Shafkat Rahman

linkedin | github | twitter | are.na | contact

Organic designs inspired by turing patterns, Georges Seurat and Wassily Kandinsky

I am an independent researcher with a BSc in Physics, passionate about exploring the frontiers of computational complexity and artificial intelligence. My research interests lie at the intersection of physics-inspired machine learning, mechanistic interpretability, and understanding the systems nature of intelligence.

My work focuses on leveraging principles from physics and advanced data engineering techniques to tackle complex problems in AI. I'm particularly interested in:

Developing physics-inspired ML/DL algorithms
Advancing mechanistic interpretability in neural networks
Exploring the emergence of intelligence in complex adaptive systems

With an interdisciplinary approach, I synthesize concepts from neuroscience, computer science, mathematics, and philosophy to uncover the hidden connections that illuminate the deep complexity of our world. My goal is to contribute to our understanding of the self-organizing principles and nonlinear dynamics that underlie both natural and artificial intelligence.

Disclaimer

💡

My writings represent an ongoing exploration of diverse ideas and phenomena where I approach topics with an open, inquisitive mindset rather than fixed positions, as my views evolve with new evidence and arguments that come to light. My goal is to prompt meaningful conversations by offering early ideas, not make conclusive pronouncements. Additionally, I tend to publish inchoate ideas rather than a perfected draft. I find value in releasing embryonic thoughts for commentary, understanding this means they won't meet my quality bar. I welcome feedback to strengthen my knowledge and communication, as through respectful discussion we can all gain wisdom, with the joy being in the journey.