GenAI - Should it be "generative" or "creative"?

I am a little late to the Generative AI party - everything seems to have already been dissected and discussed on LinkedIn. As I struggled with the challenges of the blank page, GPT-4 is out, after months on a waiting list I have gotten access to Google Bard and I am seeing articles about how AGI (Artificial General Intelligence) might become a reality in a few months. Also, Elon Musk (who was an early investor in OpenAI) and 00s of other experts are calling for a halt to trainings of AI beyond a certain capacity. So, struck with a serious case of FOMO, I needed to strive and find my relevance in the zeitgeist.

As I started to think about what I wanted to write, I looked at the discussions happening around GenAI. There has always been a discourse around the implications of AI to the “human effort” and this has accelerated even more with GenAI. A lot has been written about AI needing to be thought of as Augmented Intelligence, where tech needs to be seen as a complement to the human effort. This is projected to raise productivity overall thereby increasing the size of the pie and as a result uplifting everyone. This is a fantastic vision and one that we should all aspire to and lean-in to make happen. But with the way we seem to be going about it, I am worried this is not going to result in this utopic future.

In my view, the biggest detrimental impact of the AI evolution, is how exceedingly hard it is becoming to economically justify the effort needed to come up with truly creative work. As I write this, the Hollywood Writers strike is entering its fourth month – and this will be seen as the defining moment in history, of the depreciating value we place on Human creativity.

Creative endeavors in all its forms, has always been about the stretch humanity afforded itself - the pursuit of which should have been weeded out by evolution. As a society, we have always incurred a “creative tax” by making space and supporting artists because they were the rare sub-set of people who could tap into something deeper and transcend and create brilliance - while the vast majority of us are slaves to our reptilian brain - doomed to predictable behavior and chasing efficiency.

A while back, in 2020, I wrote about, the trend of AI over the past many decades - periods that were characterized by "peaks of inflated expectations" were followed by periods that were "troughs of disillusionment" (aka AI Winters). Back then, even in my wildest imagination, I did not have any inkling of anything resembling GenAI as we see now in 2023. But even pre-GenAI, I had hoped to convey a sense of urgency, about the need for us to "remain Human" and how Humans will find our relevance when we make AI relevant. This new age of GenAI is very quickly seeming like a period where we are cresting the peak of inflated expectations, and there is an urgent need to unpack what it means to remain Human in this scenario. Unfortunately for us, the trajectory of Human evolution has not really prepared us to constantly re-evaluate our place in a fast-evolving environment.

I was reading this book, The Alphabet versus the Goddess, by Leonard Shlain. The author has an interesting hypothesis - how the rise of alphabetic literacy rewired the human brain and adoption of the written word resulted in a decline in feminine values and an uptake for more patriarchal, masculine values. For the perception of things like trees and buildings through images delivered to the eye, our brains use wholeness, simultaneity, and synthesis. When we view an image, there are many aspects that blend seamlessly to give us a sense of what it is about - we actually process aspects like depth, color contrast etc. Even with Speech, it is a whole Brain experience for both the speaker and the listener. If it is an interaction between 2 people, each is fully engaged in not only hearing what the other person has to say but is also focused visually on the non-verbal component of the other person's emotions. While reading text, we take a reductionist approach - breaking down whole sentences into its constituent words and each word is broken down into its constituent letters. For the perception of the written alphabet, the brain relies on sequence, analysis and abstraction. In fact, we do not even need to read every single letter and can skim through text, decoding in an efficient manner what the message is.

Now with GenAI, we seem to have crossed another threshold, where with a few text lines as prompts, we seemingly can generate an infinite number of outcomes, both in text and visual form. In many ways we have already lost the ability for abstract thinking with the reductionist approach to reading letters. This same core idea is reflected in Orlando Wood’s recent books – Lemon and Look Out – where he calls out the rise of the Left Brain, and this is resulting in the “crisis of creativity” – reflected in Art and other creative pursuits, including Advertising. So, while we discuss all the applications of GenAI that will significantly impact Marketing – we need to be clear on what we are looking to keep constant, what needs to evolve and what needs to change significantly. Even before GenAI, there was research that said that 30000+ pieces of content were needed for a new Brand innovation to stand-out in the media clutter. That in itself is, symptomatic of all the clutter being created – that then requires more clutter to have a chance to be discovered. ?

I would wager that when we “generate” - essentially type out the words describing what we want to see as the outcome, we value the outcome less. Because it has inherently already been defined – we cannot look beyond the description to seek out deeper meaning. When we have GenAI art exhibitions, where the description below the painting will be the MidJourney prompt used to generate it – will it ever result in art that will be considered “classic” by future generations? If we think of what makes something a "classic", as in something that stands the test of time - a classic is anything, our experience of which changes as we ourselves evolve. With a classic we start to see facets we never saw before, unpack nuances - and while the object has remained the same, our perception evolves.?The same goes for Marketing content – advertising is a balance of the Art and Science. We can take all the creative best practice learnings, all the data-led performance details, break down ads frame by frame to understand what drives impact – but ultimately, there is an art to bring it all together. When we over-optimize the individual parts, we damage the whole.?

So, what is the answer to all this? Are we doomed to descend into a world of simplistic narratives, art that lacks depth, music that lack soul and Marketing that increasingly feels lackluster and with efficiency ruling all? I do not think that calling for a slow-down in AI development is realistically feasible. This is the new frontier that every Company and even Nations are looking to occupy and define their future state competitive edge on. All we can do, is to be mindful of Remaining Human, as we go along for the ride.

