One of the toughest concepts to grasp about evolution is its lack of direction. Take the classic image of the evolution of man, from knuckle-walking ape to strong, smart hunter:
We view this as the natural progression of life. Truth is, there was no guarantee that some big brained primates in Africa would end up like we are now. It wasn't inevitable that we grew taller, less hairy, and smarter than our relatives. And it certainly wasn't guaranteed that single celled bacteria-like critters ended up joining forces into multicellular organisms, eventually leading to big brained primates!
Evolution isn't predictable, and randomness is key in determining how things change. But that's not the same as saying life evolves by chance. That's because while the cause of evolution is random (mutations in our genes) the processes of evolution (selection) is not. It's kind of like playing poker - the hand you receive is random, but the odds of you winning with it aren't. And like poker, it's about much more than just what you're dealt. Outside factors - your friend's ability to bluff you in your poker game, or changing environmental conditions in the game of life - also come into play. So while evolution isn't random, it is a game of chance, and given how many species go extinct, it's one where the house almost always wins.
Of course chance is important in evolution. Evolution occurs because nothing is perfect, not even the enzymes which replicate our DNA. All cells proliferate and divide, and to do so, they have to duplicate their genetic information each time. The enzymes which do this do their best to proof-read and ensure that they're faithful to the original code, but they make mistakes. They put in a guanine instead of an adenine or a thymine, and suddenly, the gene is changed. Most of these changes are silent, and don't affect the final protein that each gene encodes. But every once in awhile these changes have a bigger impact, subbing in different amino acids whose chemical properties alter the protein (usually for the worse, but not always).Or our cells make bigger mistakes - extra copies of entire genes or chromosomes, etc.
These genetic changes don't anticipate an individual's needs in any way. Giraffes didn't "evolve" longer necks because they wanted to reach higher leaves. We didn't "evolve" bigger brains to be better problem solvers, social creatures, or hunters. The changes themselves are random*. The mechanisms which influence their frequency in a population, however, aren't. When a change allows you (a mutated animal) to survive and reproduce more than your peers, it's likely to stay and spread through the population. This is selection, the mechanism that drives evolution. This can mean either natural selection (because it makes you run faster or do something to survive in your environment) or sexual selection (because even if it makes you less likely to survive, the chicks dig it). Either way the selection isn't random: there's a reason you got busier than your best friend and produced more offspring. But the mutation occurring in the first place - now that was luck of the draw.
Mistakes made by genetic machinery can lead to huge differences in organisms. Take flowering plants, for example. Flowering plants have a single gene that makes male and female parts of the flower. But in many species, this gene was accidentally duplicated about 120 million years ago. This gene has mutated and undergone selection, and has ended up modified in different species in very different ways. In rockcress (Arabidopsis), the extra copy now causes seed pods to shatter open. But it's in snap dragons that we see how the smallest changes can have huge consequences. They, too, have two copies of the gene to make reproductive organs. But in these flowers, each copy fairly exclusively makes either male or female parts. This kind of male/female separation is the first step towards the sexes split into individual organisms, like we do. Why? It turns out that mutations causing the addition of a single amino acid in the final protein makes it so that one copy of the gene can only make male bits. That's it. A single amino acid makes a gene male-only instead of both male and female.
Or, take something as specialized as flight. We like to think that flight evolved because some animals realized (in some sense of the word) the incredible advantage it would be to take to the air. But when you look at the evolution of flight, instead, it seems it evolved, in a sense, by accident. Take the masters of flight - birds - for example.
There are a few key alterations to bird bodies that make it so they can fly. The most obvious, of course, are their feathers. While feathers appear to be so ideally designed for flight, we are able to look back and realize that feathers didn't start out that way. Through amazing fossil finds, we're able to glimpse at how feathers arose, and it's clear that at first, they were used for anything but airborne travel. These protofeathers were little more than hollow filaments, perhaps more akin to hairs, that may have been used in a similar fashion. More mutations occurred, and these filaments began to branch, join together. Indeed, as we might expect for a structure that is undergoing selection and change, there are dinosaurs with feather-like coverings of all kinds, showing that there was a lot of genetic experimentation and variety when it came to early feathers. Not all of these protofeathers were selected for, though, and in the end only one of these many forms ended up looking like the modern feather, thus giving a unique group of animals the chance to fly.
There's a lot of variety in what scientists think these early feathers were used for, too. Modern birds use feathers for a variety of functions, including mate selection, thermoregulation and camouflage, all of which have been implicated in the evolution of feathers. There was no plan from the beginning, nor did feathers arise overnight to suddenly allow dinosaurs to fly. Instead, accumulations of mutations led to a structure that happened to give birds the chance to take to the air, even though that wasn't its original use.
The same is true for flying insects. Back in the 19th century, when evolution was fledging as a science, St. George Jackson Mivart asked "What use is half a wing?" At the time he intended to humiliate the idea that wings could have developed without a creator. But studies on insects have shown that half a wing is actually quite useful, particularly for aquatic insects like stoneflies (close relatives of mayflies). Scientists experimentally chopped down the wings of stoneflies to see what happened, and it turned out that though they couldn't fly, they could sail across the water much more quickly while using less energy to do so. Indeed, early insect wings may have functioned in gliding, only later allowing the creatures to take to the air. Birds can use half a wing, too - undeveloped wings help chicks run up steeper hills - so half a wing is quite a useful thing.
