Here I document my 10-week experimental journey into AI/ML.
I have only two rules:
- Show up each day.
- Post what I did, however little or much.
Days 1-2
Key points: supervised vs. unsupervised; overfitting; prediction vs. inference; parametric vs. non-parametric; linear vs. nonlinear. Discuss analogies, pros and cons, and give examples.
[Day 2] p. 9-23 of ISLP
[Day 1] p. 1-8 of ISLP (An Introduction to Statistical Learning: With Applications in Python)
Day 0
August 1, 2025
So… why am I doing this?
I admit that I don’t have a super clear, “ultimate” goal for why I want to learn AI/ML. I wouldn’t even describe myself as passionate about coding. And yes, the field is highly saturated and competitive. But I also don’t want to be discouraged by such fixed mindsets and things not within my control. All I know is I’d regret it if I keep delaying.
With that said, below are some initial ideas and goals I have, but these could change depending on how the experiment goes. Note that my best and worst habit is letting my time vanish into endless math rabbit holes, so I’ve included some ways to stay grounded.
I want to:
- understand the core ideas of ML
- be able to teach AI/ML concepts clearly to others
- be “literate” in ML code: able to read code without being lost, and more than that, be able to translate code into plain English.
- anchor each new concept to code. “If I can’t make it run on data, I haven’t really learned it.”
- track progress in visible outputs; cultivate breadth first, depth second; measure progress by applied skill, not by hours
- decide whether I’m more of a researcher, an explainer, or an applied builder in AI/ML. Possibly a hybrid, or something else entirely.
- connect with others in the process: build a network, have fun along the way, and have interesting conversations and awesome learning experiences.