In terms of knowledge sharing and gathering hard-won human context I agree, sort of. An LM review can at least prompt some reasonable changes, catch performance issues, etc.
Thanks for putting into words what I have been seeing a lot at work and haven't been able to put my finger on. We tend to have quite diverse _workflows_ between devs at my company, and success seems to correlate with injecting better context earlier in the process.
I like to chat with Claude about how to approach a given problem, bring in extra context, etc, before even really drafting up a plan, while other people dive into implementation immediately and go on wild goose chases.
90% of the time we end up in the same place in roughly the same amount of time, and there are obviously tradeoffs to spending more time planning vs implementing. I'm oversimplifying as well.
A simple one - a small MacOS status bar widget that shows me if my Ethernet adapter is the default connection or WiFi. It also shows the reported link speed. I have a _very_ flaky USB/Ethernet adapter so seeing "Eth 10" in the bar is enough to prompt me to unplug/replug and get back to "Eth 100" (yes, it isn't gigabit).
This seems somewhat sensible to me - the genie _is_ out of the bottle, and students absolutely will use AI agents to finish assignments without learning a thing, but there is some value to showing how agents can be used as teaching tools and what healthy use _can_ look like
AI has made my experience as a fully remote worker in a not-fully-remote company worse. It is adding a layer of indirection between myself and the juniors I'm mentoring, or the product manager I'm working on a feature with, and many other mundane facets of work. The opportunity to passively pick up on my coworkers' idiosyncrasies is now gone and all I can do is guess at what prompt they might have used.
I'm trying to have more face-to-face calls, and to talk to people without a bot involved, but can be difficult, frustrating, and not "productive" either.
Man, I really miss Terry Pratchett too. He has been my favourite author for as long as I can remember (maybe Roald Dahl before that?). It helped having such a volume of work to go through at the time in my life where I was reading the most. I swear he has the most re-readable books too; so many small details and jokes that would be missed on a first pass.
As someone whose work enforced a switch from Cursor to Claude Code, I do keep on top of the code by pairing it with an IDE, tracking/viewing changes etc. There's no real obstacle to using an IDE as you normally would, with Claude Code as a sidecar.
I don't know exactly how to measure bang for buck, but my Sony XM4s have been holding up well, sound good, are decently comfortable for a day's work, the battery life is good, etc.
They do have some annoyances like not always sleeping correctly when left connected to my laptop, but overall they are easy to recommend
I'm building a very casual daily price-guessing game for my mum. Every day she gives me feedback, and I'm using it as a chance to de-rust my CSS/React + see how daily games tick.
Peter was quite vocal on twitter about _only_ using Codex to develop OpenClaw, but Claude is what a majority of people were (are?) using to run the tool itself.