Our experience with AI …
Craig Wallace
Business Strategist | Futurist | Resilience | Agility | Sustainability | Digital Transformation | #1
… and will it improve your strategy?
Ever since we launched our strategy platform in March this year we have been asked, what are you doing about Artificial Intelligence (AI)?
Over the last two weeks we have designed, built and tested a human-centred AI augmentation user experience, with integrated access to commercial LLMs (for now, OpenAI’s ChatGPT LLM line-up). Essentially, the platform now provides the strategist with an AI assistant that can provide additional input for evaluation.
?
We thought we would share the experience.
Our position before embarking on this journey was:
?
OUR EXPERIENCE
Our venture into AI augmentation for the strategist within our strategy platform has been an eye-opening journey. One filled with insights, innovations and an appreciation for the possible. The inclusion of ‘whole-world’ LLM access (ChatGPT 4) represents a significant milestone in our quest to enhance strategic decision-making processes through cutting-edge technology.
Iteration One (Diagnosis)
?We started by using AI to help us with the strategy diagnosis. Using the company name and industry classification, we asked the AI to provide the following
After a few iterations of refinement with the prompts we started to receive good responses. We reflected on the time it used to take to obtain this information prior to client workshops, this information was not only being retrieved at speed but the platfrom positioned it on the various analytic models ready for discussion and debate. A promising start.
?Rule 1. Positive Human Selection. Although the AI brings back some great data it does throw up some odd responses now and then, including a few hallucinations that a context-knowledgeable strategist can easily spot and weed out. A human should always positively decide to select the data that goes forward.
Iteration Two (The Challenge)
The real value from the diagnosis stage is the definition of the challenge the organisation faces. We had already analysed and defined the challenge in our test-case, it was now time to see how the AI performed.
The results surprised us, the AI identified 75% of the challenges we had defined and three of the four MECE sub-challenges. It also identified two addition subsets we had not identified. Had it just called out our own bias and blind sides?
Rule 2. Bias Champion. We all have bias, even though you work hard to keep an open mind, be guided by the facts and data, as you progress your analysis you start to form hypotheses which start to introduce bias. Maybe AI can help in keeping our analysis open and honest for longer.
Iteration Three (Ambition)
Formulating a purpose, vision and mission for the business. We started this ambition stage by firstly asking the AI to identify relevant new technologies as well as applicable innovation use-cases from other industries. We also asked the AI to identify the organisation’s competitive advantages.
We were then curious on how it could help with strategy formulation. In the platform, we use a methodology called Appreciative Inquiry and always start with the question “You wake up from a long sleep and the organisation is working exactly as you always hoped it would, what do you see and hear?”. This is always a very challenging exercise as workshop participants often struggle to describe what they ‘see and hear’, instead listing the things they want based on their pet initiatives.
So how did the AI do? We asked the AI to describe (see and hear) the vision of the organisation’s current strategy., It came back with a good list of words to describe the future state. Now, whether you can actually ‘see or hear’ them was often questionable. Having run this exercise hundreds of times with humans it certainly was a good starting position and, more importantly, as good as we have seen from most leadership teams.
Rule 3. Augments Experience. The challenge with strategy formulation is the lack of experience most leaders have, after all, this is often an activity done on an infrequent basis. When running a workshop exercise, asking leaders to participate in an unfamiliar task they often struggle to process the new exercise and the information required. Using AI can help augment this learning process by providing suggestions that get the exercise started.
We went on to use the same approach to answer the question “How did this happen?”, identifying the key pivots. This is starting to become exciting.
?
Iteration 4 (Plan)
?Initially, we had thought AI would just help with the diagnosis. Now we had started to us it to augment our ideas on crafting a vision. We were now keen to see how it would perform on planning.
We asked the AI to recommend the following, based on the organisation’s strategy
领英推荐
The best way to summarise the results is a “great starter for 10”. Is is common to see many gaps in defining actions from a workshop session, these are normally filled after by the consulting team. Using AI introduces many of these directly into the workshop encouraging a broader discussion and greater appreciation for what is required.
Rule 4. Accelerator. Asking the AI to recommend your action plan won’t provide you with your final plan, or even a comprehensive first draft. It will accelerate your planning efforts though and act as a great checklist to ensure you have covered all the bases.
