Highlights of the Week in Creative Leadership: May 9, 2025

The promise that AI will automatically and democratically boost everyone’s creative work isn’t automatically true, according to rigorous research from USC, Temple University, and Sichuan University scholars. Nan Jia, Xueming Luo, Zheng Fang, and Chengcheng Liao’s telemarketing study demonstrates that when AI conversational bots handled routine lead generation, human agents could focus on complex sales persuasion requiring creative script development. The sequential division of labor freed cognitive resources, but the benefits were starkly uneven: highly skilled agents were 2.81 times more successful at answering unexpected customer questions and reported positive emotions, while lower-skilled agents struggled with increased cognitive demands and experienced negative psychological outcomes. The research challenges the utopian vision of AI as creativity’s great equalizer. Creative leaders must recognize that AI augmentation is fundamentally skill-biased: they need to invest heavily in upskilling programs before AI implementation, create mentorship systems pairing high and low performers, design collaborative AI rollouts that gradually build capabilities rather than immediately exposing skill gaps, and measure both performance outcomes and employee wellbeing throughout AI integration processes.

Nan Jia, Xueming Luo, Zheng Fang, and Chengcheng Liao (2024) “When and How Artificial Intelligence Augments Employee Creativity,” Academy of Management Journal, Vol. 67, No. 1, 5-32.

British philosopher Galen Strawson points to a critical blind spot in leadership (and, more broadly, personal) development: the assumption that everyone must construct coherent life narratives for authentic self-knowledge. In contrast to thinkers like neurologist Oliver Sacks (‘this narrative is us’) and psychologist Jerome Bruner (‘self is a perpetually rewritten story’), Strawson celebrates figures like Montaigne, who wrote ‘the unstoried life’ with radical honesty, and artist Paul Klee’s recognition of his ‘dramatic ensemble’ of inner selves.’ This distinction between ‘Narrative types who naturally construct life stories and ‘non-Narrative’ individuals who experience existence as fragments challenges the dominant assumption that everyone experiences life as a coherent story. Among the manifold takeaways for leaders are to design team reflection processes that accommodate both story-builders and fragment-processors, allowing some to craft coherent narratives while others contribute discrete insights without forced connections; to value fragmented, present-moment awareness alongside strategic storytelling; and to resist the pressure to retrofit experiences into neat origin stories or linear career narratives that may sacrifice authenticity for imposed narrative coherence.

Galen Strawson (2015) “I Am Not A Story,” www.aeon.com, September 3, 2015. 

Researchers led by MIT FutureTech’s Peter Slattery have created the fist comprehensive AI Risk Repository exposes a sobering reality behind Artificial Intelligence’s promise, cataloging over 1,600 specific risks from discrimination in hiring algorithms to autonomous weapons development and environmental damage from energy-intensive data centers. Their groundbreaking database reveals that 41% of AI risks stem from AI systems themselves (with AI system failures constituting 26% of the risks), while 39% originate from human decisions (including 16% of risks from malicious misuse), with most emerging post-deployment rather than during development. The April 2025 update dramatically expanded the repository by 60%, adding 9 major frameworks including analysis of 200 real-world AI misuse incidents and the first comprehensive taxonomy of multi-agent AI risks – revealing how AI systems can fail through miscoordination, conflict, and collusion when interacting with each other. Notably, the research shows most existing AI risk frameworks focus narrowly on large language models, leaving other AI applications dangerously under-examined. Leaders should leverage this taxonomy to conduct comprehensive risk assessments before AI adoption, prioritize mitigation strategies across all seven domains, and establish ongoing monitoring systems and cross-functional teams that address both human and AI-generated risks throughout their innovation processes. 

Peter Slattery, Alexander K. Saeri , Emily A. C. Grundy, Jess Graham, Michael Noetel, Risto Uuk, James Dao, Soroush Pour, Stephen Casper, and Neil Thompson (2025) The AI Risk Repository: A Comprehensive Meta-Review, Database, and Taxonomy of Risks From Artificial Intelligence, March 2025; https://doi.org/10.48550/arXiv.2408.12622.

Jess Graham, Alexander Saeri, and Peter Slattery (2025) “April 2025 Update of the AI Risk Repository,” https://airisk.mit.edu.  

Behavioral scientist Michael Hallsworth’s decade-long reflection on the EAST framework (Easy, Attractive, Social, Timely) reveals why simplicity often trumps the sophistication and precision pushed by machine learning, adaptive science, and behavioral science applications. Despite advances in AI and personalization, this four-part acronym remains one of the most widely adopted tools because it embodies its own design principles – much like the BIC Cristal pen, a Museum of Modern Art design classic that succeeds through elegant functionality. Hallsworth warns against the ‘toothbrush problem’ where 79% of specialized psychological tools are used fewer than three times due to fragmentation. A key insight here is that simple tools enable sophisticated application when wielded with practical wisdom, particularly the skill of judging what to exclude. Particularly as generative AI shifts roles toward assessing rather than producing ideas, this everyday ability to curate becomes paramount. Creative leaders should resist over-engineering solutions, design tools their teams will actually adopt rather than admire from afar, and remember that the best frameworks are memorable and accessible, and – perhaps most difficult – live their own principles.

Michael Hallsworth (2025) “Why Simplicity Can be Strength in a Complex World,” Behavioral Scientist, May 4, 2025.

Bestselling author and ‘The Daily Stoic’ entrepreneur Ryan Holiday’s leadership philosophy centers on both recognizing its key roots and developing its next practitioners, as illustrated through his ‘coaching tree’ concept. Transforming American football coach Pete Carroll’s ‘Win Forever’ philosophy into actionable leadership principles grounded in operational excellence, he draws on the examples of model leaders for actionable principles. Drawing from Thomas Keller’s restaurant kitchens, Holiday champions a ‘sense of urgency’ while balancing it with the Latin maxim ‘Festina Lente’ (make haste slowly). Practicing George Raveling’s wisdom: ‘be there when they’re losing, not when they’re winning,’ he focuses support during struggles rather than celebrating successes. Implementing David Allen’s ‘don’t touch paper twice’ rule and regularly asks teams ‘What’s eating your time?’ Recalling IBM’s legendary CEO Tom Watson, he tolerates expensive mistakes but refuses repeated errors. The framework includes questioning inherited processes (‘Why is it being done that way?’) and adapting successful ideas from other industries. Creative leaders can borrow from Holiday’s approach and identify their own ‘coaching trees’ of leadership philosophy and practice, as well as consider applying specific steps like identifying team members’ aspirations, delegating completely with clear ownership, establishing ‘clock-starting’ protocols for long-term projects, and creating systematic ways to challenge existing processes before they become unproductive habits.

Ryan Holiday (2025) “These Are Leadership Ideas I Try To Apply Every Day,” www.medium.com, April 17, 2025.