As I write, my meat is earthbound in Tanzania, at the Technology, Entertainment, Design (TED) Global 2007 conference, but my avatar, Xan Hazlitt–pictured above–freely roams the virtual world of Second Life.

Like many technologists my age, I first encountered the idea of virtual worlds in William Gibson’s 1984 classic of “cyberpunk” science fiction, Neuromancer. I was at boarding school in England, and it was an exeat weekend(that is, vacation). The British boys had gone home to their families, and the foreign students were marooned in the school’s houses. But I was neither homesick nor lonely, because on the previous day the school’s bookshop had delivered a brand-new, hardbound copy of Gibson’s novel.

I remember my excitement when I read how Case, a “cowboy” or criminal hacker, jacked into the matrix after a long, chemically imposed exile from cyberspace, his “distanceless home”: “Inner eye opening to the stepped scarlet pyramid of the Eastern Seaboard Fission Authority burning beyond the green cubes of Mitsubishi Bank of ­America, and high and very far away he saw the spiral arms of military systems, forever beyond his reach.”

Wow! I thought. Now that was poetry!

My interest in what Gibson memorably called the “consensual hallucination” of cyberspace was inflamed by two later books: Neal Stephenson’s 1992 cyberpunk novel Snow Crash and David Gelernter’s 1991 Mirror Worlds: Or the Day Software Puts the Universe in a Shoebox … How It Will Happen and What It Will Mean. (Gelernter, a professor of computer science at Yale University, writes in this month’s essay about why he believes humans will never build a fully conscious artificial intelligence.)

In fact, the cyberpunks and Gelernter were imagining two related phenomena. The first is that of the virtual world, a shared, 3-D environment where people and organizations communicate–an environment that is related to our own world but is fictive. By contrast, Gelernter’s term “mirror worlds” conveyed the idea of geographically accurate software models of real terrestrial places.

My juvenile enthusiasm for virtual and mirror worlds was shared by the pioneers of the Internet. Almost all early descriptions of the Net make some appeal to the glamour of a social, 3-D cyberspace. And yet until recently, neither virtual worlds nor mirror worlds existed outside the heated imagination of science fiction writers and futurists.

For years, attempts to create virtual and mirror worlds were frustrated. When I was the editor of Red Herring magazine in the 1990s, we promoted a new standard called the virtual-reality markup language that was to have given programmers and website designers the means to bring a third dimension to the Internet. (“VRML: The LSD of the Internet!” the May 1996 cover of Red Herring exclaimed.) But nothing came of such early technologies.

Today, the virtual world of Linden Lab’s Second Life, which was launched in 2003, has seven million registered users, 30,000 to 40,000 of whom are online at any one time. The mirror world Google Earth, which is only two years old, has been downloaded 250 million times.

Mere numbers, however, do not convey the beauty, richness, and social complexity of today’s virtual and mirror worlds. Nearly everything that human beings can do, they do in Second Life. Dozens of companies, including IBM and Sony Ericsson, are doing business there. And Google Earth has become much more than a hawk’s-eye view of the globe. Call up any spot where humans live, and the visitor to the mirror world will see a multitude of layers of interesting or useful information. Second Life and Google Earth have many of the features of Gibson’s matrix.

So what changed? First, technology. Most computer users now have the graphics cards and broadband connections necessary to explore virtual and mirror worlds. Storage and processing have become cheap enough to let companies readily purchase the servers necessary to render virtual and mirror worlds in complex detail.

But there’s another, more interesting explanation for the growth of Second Life and Google Earth: the companies that created them understood that virtual and mirror worlds are social environments. The most important function of such worlds is communication and personal expression. Therefore, Linden Lab and Google gave control to users, preserving for themselves only the godlike task of maintaining their universes. Second Life avatars can build whatever buildings, clothes, or flora they wish. Anyone willing to learn the open standards of geocomputing can tag information to locations in Google Earth.

In this issue, contributing editor Wade Roush explores how virtual and mirror worlds will merge into what’s been called the Metaverse (see “Second Earth”). The Metaverse, he writes, “will look like the real earth, and it will … [function] as the agora, laboratory, and gateway for almost every type of information-based pursuit.” Do you agree? Write and tell me what you think at jason.pontin@technologyreview.com.