Thing is, short of mathematical definitions, everything is subject to some bias, any sufficiently-complex statement is just an interpretation of various signals. Moreover, the threshold for "sufficiently-complex" isn't really so high.
In other words, there aren't really facts (in the sense of being "absolute truths") but just Justified True Beliefs, which means Google essentially has to solve the Gettier Problem (https://iep.utm.edu/gettier/).
Examples:
There is no medical fact that smoking is bad for you but there is medical consensus that smoking shortens your life expectancy. Of course, the consensus does not preclude that there might be some nonagenarian in Fiji or Japan who's smoked a pack a day since he was 20 and is still chopping woods for fuel.
I don't know what exactly GP is saying about Snowden but pre-Snowden, it's a justifiable belief that, among other things, your webcam will only turn on when you tell it to. Post-Snowden that belief might no longer be so justifiable.
No, but isn't that where confidence levels come in? Consensus is formed because of research showing statistical significance of however many sigma or something
The terminology has gotten really muddy, and people talk of statements that are true, i.e. in accordance with the facts as though they were the facts themselves. The facts are what a true statement agrees with that makes it true.
Facts are, epistemologically, the way things are. That statement wasn't a fact, but it was true.
There are a vast many important and meaningful questions someone can type into Google that don't have a single factual answer. At that point, credibility, sourcing etc all matter a lot and misinformation is dangerous.
