Well, as Bertrand Russell wrote in The Problems of Philosophy, one thing I can be sure of is that "I" am receiving sensory information about things, as when I "see" a brown table in front me; although I cannot prove, beyond all possible doubt, that the table exists, there is clearly information being received by me about something, namely this brown table, and that implies, if not matter, per se, the existence of *something.* It doesn't solve the problem of solipsism or of the many versions of the brain-in-a-vat thought experiment, but it does show that there is a thing--"I"--and that this thing, the "I," is receiving data about something, suggesting that there may be something outside of the individual. Of course, reality could have been created five minutes ago, and this is all in my imagination, but that's as unfalsifiable as the existence of hundreds of undetectable invisible elephants in the room I'm typing this in right now.
At some point, one has to draw a line and take a stance based on likelihood, while acknowledging that problems like solipsism, the five-minute theory/the Omphalos theory, etc. cannot be 100% disproved.
I don't personally think we "create" reality, though. While I am not 100% certain of the existence of material reality around me, I assume that there is, in fact, something real and material around me (and that I, too, am a material object, my self--the "I"--simply a production by my material brain). Perhaps we all modify reality to some degree, filtering it through our individual and human limitations--after all, reality to a bat, as Thomas Nagel noted, is similar to but not the same as what a human would interpret things as--but there is still something there. One might see the table in front me as a lovely shade of beige, while another person sees it purely white, and another person, quite at odds with the others, asserts it is no less than the shade of MAC's Russian Red. There are reaons why these variations in colour interpretation might exist; very probably one of the persons has a form of colour-blindness, and if majority consensus for people who see colours normally is that the table is beige, well, it would be sensible to believe the table is this shade. It would not confirm it beyond any possible doubt, but we can't really do better, and if everyone appears to be at very least seeing a table, shade regardless, well, that's a good start. If we are all being deceived by an outside agent, being made to see a table in varying shades--well, that still implies outside existence, as well as a deceiving agent with a bit too much time on his or her or its hands.
I realise this is a simplification, but without succumbing to solipsism, I'm not sure how much better we can do.
