Silence - the bane of A.I.

I have recently been thinking about Silence and how it is a rare commodity in the world we live in. And I do not mean Silence as the absence of noise, rather just less external stimuli that keeps our Brains on the predicting hamster wheel.

There is research to prove that our brain which is cocooned in our skulls, has no real interaction with the outside world. Everything we perceive are electrical neuro signals from the various sensory organs that reach the brain. And to be efficient, evolution has resulted in our brains becoming predictive engines. It is constantly predicting what will come next - and when the prediction is right, there is no dissonance. Lisa Feldman Barrett in her book, Seven and a Half Lessons about the Brain says - "When your predicting brain is right, it creates your reality. When it's wrong, it still creates your reality and hopefully it learns from its mistakes". So when we are constantly bombarded with external stimuli, our Brains are on hyper predicting mode. And when the outcome is in line with our prediction, we do not even notice it.

But this alignment of the outcomes with the predictions, also means superficial experiences. Experiences where we did not engage any conscious thought, never really experienced the moment more deeply, did not learn anything new and overall quite forgettable.

There is an interesting angle, of using less stimulus, to enhance the experience. This introspection came by when I was watching a review of the movie Sicario (part 1) where they deconstructed the narrative of the movie and the choices made in its making. I found Sicario (part1) a fascinating watch, where the characters, every one of them has depth, however small their role in the movie. What makes it compelling is the pared back approach to story-telling - which in this movie is achieved by presenting the characters as who they are, without showing why they are the way they are. Apparently the initial script had a lot more scenes for the titular character Alejandro (played by Benicio Del Toro) - and then this was reduced by 90%. What we are left with is "silence" - moments intended to make us lean-in, build the narrative in our own minds and truly experience the situation the characters are in. This also means, we can watch the movie multiple times, and still be engaged because these moments, make each viewing a new experience.

We can all agree that delayed gratification - where we choose discipline over distraction - is an important life choice for all of us. (proven by the Stanford Marshmallow experiment). In the world of entertainment, Netflix is the bane of delayed gratification - they pioneered the release of entire seasons all at once, and seem to be on a constant hamster-wheel of new content releases (estimated to spend 17 Bn$ in new content in 2024). It seems that we never learned from Behavioral Sciences which have shown that more choice is not always what is needed - and we can all agree that we spend more time deciding on what to watch on Netflix, than we spend watching actual content. And even when we do watch an entire season of a new series it is not the fully immersive and enjoyable experience we hope it will be. We come away strangely dissatisfied with the experience.

Most movies and other entertainment series are chockful of end to end dialogs, there has to be something happening at all times, they do not let people into the story, but takes them on a roller-coaster ride, whizzing past the landscape at high speed. It seems that technology allows for creation of fantastic new worlds, but they are made so vast and complex, that they never invite the viewers in. The mark of a great Director, was always the ability to aid the audience to focus attention on specific part of scenes - but when the landscape is so full, how can this even happen? We also never seem to form a connection with the characters. And characters are always more important than the story.

There is another aspect of "silence" to consider, that has been substantiated by research. Two groups of people were studied - one group watched a movie in its entirety with no intermission; and another group watched the same movie, but with an intermission. The group with the intermission, had an overall more positive experience. This is possibly the best case validation for advertising!! Ad breaks enable a better viewing experience for the content they interrupt. There is something about taking a break from the activity that elevates the overall experience. The intermission allows our minds to take some time to better process what we have just experienced, potentially relive it in our minds and it makes us build up our expectations for what is to come after.

From the early days of AI, the narrative has always been about what can we do more of. The discussions and applications have always been additive, never reductive. And this has been further put on steroids with GenAI, where anyone's random thoughts can become the basis for more content.

The fact that we can do so much more, should not mean we do all those things. But is it even possible to take a reductive approach with AI? What would the prompt be, where we want to remove excess details, to enhance the overall experience? Can meaningful silence even be defined in the training data, that the GenAI tools are built on? At best we seem to be able to summarize - mostly resulting in a simpler narrative, not necessarily a better one.

Rick Ruben's philosophy to unlocking artists' creativity, was to get them to focus on what made them happy. The creative pursuit has to fundamentally be self-serving. Every successful artist, creates art that appeals to themselves and by staying true to this, their art finds wider appeal. The artist comes first and the audience who will experience the art, come last. This aspect of being true to oneself, is fundamentally a Human trait. All AI is trained on external validations - it is looking at what people liked/disliked/engaged with - and tries to recreate that same experience. And as such the whole creative equation is the wrong way around. The estimate is that 25-35% of all movie releases are sequels/prequels of existing franchises. (How many Godzilla movies is too much - 39 at last count?).

We also see this approach prevalent in tech that aids advertising creation - the focus is on past performance metrics, to guide future ad creation. Is it really the number of iterations (000s) of ads that can be generated, that will revive the role of creativity in driving business impact?

When was the last time you were surprised by a piece of content - be it a piece of TV advertising, a Creator's video on YouTube, a print-ad that captivated with its headline....?

A feeling of surprise is just the dissonance we feel when something happens that is counter to our brain's prediction. This is when we pay active attention, and this enhances our experience of the moment, builds memory structures for the brand etc. In the pursuit of Efficiency - doing more faster and cheaper - we are losing out on Effectiveness. Everyone seems to be using the same tools, and in trying to game the system, the result is a sea of sameness.

Ultimately, there seems to be an urgent need to relook at the role of tech in creativity. How much should the past dictate the choices to define the future? Is more always better, or should we embrace a reductive approach to enhance creativity and enrich the human experience?

STAY RELEVANT!