Delegation refers to what work you as human would do and what to delegate to AI models to be done which involves understanding the problem we are working on and what capabilities the AI tools have.

Effective Delegation involves breaking down complex work into smaller parts into mangeable pieces and making strategic decision who does what either through automation, augmentation or agency.

The corner stone of effectiveness of delegation does not depend on AI, but on human. Before we provide any work to AI we must ourselves evaluate -

1. What success look like or what exact output we want? This forms the basis of delegation.

2. What kind of thinking and work is needed to get to success.

3. AI fluency depends on human expertise.

Problem Statement: Send an email to your colleague explaining them about new HR application onboarded in your organization.

Problem Awareness

It means clearly understanding the goals and the nature of work before involving AI.

In this example the end goal is to educate our colleagues about the new platform through an email.

Platform Awareness

Understanding the AI landscape which is evolving daily. Understand the uniqueness and strength of each model, understand the speed, context and accuracy of the model to decide which model to use.

In this example we have to understand which model can generate better email from given capabilities of the new platform and summarize it better.

Task Delegation

The strategic process of dividing work between humans and AI.

• What could be automated

• Where would augmentation create more value than working separately.

• What should be done by human

• What could be done by agent on your behalf

Here we can ask the AI to generate an email on our behalf while we provide the capability context to the model, once generated we should refine the email if necessary and we should send the email to the required participants.