Ahmed Bouzid of Witlingo interviews brothers Ben and Sam Martin, two philosophy graduates who are innovating in the Natural Language Processing and AI in general by building a system that is based neither on the traditional rules-based strategy, nor on the newly data-heavy methodology, but rather focuses on leveraging concepts and methods inspired by Edumund Husserl's work in phenomenology to create a new unsupervised machine learning technique called Random Contrast Learning (RCL). A fascinating conversation and certainly a great example of how those steeped in the humanities can bring something news to the AI Table.
In Part One, Ben Martin lays out the foundation to how one can build an AI without having to rely on gobs and gobs of data.
In Part Two, Sam Martin does a deep dive into how they have leveraged insights from phenomenology to build an actual real life AI system.
Websites to check out:
- https://eidetic.notion.site/eidetic/Eidetic-Ventures-59cd7f26530f41e8a5a069ff22c90a57
- https://lumina-research.com/
Worth reading: Press release announcing Random Contrast Learning (RCL)