I have already answered your question a half a dozen times, but I'll answer it again.
There is no reliable way to determine which liquids are safe to drink and which are unsafe to drink.
You can buy 100 Pepsi's and have no problems whatsoever. That doesn't mean that the 101st Pepsi couldn't be tainted.
Knowledge requires certainty. Induction does not provide certainty. Therefore, induction does not provide knowledge.
And this is precisely where most people disagree with you. To have knowledge is NOT the same as to have certainty. The problem of induction notwithstanding, the *reasonable* hypothesis is that the 101st Pepsi will not be contaminated. In fact, if it was, we would want to go find a *reason* for that.
On the other hand, we can know things about the world a priori. If I am chasing a rabbit and that rabbit ducks out of sight, I know that the rabbit still exists. It has not ceased to exist.
How do you know this? Only via induction. You have *assumed* something like the conservation of rabbits and the *only* way to know that is through observation and the hypothesis of object permanence. So, even in the example you give, you use the very methods you criticize in others.
Similarly, I know that there are no atheists in foxholes. I don't need to observe a large number of atheists and foxholes to know that.
Only one atheist in a foxhole will show you to be wrong. And therefore, because such have been exhibited, you are wrong.