I want to propose a new iteration of my earlier theory of what Remaining Human means in a world of GenAI - Artificial Constraint Thinking (ACT).

The constraints will be artificial – because there are no real-world constraints – ChatGPT is free and open to use, after all. And I am also not talking about limiting access – rather it is about whole-heartedly embracing? the change. In a world that is seemingly full of possibilities, where a “click of a button” is as much effort we want to expend to get outcomes, ACT requires us to implement artificial constraints, so that the outcome is much better. The constraints will come from keeping sacrosanct that which matters in all we do – Human evolution is millions of years in the making – and GenAI does not necessarily change that. The inspiration for this came from this book I am reading – Once upon a Prime, by Sarah Hart - which looks at the mathematical structures underlying in popular works of literature/poetry. I wish I were gifted enough to truly appreciate the depth of intellect the Author demonstrates in the book – but even with my limited understanding, there are certain facets that stood out.

Sarah Hart clearly describes the many similarities between Mathematics and Poetry – both have a structure to them, both can have multiple interpretations, both have a sense of universality and in both brevity is the clear need. Every style of poem has an underlying mathematical structure and pattern. Example a Haiku, a Japanese poetic form consists of three lines - 5 syllables in the first line, 7 in the second, and 5 in the third. And there is a description in the book of why a 5-7 split is better than a “dull” 6-6 split. And there are other such examples, of entire novels written without the letter “e”. These are examples of “artificial constraints” – when submitting oneself willing to a particular constraint, spurs creativity. The constraints are intended to work like how handcuffs and straitjackets worked for Houdini – the constraints are the reason for the genius to exist.

In practical real-world Marketing terms, ACT calls for a balance between exploitation and exploration. It is about celebrating small incremental improvements, done at scale that have long lasting impact; rather than significant change expected overnight that is short-lived. It is about always keeping in mind what matters, which for Marketing means delivering to the business objectives. At its very essence, ACT is

Summary

There is a lot that has been said about what AI can and cannot do – as well as how the Human effort will find its relevance in driving creativity. But what does, being creative really mean, and more importantly are Humans wired to be creative? Our Brains are hard-wired to be efficient – so every one of our past experiences becomes the filter through which we seek to make sense of the world. We always seek to classify every new aspect into our existing framework of past experiences. As such, every new technology was made to do things we already did, but faster, cheaper, more efficiently. This is most apparent with AI, even GenAI – where the foundation is built on models of training datasets, to then efficiently replicate new themes - this pretty much mimics the way our Brains work as exemplified in the theories of Left/Right Brain or System1/System2.

Fundamentally this is not a bad thing – because there is no such thing as true creativity. Everything creative is inspired from what came before. The important word here is “inspired”. When we take inspiration from the past, and look for new interpretations, new perspectives, our own unique take – we are looking to elevate the Human experience. And this is the essence of Creative Augmented Inquisitiveness.

I was reading this book called The Creative Act : A way of Being, by Rick Rubin. It is a refreshingly good read, in equal parts philosophical and real-world wisdom. Rick Rubin can lay claim to being a “creative person”, as he is credited with helping world famous artists across multiple genres of music (from Johny Cash to JayZ) find their mojo. A lot of what Rick talks about overlaps with the core tenets of Mindfulness – the ability of being able to just observe, without judgement, so we start to see facets that we would otherwise not be made aware of. Rick also talks about how being creative means we allow creativity to flourish. Nothing fundamentally new and different, but a reminder for what matters and how we need to try and be. But have we ever explored any new technology development through this lens? Ever since the Industrial Revolution the approach has been to exploit as much as we can and as soon as we can, rather than to explore the new state in more depth. And the same seems true of AI overall and GenAI in particular now. We seem stuck in an endless loop of driving more efficiencies by prompting. The new lucrative career path of Prompt Engineering exemplifies this thinking. The very term, ‘Engineering’, has a very scientific connotation to it and the emphasis is on generating, rather than creating.

It is not about being the naysayer or purporting doomsday conspiracies as has happened with every technological advancement when they first came about. The printing press and more recently photocopying, allowed for information dissemination – truly democratizing knowledge and aiding discoverability creators and their creativity. Cover songs, remixes and other creatively inspired adaptation of popular music, allowed newer generations to discover the original artists and content. There are signs even before GenAI, that we are too distracted with the sea of mediocre content that envelops us, that makes it difficult to discover good work and keeps us captivated via ephemeral feelings of juvenile fascination (my take on dopamine rushes as well as cat videos).

I saw this quote - "Knowledge is having the right answer; Intelligence is asking the right question".

With GenAI, we need both Knowledge and Intelligence, which is another way to think of Creative Augmented Inquisitiveness - "Intelligence is prompting with the right questions as context, Knowledge is knowing if the answers have any real-world merit".

Take a break from all the bleakness and be inspired by an exemplary example of human creativity.

https://www.youtube.com/watch?v=4QtCxjyuA08 – Beatles Get Back

STAY RELEVANT. STAY SAFE.

APPENDIX

1. https://www.bbc.com/news/technology-65110030
2. https://www.freethink.com/robots-ai/idea-amplification-is-the-really-exciting-potential-of-generative-ai

1. https://www.everythingisaremix.info/
2. https://www.adweek.com/brand-marketing/these-6-charts-show-how-marketers-are-adopting-genai-and-what-worries-them/
3. https://www.bcg.com/publications/2023/generative-ai-in-marketing
4. Someone did prompt to get a version of Sultans of Swing, originally by the Dire Straits in the voices by Donald Trump and Joe Biden. (https://www.youtube.com/watch?v=SIWKf2sZfok-)