Mind Toy
Dream with AI | Mint With Flow | Made for You
Demo
About
The problem it solves # Mind Toy #### A brand for co-creative product design and on-demand manufacture. **Customers** can design and receive physical collectibles, each unique to them, co-created with their favorite brands & IP. **Brands / IP / Creators** , can connect with customers like never before via co-creative engagement, and sell products with a margin comparable to other on-demand printed goods. Collectibles from your mind - Cool huh! Sell your collectible to others - ***hang on... WHAT?!*** With FLOW, we're designing an innovative *co-licensing* business model using NFTs & split royalty smart contracts to allow fans to earn revenue from sales of their design. This is a whole new way to think about fan/creator relationships, made possible by pairing blockchains with on-demand manufacture. All wrapped in a simple interface that encourages **virality**, ignites **creativity**, unfuddles the **blockchain**, and integrates **payments** - in one place... ***DISCORD*** #### How do we do this? Two core technologies, 7 years of industry knowledge, 2 years of focused R&D - AI generative technology is changing the world, and it changed our business from “Unique toys” to “Unique toys *designed by you*”. We fine-tune AI models to output unique, beautiful images that represent a brand/creator's vision and then function within our pipeline to create a 3D model that is ready for... - Colour 3D printing, using the cutting edge machines we manufacture items in high quality color resin. We’ve been working with these machines for over a year, developing our pipeline and best practice for designing the best quality products.
How it's made
Challenges we ran into When working on this sprint, we had 2 main hurdles to overcome. ## Integrating Niftory into our bot Their API offers many handy features which we wanted to use, but they have a lack of documentation for some features, which made it difficult for us to navigate the platform. We had to explore and reach out to their support team on Discord for assistance. Despite these challenges, we found the support team to be very responsive and helpful. We also encountered some difficulties with the platform's hierarchy, particularly with regards to the "Set, NFT Model, and Mint NFT Model" system. We learned that a set is a collection, NFT models are models of the NFT, and that multiple identical NFTs can be minted from a single NFT model. Additionally, we discovered a few bugs with the platform's GraphQL API, but were able to work with their team to reproduce and troubleshoot them. Overall, while we encountered some challenges, we were impressed with the platform and appreciated the support we received from their team. We look forward to continuing to work with them in the future. ## Solving Stable Diffusion API and tuning Our partner Scenario.gg is a platform to train, use and host SD models with API. We got the first access to beta for this project (!) and so we had the challenge of working with a new, untested API. Florian spent hours working to test generations for the right settings, which have many layers of complexity to get right generations. In parallel, I was fine-tuning the AI model for our 3D pipeline. For this hackathon we wanted to make something FLOW themed, so our model is for collectibles that use your vibrant colour pallet, flowing design and cute character aesthetic. (Cryptokitties, Doodles, Flovatar). Using our 'Art Toy' model as a base we fine tuned a model using flow branding, colours and designs and after 5 iterations, we now have a Flowaii, Shadow Flow & Headz model, made just for you!