This cloud-hosted platform utilizes Clarifai and Open Source Llama 2 models to deliver a revolutionary AI experience. [Conceptual Foundation] At the core of this endeavor are dual Large Language Models (LLMs). These are not just any AI models; they are purpose-built to emulate the two hemispheres of the human brain. One LLM excels in analytical and logical reasoning, mimicking the left hemisphere's capabilities. In contrast, the second LLM focuses on symbolic understanding and creative interpretation, akin to the right hemisphere of the brain. [Harmonization Mechanism] To ensure these two divergent models work in concert, we reintroduce the foundational model as a mediating model. This simpler AI serves as a bridge, deciding when to utilize logical analytics and when to engage in artistic ideation. It integrates the outputs of both LLMs into a cohesive and nuanced chain of thought, thus creating an AI that can think dichotomously. [User Interface] The Web User Interface (WebUI) serves as the touchpoint for human interaction. It allows users to manage and interact with both LLMs and the Mediating Model. Designed with accessibility in mind, the WebUI offers a transparent look into how the AI thinks, reasons, and makes decisions. [Technical Integrity] As a full-stack project, we've designed both front-end and back-end components using standard web technologies and machine learning frameworks. This ensures a robust, scalable, and adaptable system capable of evolving as AI and web technologies advance. [Objectives and Impact] The ultimate goal is more than just technical achievement; it's to craft an elegant solution that balances the analytical and creative facets of thought, much like a human brain. The project reflects both the scientific rigor and artistic creativity inherent in complex problem-solving. Your engagement with this project offers a glimpse into the future of AI—a future where machines don't just calculate and sort but truly think and create.
Category tags:Muhammad Awais Akhter
Project Manager
MIND INTERFACES
Life Science Research (AI)
Shaharyar khan
Juushy Juush
solo dev
Team member not visible
This profile isn't complete, so fewer people can see it.
"Overall I really enjoyed the presentation of this project, very cool and well made. I think the use of the 3 models was very smart and the reasoning presented for how they work together was also very interesting! I have personally heard the problem presented that AI models lack the ability to combine these types of logic and creativity together but never heard a solution presented until today so i gave highest marks for originality. The only place I took off was for the business value - assuming this tech would be able to bridge such an important gap then I don't think you did yourself justice in showing me the actual value. Even in your slides where you address Business Value you merely say "subscriptions, advertising, and partnerships" but you don't dive into how any of those would be successful or repeatable and instead force me to imagine, which is something as a business I'd be paying you to do. You're really onto something here and I think with a bit more exploration around solving a specific business use case you'd have something very special!"
Michael Tolbert
Labeling and Learning Manager