[Empathy with AI] Preparing Interview Session Using ChatGPT

In the previous story, we tried to incorporate AI to find problems, competitors, and target users for our future products. The results are pretty good, but we must be mindful of AI's limitations, so fact-checking and cross-referencing are mandatory steps.

At the next stage, we can use AI tools to simulate a conversation with a made-up persona. We can start with these to map out the main themes, but keep in mind that we will probably get very generic and stereotypical data.

Hopefully, research tools will advance to accept real past research and then utilize it to generate more relevant, realistic, and meaningful insights for us. Until then, with the right mindset and a deep awareness of the limitations, we can still pick ideas worth digging further based on a fictional interview. For that, we’re going to ask GPT to generate an interview script and the responses, and from now on, we can conduct an actual interview session.

We will see how that turns out. We can just ask GPT to create the interview script.

“Can you provide me with an interview script that would reveal the factors that make MetaQuest 3 appropriate for everyday usage as a virtual gym application?”

The result:

We can see the AI is performing fairly well. It is important to remember that real users cannot be replaced in the process and that GPT interviews will only provide you with largely generic and stereotypical data.

Let us observe how the interview may run.

“Based on the interview script you provided, can you change the script into an actual interview conversation with potential users of the VR gym application?”

The result:

The majority of these generated interviews lack the nuance and detail that would be present in an interview with a real person because the results are based on demographic data, statistical interpretations, and other sources. Also, the conversation does not match the script that was provided earlier.

This is why it is crucial to always involve actual users in the process, and its good to even use AI-generated data to supplement the actual data from interviews. AI-generated insights will not be able to fully replace human dialogue yet. For now, consider them as hints or simply interesting directions.

Tips

Here are a few tips to make the research still sharp when collaborating AI with our research workflow, as stated in this article

Ideally, have a human research expert review your final list of ideas. If you’re an expert, that could be you. If you’re new to research, contact a more experienced researcher for guidance.

When asking AI tools to complete research-related documentation, provide them with a template as a starting point.

Watch out for mistakes in how the system completes the documentation. For example, double-check that the correct data collection permissions are outlined in your consent form.

Consider using an AI assistant for notetaking during interviews, especially if you are a UX team of one.

However, live notetaking isn’t necessary if you’re using an analysis tool that provides transcription

Double-check summaries: AI-powered transcription and summarization features can misunderstand context.

Watch out for missing summaries for sections of the session or interview.

To get better results at this stage, provide context. Tools like Dovetail allow you to provide your research questions, which are critical for the system to return codes of any value at all.

Make sure your notes, stickies, or highlights are complete enough that the feature can make sense of them

References:

Previous story about empathy stage of design thinking, https://medium.com/@ahsfalod/empathy-through-ai-237a5a160cee

Incorporating AI in preparing interview session https://www.nngroup.com/articles/research-with-ai/

🤓

⛳ If you found this helpful, please: Follow me on Twitter | LinkedIn | Instagram | Medium