How I'm Learning AI in Public: My Roadmap

When the author decided to dive deeply into Artificial Intelligence, they quickly discovered that the challenge wasn't finding resources, but deciding where to begin. With an abundance of courses, YouTube playlists, roadmaps, tools, frameworks, and technologies, it's easy to feel overwhelmed. To combat this, they created a roadmap for themselves, one that reflects their own journey as a Computer Science student studying Artificial Intelligence.

This roadmap is being shared publicly so they can document what works, what doesn't, and how their understanding evolves.

The author's motivation for learning AI stemmed from its increasing relevance in the field of Computer Science. They recognized that while they knew AI was important, they lacked a clear approach to learning it. After a year of not doing enough, they realized they needed to start somewhere and stop waiting for the right time. Thus, they focused on building a strong foundation instead of trying to learn everything at once.

Their roadmap consists of seven steps:

1. Strengthening their programming foundation, specifically Python, C++, and Java, to become more comfortable with programming and understand programming concepts well enough not to let the language become a barrier.

2. Learning the mathematics behind AI, including linear algebra, matrices, determinants, probability, statistics, calculus, and how these concepts apply to Machine Learning.

3. Becoming comfortable working with data, including using tools such as NumPy, Pandas, data visualization, data cleaning, and data preprocessing. This is crucial since real-world AI work involves understanding and preparing the data more than just training models.

4. Understanding the concepts behind Machine Learning algorithms, rather than just using pre-trained models. Topics covered include supervised learning, unsupervised learning, regression, classification, clustering, model evaluation, feature engineering, overfitting, underfitting, and model optimization.

5. Moving toward Deep Learning after establishing a solid Machine Learning foundation. This includes understanding neural networks, CNNs, and other deep learning architectures.

6. Learning through projects, building things to understand how different concepts come together, such as data, models, backend development, and real-world applications. The author shares that this is the most important part of their roadmap.

7. Learning in public by documenting their progress, sharing their thoughts, concepts they find difficult, projects they are building, mistakes they make, and resources they find helpful. This keeps them accountable and prevents them from delaying action.