Aside from what I do for a living, I've got several sites up that make more in income than they cost to keep up — if you don't count my time, which could be considered very expensive indeed.
Other comments have emphasized that you should love what you do; I would say that this is the most important type of remark in the entire comment set to this point. The sites that have done the best for me thus far are sites that I am passionate about.
That's not to say I don't have some sites out there that I put my heart into, but didn't raise a following — I do — but the two that are doing the best could cover the costs of the entire rest of them without breaking a sweat.
I have found that buying ads on Google without a concrete product to offer is a poor idea. However, displaying Google ads on your web sites works, as your traffic builds. With typical hosting that supplies adequate bandwidth for a mostly-text site going for $30/month, and domain registration for $35/year (cheaper if you want to put up with smaller registrars), you have to bring in about $40/month just to break even. It goes without saying (though I'm going to say it) that you won't make this right out of the gate. So, again as has beeen said, you have to plan for this, expect to lose it, and build solid, useful content. If you do, income will rise and eventually, the site will make its costs back. This is cause for celebration, but certainly not for quitting your day job.
The sites that will make the most money are the sites that have something to offer. The more concrete it is, the more they will make. To that end, I will offer an up example of mine:
This is a site that is of use to those who are interested in genealogy in general, and what mom, grandpa, and great-uncle Frank went through in their lifetime. It generates timelines — you put in a name and a birth+(optional death) date, and it returns an HTML page with historical events plotted out against the submitted person's lifespan. There are some other services, such as who was born in your birthyear (any birthday) and who was born on your birthday (any birthyear) but these aren't the core of the site's attraction, I can say for certain.
We take paypal donations (and receive, very, very few) and we host Google ads, which make what I consider to be an astonishing amount of money and this is, in my opinion, because it (a) offers something concrete and unique to the visitor, (b) the site is exactly what it seems to be, (c) there is enough text on the site's main pages so that Google can serve relevant ad content that will actually get clicked. Aside from this, as I said above, I was very interested in the site early on, and this kept it up while it wasn't making money. And, we have lots of traffic as a result of (a) and (b).
This theme is also reflected in my other sites that make money, as opposed to consume money.
I stick to a few basic principles: I don't use any client side technology; that means that no matter what your browser, my sites will work for the visitor. I try to provide a reasonably wordy commentary for whatever is going on at the site — this works well for Google's ad services, but it also works well for the visitor who is actually interested — pretty pictures may be worth a thousand words, but they're typically only 2-letter words, IMHO. :) I used to write my server-side features in perl, but I found python and now I can write stuff about five to ten times faster (and I suggest anyone else who thinks they're in love with perl to take a few weeks to learn python, it's an amazing language.) There is no question that the more efficient the coding process is, the better it is for you, so I consider this a significant tip.
Aside from that, I sell software along the lines of Photoshop in some ways, a concrete product and by anyone's standards a major, major application, and that, of course, is the holy grail of how to make money on the web. Compared to the income derived from that, all my website efforts are just noise. But again, I really enjoy those other sites, easily as much as I enjoy the day to day job, even though the money's not even remotely comparable.
So. Don't quit your day job. Offer something of concrete value. Bring passion to what you do. Roundly ignore your critics and keep doing what you enjoy. Don't worry about linking. If you create something wonderful (or even significantly useful), other people will do the linking for you. Put your efforts into improving what you have to offer, instead. Links come and go; valuable content is a permanant boost for your site. Try to make the value you offer timeless; current events and commentary on same, by their very nature, lose value over time and unless (like Kuro5hin and slash) you have people who both create content and add value to it, you'll kill yourself in a heartbeat if that's all you do. It'll eat all your time, all the time, and you'll have to be amazingly lucky to stand a chance of breaking even.