AI as Devil’s Advocate

Using ChatGPT to Create a ‘Team of Rivals’

#ai #chatgpt #llm

Historian Doris Kearns Goodwin popularized the term "Team of Rivals" in her 2005 book of the same name1. It refers to President Abraham Lincoln's strategy of appointing a diverse group of individuals to his cabinet, the most prominent of whom were his rivals during the 1860 presidential election. Despite their initial animosities towards Lincoln and each other, these individuals accepted Lincoln's invitation to serve in his cabinet.

The brilliance of this strategy lies in its multifaceted nature. First, by bringing his rivals into his cabinet, Lincoln neutralized potential sources of opposition. Second, he was able to benefit from a wide range of perspectives and approaches to policy, which was particularly valuable in dealing with the unprecedented challenges of the Civil War.

Lincoln demonstrated exceptional emotional intelligence and political acumen in managing this team. He was patient, inclusive, and willing to listen to differing viewpoints. He wasn't afraid of disagreement and welcomed debate in his cabinet meetings. This approach allowed him to leverage the collective intelligence of his team to make more balanced and informed decisions.

In essence, the "Team of Rivals" concept is a testament to Lincoln's leadership philosophy, which prioritized the good of the nation over personal ambition or rivalry. His ability to manage and harness the strengths of this diverse team played a crucial role in guiding the United States through the Civil War.

Sadly, implementing a "Team of Rivals" approach, like Abraham Lincoln did, is extremely rare, and can be quite challenging for reasons mostly having to do with being human:

** Ego and Competition**: Surrounding oneself with rivals requires a secure leader who is not threatened by other strong, opposing voices. Most people feel uncomfortable with this level of open disagreement or challenge.

Conflict Management: If not properly managed, team conflict could lead to paralysis and damaging internal strife. Lincoln's unique skills as a mediator and his emotional intelligence allowed him to manage these conflicts, but how many of us possess this exceptionally rare combination of qualities?

Cohesion and Consistency: A team of rivals will struggle mightily to present a united front, which can be confusing to our customers, employees, stockholders, and other stakeholders.

Trust Issues: Trust building is an important aspect of any team. In a team of rivals, individuals might be hesitant to fully trust each other; with so many rival agendas at play can the leader truly trust the motivations of those giving advice?

Effort and Time: It takes significant energy and time to manage a team of rivals effectively. Even for Lincoln, petty jealousies between the rivals on his team required enormous amounts of energy to mediate.

These factors make the "Team of Rivals" approach a complex and risky leadership strategy. It requires a strong leader with exceptional interpersonal skills, conflict management abilities, and a willingness to take on the considerable challenges of this kind of team. These requirements may explain why this approach is not widely adopted, despite the compelling example set by Lincoln. Few of us have been given or developed all of these necessary skills.

But what if you could get a lot of the benefits of such a team without the drama, conflict, and energy-sapping ego management? What if you could get these perspectives without worrying about motives, hidden agendas, or even the rivalries themselves?

Using current Large Language Models like ChatGPT to challenge your own ideas has the potential to deliver at least some of the benefits of having rivals criticize your work, while mitigating issues related to ego and competition in decision-making processes.

Neutrality: AI doesn't have personal ambitions (that we know of!) and this ego-less perspective can help us accept challenges to our ideas without triggering our egos or competitive reactions.

Emotion-free Feedback: AI can provide feedback without emotional charge, reducing the likelihood of us having defensive reactions so common when receiving human-to-human feedback.

Consistency: AI can consistently apply a set of principles or criteria when evaluating or challenging ideas, avoiding human biases that can sometimes creep into decision-making processes.

Data-Driven Insights: AI can analyze large amounts of data quickly and accurately to inform its challenges or suggestions, potentially revealing insights that a human team might miss.

Now that we have powerful AI tools at our disposal, perhaps we can use them to form our own, personal “Team of Rivals” -- without the annoying baggage that dealing with pesky humans can bring.

Crafting a prompt to invite challenge from ChatGPT involves clearly stating your idea and then directly asking the AI to critique, question, or provide alternative viewpoints. Here are some examples of simple prompts you could use to demonstrate the idea:

Direct Challenge: "I believe that renewable energy is the only solution to climate change. What are some counterarguments to this idea?"

Request for Critique: "I propose that all education should be carried out online. Can you critique this proposal?"

Playing Devil's Advocate: "My position is that artificial intelligence will replace all human jobs in the future. Could you play devil's advocate and argue against this?"

Exploring Consequences: "I argue that we should abolish all forms of taxation. What could be some potential negative consequences of this?"

Soliciting Alternatives: "My plan is to solve traffic congestion by building more roads. Can you suggest an alternative solution and explain why it might be better?"