On the human need for categorisation
In the crucible of human ingenuity, knowledge has always been the kindling that sparks innovation. Our ancestors harnessed the power of storytelling, inscribing tales on cave walls, and later, on the papyrus of the ancients. It was our first attempt at knowledge management—a primitive file manager if you will.
Sidenote: the most expensive transcription per word in the world!
From Storied Traditions to Organized Chaos
In a world where predators and risk lurked around many corners, our survival hinged on knowledge sharing, and the vital facts. Fast forward to the ancient libraries of Alexandria and the early Dewey Decimal whispers in monastic corners; these were our first clumsy steps in cataloguing human thought. Every index card, every leather-bound tome on a wooden shelf, was a piece in the grand mosaic of our collective consciousness.
But are we ultimately doomed to slap Hashtags on everything?
If you've spent any amount of time with a product team, you've probably heard about the 'double diamond'. It's a great tool to show the high-level phases and simple steps of working through solving a problem; lately I've been observing a similar pattern with the flow and summary of information that is being collated mostly by LLMs.
This is fundamentally the weird loop at play.
Diverge aka GO WIDE LLM!
AI technologies are generating an unprecedented torrent of new information, ideas, and insights, effectively widening to a breadth once unimaginable. I asked perplexity.ai what the largest context window is currently, to which it summarised Google Gemini 1.5 (Pro) with up to a million tokens. In short, plenty, and by plenty honestly, I mean plenty of filler.
Converge aka MAKE THIS LESS OF THIS!
We can diverge until our heart is content, but we all know inevitably we'll have to converge all that information into something coherent. Right now I'm observing this in varying degrees of completeness. No matter what your process, there is a level of distillation.
Diverge and Converge AKA RINSE AND REPEAT
Then, that loop happens again, but on the consumer of the information's side. Right now, most of the time, that is a human, potentially using another LLM to translate this information into another context to again, distil it back down with its new coherence.
"And Ash?" you might be thinking.
Well, it feels slightly redundant when you think about it in that context.
Why bother
Serious Business: Innovating the User Experience
Jokes aside, the crux of the matter lies in the extraordinary leaps we are about to make within user experience (UX) domains. The narrative is shifting from flat, static screens to dynamic landscapes where technology adapts to us, the protagonist in our own digital story.
Interfacing with the Building Blocks of AI
Imagine, if you will, a platform where you, the user, can dance with tensors—the very fabric of AI's dreams—as easily as swiping right on your favorite dating app. A user interface so intuitive it feels like telepathy with your smart devices.
The Capstone: Where We Stand
As we stand at the edge of tomorrow, poised to step into a world where our interfaces will be canvases for our imagination, let's not forget the roots from which this tree of knowledge sprouted. For as we sculpt the future with our ideas, it is the well-managed past that holds the clay we need.
In Closing: Your Thoughts
But this is just the starting point, a single brushstroke on the canvas of possibility. I’m keen to hear your musings—how do you envision the fusion of knowledge management and cutting-edge UX transforming our tomorrows?
