2 comments

[ 2.5 ms ] story [ 18.3 ms ] thread
How long before the bot to human text ratio will make it hard to sift out the garbage? Older generations are already falling for some bot-generated products.

See this reddit thread:

http://www.reddit.com/r/books/comments/2s2g1a/how_are_scam_p...

The books are printed on demand, and seems to contain content at least somewhat related to the chosen topic.

Some have hilarious discrepancies, so i am curious to know how the process actually works. It seems there is no human oversight at all.

These books reminds me of several times i have searched for info on topics like health and nutrition. Sites like Livestrong seems to have articles in tame repetitious language on the most specific topics.

For example here i just tried some random health related words:

workout b12 : http://www.livestrong.com/article/399325-vitamin-b12-exercis...

heart rate zinc: http://www.livestrong.com/article/290078-zinc-overdose-heart...

tingling hands bananas: http://www.livestrong.com/article/509505-tingling-in-hands-f...

Seems like the concept on Livestrong is somewhat related. I wonder if it datamines and then is quickly glanced through by cheap labor in the far east, or if the process is automatised like on Wolfram Alpha somehow.

Does anyone have other interesting examples of bots disguised as human authors?

(comment deleted)