I don't particularly enjoy online courses. The density of technical information is very low. I prefer reading papers and then doing a nose dive on related paper

I don't particularly enjoy online courses. The density of technical information is very low. I prefer reading papers and then doing a nose dive on related papers to fill in the knowledge gaps. Papers are like nodes of knowledge with edges to relevant information (references). For things usually not covered in universities, papers are pretty much crucial. For example, I found it a lot more useful to read the original or early b-tree and octree papers as opposed to looking them up in text books. And these directly affected my code. The DynamoDB paper was another one that affected my work thinking, even though we don't do a lot of distributed systems. There are many famous papers out there (search on HN, the topic gets posted frequently). I highly recommend you choose a topic, and just look up the important papers in that topic (google scholar is a great tool for checking citation numbers).