This is actually a very interesting and difficult question. One of the biggest problems with true multicellular, differentiated organisms (as opposed to pure colonies) is how the cells all manage to work together to support the whole organism rather than duke it out with one another competing for resources. Since there is no one boss cell that tells all the cells what do to, how do they come to cooperate? The answer (besides that the don’t always, even in the same organism) is that cells started off by developing a high degree of internal differentiation, i.e. ever increasingly complex organelles, the functions of which then became more pronounced in some cells than another, giving rise to differentiation between whole cells, i.e. this one masticates food, that one provides locomotion, the other one receives and converts external stimuli.
The key to all of this are the energy-producing endosymbiotes, i.e. mitochondria and plastids, which freed cells from having to produce their own energy. The now-broadly accepted endosymbiotic hypothesis states that these organelles (which contain their own, entirely seperate DNA) were once free-living prokaryotes who tagged up with other cells in a mutually beneficial relationship. (This has been presented as an argument against, or at least strongly modifying Darwinist selection, although in fact it fits hand-in-glove with the underlying sentiment of both Modern Synthesis and so-called neoDarwinist gene-centricism.) Once cells are able to use a universal energy source, they don’t need all need to provide food for themselves; they can then produce other cells which do the business of collecting and breaking down food and focus on more important business, like watching Fassbinder films and talking about philosophy in French cafes.
Some molecular biologists consider this a great conceptual leap (although it has happened at least twice, and possibly more), while others consider it to be an almost trivial assumption. Without this occurring, it is highly unlikely that multicellular organisms (and thus anything like intelligence) could have evolved, and so is a big issue in the debate over whether advanced extraterrestrial life could exist. (I’ll note that it hasn’t led to a big push for sentient rights for plants, but whatever.)
There are organisms that sit on or near the boundaries of this; Mycetozoa (so-called slime molds) are colonies of differentiated organisms, while Scyphozoa (true jellyfish) are clearly distinct organisms with differentiated cells and an overall structure, but do not have specialized organs or nervous/respiratory/digestive systems. So it’s not all that black and white as it might seem.
The real question about evolution of complex organisms is sex. (Isn’t it always?) Although the post-hoc benefits are clear, it is such a strange joining and exchange that seems to have little impetus to develop or recommend it. Even aside from the whole dating mess, it’s just hard to imaging why a species would differentiate into two different types of codependent gametes, which then shuffle genes like a deck of cards. Weird stuff…but good when you have it.
Stranger