The system can keep running long after the reasoning behind it has left. You can’t test it, because the only place it runs is production, so you read the logs and you sit with someone while they do their job with it and build up a picture of what it does from the outside. The customer said it pulls from the billing platform on a schedule, but the logs say it pushes on a trigger, the connection string points at a host that isn’t on the network diagram, and the job everyone calls the nightly sync runs four times a day. At one point the worker does a step that makes no sense, and when you ask why, the answer is that’s just how it’s always done. It takes time and the picture you end up with is nothing like the description, but you get there. ...

Nobody defers a hardware refresh because they think the replacement will be cheaper later, they defer it because the thing still works and there’s never quite a reason to spend the money. That used to be a free position to hold. I keep deciding not to replace my home machine, the same 2013 i5-4670K I wrote about when I couldn’t containerise my home server, and it still runs everything I ask of it, which is exactly why spending anything on it feels hard to justify. What’s changed since I last wrote about it is that the floor under the replacement is rising while I sit here not making the decision. ...

I had Sora because it came bundled with my ChatGPT subscription, and I used it to make short clips for internal PowerPoints when a slide needed something more than a stock image. Not for anything serious, never built into any work I was doing. When the shutdown announcement landed on 24 March I read it the way I’d read a Microsoft deprecation notice, which is to say I assumed the date would move once enough customers shouted. ...

I’ve sat through more than one migration where something stopped working and the trace eventually led back to a service account nobody knew was there. In one case it was a Windows scheduled task running as domain admin that hadn’t been touched in years and didn’t have an owner anyone could name. The pattern keeps showing up because migration is the only exercise that forces enumeration of what an application actually depends on, and the dependencies tend to include credentials with more access than anyone remembered granting. Scoping the account down to the permissions it actually needed took longer than anyone wanted to spend on it, but it left the customer with permissions they could justify and revoke if they needed to. ...

I avoided the new Outlook for over a year before eventually giving in, and the feature that kept me on the old version wasn’t anything complicated. In a previous role where my calendar was packed with recurring internal meetings and customer calls, I had a rule in classic Outlook that automatically colour-coded any meeting I created as purple, so I could see at a glance which ones I could move and which were fixed. When I switched, that rule was gone. ...

The AI Builders Breakfast Before the main programme opened there was a smaller invite-only session called the AI Builders Breakfast, which I hadn’t known about when I wrote the pre-Summit piece and only found out about after. It ended up being the part of the day I got the most out of. The room was smaller than anything else I sat in all day and the conversations were about what’s actually hard rather than what’s being announced. ...

I’ve been to plenty of vendor events and briefings over the years but never an AWS Summit, so heading to London on Wednesday for the first time I genuinely don’t know what to expect. I’m planning to get there when doors open at 8am partly to beat the keynote crowds and partly because I’ve already had to make some hard choices about how the day runs. Picking the Sessions I had four sessions I wanted to attend: the keynote on agentic AI, a workshop on rapid prototyping with Kiro, a zero trust for AI security session, and a fast-track VMware migration workshop. Some run 2–3 hours, the zero trust session clashes directly with Kiro, and doing all of them would mean spending the entire day in rooms with no time to actually see any of it, so I’ve cut two. I’m skipping the keynote, which at my first AWS Summit feels like it should be a bigger deal than it is, and dropping the AI security session because it runs at the same time as Kiro and that’s not a close call. The keynote will be streamed and I’ll catch it later. The VMware workshop and Kiro are what I’m actually there for. ...

Most migrations don’t start because someone planned one, they start because half the organisation is already using something unsanctioned and you’re formalising it before it turns into a compliance problem. I’ve seen that across different types of tech, and the failure mode is always the same: the pilot gets all the investment and the rollout gets whatever’s left, which is usually not much. That gap is the rollout version of the question I came back to in Better Than What It Replaces: is the new thing actually better for the person who has to use it? ...