Dear Reader,
I recently had to write an essay.
There was a word count. References. An assessment. And, very much of the moment, an AI declaration form.
I genuinely couldn't remember the last time I had written a proper essay. University perhaps?
That would make it more than 30 years ago, which is a slightly alarming thought in itself.
But what interested me most wasn't whether I could remember how to write an essay. It was how I was going to use AI while writing it.
Could I do the thinking first?
I use AI a lot now.
It helps me challenge my thinking, structure ideas, refine things I have written, analyse information, prepare for meetings and generally work more efficiently.
When I write, my usual approach is to get my thoughts down, use AI to help me refine them and then keep editing until they sound like me.
For the essay, I wanted to do something slightly different.
I wanted to write more of it myself first.
Not because I think using AI is cheating. We were allowed to use it, provided we declared how. And I certainly wasn't going to pretend it didn't exist.
I was more curious about whether I could still do the underlying work without reaching for it too quickly.
Could I sit with the question? Develop the argument? Work out the structure? Write something that wasn't quite right and keep working at it rather than immediately asking AI to improve it?
It was harder than I expected.
It reminded me of Tipp Ex
It took me back to another technological transition.
At university in the late 80s and early 90s, I was an early adopter of word processing. I had access to one of the first Apple Macs.
I couldn't touch type, but suddenly that didn't really matter.
There was a delete button.
I could move things around. Rewrite a sentence without rewriting the whole page. Change my mind. Experiment.
No Tipp Ex required.
Gradually, something changed.
I stopped writing things out by hand and then typing them up.
I learnt to think directly onto the screen.
Then I started work and, bizarrely, went backwards. One of my first jobs involved responding to customer complaints and I had to handwrite the letters, correct them with Tipp Ex and then give them to someone else to type.
I remember how frustrating that felt because the technology had already changed how I worked.
And I realised writing this essay that AI may already be doing something similar.
It isn't simply making some tasks faster.
It is starting to change the process I use to think.
So I removed the temptation
My solution was quite practical.
I started writing the essay on an eight hour flight back to Nairobi with no WiFi.
No ChatGPT or Claude. No quick question when I got stuck. No asking for alternative structures or a better way of expressing something.
Just me and a blank screen.
And it worked.
By the time I landed, I had most of the structure and a significant amount of the content.
Was it beautifully written? No.
There were clunky sentences. Bits that didn't flow. Arguments that needed tightening.
But they were mine.
And I was strangely proud of that.
Of course I eventually used AI
I didn't turn this into some heroic experiment in doing everything without technology.
Once I had the substance, I used AI.
I asked it to check things. Challenge parts of the argument. Spot repetition. Help me see where something wasn't clear.
But I tried to use it much more as a corrector and challenger than a creator.
That distinction was useful for me.
It made me think more carefully about when AI genuinely improves my work and when I might be reaching for it simply because it is easier than sitting with something difficult for a little longer.
What capability do we want to keep?
I don't think the answer is that we should deliberately make our work harder.
I certainly don't plan to start handwriting things again simply to prove I still can.
Technology has always changed the skills we need.
I no longer need to remember phone numbers. I rarely navigate somewhere new without Google Maps. I don't need to know how to use Tipp Ex, thankfully.
And perhaps in a few years we will think it equally strange that we used to spend hours creating first drafts ourselves.
But I do think there is a question worth asking:
Which capabilities are we happy to outsource, and which do we still need to exercise ourselves?
For me, writing isn't only about producing words.
It helps me work out what I think.
The messy first draft, the argument that doesn't quite work, the paragraph I rewrite three times: all of that is part of the thinking.
If AI removes too much of that friction, do I become more productive but slightly less thoughtful?
I don't know yet.
And perhaps that is the point.
We are still figuring this out
I don't think there is a perfect formula for how much AI we should use.
For different tasks, I use it differently. Sometimes I want it heavily involved. Sometimes I deliberately hold back.
What writing the essay reminded me is that how we use these tools matters as much as whether we use them.
The question for me isn't: Did AI help with this?
It is increasingly:
What do I want AI to do here? What do I still want to do myself? And why?
Apparently even typing is becoming outdated and I should now be talking into Whisper rather than tapping away at a keyboard.
So perhaps I am already behind again.
But for now, I am quite pleased that I can still sit on an eight hour flight, stare at a blank screen and eventually work out what I think.
Even if the first version is considerably clunkier than I would like.
👉 I would love to know how others are approaching this. What do you now use AI for almost automatically? And are there parts of your work or thinking where you deliberately choose not to use it?
Warmly,
Liz
Strategic Advisor | Former CEO | Founder, Volante
Based in Kenya, available globally
