This is gonna be a tough one. I mean, writing without using ChatGPT, Claude, etc, to generate this blog post for me. I'll create an article on how I am using AI in my studies and work
I'm currently pursuing a Master's in Artificial Intelligence with a Robotics specialisation, which I genuinely love, but let me tell you: learning in the age of fast-moving AI is both exhilarating and disorienting at the same time.
What does "fast-moving AI" even mean?
Now?
We have tools
We have automation
We have AI helping us build AI
Here's the thing, AI has been practically real for about the last two years now. And it's getting crazier every single day. Right now, AI isn't some distant future thing anymore, it's embedded in almost every industry. Discoveries drop constantly. Everything is accelerating. It's like trying to read a book while someone keeps flipping the pages faster and faster.
For me I am studying both the present and the past simultaneously, and they're moving in different directions. The past stuff is the foundational theory, which is crucial because it's what gives you actual understanding. You need to know why things work, not just that they work. But the current stuff? That's where it gets wild. You're learning cutting-edge applications while also reverse-engineering them: What's the origin? How did we get here? What does this mean for what comes next?
The generational shift in AI education
If I compare my experience to someone who did a Master's in AI ten years ago, the difference is huge. Back then, you had to build almost everything from scratch. You did the research, prepared your own datasets, and wrote your own implementations. Speed wasn't even part of the equation; research took time.
Now? We have foundational models, pre-trained systems, automated pipelines, and off-the-shelf tools that would've taken years to develop a decade ago. But here's the double-edged sword: we still need to understand how to build these things manually, so I still do the old model stuff, but using the current baked tools, fancy for me. The difference is that we also have the option to leverage automation. So the game has shifted, it's not about reinventing the wheel anymore; it's about knowing when to build wheels, when to use existing ones, and how to integrate them intelligently.
I'm planning to write deeper dives into exactly what I'm learning and how it's different from the "old way"
Robotics is its own beast
Now, robotics is a different animal entirely. Unlike NLP, where pre-trained models and massive datasets have democratised the field, robotics hasn't had that same tooling revolution. You can't just download a pre-trained robot. You have to do the practicals, actually build, test, and iterate. With lots of practicals, robotics isn't just one field, it contains multiple fields working together. I'll link to a Getting Started with Robotics article here.
The foundation matters even more in robotics because you're combining computer vision, control systems, mechanics, and AI reasoning all at once. The good thing right now is that there are excellent simulation environments that help to understand the practical part.
Keeping up without burning out
There's this underlying anxiety in AI right now: if you stop learning for a week, you've missed something important. But here's what I've learned: keeping up is actually manageable if, and this is crucial, you have a strong foundation. Every breakthrough, every new model, every new technique. They're all built on the same fundamentals of computer science, mathematics, and physics. The basics haven't changed. The applications have exploded, but the principles are solid
These are just high-level thoughts on how I personally learn AI while immersed in the AI spree. An article on how I am using AI for learning and work would be awesome to break down into low-level details
Top comments (0)