Published
Weekend Reading — Eat plenty of fiber
a beautiful bitch "This is why i make sure i eat plenty of fiber"
Tech Stuff
kompressor macOS image convertor with a very interesting UI: drag & drop images into the wormhole, then toggle the kind of processing you want, and away we launch!
secretlint/secretlint Linting tool to prevent committing credentials to Git. Somewhat tricky to setup, but does include instructions for running it as git pre-commit hook, GH action, etc.
ProjectionLab Financial analysis and charting tool for your personal financials. There's a free plan to get you started.
Æ 🤦
Recently one of my professors apologized for her laggy computer "because I have so many PowerPoint presentations opened". This is a terrible indictment of modern commercial tech. No reasonable laptop made in the last decade should struggle with ANY humanly possible amount of basic slide decks. These university slide decks aren't any more complex than ones made in 1995 on a 66MHz Pentium 1 CPU with 16MB of RAM. The wastefulness on display here in modern software is staggering.
bigattichouse/waitlock A mutex lock for your bash scripts (or any other shell langauge, we're not judging).
Bench If you're looking for an executive assistant that you can use on a whim, Bench is pretty nice. This AI follows your instructions and can do a lot of useful tasks — search the web, generate a Word doc, etc. I got Bench to summarize news articles about AI and to pull a list of VCs, both tasks together for $10. Not free, not perfect, but quite affordable.
Programming Quotes The immutble law of source code:
Well-designed components are easy to replace. Eventually, they will be replaced by ones that are not so easy to replace.— Sustrik's Law
Comparing Claude Code vs OpenCode (and testing different models)
The best tool I’ve tried so far remains Claude Code, but I have to admit that OpenCode with sonnet-4 could replace it soon (especially considering it was released less than a month ago). Even gpt-4.1 wasn’t bad, and honestly, using it through OpenCode yields much better results than using it from within VSCode.
As for Google's Gemini Pro:
I honestly didn’t try a second iteration; I just scrapped the whole thing.
Moore’s Law states that every two years, we double the amount of time we waste on computers.
Marcel Waldvogel If you understand why Quantum Computing is just a theoretical way to increase shareholder value, this doesn't come as news to you, but I like how they made the point easy to comprehend with their research paper:
Earlier this week, Peter Gutmann and Stephan Neuhaus published a bombshell paper: The looked at all the reported factorization "breakthroughs" in quantum computing. And found that all of them essentially were magician's tricks, "sleight of hand". The two reconstructed the algorithms used on a 1981 home computer (a million times less powerful than what you hold in your hand now), an Abacus and with a dog.
Today we remember the three founding fathers of logic: George Boole, François d'Intègre and Bob Float
RateMyPrompt Discover and share the best AI prompts, as rated by AI and humans.
Eye for Design
I’m hiring a Product Designer who loves to write. When you have a storyteller that writes the job listing for a product designer — it's informative, delightful, and how come you're not applying for this job?
There have been many examples of people rallying around a "free and open" version of a service. They fail to realize that the end consumer barely cares. Look at voat (Reddit), app.net (Twitter), Diaspora (Facebook), even ycreject.com (Y Combinator) tried to be a thing for a while.
…
If someone is able to make it "free and open" while also making it a better experience than the alternative, then it'll be a big success. But so far everyone gets that wrong.
Called customer service, which directed me to the online chat, which directed me to the web form, which directed me to email, which directed me to CALL THE PHONE LINE AGAIN
Martin "Never forget what Jonathan Ive took from us"
Peoples
being too ambitious is a clever form of self-sabotage
Creation is not birth; it is murder. The murder of the impossible in service of the possible.
Jay Van Bavel, PhD Mind you this statement is from a thought leaderspeech post on LinkedIn:
In a series of experiments, we recently paid people a few dollars to unfollow the most divisive accounts on X. After a month, they felt 23% less animosity towards other political groups.
Their experience was so positive that half declined to refollow those hostile accounts after the study and they felt less animosity 11 months later.
Business Side
The Pulse: Section 174 is reversed! Mostly, that is Section 174 is reversed with one notable exception:
I expect US companies to hire more in the US, and less outside of it. The updated Section 174 very clearly incentivises doing so. If you’re in the US: this is great news! If you’re outside, prepare for US-based companies to be incentivized to make cuts abroad, and to hire less outside of the US.
