Attention is All you Need. Choice is All we Want.

2023 will go into the annals of AI history as an inflection point. But what is lesser known is that Gen AI would have been inconsequential had it not been for one paper that was written in 2017.

The paper, titled Attention is All You Need, was in many ways a map of hidden treasure to AI through a concept that humans had known and used for eternity. The authors introduced something called the Transformer architecture, a blueprint for machines to understand our world and our words with an elegance previously unseen. At the heart of this architecture was a mechanism called “self-attention,” which enabled AI to dissect and comprehend the intricate nuances of language, to see the delicate interplay between words and phrases in many ways as a maestro sees the notes on a score.

But as we head into 2024, I find myself pondering over a curious irony.

The very skill that AI is honing to near-perfection — self-attention — is the one we humans seem to be relinquishing.

Our focus is fractured, splintered by the barrage of information that bombards us daily. We’ve become passengers in our own minds, ceding control of the wheel to algorithms that decide what we see, what we consume. A good example is the endless scroll of information on LinkedIn and Twitter (I see the irony as I write this) concocted and curated carefully by AI and algorithms to captivate our attention.

Curiously enough, the arms race for eyeballs, attention, and engagement started a few years ago with a fairly innocuous notion — personalization. In an economy of ads, where the views and likes convert to cash, some companies figured out that if you feed users with what they like, they tend to engage more. This led to harvesting two kinds of data — demographic (who you are) and behavioral (what you do) to start personalizing “feeds” that catered to more of the same. Thus came the era of echo chambers where algorithms started feeding humans more of what they are likely to engage with based on the two different kinds of data about them.

Yet, I have hope as amidst this cacophony of innumerable mindless yet “personalized” content, there lies a choice.

It’s the same choice that has defined the human condition since time immemorial — the choice of where to cast our gaze.

The algorithms may suggest the path of least resistance, the road well-traveled, but we can choose the road less taken. We can choose the nooks and crannies of the world that speak to us, the stories that resonate with our spirit, the knowledge that feeds our insatiable curiosity.

I have hope because with Generative AI, there now seems to be a re-established emphasis on Conversational User Interfaces. This puts the control back into the users, the humans. I am more likely to ask ChatGPT what I am interested in versus a passive scroll on social media where AI has decided what I should consume.

As we stand at the dawn of this new year, my wish for all of us is not just to find our focus but to relish the act of choosing it.

Let’s not forget that in a world where attention is currency, where to direct that attention is our greatest power.

In 2024, may you wield that power with intention.

May you find the strength to turn away from the noise and toward the intention of your attention. Because if attention is all AI needs, then choice — that beautiful, defiant act of selection — is all we want.