Depends on where you're starting from

1. Absolute zero?

Find the top rated short book in the language you're interested in. If you don't know which language, since you eventually want to get into AI, Python probably would be best bang for buck. IMO the more pictures, the better. Also, be aware that sometimes setting up the environment for coding can be half the grunt work, and to avoid that, often something like using Google Colab works just fine. [0]

2. If you have some coding knowledge and want to improve that first before thinking about AI? Here are some tricks I would try that I think are less common:

a. Find a youtuber who meets these 3 criteria:

- 1. is a great coder

- 2. is a great communicator

- 3. has posted videos of them solving the Advent of Code challenge series while they think out loud about their reasoning and solving.

Simultaneously try to solve the AoC problems yourself, while also typing out their solution. Observing tacit knowledge is very effective for skill transfer.

Pause the video whenever you think you know the way forward etc. so you get a chance to try yourself.

b. Get very used at quickly digging through codebases. Best way to do that is to find a lot of codebases to click through.

Github is pretty good for this and the code browsing feature actually lets you jump through code reasonably well enough to traverse codebases just by clicking on any symbol that looks interesting to jump to its usages or definition.

Look for anything that sounds cool on Github (browse trending repos etc), and try to figure out how it works. This is a quicker way to introduce yourself to a variety of patterns without having to whip everything from scratch - you can learn those planning skills at a later time.

3. For getting into AI, once you feel comfortable with coding generally, try the Deep Learning For Coders youtube lecture series from fast.ai

[0] I just point this out to give you a heads-up that basically it's quite easy to run into situations where for some strange reason if you just keep installing stuff, things can get borked, and trying to solve that saps energy and educational momentum