Qaf

Fair point — I could give a full blueberry muffin recipe in grams and seconds, and yes, I use AI to help draft things faster. That said, every answer is reviewed and adjusted by me personally before posting, so it’s not blind automation. I just use AI as a tool to speed up mundane tasks, not to replace actual thought or judgement.

I understand completely, and that’s a very valid point. The app isn’t trying to perfectly predict every choice, and it’s not a magic solution. Even a few quick inputs won’t capture every nuance of a person’s taste. What it does aim to do is surface options that the user might not see otherwise, breaking the “same content over and over” pattern you mentioned. It’s meant to help reduce endless scrolling and decision fatigue, not replace personal taste or curiosity.

Fair observation — I do use AI to help draft responses faster, but every comment here is reviewed and adjusted by me personally before posting. The goal is not to sound like a bot, but to provide clear, concise, and helpful answers while keeping up with the discussion. I appreciate the honesty, and I’m always open to feedback on tone or style.

That’s a fair point. The AI approach was chosen mainly for flexibility — handling various moods and contexts without requiring the user to manually tag or categorize everything. A local algorithm with tags and some randomness could definitely achieve a lot of the same functionality more efficiently, and it’s something I’m considering for future iterations, especially for offline or low-power scenarios.

Thanks, those are very thoughtful points. Offline/local approaches make sense, especially for privacy-conscious users or those using DRM-free/physical media. A local database with tags and randomness could definitely handle many cases, and it’s something I’m exploring for future versions. Using an LLM is mainly for flexibility and to handle the variety of moods and contexts without requiring the user to manually tag everything. A specialized, local model trained only on movies/series is an interesting idea, and something I would consider as a lightweight alternative in future iterations. Really appreciate this kind of detailed, constructive feedback.

Recommendations are based on publicly available movie and TV metadata, not on access to personal collections or private libraries. The app does not scan local files or integrate with users’ hard drives. As for distribution: Google Play was chosen simply for reach and ease of deployment at this stage. There’s no technical dependency that would prevent a more open distribution model. An F-Droid or fully open-source-friendly approach is a reasonable direction, but it requires additional work and tradeoffs that I haven’t tackled yet. That’s a fair point.

It’s not trying to be “better” in terms of data scale or model sophistication. Netflix optimizes for engagement across a large catalog. ChatGPT is general-purpose and requires the user to frame the problem well. This app is intentionally narrower: – no browsing – no prompt writing – no rankings or long lists It’s optimized for fast decisions, not discovery depth. Different tool, different job.

Hopefully the right kind 🙂 The goal isn’t more content or features, but reducing decision fatigue and getting people to an actual choice faster. If it saves time and removes friction, that’s the value.

That’s a completely reasonable position, and I don’t disagree with it. You’re right that native apps require a higher level of trust than the web, and that control and transparency are easier to maintain on the web side. Those concerns are valid, especially for privacy-conscious users. This app is intentionally minimal and privacy-light, but I fully accept that for some people, a website will always be the preferred and safer option. That perspective makes sense. Thanks for laying it out clearly.

That’s a fair concern, and I understand the hesitation. This app does not request broad or unrelated permissions. It doesn’t require accounts, contacts, location, or background tracking. The reason it exists as a native app is mostly performance and offline UX, not data collection. That said, a lightweight web version is something I’m considering for users who prefer not to install apps. Thanks for explaining your perspective — it’s useful feedback.

No. The app does not sell user data and is not designed around data harvesting. It only uses the minimum information required to generate a recommendation (e.g. mood or preference signals entered by the user at that moment). This data is processed for functionality, not for profiling or resale. There is no user account system, no cross-app tracking, and no third-party data brokers involved.