In my experience, you gotta be careful with those monthly subs... they add up fast!! For a budget-friendly start, I'd check out Mimo: Learn Coding/AI vs Enki: Learn Code & AI. * Mimo: Learn Coding/AI: Best for total beginners, very gamified.
* Enki: Learn Code & AI: Better for actual ML concepts and bite-sized logic. Tbh, Enki: Learn Code & AI feels more reliable for deeper AI stuff. Just watch out for the auto-renewals lol.

Seconding the struggle mentioned above!! Commute learning is honestly hit or miss. If you want actual hands-on coding, DataCamp: Learn Data Science is basically the gold standard imo. It has a legit mobile IDE for bite-sized ML exercises. For the math behind neural networks tho, Brilliant.org is way better at making it intuitive without just watching videos. Both work reallyyy well for keeping things interactive on a train... gl!

Re: "Seconded!"

• I found the suggestions so far a bit disappointing if you are aiming for actual engineering depth. Most popular apps unfortunately treat neural networks like a logic game, which wont help when you need to debug a real-world gradient descent issue. Based on the thread, people seem to like gamified options or math theory, but for long-term technical mastery, these often fall short. I have had issues with how these platforms oversimplify the backpropagation process. Here is what I recommend for a more rigorous mobile setup:
• Coursera Mobile App specifically for the DeepLearning.ai Machine Learning Specialization. Its less snackable, but the content quality is significantly higher than what you find on Enki.
• Pydroid 3 IDE for Python 3 for running local scripts. This forces you to actually handle dependencies and environment issues, which is crucial for actual development. The issue with bite-sized challenges is they strip away the complexity of data preprocessing. If you really want to learn AI properly, you need to be writing actual code in a mobile IDE, not just tapping buttons. Tbh most of these apps just sell a feeling of progress without the actual skills.

ok so, i feel u on the commute struggle... i literally spent most of last year trying to master neural networks while shoved into a train car, so i get it!! honestly though, i have to start with a warning because i wasted a lot of time on stuff that didnt actually stick. WARNING: avoid apps that rely too heavily on multiple-choice quizzes or simple 'drag and drop' code blocks. it feels like you're learning, but you're actually just getting good at the app's game mechanics. i highkey regret spending so much time on Sololearn for AI stuff because, while it's okay for basic Python, it totally misses the depth needed for actual machine learning foundations. it was reallyyy frustrating to finish a module and realize i still didnt understand the actual math behind it. In my experience, here is what actually worked for me: - **Brilliant**: This is literally the gold standard for the 'why' behind AI. Their neural networks course uses these interactive visualizations that make the math click. It's not passive at all, but unfortunately, there isn't much actual coding.
- **DataCamp**: Their mobile app is surprisingly solid for bite-sized practice. It's better than most, though typing on a phone is still a pain... i think they handle the interactive coding part better than others.
- **Enki**: I tried this for a bit... it's okay for quick refreshes but honestly not as deep as I expected for neural nets. Basically, i found that the 'interactive' part is usually just a gimmick unless the app forces u to visualize the data. i was super disappointed with most 'coding' apps on mobile because they dont let u experiment enough. i guess my lesson learned was to use Brilliant for the logic on my phone, then wait till i got home to actually touch a laptop for the heavy lifting. gl with the learning journey tho, it's a grind but worth it!! peace 👍

Seconded!

I'm gonna have to go against the grain here, even though the apps mentioned so far are pretty popular. Having been in this field for a long time, I find that most mobile-first platforms tend to oversimplify the math behind neural networks just to make it fun, which ends up biting you later when you try to do real work. Most of these interactive courses are basically just syntax puzzles rather than true engineering. Before you settle on one, I gotta ask a couple things to narrow it down. Are you trying to learn how to actually build and optimize models from scratch using a phone-based IDE, or are you looking for more of a high-level conceptual overview to use in a professional or management setting? Knowing your end goal is huge because some brands focus strictly on logic while others are just glorified flashcards for syntax.

I have been quite satisfied with the technical depth I have found by moving away from standard app stores recently. If you want the real data and specifications on these interactive platforms you should honestly just check the wiki on the machine learning subreddits. There is a massive community-led guide there that is updated constantly and it works well for finding specific modules. Just search for something like mobile ai learning roadmap on reddit and the first few threads have everything mapped out perfectly. I also saw a very detailed video on this a while back... if you search for interactive ai course technical review on youtube it should be the first hit. I have no complaints about the quality of those community-vetted resources. Basically just let the enthusiasts do the sorting for you and you will find much better hands-on material than what you see in ads... saves a lot of time tbh.