Show us how you explain AI
and how you use it.

This process is designed to understand how you think, how you communicate, and how you apply AI in practice. We are interested in your curiosity about different tools and ways of working, along with your understanding of the process in the use of AI.

01

The Process

The process has four stages. Across these interviews, we want to understand your ability to learn, understand, and communicate how AI is used, along with how you apply that in practice.

Stage 1

First Interview

30 minutes

A short conversation about your background, your AI experience, and how you are using these tools today.

Stage 2

AI Tool Coaching Session

1 hour

A coaching session on Codex or Claude coding tools, tailored to how we work and what would be useful to us.

Stage 3

Show Something You Built

1 hour

A walkthrough of a tool, workflow, assistant, or system you have built using AI, with time for questions.

Stage 4

Founder Interview

In person

A final conversation with a founder in our office.


02

The Interview Journey

Stage 1

First Interview

30 minutes

This is an introductory conversation about your background, your AI experience, and the way you are currently using AI tools in your work.

Stage 2

Coach Us on a Coding AI Tool

1 hour

In the second interview, we will ask you to run a coaching session on either Codex App or CLI, or Claude Cowork or Code. You can assume we are regular ChatGPT users who are curious about what these tools could do for us.

  • 1
    Understand how we work Ask questions about our current workflow, where we use ChatGPT, and what kinds of work we are trying to improve.
  • 2
    Explain the benefit Help us understand where Codex or Claude coding tools are different from a general AI assistant.
  • 3
    Show relevant features Focus on the parts of the tool that would matter for the way we work, rather than giving a generic product tour.
  • 4
    Coach us through usage Talk us through how to start using the tool well, including how to check outputs, refine instructions, and build confidence.
Example Coaching Setup
Imagine you are speaking to regular ChatGPT users who are curious about Codex App, Codex CLI, Claude Cowork, or Claude Code. Help us understand what the tool could do for us, starting from how we work today.

After the coaching session, we will have a discussion about AI more broadly, including your current knowledge of the space, the tools you are aware of, and how you think about the subject today.

Stage 3

Bring Something You Have Made

1 hour

For the third interview, come prepared to show us something you have created using AI, or something you have made for an AI tool such as a custom GPT, skill, workflow, or assistant. Ideally, it should improve human productivity or effectiveness, and show how you think about making work better through the use of AI.

  • 1
    Choose a real example This could be a workflow, assistant, automation, lightweight tool, or system that helps you or others work more effectively.
  • 2
    Explain the problem Tell us what you were trying to improve, remove, or speed up.
  • 3
    Talk through the approach Show what tools you used, what decisions you made, and what tradeoffs mattered.
  • 4
    Reflect on what happened Tell us what worked, what did not, and what you would improve next.
  • 5
    Try a new tool Talk us through learning a new tool to help achieve this, and what that experimentation taught you.
Stage 4

Founder Interview

In person

The final stage is an in-person conversation with a founder in our office. This is a chance for us to understand you more broadly and for you to ask questions about the role, the team, and how we work.


03

Good to Know

Expected prep time is a few hours. Across the interviews, we want to see your capability to communicate clearly and show your skills and experience in AI.

  • Show your understanding of the tool For the second interview coaching session, the focus is on demonstrating your understanding of the tool you choose to show.
  • Coach us through it live We would like to see the tool in use, understand where things could be improved or where issues were faced, and hear about the parts that are working well.
  • Talk us through your choices We will be interested in why you approached the work the way you did, not only in the final output.