Simon Griffiths

Earlier today, I looked at some work I did in ChatGPT in late February — about eight weeks ago — and it immediately struck me that the work needed to be redone. Not because it was bad at the time, but because the models have made huge steps forward in just a few weeks. The prompts I wrote then were less refined, and the outputs that seemed impressive two months ago now look noticeably rough around the edges.

What’s more, if I recall correctly, I was using my own account on the web browser version.

Now everything has changed. My employer has given us access to OpenAI in both Chat and Codex. I have and use daily native desktop versions (on Mac) of both ChatGPT and Codex. At some point in those few weeks I used the CLI versions of Codex and Claude Code, but I seem to have settled on the desktop version of Codex — which is rock-solid for me, by the way.

Security has become a big topic for everyone in my organisation. Desktops and laptops are getting locked down, servers are hardened, and non-standard services are increasingly blocked. The shadow of the Mythos release is everywhere at the moment.

The laptop lockdown means I’ve just bought my own Mac Mini for personal and development work. It was becoming difficult to even edit the occasional letter in Word, as my organisation automatically puts confidentiality statements on any document created. My Mac Mini sits on my desk and I switch my screen and devices between the two computers a couple of times a day.

But my key point is none of these things — it’s the rate of change. In eight weeks my workflow has transformed. Many of the things I did before have gone and been replaced by new AI-related tasks, and I expect all of this to change again in the coming weeks.

I’ve worked in IT since about 1985, and I started programming in 1978. Through that time the IT world has gone through many cycles, but I don’t think I’ve seen a rate of change as fast as this. Mainframes became minis, minis became PCs, we had client-server, then browsers came along. Java became the language of the enterprise, only to be challenged by microservices — which may already be a little past their sell-by date.

But this is different. Every day brings something new. A new challenge. A new model. A new tool. Every change has an impact, and it’s increasingly hard to just be aware of what’s happening, let alone adapt and gain benefit from the newest, shiniest thing.

Anyway — back to reading blogs and watching YouTube videos. I used to read books to learn, but there’s little point now when any book on specific tools or models is out of date before it hits your Amazon basket. So that’s how I attempt to stay in touch. What about you?