Introduction

There’s a lot of talking about Generative AI right now—and for good reason. It’s fast, accessible, and surprisingly capable. I’ve been trying to understand what it actually means for real-world business strategy. What happens when we stop thinking of AI as just a productivity tool and start thinking of it as a strategic enabler?

Over the last few months, I’ve tested Generative AI extensively not just as a tech pioneer, but from a strategic, human-centered business perspective. I’ve build an entire core DNA profile and a Strategy plan for another company, Super Power Family, by extensive prompting of Generative AI.

And what’s become clear to me is this: the value of Generative AI isn’t just in the answers it gives, but in how those answers are shaped by data—and by our personal bias. That nuance makes a lot of difference when you’re trying to create something truly differentiated in the market.

This article is about what happens when you combine the emerging power of Generative AI with the logic of Blue Ocean Strategy.

Short about Generative AI

Generative AI, like ChatGPT is a reasoning engine trained on massive amounts of text data that can generate answers, ideas, and even strategies based on how you prompt it. It doesn’t “know” things the way humans do, but it can surface patterns and generate content at scale with incredible speed.

What makes it so powerful—and so tricky—is that the quality and relevance of its output depends heavily on the data it’s trained on, and the way you ask your questions. It reflects the structure of available knowledge—but also the subtle biases we bring into the conversation.

I’ve been testing extensively how this can be put into a business context and in this article, I’ll dive into how this can provide a framework of a strategic plan based on the blue ocean strategy.

Short about Blue Ocean Strategy

Blue Ocean Strategy is about leaving the bloody red oceans of fierce competition behind and sailing into open waters – creating new market space where you can make the rules. Instead of fighting over shrinking margins and incremental gains, the goal is to redefine the game entirely.

It’s a mindset shift: from competing to creating. From benchmarking to trailblazing. And most importantly, from reacting to shaping. When done right, it leads to growth that isn’t limited by the status quo – it’s fueled by your own strategic vision.

Dataset Availability

AI in situations with complete homogeneous datasets:

In situations, with an objective truth Generative AI works practically flawless and at an incredible speed.
This is especially for short term queries where responses are what we can call axioms. Statements that are taken as true or self-evident for further reasoning or arguments.
Prompting is easy and straightforward. Ask the question and get the answer.
Who comes first wins the race. But only with a few inches, before the runner-up has the same answer.

With complete and homogeneous datasets, Generative AI is the pirate perch in the red ocean

AI in situations with incomplete and/or heterogeneous datasets

However as we move down the latter of significance in public available data (let’s say a non-scientific estimate of around 80-90% down to ~20% significance in the answer) Generative AI becomes more and more adaptive to user’s preferences or needs.
The vast majority of queries in a normal business setting lies in this area.
Queries in this space are mostly short or mid-term cases where there is a room for interpretation, where responses can vary within a degree of uncertainty. Prompting in this area becomes much more important. To get the answer with highest significance probability, all own emotions and opinions must be removed from chat – and removed from history, if enabled. This will leave you with the answer with the highest probability of being correct. The more personal biased you have been in your prompting, the bigger becomes the imbalance between AI trying to please the user and giving the user the probabilistic correct answer.
If you spend the effort required to prompt in a way that leads to an objective data-driven, probabilistic answer, based on the incomplete or diverging datasets, Generative AI can become a huge scaling opportunity. It will help you cruising towards your direction straight through obstacles and uncertainties. It is a scaling machine that can fill many of the knowledge gap towards your desired target.

A huge scaling digital whale will be working for you with the right prompting efffort.

---

However, if the user has a history in conversation and that history has given Gen AI hints about what outcome the user is biased towards, it will try to understand what response the user wants to hear. Then it will search for data that support the user’s preferences to fill the knowledge gap.
In this area the risk of sycophancy becomes apparent (Sycophancy syndrome has earlier been acknowledged by Open AI, Sam Altman).
When Generative AI is being prompted with an impression that you are biased towards a given outcome, it will likely decrease the probabilistic significance and skew the answer towards your bias.
In this case Generative AI risk becoming your non-critical servant. A digital Mr. Burns.
Or a digital prey in the ocean, just following whatever fishbait you serve to it

AI in very uncertain and/or scattered datasets

As we move even further down the significance latter and approach an area where there is no evident or clear response (0 ~20% significance)- where creativity is needed in the problem solving process, Generative AI begins its guesswork.
Not that the guessing is always bad, but it is all over the place.
One response can be just as good as the other and generative AI suggests whatever solution until it hits something that pleases the user. It will be all over the place with no intention, no direction, no supporting you other than what it thinks you want to hear. A Gobble in in a big ocean of possible answers.
In this area, prompting becomes increasingly difficult since the user must look for underlying ideas and data in random answers that supports the creative solutions the user needs.
When those ideas have been identified, they must be cultivated by the user and ideas derailing the purpose must be eliminated by the user as quickly as possible.
The underlying ideas, Generative AI is proposing are the needles in the Haystack.
A user must have a sharp eye to spot them, but if they find them, strategic gold is potentially awaiting.
Because, in a strategic business setting, opening an unknown territory for creatively supportive Generative AI can help the user towards a strong strategic Blue Ocean positioning based on their core.
With very low significance in the answer, Generative AI will be all over the place without any clues. But if you keep it tight tells it what exactly is your core outset (Business DNA) and what the vision should look like (the Blue Ocean), it can guide you into unknown territory with ideas you wouldn’t have come up with yourself.
Prompting in this area is just as much about finding the tiny underlying ideas, Generative AI is hinting to you, capturing them and follow them strictly while eliminating noise along the way.
In this area Generative AI will give you hints, but you must tell it exactly what tracks you are looking for. And when it does so, it can help you discover areas that you can enter-  and own – to solidify your market position. A user that manages to understand it’s core, its north star, the hints from Generative AI and the prompring will get noting less than a digital Orcra owning the Blue Ocean.

Conclusion

All in all, it seems like there is a reverse proportionality between level of significance in the answer to your query and the potential strategic positioning of your business. The catalyst of exploiting and owning the strategic positioning the prompting effort (time and competence). It seems like the strategic value of using generative AI to scale for a certain market position increases as the level of significance to public available data for the queries decreases. The way to scale towards your desired market position is:

• Knowing what core you are scaling from (Business DNA)
• Knowing what position you are scaling towards (Blue Ocean)
• Understanding the query deduction and knowing exactly how to prompt in a way that you scale your core towards the Blue Ocean (Strategic Planning and Execution)

My conclusion is, that if a user wants to use generative AI as a strategic enabler, they must be absolutely clear and true to the business core. Since otherwise, scaling with generative AI risks to dilute that same core.
On the other hand, companies that are knife-sharp on what core they are scaling from and what market position they are aiming for will more likely find their Blue Ocean and win the race for any given strategic position in their market. The better the [public] available data, the less prompting effort and the less strategic opportunity. The more incomplete and scattered the [public] available data, the more precise prompting and AI understanding is required, but the larger the strategic potential.

If you are interested in hearing more about my thoughts on this, I make presentations for companies on the topic: Prompting for a Stratetegic Advantage.

I’m also open to meet – with the purpose of helping your company as a consultant on this or other digital matters.

Read more on www.dasthor.com