• People might criticise you because they read your unfinished stuff (this of course only applies if you are considering to upload unfinished articles).
• It takes time, and you should rather use your time to write or read.
• Academia.edu, etc., are for-profit and one should rather not help them (however, they also retain a free version and moreover see No. 3 below).
• Are there other arguments I forgot?

Reasons for doing it:

1. You can reach more people.
2. You can interact with more people, especially scholars you would not meet at your usual conferences.
3. You can reach people who would not be able to read your articles because they are not affiliated to a university and/or their university does not possess a library and/or their university’s library does not have access to the journals where you published.

I think that especially No. 3 is relevant and cannot but lead one to conclude that the ethically correct choice involves uploading one’s work. What do you think?

Now, if you agree with me, you’ll probably also have to consider a next question, namely what can one upload? Some things are safe to upload, namely:

1. Articles, etc. for which you have retained the copyright (e.g., my book Adaptive Reuse, which was published open access).
2. Articles, etc. for which the copyright has expired (usually after one or three years).
3. Articles, etc. which you never published (e.g., my articles on intellectual intuition or on Hayagrīva).
4. Articles, etc. which have been published and of which you could not retain the copyright, but of which you upload a pre-print version (e.g., my book on Duty and sacrifice).

The last case is the most complex one. What does “preprint” mean? I usually upload a version which is just mine, in the sense that it does not entail anything added by the journal’s (etc.) editors. No corrections, no formatting, no copy-editing, etc. I often discussed it with journals (etc.) and they were fine with it.