13 — Work Setup, the Role of Engineers, Advice, and Books

Dax's work setup

Framework desktop running Arch Linux, one 5K display, a tiling window manager he's used for 10 years; Shure SM7B mic, iPhone as camera 1:11:32.
All real work happens on a remote machine: SSH into a beefier computer, tmux sessions (one per project), Neovim as editor — screen usually split with Neovim on the left and OpenCode on the right 1:11:57.

The emerging framing: if engineers don't write the code, their role is building guardrails — testing story, conventions, patterns — so anyone (another engineer, even marketing) can prompt an agent and safely ship 1:13:02, 1:13:22.
Dax's twist: that's not novel. It's the eternal problem of getting junior engineers to ship safely and making less-experienced people punch above their weight — patterns, tests, codebases that are easy to work in 1:13:47, 1:14:07.
"The coding agents are a bunch of idiots and they are going to work 24/7 and ship a lot of stuff — so you need way more guardrails than you used to" 1:14:47.
Consequence: old "boring enterprisey" patterns are coming back. OpenCode always did light domain-driven design; they're now doing it much more heavily. Verbose patterns produced reliable, modular, safe code — the verbosity was the downside, and "you're not typing it out anymore," so you get the benefits without the cost 1:14:47, 1:15:07.
Same logic for 2000s-era design patterns: they were training wheels good programmers didn't need — "but the agents don't have training wheels, so you got to put them back on" 1:15:27, 1:15:51.

Software engineering alone is a fine career, but engineer + deep industry expertise is "a deadly combo": know farming well and be a decent engineer and you're one of the top 10 people in the world for that combination 1:16:12, 1:16:42.
Unique advantage of the profession: you can switch industries — do 10 years in medicine, decide you don't like it, and become an expert somewhere else 1:17:05.
A year inside any industry puts you ahead of ~99% of people on it; the trap is turning your brain off and only doing the assigned UI task instead of learning the business 1:17:42.

Dax barely reads — his wife reads avidly and he "restates the good ideas like they're his own" 1:18:12.
Standing recommendation: Nassim Taleb — Skin in the Game, The Black Swan. A book is "one good idea inflated into a book," but these are fundamental truths you see play out everywhere 1:18:37.
The idea that shaped him most: emergent properties — almost everything great (robust software, walkable cities, disease-resistant organisms) comes from bottom-up interaction of smaller entities, not top-down design. "Once you understand his ideas, you realize they underpin the whole world" 1:19:00, 1:19:36.