This idea of trust is a very strange thing, to me. If I present an argument or point logically and it is accepted, then there is no problem. If I do so and it is rejected logically, there is no problem. If I do so and are assailed with noise and rubbish, then it is usually pretty clear that it's just garbage and to be ignored.
If I seek affirmation and reassurance from people online instead of flesh-and-blood folks I can talk to directly, I being to wonder if this is actually an okay state of affairs. If my support group could be replaced with a bunch of clever Python scripts, Markov models, and machine learning algos, I would suggest that something is deeply wrong--and that is exactly the sort of inscrutable interface forums provide.
Only when you start to "trust" the personas you encounter online and become invested in your model of them do you actually run the risk of getting hurt. Treat it like a game, learning what you can when you can and ignoring the rabble, and things go best.
:)