As the quality of the input will largely determine the quality of the output, it would be wise to feed it data that has already been cleaned, analyzed, and prioritized. Let’s see how our current exploration and analysis might be reflected in a problem statement.
Based on the insights GPT has generated so far, we will ask GPT to create a problem statement for us. We will try to be as specific as possible when asking GPT to write a problem statement, also stating the objectives or even the format we are looking for.
“Based on our discussion and the insights that we get so far about the comfort of VR in daily use, considering the VR gym application will be used daily by the targeted user, the comfort and immersion that is experienced by the user is one of our goals in developing the VR gym application. Can you consolidate everything into a single problem statement that will serve as the design goal we set to solve? as we continue with the ux process.”
The result:
Â
This response is fairly good, as we can see, but it is too long. So, we will attempt to make it shorter by giving GPT instructions to do so. You can very quickly ask GPT:
“Can you shorten this to make it easier to socialize and grasp by other members of the product team?”
The result:
Â
As we can see and feel, the problem statement that GPT came up with is fairly generic and broad, so we would need to make significant revisions or even rewrite it entirely. This is where critical thinking comes in. We cannot count on it to provide us with useful guidance during the design phase. But so far, the process has been enjoyable, and the goal of this adventure is to comprehend the potential and limitations of AI tools.
We will now formulate how-might-we-use and user stories to investigate how we can transform the problem statement into more concrete design questions and action points. The prompt is pretty simple:
“Based on the problem statement, can you generate 5 how might we's?”
The result:
For a response produced by AI, these are pretty good answers. These how-might-we questions can be used as a springboard for coming up with ideas for solutions that deal with our future users. Let us take a look at the potential user stories. It can be as simple as instructing GPT to transform these into user stories.
“Please transform this into user story.”
The result:
It appears to be performing pretty well. The next step would be to add these as post-its to a FigJam or Miro board, which would make it simpler to move them around, prioritize them by sentiment or theme, and above all socialize within your team and business. Here is a summary of what else AI can accomplish with the research report.
Keep in mind that design involves collaboration and discussion. Even if AI can serve as a partner in co-creating design, certainly with real people, the product will fly to the moon. 🚀
![[Define with AI] Problem-Solving: Define the Problem Statement Using ChatGPT](/_next/image?url=https%3A%2F%2Fwww.notion.so%2Fimage%2Fhttps%253A%252F%252Fwww.notion.so%252Fimages%252Fpage-cover%252Fgradients_2.png%3Ftable%3Dblock%26id%3D18b20fbb-a154-803a-9108-e424b8ba6d8e%26cache%3Dv2&w=3840&q=75)
