Haorui Mind is an experimental AI project exploring what a digital representation of myself would look like. It combines a large language model with a custom personality, long-term memory, and a conversational interface designed around a central idea: How do you make an AI as human as possible?

‍
Contrary to a traditional assistant that is always focused on the user's request and tries to fulfill it at all cost, the main idea behind Haorui Mind is to create an agent that will have preferences and will be able to make certain decisions that might include refusing the request or ignoring it sometimes. Starting from analyzing my way of communication and decision making, I started creating my conversational AI.

‍
Another aspect of my research concerns the question if an AI system can have a continuous personality. For that purpose, I created a memory module of the system where all the necessary information from the past conversations can be found and used thanks to memory timeline and dashboard. Voice interaction feature is also included in the system to make the communication more natural.

‍

Haorui Mind has allowed me to explore computational agency, personality design, memory architecture, human-computer interaction and conversational AI—while investigating how design decisions can shape the behavior of an artificial mind.

‍