In light of all the confusion surrounding "evolution", I'd like to clarify a few points that are in desperate need of illumination.

First off, let's define a few terms (as taken from [dictionary.reference.com]):

Evolution: "Change in the gene pool of a population from generation to generation by such processes as mutation, natural selection, and genetic drift."

Evolutionist: "A person who believes in or supports a theory of evolution, esp. in biology."

Mutation: "A sudden departure from the parent type in one or more heritable characteristics, caused by a change in a gene or a chromosome."

Naturalism: "The view of the world that takes account only of natural elements and forces, excluding the supernatural or spiritual."

Creationism: "The doctrine that matter and all things were created, substantially as they now exist, by an omnipotent Creator, and not gradually evolved or developed."

First of all, there is this incredibly common sentiment among creationists that evolution (as driven by mutation) only occurs "backward", and never is "positive". (The definitions for these terms are, of course, never given). First of all, evolution can result in changes that are detrimental (and by this I mean changes that result in an organism or a population being less viable in an identical setting than the parent) as well as advantageous.

One of the favorite arguments used by radical creationists is the "Sickle-cell anemia" straw man. This argument has it's roots in a misunderstood example used by an evolutionist demonstrating how a particular trait, although disadvantageous in one setting, might actually confer increased viability given the proper environment.
The proper framing of the discussion is that, indeed, sickle-cell anemia (SCA) in an environment such as mainland america results in a serious medical problem 1/2 of the time. However, when the same condition is present in an area where you have a 75% likelyhood of contracting malaria, the fact that SCA renders you immune to the effects of the parasite 100% of the time, you have a case where statistically, it's better to have SCA, at least from a viability point of view.

We need not, however, restrict ourselves to such strictly context dependent examples as SCA. One of the reasons antibiotic resistance is brought up so many times as an example of advantageous mutations is because it's relatively simple. It's not so simple that it is readily understood by someone not exposed to it on a regular basis, so I'll try to get you up to speed relatively quickly.

Most antibiotics work by selectively inhibiting a crucial bacterial pathway. What makes an effective antibiotic is something that doesn't harm the host (i.e. you), but that rapidly and efficiently kills off bacteria. Sodium azide, for example, kills bacteria very well, but also does a fairly good job of killing humans, which would make it a rather poor antibiotic.

A convenient example of a bacteria developing an immunity to an antibiotic is that of the emergence of tetracycline resistance via the TetB gene.

Tetracycline is a naturally occurring antibiotic produced by the streptomyces bacteria (as contrasted to purely-synthetic antibiotics like the sulfonamides). It inhibits the 70S ribosome of bacteria, and therefore has no effect on the human 80S ribosome.

The way that bacteria are able to become resistant to tetracycline is through the use of membrane transporters. By actively pumping tetracycline out of the cell, they manage to keep their ribosomes free of the interfering antibiotic. (Other bacteria achieve the same effect with different antibiotics by covalently modifying the offending substance, or by mutating the target of the antibiotic in such a way that it no longer binds).

Here is the important thing to note. The tetracycline transporter (an antiporter, to be exact) is derived from a naturally-occuring metal and proton transporter. As a matter of fact, it [has been found] that the TetB gene is able to partially complement (i.e. fix) strains of E. Coli that have lost their ability to transport potassium.

What we have, therefore, is an example of a bacteria adapting by taking existing genetic information and modifying it so as to be viable in a new environment. In short, we have new genetic information that was not present in the parent strain. Yes, part of it might be a copy, but it has been modified such that it is now doing something not originally described in the original code.

Most prokaryotes have built into them the ability to rapidly change their genetic makeup when presented with a fatal challenge. They actually have what are known as error-prone polymerases whose sole purpose is to, surprisingly, make errors while duplicating their genetic material.
By doing this, the bacteria hope that at least of their number will pull off something similar to what the first bacteria did with the TetB gene: modify or create a gene that will confer resistance against whatever stressor they're being challenged with.

