AI as a kaleidoscope into business complexity

I recall those many mesmerising hours as I pressed the tiny eyepiece against my face. A piece of seemingly simple plastic, when pointed at any ordinary light source revealed what felt like a world hidden from our normal view.

And when we point our little kaleidoscope towards businesses today, we get something not too dissimilar. What seems relatively simple on the surface, uncovers a detailed web of complexities hidden underneath.

Unravelling complexity in business has continued to be a passionate interest of mine. Particularly in fitting new some of the newer puzzle pieces (i.e. AI and its various other forms), into the interlinking of business norms and processes.

With disruptive technologies like this, the challenge is not whether there is a place for AI, but in moving the rest of the pieces away so that the AI piece can fit nicely. And the thing about enterprises is that this shuffling of the pieces is made even harder when the technology itself is not so easily understood. We can't all have PhDs, although Dr. William Huynh does have a nice ring to it...

Thus the challenge becomes one of creating simplicity out of one of the most enigmatic technologies out there. However, finding the right balance of complex and simple is like walking a tightrope. Too complex, and the ideas are paralysingly confusing. Too simple, and we risk not grasping the full value and risks in implementing this technology.

This pursuit of balance can be observed throughout business history as technology advancements come in waves. Whether it is 80/20 principles or the numerous one-page "canvas" that we have come across or made to aid decision-making at the strategic levels. Humans themselves are neurologically geared towards simplification, deriving patterns from our experiences, and subconsciously applying cognitive heuristics and biases in our decision-making.

But how does this all relate to AI?

I believe that AI has opened up new ways for us to think about organisational complexity and the way we might deal with it.

AI and its recent breakthroughs in generative AI and Large Language Models (popularised by ChatGPT) have moved the fulcrum of how much business complexity that we can handle. To bring back the kaleidoscope analogy, it is an improved kaleidoscope with new colours and angles. Acting as a new and improved interface into the depths of organisational complexity.

This notion that AI is an interface of business complexity can be more easily recognised in the current trends of generative AI. Particularly as AI makes use of large corpora of knowledge and data which has formed a key portion of organisational complexity that was previously not well understood or easily utilised. A concept space that is filled with content, connections and patterns at a scale incomprehensible by any normal human brain.

However, it is not feasible (currently) for every organisation to develop or own their version of ChatGPT due to significant costs training and development costs, but given this newly accessible corpus of business know-how, a question surfaces that needs some strategic thought. To what extent would the ability to extract and handle organisational complexity generate transformative value and competitive advantage?

I argue, that for most organisations of scale, this opportunity is not something to let go of. Companies with decades of experience providing services and products to customers, with thousands of employees across many verticals have built over years of know-how, goodwill, and many intangibles that are asking to be used.

Would the ability to transmute complexity into business value be a new frontier of competitive advantage? The *Complexity Transmutation Effectiveness* of an organisation.

Three aspects have surfaced in the discussion of AI: the ability to extract knowledge and patterns, the ability to continue to build on the complexity, and the ability to manage the balance of complex and simple. The first is being explored the most via generative AI in applying natural language to large corpora of data and documents owned by organisations. A simple example is Generative AI’s use to facilitate document and information retrieval in law firms.

But with the recent advances in generative AI, reflection and memory, AI will soon mature its ability to learn and construct new knowledge of the world and businesses in ways that humans will not be able to comprehend. Not only because of the volumes of data, but rather because the complexities of the patterns and logic could be formed in ways relevant to machines but not relevant to humans.

And as scary as it may become, it does elevate AI into a role where it may become a shepherd of organisational knowledge and complexity that we humans may interface or do business with. When I think of this, an image comes to mind.

A mysterious void that is all around us, and when we reach into depths, we pull out something useful and meaningful. And what lives inside the void is something we cannot comprehend - something significantly different enough to our worldview that we are unable to explicitly understand. But because it is meaningful, we engage. Despite lacking full understanding.

I'm not sure the phrase "black box" is an adequate description of what AI could become.

To follow for more musings around AI and emerging technology, feel free to subscribe to this newsletter below.

Thank you for reading No Playbook by Will Huynh. This post is public so feel free to share it.