HN Slop Uses AI to generate startup ideas from the Hacker News front page.
Machine Intelligence
Max Leibman Show I'd watch:
America’s Next Top Large Language Model.
Essential Reading for Agentic Engineers Everything you wanted to know about AI agents but were too afraid to ask your LLM.
Practical notes on getting LLMs to generate new ideas Does it work? No. But is it an interesting concept? Yes. I think day dreaming will happen in the near future:
It’s tricky to get models to generate random non-new ideas. Gwern’s day-dreaming loop requires getting the model to come up with the ideas that it then combines. But I found this surprisingly hard. A single “tell me a concrete fact” prompt gets you a small set of facts. I couldn’t find a prompt that would reliably pull a random piece of knowledge from the model (in other words, a prompt that could be run over and over in the loop to generate new ideas from). I ended up manually prompting Claude Opus to “tell me twenty facts about [discipline]” a bunch of times and then stitching them together in a yml file.
Vast Numbers of Lonely Kids Are Using AI as Substitute Friends
Of the 1,000 children aged nine to 17 that Internet Matters surveyed for its "Me, Myself, and AI" report, some 67 percent said they use AI chatbots regularly. Of that group, 35 percent, or more than a third, said that talking to AI "feels like talking to a friend.
Perhaps most alarming:
12 percent said they do so because they don't have anyone else to speak to.
More evidence that generative AI is most useful for tasks involving generation or analysis of formulaic content. According to LinkedIn's latest Workforce Confidence survey, nearly 4 in 10 U.S. employees (38%) say AI helps them complete tasks faster and boosts their productivity. That percentage is even higher for marketing and HR roles, where 58% of professionals agree AI is already helping them on the job.
Insecurity
Everything Else
London library forced to briefly close after fox 'made itself comfortable' inside 😮
Then they came for the sarcastic people and I was like oh great that's exactly what we need right now well done to all concerned.
'English is what happens when Vikings learn Latin and use it to shout at Germans, and then the French shout back.'
teledyn "On the other hand, humans can complicate anything…"
It’s always OK to lie on the floor doing nothing for a long time
Not a fan of Waymo, but they are one of the few cars that fully stop for every stop sign in my neighborhood.
Etsy Witches The hustle is real.
I bought a second hand time machine next Sunday. They don't make them like they're going to anymore.
Before you marry a person, you should first make them use a computer with a slow Internet connection to see who they really are.
This gigantic AirPods backpack actually holds and organizes your items
if a cafe has a no laptop policy i think they should also be required to loan you a friend skilled in the art of conversation
A technique I've been using a lately for non-fiction, including work stuff, is write a really rough first draft and put it on the left side of my screen--window managers will do that--and then have a fresh document on the right side and essentially summarize my own draft.Instead of tinkering with crap prose, I write fresh new prose, and the result contains all the same information but is way clearer and shorter. It's great.
Want to Live Like a Hobbit? This $775K SoCal Home Is Straight Out of Middle Earth
Learn to love the Moat of Low Status
The Moat is effective because it’s easy to imagine the embarrassment that comes from being in it. It’s so vivid, it looms so large that we forget the novel upsides that come from transcending it. Easy to imagine the embarrassment from your first months of singing lessons, because you’ve faced embarrassment before. Harder to imagine what you’ll sound like as a trained singer, because that’s never happened to you before.
'Someone's always watching': Gen Z is spying on each other
“These location services, the ability to use your cellphone to locate something, have shifted from being informational, in other words, where is that ice cream shop I’m interested in, into being social,” Albright said. “In other words, where are you now?”
What to know about mystery benches popping up in San Francisco
The bench collective has already installed dozens of benches across the East Bay. The city of Richmond even passed legislation to legally permit their DIY installation.
Are a few people ruining the internet for the rest of us? The annoying minority:
Similar patterns can be observed across the internet. Only a small percentage of users engage in truly toxic behaviour, but they’re responsible for a disproportionate share of hostile or misleading content on nearly every platform, from Facebook to Reddit. Most people aren’t posting, arguing, or fuelling the outrage machine. But because the super-users are so active and visible, they dominate our collective impression of the internet.
It took 45 years, but spreadsheet legend Mitch Kapor finally got his MIT degree Better late than never.
Max Leibman "The U.S.S. Toilet Flapper, laughingstock of Starfleet."
