That's almost part of the problem though. My feelings about LW/the Sequences/SSC/etc. are complicated with lots of positives and negatives-- way too much to summarize in an HN comment when I'm supposed to be working ;) -- but the main negative I was referring to here is that they are incredibly potent "insight porn". That term was originally coined to describe the work of hacks like Malcolm Gladwell, and while LW et al are definitely on a level above Gladwell's crap, something can be true and still be insight porn, with all the addictiveness and consequences to your mental faculties that implies.
Fair enough, I have had to make sure I don't lose myself as an insight junkie. When it comes to the consequences are you thinking of it damaging our ability to seek out insights for ourselves if they are so easy to read?
The reason these works feel so insightful is that they link all their material together into a massive, cohesive worldview. Each additional detail makes it sound more true, even though rationally each detail is one more thing that could be wrong (ironic LW link: http://lesswrong.com/lw/ji/conjunction_fallacy/).
What you're left with after consuming the work is a single giant "fact" that's actually composed of a whole bunch of genuine knowledge and a whole bunch of nonsense, and it's mentally draining to toss the nonsense because it feels like it's somehow a key part of something valuable. The big cohesive viewpoint is a lot more seductive than the actual truth - that all of the material needs to be evaluated separately and there are no shortcuts for sorting it out.
Or another way to put it: You really need to keep a critical eye towards each new fact that's introduced, but the entire appeal of this sort of thing is the idea that it's all one super-fact, so your critical thinking is at odds with your desire to enjoy the big picture.
That's an interesting way of thinking about it, in some cases there is overlap. Both do transmit dopamine and reinforce neural connections. However this is just one of many types of "addiction", so I think the relationship goes both ways with learning having the potential to be one of many types of addiction.
The replication crisis definitely leaves a lot of room for questioning. However, even if all the studies we have were properly replicated we still wouldn't understand addiction perfectly. We still have a lot to learn, just look at how vague definitions of addiction are still http://www.dsm5.org/documents/substance%20use%20disorder%20f... But both addiction and learning involve the brain's dopamine circuitry; ironically when responding to your point I can't read the relevant articles I found because they are behind paywalls. This article has some interesting commentary but the sources are paywalled: https://www.thieme-connect.com/products/ejournals/html/10.10... Most likely there is a connection, but anyone who claims to understand it entirely at this point is overconfident.
I don't think I have ever learned anything from some such sites. Distilled, prepared wisdom of people with limited scope and understanding. Who is your source? It this source trustworthy? What are his/her biases and prejudices?
At least with a good book you get an expanded argument, for or against which you can argue yourself. These feel-good pages, as I call them, are as good as a fart against the wind.