?
OUR CONCLUSION
?Can Generative AI formulate your strategy or do the analysis? We believe, with some practice, it can help a lot. It is not perfect, as we have mentioned, but the ability to accelerate research and inject content into the workshops and decision making discussions helps create a broader and more inclusive debate.
Our initial scepticism about the potential of AI in strategy was quickly overshadowed by the richness of opportunities it presented. From enabling swift strategy diagnosis to facilitating deep market analysis.
Through iterative prompt refinement and human-in-the-loop oversight, we leveraged AI to illuminate blind spots and challenge our biases. By identifying key challenges and generating broad perspectives, AI emerged as a valuable ally in fostering more objective and comprehensive strategy analysis, formulation and planning.
Our biggest concern are not the AI’s hallucinations or mistakes.? Alert humans can spot those. But human brains have evolved to look to past experiences for the answers to new challenges, a fast-track to a solution, preserving brain energy. With the AI capabilities we are seeing today for strategy formulation (and things will only get better from here, and fast), will we stop thinking and challenging, and just take the easy option, whatever the AI says.
We do hear the following arguments
?Generative AI is not forward looking
We agree, human innovation and creativity is invaluable. However, most strategies are dealing with a mix of the current reality and the future. AI can certainly assist with the current reality.
Generative AI will not create a differentiating strategy
If we are looking for that new unique proposition then we agree. But the reality is most organisations will move their dial by aligning their organisation on the advantages the company has today and leveraging them by setting aligned objectives and measures.
With the capabilities we are seeing with AI in strategy formulation, will we stop thinking and challenging and just take the easy option, whatever the AI says.
This is a real concern. We must redesign our methods and approaches to ensure humans are still forced to think, challenge and innovate.
?
The real challenge is understanding and ownership of the strategy, the cause of the strategy-execution gaps most organisations experience. Relying on external strategy consultants will not address this. Hence the workshop and alignment sessions most consultancies run with their clients. Surely, anything that helps leaders develop and sharpen their strategic decision-making skills is a positive.
Our firm belief in the potential of Generative AI to enhance strategic decision-making is underlined by the diverse perspectives it offers. As we navigate the dynamic landscape of strategy and strategic planning, AI's ability to bridge current realities with future trajectories is emerging as a critical asset in organisational evolution.
?
THE FUTURE – THE BIGGER (AI) PICTURE
?What we trialled over the past few weeks, and what is now available as a prototype AI Assistant in our platform, is just the beginning. It relies on the power of “whole-world” foundation LLMs such as ChatGPT.
?A logical next step is to refine and retrain such a foundational LLM with a trusted, human-curated industry data set to give the AI Assistant industry context.
?The final frontier though will be to refine the industry-trained LLM with an organisation’s own strategy, planning, execution history. Unfortunately, today strategies and plans are often still kept in a variety of disparate and often non-digital sources, from PowerPoints to Excel spreadsheets. Organisations will have to consolidate their strategy, planning and execution value information into one unified information model (corporate memory) and improve the data integrity and quality (by switching to a digital strategy platform, like MyCoStrategist). ?With years of proprietary strategy information accumulating the organisation will eventually be in a position to train their own proprietary company-aware LLM. ?We can envision such an AI assistant and sentry continuously helping with guiding the company through the uncertain and dynamic world ahead.
?
Imagine a platform where the AI is continuously monitoring the macro-environment, competitors, trends, etc. and prompting you when your strategy and plans need refreshing.
By leveraging AI to guide organizations through market dynamics and strategic recalibrations, we pave the way for a more sense-and-respond, continuously adaptive strategy process.
In an uncertain, complex, fast-changing world this would be far more valuable than a periodic (quasi-static) strategy process, would it not?
At MyCoStrategist, we created our platform with this vision in mind.? We will continue to experiment with AI as it races to change all aspects of our lives.?
Today, the nascent AI capabilities of the platform all but offer a glimpse into a future where the AI Strategy Assistant continuously monitors and frequently proposes refinements to ?strategic initiatives, guiding organisations towards sustained success in an increasingly complex and competitive environment.
We want to redefine how we solve today’s challenges, helping organisations and leaders find their inner strategist.