Lean and AI
Implementing AI effectively goes beyond deploying sophisticated models, it's about learning swiftly and delivering measurable value. Lean experimentation provides a structured approach to achieve this.
A notable example is Johnson & Johnson, which recently recalibrated its generative AI strategy. Initially, the company supported nearly 900 AI use cases across various functions. However, upon evaluation, they discovered that just 10-15% of these initiatives accounted for 80% of the value. This insight led to a strategic shift towards high-impact areas such as drug discovery, supply chain optimization, and internal support tools like the "Rep Copilot", an AI-driven assistant aiding sales representatives in engaging healthcare professionals.
The company has shifted its generative AI strategy from broad experimentation to a focused approach, prioritizing only the highest-value use cases while cutting projects that are redundant, ineffective, or better served by other technologies.
This shift highlights a critical lesson: experimentation alone is not enough. Success comes from focusing on initiatives that generate real impact, learning rapidly from each iteration, and strategically applying AI where it can truly transform outcomes. It's a pragmatic approach that balances innovation with measurable value.
Attention and Intention
Phantom Obligation
Hero
Shadow AI