Instead of taking the voices' advice to flee, I simply stop the madness. I remove the virtual reality goggles and place them by the Apple PowerBook running the simulation. My tour is over. I don't necessarily feel more empathy for schizophrenics -- the five-minute program's goal, according to its sponsor Janssen, a pharmaceutical company -- but it's hard not to be a bit intrigued by the unexpectedness of virtual insanity.

Virtual reality's original prophets saw their technology as a holodeck of pleasure, not pain. Pioneers like Jaron Lanier and Char Davies aimed for beauty, art and emotional rapture. They created computer-generated amusement parks where people floated through objects and lost track of time. Even now, during an era when virtual reality no longer commands the regular headlines it did in the early '90s, veteran visionaries like novelist Richard Powers remain obsessed with the ecstatic promise of the unreal world.

The virtual-schizophrenia booth looks nothing like such dreams. It's a nightmare -- only the latest example of a much larger trend toward building environments that frighten rather than fascinate. Call it real therapy through virtual means: In the name of mental health, scientists are exposing people to virtual Vietnam battles, tarantulas and, for those who fear public speaking, raucous crowds. Hell on earth has been transferred to hell in a head mount. And with 20 million Americans suffering from mental problems that some therapists believe could be solved with V.R. -- now cheaper and more accessible than ever before -- many researchers believe that psychiatry will soon do what the V.R. pioneers have not, and introduce virtual reality to the masses.

"It's going to go beyond the university to private practice and to the arena of public health," says Ken Graap, CEO of VirtuallyBetter, an Atlanta company spun out of research at Georgia Tech that sells V.R. therapy tools. "This could be one of the first applications that brings low-cost V.R. to a broad base of users."

Not everyone agrees that V.R.'s technological sleight of hand is the answer to mental illness, but proponents are increasingly enthusiastic. V.R. therapy should be welcomed, they say, because it's safer and cheaper than "in vivo" options -- and because it works. Patients overcome their fears, moving on to mental health in part because they experience the computer-generated world as completely real. Isn't that what V.R. was supposed to be all about to begin with?

The Free Software Project
Read Andrew Leonard's book-in-progress on Linux and open source -- and post your comments.