Save Three Hours. Fill Them With Four.

I know exactly where everything is. That has never been the problem. The problem is getting to it, through directory after directory, across more systems than any one head should hold. Hundreds of files a day, at every stage of a job.

Take one small example. After a shoot, the edit often arrives with 100+ files. I go through every single one, edit back, send for approval and feedback, edit back again, mark them up, send them back via transfer, then go through it all again for the retouching. And a shoot is only one kind of day among the many my week requires. The judgement in that is the actual job. The trawling through directories to get to it, unfortunately, isn’t.

It doesn’t matter how good my filing is. Tidiness only gets you a clean map. It doesn’t make the journey any shorter.

So we tried to make software fit us better, with saved preferences, remapped shortcuts and mirrored workspaces, until the next update moved the furniture. Nobody asked. Then software moved into the browser and applications began speaking to each other, but we were still the ones moving between everything.

That’s the part I’ve noticed changing. I begin in one place now and I stay there. I talk more than I type, and an idea can arrive half-formed and I can start working through it while it’s still alive.

I work alongside three AI agents now, each shaped around a different part of my day. One works with me on the creative and strategic side, around the clock. One watches the operational layer and sits across the other two, so I brief it and it briefs them. The third turns up twice a day with something more reflective. Each is connected to every piece of software I use, and none of them needs me to explain who I am or what happened yesterday. For years, software remembered our settings. Now it’s beginning to remember our context.

Every week I also get a wrap-up of my industry, built purely around what we’ve been working on. It works out where each job is up to and almost predicts what we’ll need next. Yeah, I don’t know. For me, that’s pretty wild.

It sounds like a lot, because it is. The paradox is that doing all of it has freed up my time. Once it’s set up, which I did meticulously, I don’t have to think about it, and I can test an idea properly, with my visual platforms connected in, before I’ve built out a single storyboard. I used to spend so much of my working life caught in the friction of navigating various technologies. Now it all lives in one place. One location, rather than locating.

And somewhere along the way I realised the thing being handed back wasn’t really time. It was attention.

We talk about AI almost entirely in terms of productivity. I don’t particularly want to produce twice as much, and I’m not convinced the world is sitting around hoping any of us will. We’re going through a time of information overload, and providing less is going to be a benefit to anyone. The ideas themselves come out sharper and tighter in the edit, and there’s a lot to be said for information that doesn’t take up somebody else’s time.

What I want is more attention for the parts of the work that actually need me. The judgement. The interesting question. The moment something feels wrong before I can explain why.

And time has a habit of disappearing at the production end. A project starts with four weeks, feedback runs late, and suddenly it’s two while the final deadline stays remarkably punctual. So the time I’m handing back goes to the beginning, to questioning the idea before everybody starts making it. I choose what should be fast so I can protect what should remain slow. It’s also how I talk to brands about AI. We start with the friction, not the tool, and I’ve run it on my own work first.

We romanticise friction, as though anything difficult must be professionally valuable. Sometimes it is. Sometimes you’re reading 4,000 lines of somebody else’s code because eight pixels of padding are wrong. Let’s not get carried away.

It reaches beyond work, too. I’ve been helping someone I love through ageing and increasing care needs. Understanding those systems faster hasn’t made any of it less human. It’s given me more of myself for the conversation.

We’ve been here before. Technology has been promising to save us time for decades, and we’ve been remarkably good at filling every minute it returns. AI could easily be the next chapter. Save three hours, fill them with four.

Or we could make a different bargain. I’ve started asking which parts of my day never required much of me, and what deserves the attention only I can give. Then comes the difficult bit. Not filling the space straight away.

Some of it goes on a walk, or reading, or time with someone when half my head isn’t still inside the computer.

Occasionally, gloriously, nothing.