AI Strategy: Distinguishing Incremental Tools From Fundamental Leverage
An executive analysis of AI's impact on business models, distinguishing between incremental efficiency gains and fundamental structural shifts. The discussion highlights the limitations of LLMs in generating novel value versus averaging existing data, and provides a framework for identifying true points of leverage in AI adoption.
The Strategic Imperative of AI Adoption
The current landscape of AI adoption mirrors the early internet era, characterized by significant uncertainty regarding fundamental business impacts. While large enterprises are actively integrating AI, the distinction between incremental efficiency gains and transformative structural shifts remains unclear. This analysis provides a framework for evaluating AI's potential to alter industry dynamics, focusing on the nature of digital automation and the limitations of large language models (LLMs) in generating novel value.
Automation and Validation
A primary driver of AI utility is the ability to automate digital tasks where responses can be validated at scale. If a process is explicit, digital, and mechanistically testable, LLMs can likely automate it effectively. This leads to faster, cheaper execution of many business functions. However, this automation does not automatically translate to market disruption. It may simply reduce costs or increase speed without altering the core competitive landscape. The Jevons paradox and price elasticity play roles here, but the immediate operational impact is often more significant than the long-term strategic shift.
Leverage vs. Tools
A critical strategic question is whether AI addresses the core point of leverage in a business model. In some industries, such as music or newspapers, the internet fundamentally changed the product and distribution, creating new markets and demolishing old ones. In others, like heavy machinery or traditional banking, digital tools improved communication and efficiency but did not overturn the fundamental business of owning assets or serving specific customer needs. For example, neobanks represent a new entry point for digital-first customers but have not eliminated traditional banks. Companies must assess if AI is a useful tool or a fundamental unlock. If it is the latter, it may create new competitors or markets; if the former, it is an operational improvement.
The Limitation of Averaging
LLMs are fundamentally averaging engines, providing probable answers based on training data. They excel at scaling standard processes, such as analyzing call center data or running cybersecurity simulations, effectively acting as thousands of fast interns. However, they struggle with tasks requiring true novelty or creativity that deviates from existing patterns. The challenge for businesses is to identify where the average answer is sufficient and where unique, non-derivative value is required. This distinction is crucial for strategic positioning, as relying solely on AI for innovation may lead to generic outputs that fail to differentiate a brand.
Conclusion
Businesses should approach AI adoption with a clear understanding of its capabilities and limitations. Focus on automating validatable digital tasks for immediate efficiency, while carefully evaluating whether AI addresses the core leverage points of the business. Avoid over-reliance on AI for novel innovation, recognizing its strength in scaling and optimization. Strategic planning should remain adaptable, acknowledging that the true impact of AI may emerge through unexpected applications rather than predicted scenarios.
Key insights
-
LLMs can automate any digital task where training data is sufficient and responses can be validated at scale. This capability drives significant cost reductions and speed improvements in explicit, testable workflows.
Impact: Enables rapid scaling of back-office and customer service functions, allowing companies to reallocate resources to higher-value strategic activities.
-
The impact of AI varies significantly by industry, ranging from fundamental disruption to incremental improvement. Some sectors see new market structures, while others experience only marginal efficiency gains.
Impact: Requires tailored strategic approaches for different business units, as a one-size-fits-all AI adoption strategy is ineffective across diverse industry contexts.
-
A critical strategic question is whether AI addresses the core point of leverage in a business model or merely improves peripheral tools. This distinction determines the long-term competitive advantage.
Impact: Helps executives prioritize investments in AI capabilities that directly impact core value propositions rather than those that offer only superficial benefits.
-
LLMs primarily generate average or probable answers based on existing training data, limiting their ability to produce truly novel or creative insights. They excel at scaling standard processes but struggle with unique strategic positioning.
Impact: Prevents over-reliance on AI for innovation, encouraging businesses to combine AI with human creativity to achieve differentiated market positioning.
-
The value of AI often lies in deploying massive numbers of agents to test combinations rapidly, akin to having thousands of fast interns. This approach is effective for problems where brute-force testing is viable.
Impact: Enables rapid optimization and problem-solving in complex domains like cybersecurity and logistics, providing a competitive edge in speed and accuracy.
Action items
-
Audit current business processes to identify digital tasks that are explicit, testable, and scalable for AI automation. Prioritize these workflows for immediate implementation to achieve quick wins in cost reduction and speed.
Impact: Delivers immediate operational efficiency gains, freeing up resources for strategic initiatives and improving overall organizational agility.
-
Conduct a strategic assessment to determine if AI addresses the core point of leverage in your business model. Evaluate whether the technology unlocks new markets or changes the fundamental value proposition, rather than just improving existing tools.
Impact: Ensures that AI investments are aligned with long-term strategic goals, avoiding wasted resources on incremental improvements that do not drive significant competitive advantage.
-
Develop a framework to distinguish between tasks where average answers are sufficient and those requiring novel, non-derivative insights. Allocate AI resources accordingly, using it for scaling standard processes and human creativity for innovation.
Impact: Optimizes the use of AI capabilities, ensuring that the technology is applied where it is most effective and that human expertise is leveraged for unique value creation.
-
Implement AI-driven agent systems for complex problem-solving tasks where brute-force testing is viable, such as cybersecurity simulations or logistics optimization. Leverage the speed and scale of AI to test multiple combinations rapidly.
Impact: Enhances the ability to solve complex problems quickly and accurately, providing a competitive edge in domains where speed and thoroughness are critical.
-
Maintain strategic flexibility in AI adoption plans, recognizing that the true impact of the technology may emerge through unexpected applications. Avoid over-committing to specific use cases and remain adaptable to new opportunities and challenges.
Impact: Positions the organization to capitalize on emerging AI capabilities and market shifts, ensuring long-term relevance and competitiveness in a rapidly evolving technological landscape.
Quotes
“LLMs are going to be able to do anything, at a minimum, anything digital where you have enough training data and can validate the response at scale.”
“Is the AI or internet or the PC getting at the key point of leverage in your business?”
“These things are telling us what most people would probably say.”