I saw this on Reddit and decided to check it out. A lot of people rejected the app outright because of its AI usage, but I’m neutral to AI so I wanted to give it a shot.I like the concept. Collect bugs, take their quizzes, get points to unlock new bugs. I enjoy that there is no “streak pressure” as the developer puts it, so I don’t have to log in daily.However, I have no interest in paying for a subscription! Maybe I might would spring a few dollars towards extra features once or twice a year, but monthly? No.I also had to dig into the terms of service to find out what kind of LLM is being used for the premium bug chat feature. It’s Google Gemini. I don’t really know enough about Gemini to comment, but while I don’t think the existence of this completely optional feature is any more destructive to the environment than a regular Google search, I still have concerns about the LLM inadvertently giving out false information. I think the idea of being able to chat with little bug friends is very cute, but maybe it would be better to do it a bit differently. I feel like having a more restricted chat with pre-written answers would be safer, but I understand that would be a lot of work too, even if you had an LLM generate them for you and you just fact-checked it before writing it into the app. Still, it’s something to think about, especially if you’re not getting many people to pay for the subscription service.
Thank you for taking the time to write such thoughtful feedback !The goal of the app has always been to make insects feel more approachable and accessible, including for people who normally wouldn’t be interested in them (or even people with fears around insects), while still keeping the educational side reliable and engaging.Regarding the AI chat feature, reliability was something I took seriously from the start. I spent a lot of time testing responses and added clear in-app notices explaining that the answers are AI-generated, educational rather than scientific, may contain mistakes, and should be used with parental supervision for children. The quizzes themselves are also manually reviewed, since they are the core educational part of the app and I wanted them to be as reliable as possible.I also don’t directly mention a specific AI model inside the app because the field evolves very quickly, and the underlying model may change over time depending on reliability and environmental considerations. I want to keep using solutions that are both useful and as responsible as possible long-term.And I’m really glad you mentioned the lack of streak pressure, as someone personally trying to spend less time on screens, it was important to me not to design the app around forcing daily engagement. I still find it interesting though, because many users also ask for streak systems, so it’s something I’m carefully thinking about as the app evolves.Thanks again for the detailed feedback, it genuinely helps improve the app over time.