

I had to do that to keep it under my control, instead of Bambu's.