This error-prone system does all sorts of strange things. It might splice two completely different genes together in the middle, or it might just drop an entire section from a gene. It might just make a few point mutations. Of course, most of these mutations will either not do anything, cripple the organism, or outright kill it. But when you have literally billions of bacteria all trying out thousands of combinations, chances are you'll eventually get it right (as in the case of TetB, and the myriads of other antibiotic resistance genes). Additionally, once one bacteria "gets it right", they can share the gene they've created through either conjugation/plasmid. We exploit this all of the time in the laboratory by putting our gene of interest in a plasmid with an antibiotic resistance gene, and having the bacteria pick both up together in a process known as transformation.

Now, some will argue that these new "mutated" bacteria that have the TetB gene are now not "as fit" as their parents. Balderdash. The concept of "fitness" only has meaning when given in a stated environment. If tetracycline is present, than any bacteria without the TetB gene (or another gene that confers similar resistance) will die. End of story. Take away TetB, and that selective pressure is gone.
Since most prokaryotes are very stingy about their genetic material (as opposed to eukaryotes, who keep around all sorts of trash), they will quickly purge out any genes that no longer do anything for them. We run into this problem all the time in the lab, as when the antibiotic is depleted from a solution, the bacteria will "revert" back to the strain w/o the gene of interest.

Bacteria are much more clever than we give them credit for. Instead of being bumbling little robots, they're more like a squadron of kamikaze engineers, always trying new things and (usually) going down in flames over them.

Also, don't try and tell me that "well, because the TetB gene is actually a derivative of a metal antiporter, it's not new information." Rubbish. that's like saying that because Wilbur and Orville Wright watched birds fly, and built their airplane from what they observed, they didn't do anything new or special. Everything is, in it's own way, a derivative.

In that regard, the TetB example isn't the best model for truly de novo antibiotic resistance genes. There are some resistance genes that bear little or no resemblance whatsoever to any naturally-occurring genes. Some of the most notorious, like the [Qnr (quinoline resistance)] genes that are causing lots of problems by conferring resistance against high-power quinolones such as Ciprofloaxacin ("Cipro").
There is no reason to suspect that genes such as these are not produced in a manner similar to that of TetB resistance.

Going back to our definitions, therefore, requires any honest scientist (or for that matter, one who has read this entry even) to be an evolutionist. Yes, it does happen, and no, it's not always "backward". Get over it.
Like just about everything in life, there is no absolute black and white that we can stick to. Just because bacteria can evolve and become resistant to powerful antibiotics doesn't mean that men came from monkeys any more then the fact that there's a town called "Galilee" means Christ was the Son of God.
A rational creationist (as contrasted to a radical, irrational one) has to go beyond the childish concept that "evolution doesn't occur so I can stick my flag here" mindset. The truth is much more complex than that.

Personally, because I know evolution occurs, I have to admit that it's at least conceivable that one species can evolve from another. For example, a sufficiently skilled biologist could easily convince me that hyenas (just to pick something out of the air) are ancestors of our domesticated dogs. Again, I'm completely out of my element when I start talking about anything bigger than a bacterium, so I might be all wet with that statement.

For me, my rubicon is that of the emergence of life from non-life. Going from "soup" to even the simplest prokaryote is something I find utterly ridiculous. A creationist biologist might not have such a problem with that due to his ignorance of chemistry, but might be appalled with the concept of whales evolving from cows. My ignorance of biology makes me think that's infinitely easier than my example of nonlife -> life.

Regardless, I've gotten tired of the whole "mutations are always negative" claptrap I hear time and time again. It's not true, and using that statement makes the speaker end up looking like a fool.
If you have any questions, or need clarifications on anything that I've said, I'm more than willing to address them. Do keep in mind, however, the definitions I started this entry with. Using nonstandard meanings for words, although sometimes reasonable, usually only further confuses the matter.