Maybe We’re Not Tired of AI. Maybe We’re Tired of Being “On.”

I have a slightly uncomfortable relationship with AI.

I work in marketing. I spend a good part of my day thinking about AI, writing about AI, strategising around AI, figuring out what AI means for businesses, and occasionally asking AI to help me write about AI.

There is something deeply absurd about that sentence.

And yet, here we are.

Every week, there is a new model. A new capability. A new benchmark. A new announcement telling us that something we thought was impressive three months ago is now obsolete.

And while AI is moving at an almost impossible speed, the rest of us are still trying to catch up with our own lives.

That disconnect has been bothering me lately.

There’s a quote by Dario Amodei that stayed with me: the intersection of AI and our political institutions feels a bit like the Hobbits and Treebeard. The technology is moving like it has somewhere urgent to be. Our institutions are… well, taking their time.

The metaphor is funny.

The reality isn't.

We are building faster than we are learning how to live with what we build

The thing about exponential growth is that human beings are spectacularly bad at intuitively understanding it. We adapt to change in increments. AI doesn't, it compounds.

Our policies, education systems, workplaces, social norms and even our understanding of what constitutes "work" are evolving at a pace that feels almost embarrassingly slow beside it.

And honestly, that's where my concern lies.

Not in the tired "AI will take all our jobs" conversation.

Not even in the "AI is going rogue" headlines. Those deserve attention, certainly. But there is another, quieter thing happening. Something much more ordinary. We are becoming exhausted.

I think we're beginning to feel the cost of constant intelligence

There was a time when being connected felt liberating. Now I sometimes crave the opposite.

A phone that does less.

A conversation where nobody is checking anything.

A book that doesn't know what I should read next.

A walk without turning it into content.

Making something badly with my hands.

Sitting over chai with a neighbour and having a conversation that produces absolutely no measurable outcome.

I've started wondering whether this is actually a form of resistance. Not resistance to technology. Resistance to constant optimisation. Because that's what AI has quietly brought with it.

Everything can be faster.

Everything can be automated.

Everything can be summarised.

Everything can be personalised.

Everything can be optimised.

Everything can be generated.

And eventually you start wondering:

Optimised for what?

Perhaps efficiency was never the whole point

I think about this a lot as a marketer. Marketing has spent decades trying to eliminate friction. Make the journey shorter. Reduce the clicks. Personalise the message. Automate the follow-up. Predict intent. Increase conversion. Now AI is accelerating all of it.

But somewhere along the way, I think we confused frictionless with better. Some of the most meaningful things in my life are wonderfully inefficient.

Learning something slowly. Having an uncomfortable conversation. Making something with my hands. Getting lost somewhere. Reading something that doesn't immediately become a LinkedIn post. Spending time with people without an agenda. None of these things scale.None of them optimise particularly well.

And perhaps that's precisely why they matter.

The irony isn't lost on me

I'm writing this on a device that can access more information than any human being in history. I'm using AI to help me think about the consequences of AI. I'm professionally invested in understanding where this technology is going. And I still find myself wanting to switch everything off and make something out of clay.

That contradiction doesn't bother me anymore. In fact, I think it might be the point.

The answer isn't to escape AI. We can't. And, frankly, we shouldn't want to.

AI will change how we work, create, learn, govern and probably how we understand intelligence itself.

But perhaps our job isn't simply to become better at using it. Perhaps our job is also to become better at knowing when not to. Because there is a difference between having access to everything and being present for anything. And increasingly, I think that difference is going to matter.

Maybe the future won't belong to people who can use AI for everything. Maybe it will belong to people who can still recognise the things that shouldn't be optimised at all.

Like a cup of chai with a neighbour.

A badly made clay cup. A long walk. A stupid conversation. A quiet afternoon. Or simply sitting somewhere without asking what it can do for you.

Perhaps in a world getting exponentially smarter, one of the most radical things we can do is remain human.