But what's really key is that if you rewound time and took one of the ancestors of modern birds, a dino with proto-feathers, or a half-winged insect and placed it in the same environment with the same ecological pressures, its decedents wouldn't necessarily fly.
That's because if you do replay evolution, you never know what will happen. Recently, scientists have shown this experimentally in the lab with E. coli bacteria. They took a strain of E. coli and separated it into 12 identical petri dishes containing a novel food source that the bacteria could not digest, thus starting with 12 identical colonies in an environment with strong selective pressure. They grew them for some 50,000 generations. Every 500 generations, they froze some of the bacteria. Some 31,500 generations later, one of the twelve colonies developed the ability to feed off of the new nutrient, showing that despite the fact that all of them started the same, were maintained in the same conditions and exposed to the exact same pressures, developing the ability to metabolize the new nutrient was not a guarantee. But even more shocking was that when they replayed that colony's history, they found that it didn't always develop the ability, either. In fact, when replayed anywhere from the first to the 19,999th generation, no luck. Some change occurring in the 20,000 generation or so - a good 11,500 generations before they were able to metabolize the new nutrient - had to be in place for the colony to gain its advantageous ability later on.
There's two reasons for this. The first is that the mutations themselves are random, and the odds of the same mutations occurring in the same order are slim. But there's another reason we can't predict evolution: genetic alterations don't have to be 'good' (from a selection standpoint) to stick around, because selection isn't the only evolutionary mechanism in play. Yes, selection is a big one, but there can be changes in the frequency of a given mutation in a population without selection, too. Genetic drift occurs when events change the gene frequencies in a population for no reason whatsoever. A massive hurricane just happens to wipe out the vast majority of a kind of lizard, for example, leaving the one weird colored male to mate with all the girls. Later, that color may end up being a good thing and allowing the lizards to blend in a new habitat, or it may make them more vulnerable to predators. Genetic drift doesn't care one bit.
Every mutation is a gamble. Even the smallest mutations - a change of a single nucleotide, called a point mutation - matter. They can lead to terrible diseases in people like sickle cell anemia and cystic fibrosis. Of course, point mutations also lead to antibiotic resistance in bacteria.
What does the role of chance mean for our species? Well, it has to do with how well we can adapt to the changing world. Since we can't force our bodies to mutate beneficial adaptations (no matter what Marvel tells you), we rely on chance to help our species continue to evolve. And believe me, we as a species need to continue to evolve. Our bodies store fat because in the past, food was sporadic, and storing fat was the best solution to surviving periods of starvation. But now that trait has led to an epidemic of obesity, and related diseases like diabetes. As diseases evolve, too, our treatments fail, leaving us vulnerable to mass casualties on the scale of the bubonic plague. We may very well be on the cusp of the end of the age of man, if random mutations can't solve the problems presented by our rapidly changing environment. What is the likelihood that man will continue to dominate, proliferate, and stick around when other species go extinct? Well, like any game of chance, you have to look at the odds:
99.99% of all the species that have ever existed are now extinct.
But then again - maybe our species is feeling lucky.
* If you want to get into more detail, actually, mutations aren't completely random. They, too, are governed by natural laws - our machinery is more likely to sub an adenine for a guanine than for a thymine, for example. Certain sections are more likely to be invaded by transposons... etc. But from the viewpoint of selection, these changes are random - as in, a mutation's potential selective advantage or disadvantage has no effect on how likely it is to occur.
Airoldi, C., Bergonzi, S., & Davies, B. (2010). Single amino acid change alters the ability to specify male or female organ identity Proceedings of the National Academy of Sciences DOI: 10.1073/pnas.1009050107
XU Xing, & GUO Yu (2009). THE ORIGIN AND EARLY EVOLUTION OF FEATHERS: INSIGHTS
FROM RECENT PALEONTOLOGICAL AND NEONTOLOGICAL DATA Verbrata PalAsiatica, 47 (4), 311-329
Perrichot, V., Marion, L., Neraudeau, D., Vullo, R., & Tafforeau, P. (2008). The early evolution of feathers: fossil evidence from Cretaceous amber of France Proceedings of the Royal Society B: Biological Sciences, 275 (1639), 1197-1202 DOI: 10.1098/rspb.2008.0003
Marden, J., & Kramer, M. (1994). Surface-Skimming Stoneflies: A Possible Intermediate Stage in Insect Flight Evolution Science, 266 (5184), 427-430 DOI: 10.1126/science.266.5184.427
DIAL, K., RANDALL, R., & DIAL, T. (2006). What Use Is Half a Wing in the Ecology and Evolution of Birds? BioScience, 56 (5) DOI: 10.1641/0006-3568(2006)056[0437:WUIHAW]2.0.CO;2
Blount, Z., Borland, C., & Lenski, R. (2008). Inaugural Article: Historical contingency and the evolution of a key innovation in an experimental population of Escherichia coli Proceedings of the National Academy of Sciences, 105 (23), 7899-7906 DOI: 10.1073/pnas.0803151105