No, I think he was just arguing that you should go to the library and anything that you don't understand you should look up on Wikipedia.
Because undirected study in a rich subject is a great way to understand it.
For example, which would you rather do: spend a few hours/days reading about existentialism or participate in a dynamic class about 19th century philosophy? Which will you get more out of?
Now switch that to machine learning algorithms versus watching Andrew Ng explain it to you in videos.
The two are not the same.
However, I think that's not the point of the post, or even what the author is saying. What I understand he means is that you can achieve some decent level of education in any subject by yourself (by some definition of decent), and that for payed education, you should focus on something that has a higher expected return of investment.
I suscribe to that chool of thought, but I was lucky enough to be interested in a lot of field of study when I had to choose what education I'd get. I could have gone for architecture, japanese translation, literature or physics, but honestly, the job market was one of the big reasons I chose to study Computer Science. I still dabble in other fields on my free time, but I chose to focus on something that would give em some stability.
If you have the cash to pay for it, then go ahead and get any education you want.
If you already have the skills you need to get a job and pay off your loan, then go ahead and get any education you want.
If you don't already have a marketable skill and the education you are going into debt for isn't going to increase your employability, then you might want to reconsider your options.
If you think of yourself as a start-up, you should get your MVP first (a skill someone will pay you for). Then when you have a paying customer (a job or a nest egg) add all the cool, extra features (a well rounded, liberal arts background)