The problem Mind Network solves

n web2, users have to give up their independent data privacy rights. For example, Instagram can access over 40 billion photos posted by users without any authentication, including private photos (https://journals.sagepub.com/doi/full/10.1177/2056305120924779). What happens if private photos are leaked? What happens if a stranger randomly posts a photo of you?
Equally, in web3, photo sharing is a must-have feature for SocialFi to become a daily dapp. For example, Twitter has already become the new reigning photo-sharing app (https://www.collegeofinfluence.com/blog/8ikxaq85o2i2xv5k5d7427aqpydhwi). But lack of user privacy protection in photo sharing could be a major blocker for massive user adoption, especially in the network state where everyone may have several digital identities.
So, a common pain point in the social networks is wishing for disruptive privacy-preserving data technology. The Mind Network offers a decentralized privacy-preserving data lake to maximize the usability of data without trade-offs on privacy protection. Mind Network is built on cutting-edge encryption and privacy-computing techniques by a group of world-class experts.
In this hackathon, we would like to demonstrate how Mind Network can be used for private photo sharing on SocialFi. With Mind Network, the user stories are reformed:

1. Only masked photos are shared and accessed publically. No original photo can be accessed, even leaked.
2. Important information (e.g. face encoding) is encrypted. Only authorised users can access important information based on permission.
3. AI Models are decentralized and secured, and the same as permission is decentralized.

Challenges we ran into

The first challenge the team encountered was "what problem to solve?", in a positive way, that there are too many privacy issues in web2 and we have to identify which one to solve in this 10-day hackathon. We experienced current SocialFi dapps and brainstormed to list out the big rocks for SocialFi revolution. Based on our personal experience, we think private photo sharing is definitely something most people will do and is a must-have feature for social networks on both web3 and web2, but privacy issues have not truly been tackled. This is an issue big enough and an exact issue Mind Network could make a difference to web users.
The second challenge is "how to solve the privacy issue in photo sharing". There are three technical issues to resolve with the existing Mind Network product.

1. How to identify users. We explored and researched DID solutions in the Starkware ecosystem, and we found .bit and starknet.id are great solutions to identify web users across protocols.
2. How to encrypt part of the photo base on AI Model. Users may not want the whole photo to get masked. We started with face identification and masking faces only if there was no authentication. We further developed our encryption algorithm to make it driven by the face identification AI model and this model will be deployed and persist on Mind Network, which means 1) no privacy leak with computation (encrypted computation on encrypted data); 2) the model is trustable as guaranteed by our proof of computation consensus mechanism.
3. How to integrate with existing SocialFI protocols. We don't want to work on a B2B SaaS business model. We want our solution to be permissionless and composable with other protocols. Therefore, by integrating with DID and social graph protocols like Lens, Nostr, Cyberconnect, KNN3, RSS3, etc., we are offering a privacy-protected photo-sharing solution to all users and developers in the community.
   The third challenge is time-limit. 10 days is not enough.