Wispr Flow: Dictation That Works EverywhereStop typing. Start speaking. 4x faster.
StoneClip
AI Rock Identifier
10 followers
AI Rock Identifier
10 followers
Unearth the secrets of stones with StoneClip, the smartest way to identify and learn about rocks, gems, and minerals using cutting-edge AI. Whether you’re out exploring or browsing your photo library, StoneClip gives you instant, accurate results.
Hey Design Pro, congrats on launching StoneClip! 👋 Taking existing source code and building a focused AI wrapper app to solve a specific problem (identifying rocks) is a great example of leveraging foundations to create something new. It sounds like AI was a key part of your development process too, which is awesome to hear.
As builders using AI extensively (@UNI AI), we love seeing how it empowers developers, especially newer ones, to explore tools and launch products faster.
To answer your question: One way we use AI is in generating diverse starting points for creative content (like article outlines or image concepts) which our users can then refine, dramatically speeding up the initial ideation phase.
Question about StoneClip: How does the AI handle variations in lighting, angles, or partial views when identifying rocks from photos? Does it provide confidence scores with its identifications?
Cool niche application of AI vision! Wishing you a successful launch! 💎🪨
@theo_l
Hi Theo. Thank you. Source code really opens the door to so many new ideas. AI was very prevalent for my work because this was my first time using PHP to create a working connection between my app and OpenAI. It took a lot of human brain power too!
It’s encouraging to hear how you use AI in your workflow to facilitate better communication with clients. So many useful areas, especially for repetitive tasks.
StoneClip’s accuracy is mostly determined by the initial prompt which explains precisely how to prepare the answer. If the AI cannot detect a rock in the image, it explains this and doesn’t use up the user’s daily free answer. As for different lighting conditions, I have yet to test thoroughly, put the in-app flashlight activation should assist with this. Other edge cases should be handled quite well (I spent many AI tokens on testing). Answers do not provide a confidence score currently, but that sounds like a great idea in the future.
Thank you for the support! I wish you the best in your AI journey. So many exciting horizons…