It's difficult to navigate what your engineering strategy should be when all everyone is ever talking about is AI, so this is something I put together to help guide where I see value in making bets and buying things, and where I don't - to help teams work out what they should focus on when everyone is telling them they can be automated.

https://davidwhitney.co.uk/blog/2026/06/08/ai_engineering_baseline/

Ruminating on why I find LinkedIn such pure brainrot, and it's because it's the tone and audience of Facebook (done corpo) but the content quality of X (pejorative).

Somehow the worst of all worlds.

Nobody is themselves.

Lads, I think I'm looking at the worst codebase I've seen in my entire career as a programmer right now.

It's like performance art.

Build this year is full of lots of useful boring work that'll push local models into mainstream computing.

The teams have to be really careful with positioning after they overzealously ruined the Copilot brand, but it's all really good foundational work.

WSL Containers is finally Microsoft getting containers right on Windows. Apple slightly beat them to the punch but between those two implementations containers are actually ergonomic now.

Finally.

If you're at DevSum this week I have an extremely limited set of t-shirts to give away. They were samples that one day I'll get around to putting out properly.

In red and black. First come, first served.