1. Kaggle is your friend
I kicked off my journey with beginner-friendly competitions like Titanic and Dogs vs Cats.
Kaggle provides an awesome environment to learn and experiment.
It's much easier to adapt existing solutions before jumping into coding myself.
2. Jupyter Notebooks are your workshop
Set up your local environment and get comfortable with this essential tool – it's where the magic happens!
3. Pair program with an LLM!
They speed you up by taking care of the burden of understanding hundreds of distracting nuances, allowing you to focus on achieving your first milestones and working through the first happy paths.
4. Exploratory Learning
Understand the current (and historic) AI/ML landscape and what others are doing.
Leverage YouTube, YouTube, YouTube, deepLearning.ai, and huggingface.co.
Leverage the existing solutions and code.
5. Structured learning
I found structured learning paths like Google Skills Boost + GCP ML Engineer very helpful in helping me understand the current AI/ML landscape and its historic developments.
Refresh your math (Andrew Ng, Yann LeCun), but don't get overwhelmed - one can spend years studying! Treat the math as helping you build the initial intuitions!
Learn the basics of data concepts: how to clean, prepare, and analyze data effectively.
6. Immerse yourself in the AI/ML community
Attend local meetups, meet people, talk! – the conversations and connections are invaluable.
Sign up for groups, newsletters and mailing lists from leading AI/ML publications and organizations to understand the current agenda and progress.
7. Focus on one thing at a time 😊 There is too much to absorb, it's easy to scatter one's energy among too many threads.
Enjoy the journey!