AI to smash decision-making overconfidence in project management

AI to smash decision-making overconfidence in project management

Successful project management is the art to take rational decisions in any situation.

Across last decades, multiple methodologies have tackled decision-related topics to optimise critical thinking, decision-making processes and project constraints. Examples are provided in ways to manage meticulous plannings in predictive approaches and continuous feedback loops in iterative ones. However, regardless of the chosen methodology, reality shows that in nonlinear or stressful circumstances, decision accuracy of project teams can be compromised by disturbing factors like unexpected challenges, approaching deadlines and conflicting priorities. These contingencies pave the way to misunderstanding and interpersonal conflicts.

On the other hand, what can be conceived as good decision can belong to a negative non-discordant culture where individuals deny themselves to speak out. In these circumstances, uncriticised groupthink, pursued group harmony and confirmation bias prompt to drive away disagreements at the expense of disrupting ideas and potential innovations.


Embrace the dissent

The two opposite scenarios converge in showing that human rational analysis needs to be counterbalanced by a neutral and efficient devil’s advocate. Having at least one disagreeing opinion in projects, major benefits arise like:

  • Improved quality of decisions with alternative perspectives
  • Stimulated creativity and innovation through new considerations
  • Fostered learning experience thanks to constructive debates

However, cultural codes can hamper the implementation of an effective devil's advocate. Challenging the status quo in a respectful and productive manner results in a hard task, and not anyone is comfortable with that.


Make AI the go-to devil's advocate

AI is extensively employed to generate new concepts. Rooted in large languages models and impressive databases of information, it can provide fresh points of view and situational awareness. On top of that, AI isn’t affected by human sentiments and can bring its ideas without being afraid for judgments or reputation.

Ethically used, it applies deterministic and objective rationality. Detached by hidden agenda, AI is better placed to stress project team purposes, come up with concrete arguments, stimulate alternative scenarios, and reveal biases of overweighted positive propositions. Moreover, and in supporting decision-making processes, it can consult and retain complex documents not always manageable by common biological brains without reveal any sign of fatigue.

In conclusion, AI can positively impact projects. To implement that, it’s necessary to keep in mind that the best benefit of involving the machine in human processes is to avoid errors. Project teams must keep the control on what decisions must be pursued or left aside, taking responsibilities for human and technological capacities and limitations.

To view or add a comment, sign in

More articles by Gianluca Odoardi, Ph.D.

Insights from the community

Others also viewed

Explore topics