While I like to think that if you asked for help you would get it, how do you normally respond when people ask you to help them with computer tasks? Now I don't know about you, but my first instinct is to flee. Helping people with computers, a lot of the time, is annoying.
This is compounded by the condescending attitude that most computer users have towards "non-computer" people. If you ask someone to help you with filling out forms, or logging into facebook, or whatever it is you want to do, your probably going to be met with an attitude analogous to "You silly idiot...". I'm not trying to take the moral high ground here, chances are I've acted on this attitude just as many times (On average for years of life.) as you may have, probably more.
It's a very damaging meme. And very hard to eradicate.
EDIT: This was originally appended to my post above. But I eventually decided that was basically just a way of replying that bloated the discussion.
I'm not going to try and rationalize away good answers. Of all the things you could say to someone, I'd say this has the highest chance of success. And if I was behind you, I'd totally go for that.
Part of the problem of what I said above was focusing too much on the totally-green scenario. I think we take a lot of our computer-know-how for granted. As it stands, were still in the age where I tell people over webchat I'm into "hacking" and they leave because they're genuinely afraid that I'll break into their computer. Regardless of how much you and I both know that it's a lot more complicated than that, and probably not going to happen even if someone "has your IP". (Which has become something of a meme among generation Z for "That thing everyone knows you need to break into computers, but not really sure why.")
This is ignoring even the definition of hacker, which I would prefer we get past abject ignorance like this before we put hobbyist hackers into the public mind.
I mean, it's not like it takes more than five minutes to explain:
"The web is made of a bunch of computers sending text streams to each other. The program that interprets these streams is called a server. If a server isn't written correctly it can be exploited to do things the owner didn't intend, and be used to break into the host computer. For the most part nobody is going to get into your computer manually unless theres a server they can exploit. Default windows installations have not had servers installed since XP. Your web browser interprets a different set of text streams that can be exploited as well, you should be more worried about this than you should be of human attackers."
But somehow, it's not happening. Instead an embarrassingly high portion of the population believes that David Lightman type figures run around terrorizing the Internet with magical-but-undefined and fuzzy powers that let them break into anything they have a network connection to.
EDIT: The above explanation is obviously a bit more simple than whats really going on, but is really all most people need to know. Though I wish they knew more, that might just be a pipe dream.
The general public is not going to know what a "text stream" is, what a "host" computer is, or what you mean by getting into a computer "manually". And I'd be willing to bet that "interprets" in this context is also going to be difficult. Explaining these things to people who don't care about computers isn't as easy as you make out.
In fact, I expect that the first four minutes would be spent defining most of the words in that paragraph.
And thats assuming you have someone who "clicks" on it the first time around. If they don't, it's going to take a while. Part of the problem is that as you put it, most people don't care. And it's just too costly (in all senses of the word) to really educate people about "computer stuff". This is why we've been advocating user education for years and it hasn't worked.
So we say that we can't teach users not to shoot themselves in the foot. Cool. What are our other options? We can try to make it impossible for users to shoot themselves in the foot. But what we've found is that users ignore warnings. If you make the security restrictions too strict, users can't figure out how to evade them and deem the machine broken. If you make them too easy to circumvent, users go ahead and shoot themselves anyway. If you use a trusted computing architecture, user freedom is lost in such a way that it'd be better for people to be shooting themselves.
If you can't stop people from losing a limb, and you can't tell them how not to do so, what is the world to do?
EDIT: The only thing I can think of is to make computer education cheaper. And I don't mean in the monetary sense, I mean the mental sense. We need to find some way to get more people to care about computers, or make the concepts easier to teach, or make the materials easier to access, or make humans smarter, or something.
This is a problem I spend mental computer time on daily trying to solve. No answers present themselves. I would actually love for someone to show me the silver bullet, or even a box of lead ones. (LINK: http://bhorowitz.com/2011/10/26/lead-bullets/)
EDIT_2: And just for clarification, I'm focusing on all aspects of computing here, not just security or usability or what have you.
As for what I would do if I was operating firefox for the first time? Try. Or ask somebody next to me (or one of those standing in line).