Exploring Human Agency in the Age of AI: How to Become a Better Collaborator with AI
This article is based on a transcript featuring Jeremy Utley, an adjunct professor of creativity and AI at Stanford University. He focuses on helping non-technical professionals collaborate effectively with generative AI.
The Winston Churchill Analogy: AI as Your Personal Assistant
Jeremy Utley begins by expressing his admiration for Winston Churchill and shares an anecdote about Churchill dictating a national address from his bathtub. He imagines having an assistant who intimately understands his context, voice, and intent, allowing him to dictate speeches even while bathing. Utley asserts that, thanks to AI, even the "poorest villager in Palo Alto" can now access a tool that provides similar contextual awareness and assistance, making it technically possible to dictate an address from the bathtub today.
Jeremy Utley's Background and Expertise
Jeremy Utley has been teaching at Stanford for 15 years, focusing on the intersection of creativity, innovation, entrepreneurship, and, increasingly, artificial intelligence. He co-authored "Idea Flow," a book on idea generation and prototyping. The release of ChatGPT shortly after his book prompted him to immerse himself in studying AI, focusing on how it impacts problem-solving abilities at the individual, team, and organizational levels.
Chapter 1: Don't Ask AI, Let It Ask You
A key principle is to leverage AI's ability to evaluate its own work. Instead of directly asking AI a question, consider asking it how to best ask the question. Utley suggests prompting the AI with a request like: "As an AI expert, please ask me questions to understand my workflows, responsibilities, KPIs, and objectives. Then, provide two obvious and two non-obvious recommendations for how I can leverage AI in my work." This approach fosters an enlightening conversation driven by AI's self-assessment capabilities.
Real-World Impact: The National Park Service Example
Utley shares a success story from a training program he conducted for National Park Service backcountry rangers. Adam Rymer, a ranger at Glen Canyon National Park, used AI to automate the paperwork required for replacing carpet tiles in the lodge. He created a tool that reduced the process from 2-3 days to just 45 minutes. This tool was shared across the park service, and it is estimated to save 7,000 days of human labor annually. This highlights the significant impact that non-technical professionals can achieve with basic AI training.
Chapter 2: Do Not Use AI, Treat It as a Teammate
Many organizations are eager to transform their businesses with AI, but they lack basic understanding of how to collaborate effectively with it. Research suggests that while AI can improve speed, output, and quality, a significant percentage of professionals are not experiencing meaningful productivity gains. This "realization gap" stems from how people approach AI.
Utley's research identified that outperformers treat AI as a teammate, not just a tool. This shift in orientation significantly impacts outcomes.
-
Tool vs. Teammate: If AI delivers mediocre results as a tool, it might be dismissed. As a teammate, you provide feedback, coaching, and mentorship to improve its performance.
-
Question Asking: Encourage AI to ask questions to gather necessary context for providing optimal responses.
-
Leveraging AI for Drills: Explore unconventional uses, such as roleplaying difficult conversations with coworkers. AI can interview you about your conversation partner, develop a psychological profile, play the role of your partner, and provide feedback.
Utley emphasizes that his students constantly discover new AI applications he couldn't have imagined, leading to unexpected and beneficial outcomes.
Unlocking Creativity: Inspiration as a Discipline
Utley believes everyone possesses innate creative capacity. Inspiration is not just a fleeting moment but a discipline. Creative individuals actively cultivate inputs to their thinking, recognizing the impact on their outputs. When using AI, the differential output comes from what the user brings to the model, including experience, perspective, and accumulated inspiration.
Chapter 3: How to Go Beyond ‘Good Enough’ Ideas
Creativity, in the age of AI, remains about doing more than the first thing you think of. A seventh grader defined creativity as "doing more than the first thing you think of." This definition resonates because it highlights a common cognitive bias: settling for "good enough". While AI makes achieving "good enough" easier, world-class results require prompting for volume and variation, which takes time and effort to process.
The definition of creativity doesn't change with AI, but human abilities are affected by the technology and their collaboration objectives. Creators should embrace AI to unlock unprecedented potential. Ultimately, Utley advocates working with AI, rather than merely using it, to unlock its transformative power.
{#