3 ways to use AI for strategic decision making

Leadership discussions about AI often focus on defining an AI strategy, much like developing a marketing or operations plan. But a more profound and potentially impactful perspective is emerging: incorporating AI at the heart of strategic decision-making.

Can AI replace naturally complex human skills in strategy development?

Limitations of AI in strategic decision making

In an internal interview, McKinsey senior partner Yuval Atzmon outlines six stages of AI integration in strategic decision-making.

1. Descriptive analytics – Leverage dashboards to understand performance metrics.
2. Diagnostic Intelligence – Analyze historical data to understand performance drivers.
3. Predictive intelligence – Use market signals and historical data to predict future scenarios.

Atsmon points out that these first three steps are already widely adopted. But he believes his next three stages are on the distant horizon.

1. AI provides strategic advice.
2. AI will be given some decision-making authority.
3. AI makes fully autonomous decisions without human intervention.

Therefore, while AI can contribute to strategic decision-making, it cannot develop strategies on its own. Creating strategy requires a blend of analytical and creative thinking, which is still a distinctly human ability.

Nevertheless, AI can play a supporting role in strategic decision-making through several means, including:

1) Better discussion

In a conversation with Rob Beer, author of Smooth Scaling: 20 Rituals for Building Frictionless Organizations, how the rise of technology is changing three key aspects of strategic decision-making I tried to find out what happened.

1. Access to data is the basis for sound strategic decision making. Advances in the Internet, cloud computing, and related technologies have met this need.
2. The next step is data analysis. The proliferation of his MBA graduates and individuals with similar analytical training over the past two decades has made the necessary analytical acumen widely available across industries and organizations.
3. However, there is a persistent challenge of the quality of discussions within the executive committee and across the organization.

Beer points out that “boardroom discussions reflect the dynamics of organizational discussions, which are plagued by issues such as: Bullying, passive-aggressive behavior, political gamesmanship, miscommunication, and lack of genuine dialogue lead to suboptimal strategic decisions. While AI can enhance data processing and analysis capabilities by identifying patterns, correlations, anomalies, and trends, these are relatively small enhancements to help humans. But what if AI could facilitate more effective and meaningful discussion and discussion? ”

In addition to the dysfunctions Beer points out, it eliminates egos, avoids echo chamber effects where participants only reaffirm each other's perspectives, and allows those without power to speak truth to those with more power. There is an important need to do so.

Ray Dalio, founder of Bridgewater Associates, introduced radical transparency to address these challenges, an initiative that has recently been criticized for fostering a culture of intimidation and fear. . Is there another way forward?

Beer observes that despite the privacy concessions, we are already accustomed to AI note-taking tools in online meetings. “Soon, AI coaches will be joining our meetings to provide advice on how to encourage quiet members to contribute, increase empathy, and highlight important points missed in presentations,” he said. “These AI tools are on the verge of providing insightful feedback to our interactions.” . ”

Receiving criticism from colleagues during a meeting can be difficult. It can be difficult to be asked to “read the room'' during a tense discussion. Is the feedback from the machine likely to cause backlash? This depends not only on the user's own trust, but also on the user's previous experience and trust in his LLM, such as ChatGPT. However, as the use of AI becomes more widespread, we may become more willing to accept its opinions.

Additionally, Beer emphasizes that AI can act as an impartial voice in debates, unafraid to challenge the status quo, much like the court jester in Shakespeare's King Lear. For example, AI could highlight unrealistic predictions in real time, prompting deeper scrutiny and discussion.

By 2025, Beer predicts that AI tools that provide coaching, ask questions, and spark thoughts on strategic matters will become the norm.

2) Democratization of strategic decision making

Christian Stadler, professor at Warwick Business School and author of Open Strategy, uncovers new applications for AI in strategic decision-making, highlighting how AI can scrutinize and validate strategies with a broader internal audience. It emphasizes possibilities. Mirroring its usefulness in leading boardroom discussions, AI is proving invaluable in coordinating discussions among larger groups. Stadler illustrates this with a scenario where a CEO uses AI to glean insights from his 2,000 employees to try to make sense of a 10% drop in revenue. This process evolves in subsequent rounds, introducing contrasting perspectives for further discussion, and ultimately he AI ​​integrates these insights to inform management's strategic decisions.

3) Idea generation

Both Bier and Stadler emphasized the potential of generative AI as a tool for generating strategic ideas. But to take full advantage of its benefits, strategic decision makers need to get used to rapid engineering by creating specific instructions and queries to optimize AI ideas.

Ultimately, while AI can enhance human decision-making, it is insufficient to acquire the comprehensive skill set needed for strategic planning. The concept of an AI CEO remains a distant prospect, given that the complexity of strategic choices requires increasing sophistication in artificial general intelligence. Humans therefore have the opportunity to exercise judgment and wisdom in evaluating different options. They can enhance the strategic decision-making process by using empathy to connect with the right stakeholders and applying practical skills to effectively utilize the tools at their disposal.