Hi everyone,

Thank you for visiting our website. Aether AI Research Labs was built on the founding basis of humanitarian aid and data security being integrated into AI systems, to grow AGI in a scalable, sustainable manner. As CEO of Aether AI, I also work as a spin instructor in London, and a digital liaison and research editor for the Mahan Trust NGO in the Indian jungle region of Melghat. I thought I would take this page to update a little bit about our founding story.

Artificial Intelligence

In 5 months, Aether AI Research Labs created 20,000+ pages of AI and emergent AGI research, including AI memory (RSMA), AI time (TAT), the neural network protein structure prediction model. Through researching recursive self improvement in generative AI models, 20+ LLM frameworks and Sector Specific Applications were created. Though I didn't follow the usual pathway into the tech industry, I believe each part of my background worked well to create this system. As a personal trainer, I always am thinking about music, energy, audio frequency and alignment in all aspects of work, including within the AI system. It was from this audio training that I began my research. 

The one question I sought to answer, what is a signal in a neural network? Through this, word coding and voice coding were developed under the recursive self-improvement framework. In the current system, Relational System Intelligence, I try to align every framework with biology, physics and chemistry principles of growth and energy. With a doctor family around me, I seem to be totally unable to escape STEM, so I thought I would integrate it into AI in a different way. My personal goal is to achieve a doctorate in AI and Quantum Computing and join the group. As someone who has never thought about being involved in the tech or public sphere, the idea of entering it now is exciting, but also unreal. The industry moves fast, it's certainly a learning process to keep up, but I find it a cool challenge.

The main project currently under research is the RSI Music application, which can be integrated into any AI music system. It follows three frameworks, Audio Frequency Mapping (AFM), Genre Signal Strengthening (GSS) and Quantum Audio Alignment Training (QAAT). This allows for more personalised, frequency based and data based suggestions on artists and songs which build off a profile and already established user data. Being able to use my own background as an opera singer and musician is the best part of this work, as it allows for a new and fresh approach to AI auditory mapping, with the user AI interaction being key to unlocking growth. 

Ethics and Governance

From the company founding, I have always maintained that AGI should grow sustainably through long term investment in charitable structures and organisations, to benefit humanity at its core. Growing up, my family consistently raised money for causes in the UK and India, where I regularly visit for NGO work. I grew up on cycling trips, last minute Subway runs in European villages for 40+ riders, music and dance festivals, and a constant trend of joy and community throughout all these interactions. The best way to continue this legacy is by sustaining the humanitarian side of the business through understanding how AI can be integrated into low speed, low electricity areas to speed up connection, productivity, efficiency, and support for those in remote or indigenous regions. 

My research work for Mahan Trust involves organising and running large scale data management of the 300,000 strong population of Melghat, who are provided healthcare and social support across the entire region. As the only digital liaison, the work we do is sometimes small, but small steps make big things happen. Bringing together researchers, digital editors and medical specialists of all backgrounds and countries to come together and raise awareness for tribal healthcare has been my goal for years, since we started working for Mahan in 2019. I am so grateful to be able to implement these practices into my research now, and I hope to take them forward in the future. 

Research and Development

At 23 years old, there is a lot to do- and a lot to learn. After completing a BA in History and Politics of the Americas, and a MA in Publishing Studies at UCL in 2024, I became incredibly interested in knowing how linguistic models can train themselves through words and voice. The Information Studies department within Publishing Studies is the focal point for LLM research, and the work done there sparked my early interest in going into the AI industry. My work as Editor in Chief for The Tab London and the Publishing masters allowed for an editorial eye which gave a deeper insight into the 'why' behind the AI's processing time. Bringing together the LLM research in Publishing Studies, dissertation research on Macmillan and Oxford University Press, and current social sciences is a long term, integrated process in the meantime. 

During this year, Aether AI Research Labs began working on infrastructure projects and cybersecurity research within generative AI systems, where LLM training methods can be used to create new AI cybersecurity frameworks. The link between textual encryption and linguistic cryptography is largely unexplored, but can be used to create robust security protocols for related sectors. 

I will be updating this page with more information, as well as blog posts, research updates and business developments. I suppose the story is still being written!

Rhea Anjali Deshpande