Thinking About Developing an Artificial Intelligence Strategy?
Here are the areas you should cover.
In this edition of the Senior Decision Maker, we take a deeper look at your AI Strategy, and in particular which areas you should focus your strategy on.
Introduction
In today's accelerating technological ecosystem, Artificial Intelligence (AI), especially generative AI such as Large Language Models (LLMs), is reshaping the competitive dynamics across industries. However, integrating AI into existing strategic frameworks is a complex challenge for many organizations. The core question we address here is: What fundamental areas should an AI strategy cover?
This article is designed as a guide, aiming to equip senior decision-makers with actionable insights and considerations. We outline key areas that will shape the outcome of your AI initiatives. Given the advancements in AI, the question is not ‘if’ to adopt AI, but ‘how’ to implement it in a timely, effective, and efficient manner. Here, we break down the essential pillars that warrant urgent and thoughtful consideration in any robust AI strategy.
AI in a Historical Context
Before diving into AI strategies, it's important to understand the broader historical context. The closest precedent to the advent of artificial intelligence may be the development of artificial muscle power. This trajectory started with rudimentary mechanical tools and animal labor millennia ago and has evolved to encompass modern technologies such as electricity and rocket propulsion. Each step along this path reshaped industries, labor, and even entire economies, but the evolution was most of the time gradual, providing time for adaptation and planning.
Contrastingly, artificial intelligence is ushering in a broad range of changes at a considerably faster rate. This isn't merely a tweak to existing paradigms. Given the rapid pace of AI development, a business-as-usual approach may not suffice. Instead, it makes sense for businesses to consider AI as a unique and significant leap that warrants its own strategic focus.
Hypothesis-Driven Strategy Framework
In corporate strategy, traditional methodologies often led to detailed yet rigid plans that struggled to move from strategy formulation to actual execution. You might remember the early 2000s, when strategies would often be created as comprehensive documents but lacked mechanisms for implementation, and ‘got stuck in the drawer’.
In contrast, contemporary approaches have increasingly adopted agile and iterative frameworks. These are often built on a hypothesis-driven model that allows for quicker adjustments based on real-world feedback. This approach is vital for areas like AI strategy, where the technology landscape is rapidly changing. By starting with a set of testable hypotheses, organizations can engage in agile sprints to quickly validate or modify strategic initiatives, making the strategy more adaptable and aligned with business outcomes. Identifying the key hypotheses is crucial for effective strategy formulation.
Below, we outline four focus areas hypotheses that you can expect an AI strategy to cover:
1. Build the foundation for using AI
2. Strengthen Key Capabilities with AI
3. Improve everyday efficiency with AI
4. Improve executive decision-making with AI
Focus Area #1: Build the foundation for using AI
The foundation is the groundwork on which every future AI initiative will be built, the framework that will enable your organization to get leverage from AI technologies.
Selecting the Right AI Tools
One of the core components of this foundation involves the selection of AI tools. Contrary to what one might think, this isn't as easy as hopping on the bandwagon of the latest technology. Instead, it's about aligning your tool selection with your strategic objectives. In the realm of large language models, for example, the highest-performing solutions are often SaaS offerings, like OpenAI’s ChatGPT. These solutions are generally best-in-class but may not align with every organization's data security requirements. Open Source solutions, like Meta’s Llama 2, while not as advanced, offer more flexibility and control over your data. The choice isn't a one-size-fits-all; it's a nuanced decision that can differ from one use-case to another within the same organization. You might find that a hybrid approach serves you best, using SaaS solutions for less sensitive but high-complexity tasks and Open Source models for operations involving sensitive data. This intricate balancing act between capabilities and security is but one example of the strategic questions that need to be addressed.
Building Workforce Proficiency in AI
But tool selection is just a part of the equation. Another foundational element is training your workforce to effectively utilize these AI tools. Skill-building benefit from being a continuous process. A single training session is insufficient; continuous learning is essential to keep your team proficient as the technology matures. Alongside this, creating a platform for internal knowledge sharing can be valuable. Just as in any other professional field, the collective wisdom of the team can surpass individual expertise.
Data Management as a Strategic Asset
Then comes the role of data, a pillar of your AI strategy that can't be overlooked. Proprietary data isn't just an asset; it's a competitive differentiator. However, this necessitates a proactive stance on data management. Organizations need to scrutinize what data they're collecting, assess its quality, and determine its relevance to their strategic objectives. Poor quality or irrelevant data can be more than just unhelpful; it can be misleading and detrimental to your AI initiatives.
Compliance and Ethics: Integral to AI Strategy
Finally, compliance and ethical considerations must be woven into the very fabric of your AI strategy. Laws and regulations pertaining to data and AI are not static; they evolve as the technology does. Ignorance of these guidelines is risky. Beyond legal compliance, ethical considerations should guide your AI practices. This might mean establishing a set of principles that govern how AI is used within your organization, ensuring that its application aligns not just with what is legal, but also with what can be considered right.
Building a foundation for your corporate AI strategy is a prerequisite for everything else. Tool selection, training, data management, and ethical considerations are the cornerstones that will support and inform all your future AI initiatives.
Focus Area #2: Strengthen Key Capabilities with AI
In Enterprise Architecture, a 'Capability' is a specific function or service that an organization delivers. Unlike resources or processes, which are what a company 'has' or 'does,' Capabilities focus on what a company 'can do' (and should be good at doing). These are the core competencies that drive customer value and differentiate your business in the marketplace. When crafting an AI strategy, focusing on enhancing these key Capabilities could be a good starting point.
