Variations on themes

Hullo! Finally, a weekend back home where I’m neither sick nor on the road. Doing chores has been… satisfying? And the weekend’s got a good mix of social time and shop time¹—truly, no complaints.

So, quite a week in world events. Tuesday and Friday evenings alike had an 8pm Eastern deadline—one existential on something approaching a global scale, the other existential for four good folks offering a beacon of hope to all us down here. Thankfully, both passed relatively uneventfully—but I’ve little doubt we’ll soon have new reason to worry, as the continued arbitrariness of world events buffets all us folks just trying to make something from this life.

Ah, but who has anything wise to say at a time like this? My turtling this week consisted of loads of reading, so let’s just get on to the links! As we sometimes do, loosely grouped with discernible but unnamed themes. (oooooh)

---

• Anil Dash’s title “Actually, people love to work hard” kinda says it all, but it’s worth reading the whole post for his repeated rearticulations of that concept as a management principle.
• What should be outsourced, and what shouldn’t? Dan Davies uses Palantir’s contract with the UK National Health Service to explore exactly that question. Three lines give you the gist: “Running a database isn’t a core competency for most companies. But managing the freaking company is a core competence, or at least it ought to be. And one of the big messages from management cybernetics is that the distinction between these two things is not as clear as you’d think.”
• Neat idea from Alex Usher, to think about tradeoffs through a menu of options, with examples given from higher education. When dealing with a budgeting situation, that menu should include both potential cuts (e.g., to wages, facilities, or services, like how often you clean your facilities) and revenues (e.g., the dollar value of a 5% tuition increase, a parking fee increase, etc), to provide a tangible sense of scale for financial discussions (and options to order à la carte, for those so inclined).

---

• “How I Came to Own A Cider Mill” is a delightful saga of taking a risk and seeing it through, with family and in community. (via Shawn Graham’s syllabus / resources for his recent fourth-year course “The Business of History”)
• Sticking to the open web means, for Alan Jacobs, little money in his pocket. Instead, they’re full of some other, indescribable, maybe better substance. (Definitely open it for the one-liner about the open web from Austin Kleon.)

---

Two sides of the same coin with everyone’s, err, favourite technology, LLMs:

• Cybersecurity implications abound: “vulnerability research is cooked”, Anthropic’s Claude Mythos model withheld because it’s too good at finding vulnerabilities—it all adds up to new ground for both the offensive and defensive sides in cybersecurity, with us, our poor devices and data, caught in the middle. (Bring back the airgap?)
• Meanwhile, it offers great promise for the mundane but essential work that underpins scholarly and journalistic projects. Dan Cohen catalogues a bunch of recent academic work that was enabled by one-off tools or data extraction built easily with the latest models. Derek Willis announces his revamped dataset of congressional press releases, with more robust scraping enabled by LLMs: “There are many, many problematic uses of Large Language Models out there. Writing scrapers is not one of them.”

---

• Love this idea of DIY-ing a printed and bound edition of favourite issues from a personal blog.
• Why many interface buttons can’t react right away.

---

• Giles Turnbull is, yet again, planting his year’s potatoes.
• How ancient wild cabbage became SO MANY DIFFERENT VEGETABLES.
• There’s a Mary Oliver documentary coming this year. GOODNESS YES PLEASE

---

All the best for the week ahead!

Lucas