Generative AI: The Speed of Curiosity

What i’ve really noticed using Generative AI in research is the speed of my curiosity. Not simply speed of retrieval, but speed of synthesis and iteration. It reminds me of the process of marking up a large illustration, where you sketch out the rough outline before filling in the fine detail. Or the way i often create a high level structure for a book before actually writing much content, to allow me to balance the effort across all the sections.

Using tools like Claude i can more easily try things out: synthesising ideas, asking for depth, or inspiration. I can mark out a landscape and then delve more deeply into it.

I find that it’s easier to retrieve half remembered theories or facts, or to dive into entirely uncharted waters. In a weird way it’s like asking the server what their favourite desert is when you can’t decide.

It’s not efficiency without cost: i notice that i’m overall reading less. Partly because i do not need to, but partly because i have become impatient. Perhaps i will lose some of the happenstance and emergence of long form exploration, but overall the landscape i traverse will be broader?

It’s hard to know: will my perspectives become superficial, or will my self critical lenses survive the convenience of my accelerated curiosity?

I am, as you know, an optimist, so naturally i feel the benefits acutely, especially when i think back to my earliest experiences of research as a postgraduate, where i still had to get my supervisor to sign a piece of paper (after i’d cycled to the campus and wandered around till i found them), which i’d take to the library who would, after six weeks usually, ring me up to tell me that a photocopy of an article i’d requested had arrived. From there to here is a journey that sees the radial compression of time – to near instantaneous, through to the radical expansion of the creative space, as i have a partner in thought at my fingertips.

I know it will make me different, but better? Hard to know: to an extent it depends on which measure you are using.

My favourite use case (which i must not therefore mistake for a broad truth) is that Generative AI lubricates our collective thinking: working with Sae this week on new ideas we have used it as a dynamic dialogic partner, in the flow of our thinking. It’s felt like an energy added to our (already energetic!) conversations.

I’m pragmatic, but also stubborn. I do not intend to write ‘with’ Generative AI, any more than i intend to stop illustrating by hand. But will i use these tools to mark out ideas, to broaden my thinking and challenge my output? Asking for feedback, critique, ideas, or where to look next? I’m sure i will.

It’s easy to get caught up in the popular debate: about bias, about validity, about influence or infiltration, about the ‘good’ and the ‘bad’. All of those things are important. But let’s not miss the potential, the excitement, the dynamism and change.

I have no hesitation in saying that Generative AI will change almost everything, and faster than most Organisations will be able to think, let alone react. Things that we feel will last forever will be tumbled into the sand (including the memories of those very Organisations who felt their intelligence, history, money, and pride would make them agile, whilst failing to actually change).

At the heart of it, Generative AI will challenge legacy notions of value, and we will need to recalibrate marketplaces to accommodate that. I look at this within the broader context of the Social Age (which is already exerting existential pressure onto our systems) and the legacy of the pandemic, which has fractured some of the pillars that we rest upon.