Identifying High-Impact Capabilities Through Customer Journeys and Service Design
The first step is to identify which Capabilities have the most impact. This requires a structured approach. Use methods like Customer Journeys and Service Design to dissect how your organization interacts with clients and how you can create value at each interaction point. This exercise provides you with a framework to isolate the Capabilities that are ripe for AI enhancement.
Prioritizing Capabilities: The Effort-Impact Matrix
Once you've identified these key Capabilities, the next task is to prioritize them. One way is to employ an Effort-Impact Matrix to evaluate each Capability based on the effort required for AI implementation versus the potential value generated. The aim is to focus on high-impact, low-effort areas first, thereby ensuring quick wins that creates momentum for more comprehensive implementations.
The Digitalization and Digital Transformation Link
This strategic approach to enhancing Capabilities is not an isolated concept but aligns closely with broader Digitalization and Digital Transformation initiatives. While Digitalization often focuses on using technology to improve existing processes, Digital Transformation takes it a step further by re-architecting the Business Model around information flows. Generative AI can be a part of either approach, whether you are looking to optimize your Operating Model or engage in a full-scale Business Model transformation.
Focus Area #3: Improve everyday efficiency with AI
The integration of generative AI into the professional sphere offers the potential of a skilled and efficient intern always ready to assist. However, merely having access to a virtual intern doesn't automatically guarantee enhanced productivity. Just as an intern might require clear instructions and guidance, maximizing the benefits of AI demands both proficiency in its use and the establishment of optimized workflows.
Leveraging AI for Meeting Efficiency
Generative AI is being progressively incorporated into mainstream productivity tools, such as Microsoft 365, making it accessible for routine office tasks. One significant area of its application is in meetings. By leveraging AI effectively, it is conceivable to drastically improve meeting efficiency. This not only pertains to creating real-time transcripts or generating comprehensive pre-read documents but also to fostering a culture where attendees come better prepared. While it might be daunting to sift through a 20-page document, AI can assist participants by summarizing, prompting questions, and cross-referencing information, ensuring meetings are more actionable and focused.
Transforming Email Communication with AI
Email communication can also benefit from Generative AI. AI can be employed to enhance the effectiveness of email communication. For instance, while one can send long, detailed emails to avoid missing out on critical context, recipients can utilize AI to extract summaries based on their understanding and perspective. Furthermore, for formal emails where precision is paramount, AI can easily be trained on your individual writing style, or adhere to a corporate standard, ensuring consistent clarity and professionalism.
Eventually, you can even envision using an AI to act as a virtual manager in certain scenarios. AI can build project timelines, and plan and track daily activities.
Focus Area #4: Improve executive decision-making with AI
The application of generative AI in decision-making could be a key part of a corporate AI strategy. Here are the essential areas to consider:
Rethinking the Decision-Making Process
Generative AI shines when decisions are complex, have multiple dependencies, and require generalist insights—common scenarios for senior decision-makers. To fully leverage AI, a change in the ways of working might be necessary. The foundational elements discussed earlier are equally applicable here. Senior decision-makers, often adept at asking the right questions, will find this skill invaluable when interacting with AI. A text-based decision-making process is preferred, given the current capabilities of AI tools. Such a process allows decision-makers to directly engage with raw data and textual information, amplifying their analytical reach.
Best Practices for Data Utilization
While AI can provide a wealth of raw data, effective use often requires guidance. Best practice prompts or frameworks can help decision-makers extract the most meaningful insights from the data. Mastering this practice offers several advantages. First, it mitigates the selective screening of facts that occurs when organizational levels try to distil key issues. Second, it provides decision-makers a more profound understanding of the bigger picture, enabling them to ask all the "stupid" questions before meetings and deep-dive into critical concepts.
Identifying Synergies and Overlaps
AI's analytical capabilities can further be leveraged to cross-examine various projects and initiatives. This analysis can reveal potential synergies and overlaps, allowing for optimized resource allocation and strategic alignment.
Recommendations
Creating an AI Strategy is a different process for every company. It is important to adjust it to the needs, opportunities and threats that are relevant to your business. Everyone needs a foundation, of some sort. But a lot come down to the ambition level. One consideration is that inaction regarding AI adoption can also be a strategic decision. There might be a case for waiting and let others do the mistakes. Still, this is a high-risk strategy. The risk of falling behind may lead to a cost disadvantage that could be challenging to overcome. In particular, there is a risk of losing key employees, that realizes that they can get a higher leverage of the work by using AI to support them. So, I would recommend everyone to, if not do a full AI Strategy, at least do a minimal analysis. Use common sense, and keep the following three things in mind:
Focus on High-Value Areas: Ensure your AI strategy encompasses the domains where maximum value can be generated: building a solid foundation for AI adoption, strengthening key capabilities, enhancing daily operational efficiency, and consider also using it to improve executive decision-making.
Adopt a Hypothesis-Driven Approach: With the swift pace of AI evolution, anchor your strategy in testable hypotheses, emphasizing those with the most substantial potential impact. This method ensures agility and adaptability to real-world feedback.
Maintain an Agile Foundation: Regularly reassess and refine the fundamental components of your AI strategy, staying nimble in response to technological shifts and regulatory changes.