I recently finished my Business Analytics studies and got hands-on experience with deep learning using Python packages. But I’ve always wanted to learn neural networks from scratch, digging into the nitty-gritty details of algorithms. My thinking is that understanding matrix calculations will help me grasp other deep learning architectures like CNN and LSTM. However, with the rapid rise of LLMs like GPT, I’m wondering if it’s still worth investing time to learn deep learning from scratch.
On one hand, it’ll satisfy my intellectual curiosity. But beyond that, is it worth the time if it doesn’t have a direct impact on my academic progress? I’m torn between the desire to understand the underlying mechanics and the fear of getting left behind in the fast-paced world of AI.
So, I’m asking you – is it still worth learning deep learning from scratch in today’s LLM age? Or should I focus on more practical, high-level applications of AI?