A friend’s reaction when I told him I was writing this gave me pause. He didn’t dampen it deliberately — he was encouraging — but he said he hoped my voice wouldn’t turn into AI slop. I slept on it. The concern is legitimate. The internet is already full of content that was written by AI for no one in particular, saying nothing specific, in a voice that belongs to nobody.
I woke up this morning and decided to go ahead anyway. With an explanation.
My condition affects how I communicate. The thoughts are there — the inner life runs clearly enough at its own pace — but getting them out, organised, in a form that does them justice, is where things break down. Before AI, that gap was widening. What Claude does is give me back a voice I was losing. Not its voice. Mine, recovered.
Everything here started as a conversation. I talked, we thought it through together, and what came out the other side was something I genuinely meant and couldn’t have produced alone. Without this tool I wouldn’t have a blog. I wouldn’t have a voice to worry about corrupting. That’s worth saying clearly before anything else.
I spent years working in community education, making IT and the internet accessible for all. I managed a team that built learning programmes, and helped people with very low skills, older learners, disabilities, chronic illness, and often cognitive challenges learn to use technology. I saw, over and over again, what made the difference between technology that transformed a life and technology that arrived too late to help.
The answer was almost always timing. People who started early — while they still had the cognitive capacity to learn — did well. People who waited until crisis hit found the learning itself had become impossible. The thing their condition had taken from them was exactly the thing they needed to get started.
I know this now from the inside. I have a neurological condition that is progressing. I am watching my abilities slowly change. And I am in a quiet race against time — building systems and learning tools now, while I still can, so they are there for me when I need them more than I do today.
That’s why this blog exists.
What AI actually does for someone like me
On a difficult morning, when brain fog makes a simple decision feel like wading through concrete, having a thinking AI partner that doesn’t get tired, doesn’t judge, and works entirely at my pace is not a luxury.
It helps me organise thoughts that won’t organise themselves. It tracks things I can’t reliably hold in my head. It reduces cognitive load on the days when cognitive load is the whole problem.
But the thing that genuinely moved me — the thing I didn’t expect — is that it gets better over time and this isn’t from me just improving my skills, though that helps. Assistive technology is fixed, it does what it does. AI improves continuously. So while my condition may progress, the capability on the other side of that conversation keeps increasing. The gap doesn’t widen the way it does with conventional tools. If anything, it narrows.
For anyone managing a progressive condition, that’s a different proposition from anything I’d encountered before.
What I hope this blog does
I’m not writing a how-to guide, though there will be practical posts alongside the personal ones. I’m documenting a thinking practice — what it actually feels like to use AI when your cognitive bandwidth is genuinely limited, and what I’m learning as I go.
I want people in similar situations to find this and know they’re not alone. I want the people building these tools to understand what this experience is actually like from the inside. And I want to make the case — quietly but clearly — that the people who would benefit most from AI assistance are very often the ones least able to afford it. That’s worth saying in public.
This might be for you, it might not. If it is, I hope it helps you get started. If not, I hope it’s just thought-provoking.
This blog is written with the assistance of Claude. That’s not a footnote — it’s rather the point.
Thinking at Half Speed 