Natural Language Tasks for Brainstorming

Superpower to Gain

With Natural Language Tasks for Brainstorming, you will gain an invaluable creative partner – always quick with a few ideas to improve on, always happy to write the first draft, zero ego.

Target Output

The goal for this module is to consider a variety of solutions to the problems that you identified in the last module. The output you should end with is a single solution statement (it should fit in a tweet). While you will use GPT-3 to help stimulate different ideas, ultimately the solution description is something that you will want to edit and refine yourself.

Steps to Take

Pick an Ambition statement and Problem statement pair that you want to work on further. (Note: you can repeat this step multiple times to think about different Problem statements).
Enter the Ambition and Problem statement into the GPT-3 Playground, in the format of:
➡️
For [ insert stakeholder description ], who want to help me achieve my goal of [ insert your ambition ], it sucks that… [ insert problem statement ] The solution is:
Probe at the responses that you receive from GPT-3. Ask questions about “how could that work?” and “I like the idea of… can you expand on that?” Treat GPT-3 like you might any other person that you engage with creatively. If it seems that the conversation is heading down the wrong path, click the rewind arrow which is just to the right of the green “Submit” button in order to roll back the last response(s) from GPT-3.

Your Solution

Jot down somewhere your solution description (<280 characters). Start with “A service that…” or “An app that…” or the equivalent.

Next Module

Image Generation

Intro Video