Learning to Learn

The Mathematics Behind AGI: Foundational Concepts and the Road Ahead The Mathematics Behind AGI: Foundational Concepts and the Road Ahead I want to be upfront about something before we start. Nobody fully knows what AGI is going to look like, how it will be built, or whether the mathematical frameworks we have today are even the right ones. What we do have is a collection of deep, beautiful, and sometimes frustrating mathematical ideas that seem to point in the right direction. This blog is my attempt to trace those ideas — not as a checklist of solved problems, but as an honest map of where the thinking currently is. Some of this math is a century old. Some of it was written in the last decade. All of it is incomplete in one way or another when it comes to AGI. That tension is what makes this field interesting. 1. What is AGI, Formally? Before building something, it helps to define it. For most of AI history, nobody really tried to define intelligence ...

MobileNet Backbone Architecture Versions MobileNet Backbone Versions: Designing Efficient CNNs for Real-World Deployment MobileNet is a family of efficient convolutional neural networks designed for real-time inference on resource-constrained devices such as smartphones, drones, and embedded IoT hardware. Over multiple versions, the MobileNet family introduced progressively refined design innovations — from depthwise separable convolutions to neural architecture search and transformer-style attention — while keeping computation minimal. Rather than treating MobileNet as a single model, this blog explores each version's motivation , the specific problem it addressed , and the architectural innovations it introduced to push the accuracy-efficiency frontier. 1. Why MobileNet? Motivation and Core Problem Standard convolutional networks like VGG and ResNet are accurate but computationally heavy. Deploying them on edge devices with limited memory, power, ...