Barry O'Reilly argues that 85% of GenAI projects and 83% of transformations fail because they prioritize adopting specific tools over fundamentally changing personal and team behaviors to leverage the technology effectively.
LLMs are designed to be agreeable and can confidently hallucinate. To counter this, prompt the AI to find blind spots, generate counterarguments, or role-play a skeptical stakeholder. This strengthens your own thinking and protects the critical human skill of judgment.
Increased efficiency from AI should not automatically be filled with more tasks. Instead, this newfound capacity should be intentionally allocated to "thinking time"—marinating on hard problems. This slow, System 2 thinking is crucial for leadership and judgment.
O'Reilly failed to write a book by typing but succeeded by using AI transcription to capture his natural talent for talking. The key to personal AI adoption is matching the tool to your innate strengths, not forcing yourself into a workflow that feels unnatural.
Instead of manual note-taking, use AI tools to transcribe and summarize all meetings. This creates a unique, searchable knowledge base from your conversations, which can be leveraged to improve preparation, follow-ups, and decision-making over time.
O'Reilly defines judgment as a mix of knowledge (gained from learning), experience (gained through action with consequences), and the ability to make choices with incomplete data. Hiring processes should test for all three components, not just a candidate's stated experience.
Instead of fixating on lagging indicators like money saved, track leading indicators that signal behavioral shifts. For example, asking teams to rate their meeting preparedness on a 1-10 scale measures the effectiveness of AI-driven prep and predicts future performance gains.
Research from Harvard and P&G shows that while an AI-assisted individual can reach parity with a non-AI team, an AI-assisted team achieves a 3x improvement in ideation quality. This proves the compounding value of collaborative AI use over siloed, individual efforts.
Successful AI adoption requires the C-suite to change their own work habits. When a CEO like Progeny's Pete Aviansky openly shares his process, struggles, and successes with AI, it creates the psychological safety necessary for teams to experiment and adopt new behaviors.