18: Serendippidus

In the last email (subject: Ducks in a row), we acknowledged that for organisations to leverage AI, they need to understand their work. Hence, work design is becoming a more valuable skill. 

So, how do we “design work”?

Well, I want to start by highlighting something important:

The requirements for reliable performance are the same whether we’re designing work for humans or AI.

Isn’t that reassuring? Because everything we’ve ever learned about performance improvement suddenly becomes even more valuable. 

Hence my optimism for those in L&D who have been building expertise in work design. Regardless of whether or not we’re AI experts (and I’m definitely not), we’re experts at the thing that makes AI effective.

Talk about right place, right time?!

So, let’s get back to the question - how do we design work?

Well, before we discuss that, let’s make sure we’re speaking the same language. Because there isn’t a globally recognised taxonomy for this stuff. 

So, when it comes to the activities within a workflow, I’ll primarily use the word task. Meaning the individual things we do to achieve an outcome. 

Now, if we were to get more precise, a task is actually only one type of workflow activity - others include decisions, handoffs and delays. But we don’t always consider these because we do them without thinking. So, for the purpose of this conversation, let’s focus on tasks as the primary type of activity. 

In my earlier email (subject: ​Guidance Inc.​), I used Making a video as the workflow example.

Tasks: 

1. Brainstorm video ideas

2. Research high-ranking titles

3. Write script

4. Design storyboard

5. Set up camera and lighting

6. Record video

7. Edit video

8. Write description

9. Publish to YouTube

10. Share on socials

Now, to effectively define the work, we need to break each task down further. Why? Because if I asked a new starter to brainstorm video ideas without any further instruction, they’d struggle. 

So, let’s break our Brainstorm video ideas task down into steps.

Steps:

1. Use the ‘Questionstorming’ technique to consider pain points

2. Search YouTube for high ranking videos on similar topics

3. Use Google Trends to detect rising interest and keyword options

But is that enough granularity for someone else to execute? Probably not. So, let’s break down our Search YouTube for high ranking videos step into actions.

Actions:

1. Enter relevant search terms

2. Review top results

3. Identify standout videos

4. List recurring patterns

5. Save the strongest ideas

Once we’ve gotten to this level of detail, we can consider what we need to support someone (or AI) in doing the work. 

And to do this, it can be helpful to map the requirements against four categories - context, instructions, guardrails and blueprints/templates. These categories translate the core principles of performance design into something practical i.e, aligning the work, guiding the action, defining the boundaries, and providing a structure for consistent execution.

So, let’s consider what that might include for Identify standout videos action.

1. Context

• Video ideas should align with audience interests

• Standout videos should outperform channel’s typical content

2. Instructions

• Review top search results

• Compare views with recent channel performance

• Identify unusually strong videos

• Record strongest ideas

3. Guardrails

• Videos must be relevant to audience

• Performance must be judged relative to channel

• Ideas should be adapted rather than copied

4. Blueprints and templates

• Table showing title, channel, views, typical views and underlying idea

When we start to put this level of structure around each action, we begin to see what’s needed to design the work. 

Not only that, but writing clear instructions forms the basis of AI prompts. So, if the tasks in question could be done by AI, we’ve actually started prompt engineering!

How cool is that?

But I can already hear the objections - how can we do this for every action?

We’ll tackle that next.

Yours,
- Ant