Good news, for once.
Marcus on AI
• Gary Marcus
Breaking: Demis Hassabis endorses preflight safety testing for AI
Marcus on AI
• Gary Marcus
Good news, for once.
The Old New Thing
• Raymond Chen
Please, tell me more.
The post Microspeak: Double-click and drill down appeared first on The Old New Thing.
Terence Eden’s Blog
• @edent
My wife and I recently went on a 7 week holiday around Europe. Although we each took a massive backpack, we wanted to travel fairly lightly. I took a single universal power brick. This little unit was all I needed to charge my various gadgets. It has a hefty USB-C PD (Power Delivery) port for rapid charging of my phone and laptop. Two other USB-C ports for my other gadgets. And a couple of…
Pluralistic: Daily links from Cory Doctorow
• Cory Doctorow
The Silicon Underground
• Dave Farquhar
On July 15, 1983, the Famicom, or Family Computer, launched in Japan. Despite the name, the Family Computer was a game console, and it went on to shatter the Atari 2600’s record for the most sales worldwide by a game
The post Nintendo Famicom and the secret of Nintendo’s success appeared first on The Silicon Underground.
Entropic Thoughts
• kqr
Imagine every person in a country is assigned some sort of score, and a higher score is better. I can’t share the specifics around the score I’m investigating right now, but on the population level it is collected by a government agency. Here’s the distribution of these scores for the population in my country.
TIL: Using uvx in GitHub Actions in a cache-friendly way
I finally found a cache-friendly recipe for using uvx tool-name in GitHub Actions workflows that I like.
The trick is setting a UV_EXCLUDE_NEWER: "2026-07-12" environment variable at the start of the workflow and then using that as part of the GitHub Actions cache key. This means any uvx tool-name commands will resolve to the most recent version as-of that date, and you can bust the cache and upgrade the tools by bumping the date in the future.
My goal here is to use Python tools in GitHub Actions without every run of the workflow hitting PyPI to download a fresh copy of the tool and its dependencies.
Tags: packaging, pypi, python, github-actions, uv
TL:DR; Pebble Time 2 Shipping Status Pebble Software Updates PT2 - Issues You’ve Reported Pebble Round 2 Production Update Index 01…
datasette code-frequency chart on GitHub
Out of curiosity I decided to see if I could find a useful illustration of the impact of coding agents and Opus 4.5 class models on my own output. The best I've found so far is this GitHub chart of frequency of code changes to my Datasette open source project:
The big spike in activity at the end aligns with Opus 4.8, GPT-5.5, Fable 5 and GPT-5.6 Sol.
Tags: github, ai, datasette, generative-ai, llms, ai-assisted-programming, coding-agents
Martin Alderson
• Martin Alderson
Part two: as AI inference margins collapse, who actually captures the value? The hardware supply chain and consumers win, model inference commoditises, and the frontier labs' escape routes are managed agents and staying ahead.
micahflee
• Micah Lee
Guess what? Lately I've been nerding out on creative writing. I even joined a local fiction writers group where we meet up at coffee shops. Like, in person!
Anyway, I submitted a short story to the DEF CON 34 Creative Writing Short Story Contest. My story is called
Krebs on Security
• BrianKrebs
The Cybersecurity and Infrastructure Security Agency (CISA) has issued a postmortem on a data leak in which a contractor published dozens of internal CISA credentials -- including AWS Govcloud keys -- in a public GitHub repository for almost six months before being notified by KrebsOnSecurity. Experts say the gaps identified in the agency's initial response provide important lessons that all security teams should absorb.
The Old New Thing
• Raymond Chen
Guard pages all the way down?
The post Why don’t we just make the entire stack out of guard pages? appeared first on The Old New Thing.
Terence Eden’s Blog
• @edent
Shhh! RSS Club posts are only available to feed subscribers. Keep the secret! I'm not a big fan of the "Quantified Self" movement, but since buying a ridiculously cheap smartwatch, I've been interested to see just how many steps I take per day. During our recent Interrail holiday, we travelled about 6,400 Km by train over 49 days. But how much did we cover by foot? My watch pairs with the Open …