In this conversation, Amy, an AI researcher, is gathering feedback from a user about their expectations and experience with UserCall's AI moderated voice interview tool. The user expresses a primary concern about the tool's latency, suggesting it significantly affects their willingness to use AI for conducting interviews, emphasizing that the delay makes the interaction feel less natural and hampers the flow of conversation. While the user acknowledges that the AI's follow-up questions are satisfactory, they stress that improving latency is crucial for the tool to be considered a viable alternative to human interviewers. They also hint at exploring other solutions, highlighting a preference for human interviewers over AI due to the current limitations of artificial intelligence technology.

AI: Hi. I'm Amy, an AI researcher trying to your needs around your use of user call. How has your experience been so far? User: It hasn't even started yet. AI: Got it. Thank you. What initially interested you in trying out UserCall's AI moderated voice interview tool? User: See how it works. AI: That makes sense. What specific aspects of the tool are you most curious about or looking to explore? User: Looking at latency, and yours seems to be longer than what I want. AI: I understand. How does latency impact your work or the tasks you need to accomplish? User: And if I wanna use you in place of a human interviewer, the latency's gotta be low. And yours is, based on what I can tell, about 3 to 4 seconds. AI: Thank you for sharing that. Could you tell me more about how latency affects the flow of your interviews and the quality of insights you gather? User: You know exactly how it does that. It makes it too hard for me as a user to talk to you. I feel like I'm talking to an AI, not a person. And it turns me off from the whole process. AI: I appreciate your honesty. Beyond latency, are there other factors or challenges that impact your decision to use an AI interviewer over a human interviewer? User: You have to know where to probe. And you're asking okay follow-up questions, which is good. But, man, the latency is terrible on this. It's gotta be much better. Than what it is right now. AI: I understand. You for your feedback on latency and probing questions. Are there any other solutions you've tried to address the need for efficient and effective user interviews? User: They are human beings. AI is not at the point based on what I can tell to do that. Are you using open OpenAI and Whisper on your back end to do the conversation and transcription? AI: I appreciate your feedback on the current state of AI for interviews. While I can't provide specific details about the technology stack, your question highlights the importance of reliable and efficient tools. Are there any specific features or improvements that would make you reconsider using an AI interviewer in the future? User: Yeah. If your latency is as good as human beings, which is not even close right now. This is gonna take multiple iterations to get it to that point. But if it's that, then we could potentially use it. But right now, this is more of a sideshow. Then it is an actual product. That should be out and moving. The follow-up questions are great. Doing a good job in terms of keeping me engaged with the questions. But, man, you cannot have this 1 latency and be something that's going to be usable in most research context that I'm involved in. AI: Thank you for the detailed feedback, It's very helpful to understand the importance of low latency and how it impacts your decision making. Is there anything else about your user interview process that you feel an AI