My train from home to work takes about 60 minutes to get me from the sleepy shires of England to the beating heart of London, and despite occasional moments of brightness, the mobile signal remains mostly dark. (Oh how I long for a slightly shorter journey time but still with no signal, roll on high speed rail!) Usually I struggle to get the weather forecast to load. This morning, that same journey produced a five-page whitepaper.


The subject was the environmental impact of AI (more on this next week), and how we might talk about it to people who don't care about the technical detail but do care about what it means for them and of course, our planet. I didn't need a fat connection. I needed just enough bandwidth to get a decent prompt into the cloud, and by the time I reached the office the draft was waiting. Ready for review, which is doing a lot of work in that sentence, because the judgement, the edit and the accountability are all still mine.


It's the fourth thing in recent weeks that has made me stop and think that would have been considered witchcraft not all that long ago.


The first was a walk with the dogs through the rolling British countryside, talking to my AI about a new keynote. Because my AI already knows my role, my keynotes and my areas of expertise, I could describe the event, the theme, the client and what they were after, and then spend about 45 minutes talking through the narrative and which bits of my material and experience were relevant. This wasn’t me dictating stuff, it was a fluid, thoughtful conversation with a colleague who knew my work and could contribute to help me get to a better result. By the time I'd got home, fed the dogs and got them settled for the day, the same chat was open on my desktop, and waiting in it was a fully drafted, more or less client-ready overview of the pitch.


The second is the arrival of the AI assistant in the car - on either CarPlay or Android Auto. When the moment strikes me, and assuming the appropriate levels of awareness, concentration and driving conditions (OK mum?), I can talk through my inbox (we’ve been promising ourselves that one for at least 25 years), or better yet, an idea for new material, or an opportunity I'm working on.


The third will be familiar if you've seen me speak recently. I usually take audiences through the four main opportunities that AI brings every organisation. The second of these is Insight and Discovery: your own personal spotlight for finding the needle in every haystack. I usually add a joke about "having dinner with your data" (and I’ll tell you here as I tell them in the room - don't judge me, I'm still married, she’s not left me yet, sometimes when she’s sitting with her mum for dinner, I work while I eat). Given the opportunity, I upload a dataset to my AI agent, put my phone on speaker, and ask it to tell me the trends, and chat with the machine as it helps me explore the value that’s buried in the data.


In Business Reimagined I defined flexible working as choosing the right time and the right location for the task in front of you. A few weeks ago I argued that we stalled because we never changed how we measure work. I think there was a second brake I'd missed. We promised people the right place, but work that needed a screen, a keyboard and a decent connection could only happen where those things were. "The right location" quietly meant "somewhere with a desk, a screen and a keyboard."


Yet the moment I'm in the right headspace to think about a keynote and the moment I'm sitting at a keyboard don’t always coincide. The ideas turn up on the dog walk, in the kitchen and in the car as often as they do at the desk. The difference now is that I can catch and develop those ideas where ever they land, irrespective of whether there’s a screen in front of me and a keyboard under my fingers.


I want to be careful, because there's a Victorian version of this story that takes me back to the heyday of the BlackBerry: A device in your pocket giving you 24/7 access to your inbox, so as a result you spent every waking moment cranking out more email (and ironically complaining about the volume of email at the same time). Oh boy were we efficient! But we all remember the burnout (and the sore thumbs). What I’m talking about is different: instead of cranking the monotonous handle of productivity for longer, you can now do something creative (shaping a narrative, finding the pattern in the data, drafting an argument) when you’re in the right place cognitively, even though you’re in the wrong place technologically.


Which brings me to Bill Buxton. Years ago I had a wonderful conversation with Bill, then a principal researcher at Microsoft Research, a pioneer of multi-touch interfaces, and a passionate builder of birch bark canoes in his spare time (surely the second favourite pastime of every Canadian? Look him up if you get chance.) He told me the point about human-computer interaction is that it's never about finding the one best interface. It's about having a selection available and choosing the best one for your circumstances.


I've often wondered why QWERTY, a layout designed for a Victorian typewriter, has survived every better idea thrown at it. Now that voice not only works but is genuinely powerful because there's an agent on the other end, perhaps that changes. But I’m betting Bill would say the question isn't whether voice replaces the keyboard. It's whether we finally have a choice.


I'm writing this on the train home, and voice would be the worst possible interface right now. The signal is patchy, and it would annoy the hell out of my fellow commuters. Though at least two of them are watching something on their phones with the sound blaring for all of us to “enjoy”. Seriously, what is with that? So I'm typing. Right interface, right moment.


That's the flexible working I described all those years ago, with one more choice added to the list: the right time, the right place, the right task, and now the right way to interact. It was never meant to squeeze more email into the gaps in our lives. It's meant to let us catch the good idea when it turns up, and do something worthwhile with it.


So here's my question. The next time an idea strikes and you're nowhere near a desk, what will you do with it?