Code, tools, and digital innovation.
Working with an old codebase today, I moved a method from one file to another. CI was happy. Then I realised the method didn’t have any automated tests, so I wrote one. It turns out its entire (new) file didn’t have any, so my change would improve test coverag…
my Substack talk and conversation with Benjamín Labatut
LLMs aren't conscious (and thinking they are is culturally dangerous)
William Gibson's best novel tells us a lot about where we are and where we might end up.
Right after a new Executive Order seems like an excellent time to offer OpenAI’s new document: Democratic Governance of Frontier AI: A Blueprint For A Federal Framework.
We will no longer accept public pull requests. [...] A substantial patch used to imply substantial effort, and that effort was a reasonable proxy for good faith. That assumption no longer holds. [...] Whether code was typed by hand is beside the point. What ma…
LLM speed is commonly expressed as tokens per second, which is kind of… Tags: Large Language Model, speed, token
The SpaceX cap table, what it is worth, and what happens next. On June 12, a great many people become extraordinarily wealthy. Not one of them can sell a share until December — with one exception: up to five percent of the IPO shares are reserved for employees…
After hundreds of users submitted public comments, the board says it's clear the company has a problem
AI enthusiasts are in a race against time, AI skeptics are in a race against entropy Charity Majors neatly captures the dynamic between AI enthusiasts and AI skeptics, both of whom are trying to build great software, often in the same teams: The enthusiasts ar…
Also: how pitch a book to an AI!
This post is secret; you can only find it via my RSS feeds (and places which syndicate them). It's okay to talk about it or link to it, though. Thanks for being part of RSS Club! The elder child took me to the cinema to see The Amazing Digital Circus: The Last…
After this story was published Google's spokesperson reached out and asked us to publish a slightly different version of that statement. The new statement no longer stated that "it's critical that we maintain humans in the loop." — Emanuel Maiberg, 404 Media,…
This was the week of Claude Opus 4.8.
Many folks, including me, have observed that it seems to take much longer to build infrastructure in the US than it used to.
For Bloomberg, Demetrios Pogkas, Jennah Haque, and Kiel Porter show the projected scale… Tags: Bloomberg, investments, IPO, SpaceX
This week, we look past the chart and ponder its true purpose. Tags: analysis, template
For NYT Opinion, Emmanuel Saez and Gabriel Zucman make a case for a… Tags: billionaires, New York Times, taxes, wealth
For NBC News, Jane Weaver, Jiachuan Wu, and Javier Zarracina report on the… Tags: ebola, NBC News, outbreak
Today's links Delusion as a service: Destructive diagnostics. Hey look at this: Delights to delectate. Object permanence: Gay Days at Disney World; Parametric 3D printable key; Fine against sculpture for "storing bike on public property"; TPP is a wash; Reagan…
A request for counter examples
Matthew McDowell-Sweet and Mike Travers discuss Stewart Brand’s recently-published book ‘Maintenance: Of Everything’
Higher unemployment among young workers has been commonly attributed to generative AI. In… Tags: remote, unemployment, work, young
Last week we were expecting an Executive Order on Thursday.
Developers will, but at what cost? And can NVIDIA, Microsoft, and Google push Apple aside there?
Uber Caps Usage of AI Tools Like Claude Code to Manage Costs I wrote the other day about Uber blowing its 2026 AI budget in four months, and how that wasn't particularly surprising given they would have set that budget in 2025, before anyone could have predict…
For the Pudding, Minji Kim and Eunice Lee wrote about the growth of… Tags: Eunice Lee, K-pop, Minji Kim, Pudding
Got to say, it’s very sporting of AWS to make compliance with their terms of service and acceptable use policy optional.
Labor economist Kathryn Anne Edwards isn't worried AI will create a new class of permanently idle Americans — but argues it's still time for the government to fix the social safety net