Web3_NST

• 

  starry night style paris

Inspiration

The rapid advancement of productive forces will inevitably catalyze a transformation in the relations of production.

In the current era, AI technology stands as one of the most noteworthy advancements, and it is highly probable that it will significantly propel the development of productive forces in the foreseeable future. However, the rapid development of AI brings with it a myriad of concerns and issues resulting from outdated production relations that are incompatible with its advancements. Philosophically, there have been numerous discussions addressing the severe challenges posed by AI, such as theories of AI Ownership(Bernes 2013; Toscano 2014; Steinhoff, James 2017), and AI Labor Concealment(Huws 2014; Moody 2018; Friedrich 2017). Recently, the public and legal communities have also shown a heightened interest in privacy issues and intellectual property matters related to AI.

Advancing alongside the rapid development of AI is another swiftly evolving technology: blockchain. However, its contribution to the enhancement of productive forces is relatively minor; in fact, it could even be seen as a certain level of "squander" of existing productive capacities. Unlike AI, which enhances productivity, blockchain focuses on transforming production and ownership relations, while striving to minimize the "cost" of human interactions. The severe challenges posed by the advancement of AI technology are closely related to these aspects.

What it does

This is a sample project designed to demonstrate how an NST (Neural Style Transfer) web project can implement web3 transaction logic. This setup facilitates the distribution of benefits to owners of stylized artworks, proposers of academic papers and AI models, as well as operators, thereby ensuring the protection of interests for all parties involved.

How we built it

Initially, I utilized Python and the TensorFlow framework to create a basic NST (Neural Style Transfer) application. Subsequently, I employed Solidity for writing smart contracts to manage and distribute funds, as well as to handle the interaction logic with Chainlink Functions. Due to time constraints, I had to reduce some parts of the code, resulting in a rather simplistic Solidity code that primarily dealt with charging fees, account splitting, and recording functionalities. Finally, despite my limited knowledge in frontend development, I managed to write some frontend code by learning on the fly, utilizing web3.js, HTML, CSS, and JavaScript. The project was just one step short of completion: setting up a backend server to facilitate interaction between the frontend and backend, thereby enabling the backend server to call the Python application and record the number of transactions. Unfortunately, not completing this within the set timeframe was a disappointment. Additionally, due to the absence of backend database records, the source code for the Chainlink Functions feature could only generate random numbers to simulate transaction records temporarily.

Challenges we ran into

This was my first relatively formal Web3 project, and I encountered numerous challenges along the way. Since I didn't invest effort in finding a team, I had to complete all tasks alone, including frontend work, which was entirely new to me. Setting up NST also presented some difficulties, but fortunately, the Keras official website offered example codes, and my proficiency in Python helped me save time in this area. The main challenge in writing the Solidity code stemmed from my lack of experience. I had never written my own Solidity code before and had only learned through various tutorials. This project was my first hands-on experience with writing my own Solidity scripts. Finally, the frontend work was a disaster. I had to scavenge for frontend knowledge from various sources to piece together a simple website. Unfortunately, I was unable to complete the backend server setup, which remains a regret for me.

Accomplishments that we're proud of

Despite not completing the project as scheduled, it still provided me with invaluable development experience. In the past, I was always lost in various ideas, but this time, I finally took the first step and practically completed a portion of the work, which brought me some solace. The burst of potential in a short period was also surprising. I never imagined I could learn frontend skills so rapidly and create a webpage, even though it was far from perfect. This experience also made me realize the necessity of systematically learning both frontend and backend development skills.

What we learned

I learned some Solidity programming techniques and gained some frontend knowledge, but most importantly, I realized this: the power of an individual is limited; a team is truly crucial!!!

What's next for Web3_NST

This endeavor was just a preliminary attempt. I will continue to ponder and explore how blockchain can address issues arising in the AI field. As I expressed at the beginning, I believe blockchain can offer practical solutions to the problems related to production relations that AI encounters.

Acknowledgements

Finally, I would like to thank the official Chainlink team in Mainland China for their technical support during the competition, as well as Professor Wang Defeng from Fudan University and bilibili: @学院派Academia (youtube: @xueyuanpaiacademia) for inspiring philosophical reflections. This project would not have been possible without your help and the thoughts it provoked.

Built With

• chainlink
• node.js
• python
• solidity
• tensorflow
• web3.js