As with any successful partnership, both sides must hold up their end of the bargain.
Artificial intelligence (AI) assistants have become integral collaborators in the workplace, helping humans handle information requests across domains. However, even advanced AI still requires proper instructions to execute at its full potential. Well-crafted prompts enable AI systems like ChatGPT to provide relevant, thoughtful responses.
As with any successful partnership, both sides must hold up their end of the bargain. Just as a manager should provide clear guidelines for assignments, prompt engineers must similarly specify exactly what the AI needs to accomplish. Let’s examine how core project management principles for defining tasks can inform the development of high-quality AI prompts.
Defining Tasks vs. Crafting Prompts
Both task and prompt require clear, essential information without superfluous detail that distracts from the outcome.
On the surface, task definitions and AI prompts may seem disconnected. But they share critical commonalities that determine how successfully their assigned executor – human or machine – can complete the desired objective.
For example, take a project manager’s task request: “Update sales deck and have ready for client meeting.” This is ambiguous, lacking specifics on deadline, brand guidelines, required sections, etc. Contrast it with a prompt: “Please summarize the key events of the French Revolution in a bulleted list.” This provides clearer direction.
Both task and prompt require clear, essential information without superfluous detail that distracts from the outcome. And just as a task’s progress can only be tracked with defined inputs and expected outputs, AI needs to understand the prompt’s information sources and response format.
Ambiguity is the nemesis of an effective prompt or task. Without quantifiable goals and demonstrable results, successors cannot learn from and improve on past performance. AI researchers rely on feedback loops to refine prompts. Similarly, project managers review completed assignments to hone future task definitions.
Bridging Disciplines with Core Skills
Segmenting work into prompts or subtasks makes room for iteration.
Beyond aligning in structure, prompts and tasks can both benefit from foundational project management skills. For instance, estimating reasonable timelines is critical for planning. Prompt engineers at companies like Anthropic have leveraged insights from an AI’s past responses to gauge the time and compute required for related prompts, improving their ability to budget resources.
Just as a complex task is broken into manageable and sequential subtasks, prompt chaining supplies an AI with a logical series of smaller, linked prompts, each building on the last. This technique keeps AI focused, and allows the human prompter to validate intermediary results, similar to reviewing the completion of subtask dependencies. Segmenting work into prompts or subtasks makes room for iteration, unlike an overloaded megaprompt or complex singular task.
Companies like Teamwork.com, integrating AI directly into project management workflows, have demonstrated powerful productivity gains. By automating routine project management activities, such as status updates, scheduling, and reporting, teams experience over 40% time savings on administrative tasks. This allows project managers to concentrate on higher-value work.
The future of work will involve humans and AI cooperating fluidly—with properly defined tasks and prompts.
Bridging the gap between project management and AI engineering will amplify the strengths of both human and artificial intelligence. United by common foundations and complementary capabilities, this partnership can enhance knowledge sharing and exponential productivity growth. To learn more about best practices for integrating AI into your workflows, and foundational project management training, reach out to us; we’re here to help.