No notes. Generalists, those who take an interest in many art forms, are stuck. They want more meat on the bone than a movie made for the 50th percentile gives and aren’t so steeped in cinephilia as to appreciate the 95th percentile Oscar nominees.
The World Economic Forum proudly proclaimed in 2023 that robots could automate 40% of domestic chores within a decade, “freeing up time — particularly women’s — for other things”. The future the Jetsons promised may finally be upon us. Goldman Sachs echoes the sentiment for office work: “AI could automate 25% of all work hours”. Follow enough headlines of late and you’d conclude we’re on the precipice of a new golden age - robots fetching cold beer from the fridge while the AI adds the replacement six pack to the grocery list it just made. We have seen this mirage before.
I was reading an essay (Auren Hoffman) about fertility and AI, because everything these days is about AI (except this essay), when one of the statistics stopped me cold.
College-educated women have an almost eight-in-ten chance of their marriage lasting 20+ years. for women with a high school education or less, the share is only 40%.
This has potential to be a driving theme of the current LLM-driven revolution. It’s not that the craft of writing will cease to exist because of AI. It’s the nature and process of writing that has changed. You spend less time typing at a keyboard. I imagine this has similar parallels to when people went from typewriters to word processors. You spend more time on the craft of communicating then on the frustrations of pounding out perfect strokes on a typewriter.
Are EMs actually the ones poised to benefit the most from agentic coding? This article makes the case they have all the right skills for the coming AI age after honing their craft managing teams of engineers. Exchange engineers for Claude, Claude, and Claude and it turns out their job is the same.
Coming off a few years of flipping between tech lead and management, I have to agree. Trusting agents with coding is no big deal after you’ve had experience handing the reigns over to your team. What’s more important is the process they (agents or humans) follow. Force agents to jump through the same engineering design hoops (design, plan, implement, review) and the results are impressively good.
This article feels like its just laid the groundwork for justifying many of the agentic apps being built. As someone building an app hinging on using “messy text” as the primary input I find this research encouraging.
This is a handy tool, makes it super easy to share how I leverage Claude to produce improved designs, implementation plans, code reviews, etc.
I’m severely tempted to put Windows in the rearview mirror and switch to Linux too. Unfortunately, my Windows install exists purely for playing the one type of game Steam’s Proton can’t run: MMO FPSes like Battlefield.
My Hugo blog is barely a month old, but I already wondered whether anyone besides bots ever stops by. I didn’t want heavyweight or cookie tracking pop ups, so I spun up Ackee —open‑source, cookie‑free, and small enough to ignore until I need the numbers.
Ackee went up on the same box that serves the blog—Node, Mongo, nothing exotic.
So What?
Because life is short and repetitive tasks are dull, here’s a straightforward setup to automatically update my Hugo-based website every time I push changes to Gitea. No glamor, just practical efficiency.
I’ve been using Hugo for this site—it’s simple, fast, and gets out of the way. Gitea hosts my Git repos because it does exactly what’s needed without fuss. Combining these two tools to automate deployments just made sense.