Inspiration
When LLM based chatbots launched, whole world is excited but the one's who are truly making them: our beloved users are not getting any benefit instead, they have to pay the fees in the current market.
What it does
It monetizes the LLMs and stores the data into Sia's decentralized cloud & incentivize users for using the chatbot and seeing the sponsored messages. Also, the chatbot can act as a widget on eth.id where we can pass the data of the user and any guest who is vising the user's profile can chat with a chatbot as if the guest is chatting with the user itself.
How we built it
ReactJs, OpenAI API, Filebase, Sia Decentralized Storage
Challenges we ran into
Getting up the Sia's docker was the most difficult and time consuming thing. We also thought of using other options like Filebase for quicker integration. Also, choosing the best LLM implementation research took our time.
Accomplishments that we're proud of
Innovative business model, setting up the docker, fine tuned LLM for better accuracy.
What we learned
LLMs are little immature for now and the problem of hallucination of AI exists. Also, integration of blockchain with web 2.0 is very difficult. Would love if everyone implements solution like eth.id where we only need to embed iframe
What's next for AdGPT
Making the app 100% decentralized and launching own token
Built With
- ai
- llm
- react
- sia

Log in or sign up for Devpost to join the conversation.