One can imagine the time in our pre-paleolithic history when formal learning consisted of two balanced parts:

During the day, people with skills would show others how to do something. "Grab the spear here," the teacher might say, taking the hands of the apprentice and putting them in the right spot. "Now practice in that sandlot over there by throwing it at that big tree. Keep doing it until you get it right. Then throw it at the smaller tree."

While at night, people around the campfire might tell of great adventures, including myths and legends. People would share ideas, and help their community expand their thinking. The best story tellers would gain bigger audiences and develop their own craft of narrative and suspense.

Then came the technology of writing. And suddenly the balance shifted. Communities were able build on the written work of the past. Written work also scaled well, where the work of one village could impact villages all around it. The disciplines of accounting and drama evolved geometrically. Meanwhile, practicing in the sandlot didn't change much. It was still a one-to-one activity.

Since the technology of writing, many subsequent discoveries have further augmented the "learning to know" skills. Paintings, theaters, printing presses and books, photographs, schools, universities, sound recordings, movies, scanners, Google (and now, the Kindle) all turned our culture into masters of linear content, enabling both great artists and our own exquisite vocabulary around such catnip as plot devices, antagonists, suspense, and the hero's journey, just to name a few. We can watch a Spielberg movie, a piece of campfire-style intellectual property that is the recipient of cumulatively trillions of dollars of investment and R&D, and evaluate it at a level of cultural sophistication that would awe citizens from a even a hundred years ago.

And yet, in the "learning to do" area, we are probably worse than our hunter-gatherer ancestors. For teaching the simplest skills, we mirror our ancestors ("put your hands here"), and for the more complicated skills, we don't have a clue. Ask a Harvard Business School professor to develop leadership (or any Big Skill) in a student and she will go into campfire mode with PowerPoint slides of grids, case studies, and so-called inspirational stories.

The advent of flight simulators and computer games, however, have introduced technology around "learning to do" that can finally scale. Today, there is a robust, if nascent, set of "sandlot" tools that is receiving a significant intellectual investment of the current community, and is able to build on the discoveries of the past. Today's "authors," often game designers, can begin to create virtual sandlots where participants can practice skills, instead of just hearing about them (the theory of nudging a pinball machine to get a better score, from a campfire perspective, is trivial; the practiced application is where it is hard).

And, correspondingly, an entirely new language is being developed. Gamers now effortlessly talk about end-of-level bosses, mapping Actions to interfaces, the attributes of Units on Maps, and what is good or bad level_design_.

During the next twenty years, the sandlot technologies (the "learning to do" through games and simulations) will successfully challenge the campfire institutions of universities, movies, and books not only for the discretionary time of the community (which we have already seen), but for help in improving their quality of life. We are already seeing glimpses of the latter through Carmen SanDiego, The Oregon Trail, Age of Empires, America's Army, Full Spectrum Warrior, Virtual Leader, and Brain Age. Will Wright, the creator of SimCity and The Sims, is the first Shakespeare or Beethoven of this medium.

In other words, people will engage in games not to play a super-hero, but to actually become more like one. And the balance between "learning to do" and "learning to know" may finally be restored.