Getting beyond "Getting Started"

A few months back I became pretty frustrated when I felt as if I'd run up against a seemingly unclimbable learning wall. I'm guessing many others have as well. Being a life long learner can be challenging if you want to get beyond "getting started".

-------

Here's the scenario - I found (find) myself very interested in Artificial Intelligence and Machine Learning. I'm a huge fan of Python, Postgres, Rails and R. I found a few 'Getting Started' tutorials and courses to see how much I really enjoyed it. I've completed quite a few of these. Salesforce has offered a few of these, I've taken them. Udemy has a bunch of classes on getting started, but one needs to be careful, quality varies. Here is the wall; on this side getting started on the other side is doing cool things with AI.

-------

Now what? How does one scale the learning wall. One way - go back to university. Nope. So, the next way is completely self guided/self taught study. Ok, that's fine. That's how I learned Python, Rails and R (learned and have lots more to learn). But the biggest challenge is the middle ground. How does one go from 'Getting Started' to self-guided building of Artificial General Intelligence which is my interest. There is a dearth of material around the 201 and 301 levels of training.

-----

What I have started doing is going to the projects I really am excited about like Salesforce , OpenAI , H2O.ai , Netflix . I head to their Research divisions, download the research they've done and learn that way. My current learning plan is taking what I've learned in Python, Postgres, Rails and R and thinking through how these AI teams are solving problems. The most challenging thing for me is the context and historical background. I find I have to do research on the research in areas where I don't have the understanding, which is the middle ground I talked about before. There are no real guide markers. So I need to reverse learn theories and maths to get there. This is a really fun and valuable way to learn without going back to uni.

-------

I'd love to hear what other folks are doing to learn new tools or technologies. Right now I'm deep in research and testing mode on building my own home assistant in Python. This has meant building my own servers and home network. This has also meant spending a lot of time reading articles and research papers on NLP and ANNs. This has also meant going back and learning some finite and linear algebra which I truly enjoy. As always if you're interested in at all nerding out about any of this I am always open to a IRL or virtual coffee.