Scryptonaut.AI Assistant

In today's world of rapid technological advancements, AI assistants have become an invaluable resource and productivity booster. They serve as a bridge between the vast expanses of information available and our need for concise, accurate, and instant answers. The problem with assistants like ChatGPT, Google Bard and GitHub CoPilot is, that they lack specialization and information about recent developments. This is where Scryptonaut comes in:

Scryptonaut.AI is a fully fine-tuned AI assistant created to help you with questions on Radix DLT and Scrypto. It is trained to have the most in-depth, up-to-date and relevant knowledge to assist you better than a generalist AI assistant can.

How does it work?

LLM (Large Language Model) assistants are designed to perform tasks, answer questions, or assist with specific needs. They are trained on billions of parameters, but this comes at a huge monetary cost. This is why OpenAI, Google and Microsoft can only train their models every so often. Fine tuning means that additional training is added on top of a pre-trained model. Fine tuning is less resource intensive and is used to teach an AI assistant about certain topics, or to behave in a certain way. For Scryptonaut.AI, a curated data set of several thousand data points has been collected for training, to cover a range of beginner to expert questions a user might have about Radix or Scrypto.

We'd like to acknowledge that the Scryptonaut.AI Assistant is currently in its beta phase. This means that while it's equipped to answer your questions, there might be instances where the responses could be incorrect or outdated. We highly appreciate your patience and feedback during this phase, as it will help refine and enhance the assistant's capabilities. Your insights are invaluable in our mission to offer a seamless experience.

Contribute

Our community-driven approach encourages users to actively participate in the assistant's evolution. If you possess data, knowledge, or insights related to Radix DLT and Scrypto, you are most welcome to participate. You can reach out to wizzl0r on Telegram and Discord to submit information that can be used in further training and enhancement of Scryptonaut.

The final dataset, model and the training methodology will be made publicly available and open source once it has matured.

More information will be made available soon on www.scryptonaut.ai - the AI assistant, while in beta, will only be available as Telegram bot: @scryptonaut_ai_bot

This project is powered by StakeBros.info - consider sending some stake our way to support us :)