AI Changes the Way the Job is Done. You Need To Change the Way You Design for It.

Digital design mastery used to require a high tolerance for technical friction.

Being a pro meant knowing how to manipulate pixels, manage layers, and execute complex workflows. It often meant spending hours on tasks like masking hair or cleaning data—tedious work that was the price of entry for creative output.

AI changes the equation. It automates the technical execution. Now, the mastery isn't in how you move the pixels. It's in what you tell the pixels to do. Now, you type “remove background.” Or “expand landscape.”

Twenty-minutes of work becomes a one-second task. But something bigger happened than just increased speed. The nature of the underlying Job itself changed. The user stopped being a pixel pusher who manipulates complex tools and became a Creative Director who evaluates outcomes.

It’s easy to miss this significant shift. You see the old task and try to make it faster. But the winning move isn’t simply removing steps. It’s understanding what new role the user is trying to play — who they want to be and how they want to feel — once those steps disappear.

The real opportunity of the AI gold rush isn’t to discover how users currently work and accelerate it. It’s to design for a different Job entirely.

This article shows you how AI changes the Job and what that means for how you design the experience. Every article in this category helps you apply JTBD thinking to AI products: how to recognize how AI has changed the user's role, how to design for delegation instead of manipulation, and how to build the kind of trust that earns repeated use and loyalty.

The faster horses trap

If you interview a junior data analyst about their struggle, they’ll tell you: “I spend too much time writing SQL joins and debugging syntax errors.”

Standard JTBD discovery says: Great. Let’s build an AI that auto-completes SQL and fixes syntax. That’s a helpful tool. But it’s a lack of imagination.

Because “writing SQL” isn’t the job. It’s the tax the user pays to get to the job.

The actual job is: Find the signal in the noise. If you design for the new job, you don’t build a SQL helper. You build a system that says: “I ran 50 queries on the new dataset. 49 were normal. One showed a 40% spike in churn for users on iOS 16. Here is the chart.”

You elevate the analyst from “Query Writer” to “Insight Auditor.”

You remove the Gulf of Execution—the gap between “I want to know what’s wrong” and “I have to type code to find out.”

The shift: direct manipulation vs. delegated intent

Traditional software design is based on direct manipulation, You drag the file. You click the bold button. You draw the line. You type the formula. And so on.  The user provides the labor. The computer provides the canvas.

AI introduces Delegated Intent.

• “Make this sound more professional.”
• “Fix the lighting.”
• “Find the anomaly.”

The user provides the goal. The computer provides the labor.

When you move from manipulation to delegation, you’re changing the user’s role description. And if you don’t explicitly design that new role, the user feels lost. They feel like they’ve been replaced, rather than promoted.

Design for the new role

Writing & Code

The Old Role: The Drafter. Staring at a blank page. Typing every word. Worrying about grammar and syntax.

The New Role:The Editor-in-Chief.

If your AI generates a blog post or a code block, the user’s job is no longer “creation.” It is judgment.

Design for the Editor:

• Don’t just dump text. Highlight the choices made.
• “I used a more aggressive tone here—keep it?”
• “I refactored this function to reduce complexity—review the logic?”
• Give controls for shaping, not just typing. (Toggles for tone, length, complexity).

The user needs to feel like they’re directing the orchestra, not just listening to a recording.

Data & Analytics

The Old Role: The Miner. Digging through rows. Building pivot tables. Hoping to stumble on a pattern.

The New Role: The Investigator. The AI does the mining. It presents the “suspects.” The user interrogates the findings.

Design for the Investigator:

• “Here are three trends that break the pattern.”
• “Why?” (Click to drill down).
• “Verify.” (Show the raw data source).

The user provides the context (“We ran a promo last week, that explains the spike”) that the AI lacks.

Design & Media

The Old Role: The Builder. Constructing layers. Masking objects. Adjusting curves.

The New Role: The Curator. The AI generates four variations of the image. The user chooses the best one and refines it.

Design for the Curator:

• Rapid comparison tools.
• “More like this / Less like this.”
• In-painting to fix specific details without redrawing the whole.

The skill shifts from “hand-eye coordination” to “taste.”

The “Gulf of Execution” is gone. Now mind the “Gulf of Evaluation.”

Don Norman famously described the Gulf of Execution: the difficulty of figuring out how to get the system to do what you want. AI collapses that gulf. You just say what you want.

But it widens the Gulf of Evaluation: the difficulty of figuring out if the system did it right.

If the AI writes 1,000 lines of code in a second, the Gulf of Execution is zero. But if the user has to spend four hours reading that code to ensure it’s secure, the Gulf of Evaluation is massive. Designing the new job means building bridges across that second gulf.

• Legibility: Make the AI’s work easy to scan.
• Traceability: Show where the data came from.
• Confidence: Flag the parts the AI is unsure about.

If you automate the doing but complicate the checking, you haven’t saved the user time.

You’ve just changed their anxiety.

Don’t ask: “How do I make the current task easier?” Ask: “If the cost of execution drops to near zero, what is the human actually here to do?” Design for that role.