Strategic Dissonance and the Rapid Development of AI

The concept of strategic dissonance, introduced by Robert A. Burgelman and Andrew S. Grove in their seminal work published in the 1990s, refers to the misalignment between a firm’s strategic intent and its strategic actions, often signalling a need for significant strategic shifts.?

Today, this phenomenon is being exacerbated by the rapid technological advancements and dynamic competitive landscape being experienced in the area of artificial intelligence (AI). One hardly needs to mention the stunning proliferation of large language models (LLMs) and AI assistants to back up this point.

Now, more than ever, understanding and managing strategic dissonance is crucial for sustaining competitive advantage and fostering innovation. This analysis explores how strategic dissonance relates to the speed at which AI is developing and how organisations can leverage this understanding to improve their strategies and minimise dissonance.

Strategic dissonance and AI

In high-technology industries, strategic dissonance arises when the pace of technological change outstrips an organisation’s ability to adapt its strategic intent and actions accordingly. Burgelman and Grove highlight that such dissonance is inevitable in dynamic environments, where the alignment between a firm’s strategic intent and its actions is tenuous. This misalignment can lead to strategic inflection points (SIPs), where organisations must pivot their strategies to capitalise on new technological and market conditions (Burgelman & Grove, 1996).

Currently, the AI landscape exemplifies this dynamic, fast-paced technological environment in which organisations are very likely to experience strategic dissonance.

Breakthroughs in generative AI, natural language processing, LLMs, and other subfields of AI are occurring at an unprecedented pace. LLM-based technologies like GPT-4 (Open AI) and Gemini (Google) have radically transformed the options available to organisations. These rapid advancements often outpace almost all organisational strategies, thereby creating significant strategic dissonance.

Even organisations that are sophisticated in the area of traditional machine learning (ML), are experiencing strategic dissonance due to the introduction of powerful generative AI capabilities. Generative AI, which involves models that can produce high-quality content, such as text, images, and code, represents a significant technological shift.?

Organisations that were previously focused on traditional ML approaches may find themselves at a strategic inflection point, needing to pivot towards generative AI to remain competitive.?

The gap between the old strategic intent (having no AI capabilities or focused on conventional ML applications) and new strategic actions (leveraging generative AI) exemplifies strategic dissonance.

Strategic dissonance as a signal for strategic renewal

Burgelman and Grove argue that strategic dissonance should be viewed not merely as a challenge but as an opportunity for strategic renewal. By recognising and responding to dissonance, top management can harness conflicting information to realign their strategic intent with the evolving external environment. This process is particularly important when it comes to AI, where continuous innovation and adaptation are necessary for sustained success.

Effective management of strategic dissonance involves strategic recognition – the ability of top managers to identify the strategic significance of emerging trends and internal actions before conclusive environmental feedback is available. In AI, this means recognising the potential of emerging technologies like retrieval augmented generation (RAG) or agentic workflows, which combine generative models with traditional information retrieval methods to enhance AI capabilities.

An example of managing strategic dissonance with respect to generative AI is through the use of LLM model fine-tuning. By fine-tuning pre-trained models on specific tasks or datasets, organisations can adapt to new applications and/or market demands without the need for extensive resource investment in developing models from scratch. This flexibility could allow organisations to bridge the gap between strategic intent and action, minimising dissonance.

Using strategic dissonance for competitive advantage

Organisations that successfully navigate strategic dissonance can turn it into a source of competitive advantage. This involves not only recognising dissonance but also fostering an internal environment that supports strategic adaptation and innovation.

Open-source LLMs potentially provide a valuable resource for organisations to mitigate strategic dissonance. Organisations can integrate powerful capabilities offered by these models while protecting IP and retaining the highest security standards.

Prompt engineering, which involves designing and refining prompts to elicit desired behaviour from LLMs, received a great deal of attention with the release of ChatGPT. However, prompt engineering alone likely does not provide a true competitive advantage. The reason is that prompts are relatively easy to construct or emulate. Prompt engineering may, therefore, represent a facile response to strategic dissonance.

Practical examples of resolving or avoiding strategic dissonance

Google’s strategic investments in AI, from acquiring DeepMind to developing the TensorFlow framework, demonstrate how recognising and responding to strategic dissonance can drive sustained competitive advantage. By aligning their strategic actions with the rapid developments in AI research, Google has maintained its leadership in the industry.

OpenAI’s clear commitment to advancing AI capabilities highlights the importance of strategic recognition and adaptation. Their development of GPT-3 and subsequent models like GPT-4o underscores the need for continuous strategic renewal in response to technological advancements.

Recommendations

To effectively manage strategic dissonance being experienced by rapid advances in AI, organisations should consider the following approaches:

Foster a culture of innovation and experimentation

Encourage a corporate culture that values innovation, flexibility, and open communication. This involves creating an environment where employees feel empowered to explore new ideas and challenge existing strategies.

Invest in strategic recognition capabilities

Develop mechanisms for continuous monitoring and analysis of technological trends and market dynamics. This includes establishing dedicated teams or roles focused on strategic foresight and environmental scanning.

Leverage open-source and collaborative platforms

Actively participate in open-source communities and collaborative projects to stay at the forefront of AI advancements. This can also provide access to a broader pool of knowledge and resources.

Focus on AI motes

Invest in the development of AI expertise that will allow true differentiation. This likely implies going beyond simple prompt engineering on common LLMs like Gemini and GPT-4.

Engage in continuous learning and development

Promote continuous learning and professional development for employees to ensure they stay updated with the latest advancements in AI and related fields.

Conclusion

Strategic dissonance, when properly understood and managed, can serve as a catalyst for innovation and strategic renewal. The rapid pace of AI development necessitates a proactive approach to recognising and responding to dissonance, allowing organisations to realign their strategic intent with evolving technological and market conditions. By fostering a culture of innovation, leveraging open-source resources, and investing in strategic recognition capabilities, organisations can navigate the complexities of AI and maintain a competitive edge in this dynamic field.

Further reading

• Burgelman, R. A., & Grove, A. S. (1996). Strategic dissonance. California Management Review
• Reynolds, L., & McDonell, K. (2021). Prompt programming for large language models: Beyond the few-shot paradigm. arXiv preprint https://arxiv.org/abs/2102.07350