From the course: Using AI in Research Projects

Doing research on AI when it's not AI research - ChatGPT Tutorial

From the course: Using AI in Research Projects

Doing research on AI when it's not AI research

- [Instructor] There are three ways that you are likely to be involved with AI in your future research. AI may show up even when you're not doing AI specific research, or you may be researching an AI-driven system likely focusing on user-driven inputs and AI generated outputs. Finally, you may be using AI as a tool to help you in your work. Let's start with the first and most basic of these AI interactions, tangential AI research. The mass adoption of generative AI, such as ChatGPT is new. However, you very well may have encountered AI, at least more rudimentary AI, in the research studies that you've done over the past few years, and may have found yourself writing up AI specific findings in your research report. For example, I was doing a web usability study focused on the particular web application. The developers had inserted a half-baked AI bot into the help section and provided it with a rudimentary set of AI learning. When participants got stuck and looked for assistance, they entered a natural language query, only to get a disjointed answer that made no sense. So while not an AI focused study, I ended up writing recommendations about the poorly constructed AI. Whether or not you intended these AI elements as part of your research, they will likely impact your findings, be prepared for this, talk to your stakeholders in advance as best you can anticipate what AI interactions users might have. Be prepared to probe and better understand the participants who don't like AI, who don't want recommendations or chat bots, and who may want to talk to a real person, not a bot, and be prepared to explore user frustration and AI bias that confuses or perplexes your participants in ways that you didn't anticipate when you prepared for the study. Simultaneously, even if not the main event of your research effort, probe about what participants like about the AI or in this ever involving technology, what participants imagine could and should happen when they interact with the AI. Ultimately, though, we've only talked so far about exploring AI when AI is not the main event but what if it is? What if our research is entirely focused on the encounter with AI? Let's talk about that next.

Contents