Inspiration Point Walker frequently meets with the Walker Digital brain trust in the seating area of the library, hoping to draw inspiration from the surroundings. Artist Clyde Lynds (known for integrating fiber optics into his work) created the intricate illuminated glass panels and many other visual elements. Walker himself designed the Escher-like tile floor, modeled after a tumbling block pattern from the Victorian age. He bought the chandelier (seen in the Bond film Die Another Day) at an auction and rewired it with 6,000 LEDs. The open book on the table features watercolor illustrations for an 18th-century papal palace that was never built. The globe has special meaning for Walker: "It was a wedding gift Eileen and I received in 1982."

Reading Room In the foreground are several early-20th-century volumes with jeweled bindings—gold, rubies, and diamonds—crafted by the legendary firm Sangorski & Sutcliffe. On the table (first row, from left) is a 16th-century book of jousting, a Dickens novel decorated with the author's portrait, and (open, with Post-it flags) an original copy of the 1493 Nuremberg Chronicle, the first illustrated history book. Second row: the 1535 Coverdale Bible (the first completely translated into modern English), a medieval tome with intricate illustrations of dwarfs, a collection of portraits commissioned at a 17th-century German festival ("Facebook in 1610!"), a tree-bark Indonesian guide to cannibalism, and a Middle Eastern mother goddess icon from around 5000 BC.

Walker shuns the sort of bibliomania that covets first editions for their own sake—many of the volumes that decorate the library's walls are leather-bound Franklin Press reprints. What gets him excited are things that changed the way people think, like Robert Hooke's Micrographia. Published in 1665, it was the first book to contain illustrations made possible by the microscope. He's also drawn to objects that embody a revelatory (or just plain weird) train of thought. "I get offered things that collectors don't," he says. "Nobody else would want a book on dwarfs, with pages beautifully hand-painted in silver and gold, but for me that makes perfect sense."

What excites him even more is using his treasures to make mind-expanding connections. He loves juxtapositions, like placing a 16th-century map that combines experience and guesswork—"the first one showing North and South America," he says—next to a modern map carried by astronauts to the moon. "If this is what can happen in 500 years, nothing is impossible."

Gadget Lab A brand-new One Laptop per Child XO, far left, sits next to a relatively ancient RadioShack TRS-80 Model 100. In back, a 1911 typewriting machine and a 1909 Kent radio. The large contraption at center is the Nazis' supposedly unbreakable Enigma code machine. The book to its left is a copy of Johannes Trithemius' 1518 Polygraphiae, a cryptographic landmark. On the right is an Apple II motherboard signed by Woz. An Edison kinetoscope sits beside an 1890 Edison phonograph (along with three of the wax cylinders it uses for recording). Nearby is a faithful copy of Edison's lightbulb. The gadget with the tubes is an IBM processor circa 1960. In front of it stands a truly ancient storage device, a Sumerian clay cone used to record surplus grain.

Walker struggles to balance privacy with his impulse to share his finds with the outside world. Schoolchildren often visit by invitation, as do executives, politicians, and scholars. Last February, the organizers of the TED conference persuaded him to decorate their stage with some of his treasures. But he's never invited any press in to see the collection—until now.

Senior writer Steven Levy (steven_levy@wired.com) profiled sci-fi author Neal Stephenson in issue 16.09.

Don't trust anyone over 30. That's the prevailing wisdom in Silicon Valley, a land once again bestrode by millionaire CEOs who just learned to shave. Many people believe that the breakthrough ideas come only from the young. And why not? Media stories constantly recite the ages of a few famous founders: Bill Gates of Microsoft, 20; Mark Zuckerberg of Facebook, 20; the Google boys, 25; YouTube's Chad Hurley, 28. Tumblr founder David Karp is 21 — and on his second successful company.

Young people rule tech innovation, we tell ourselves, because they have several key advantages. They're fearless and naive, so they'll try anything. They can spy markets that elders, with their locked-in views, cannot. And without dependents or spouses, twentysomethings can work the sort of pyramid-building hours necessary for a startup. It's a kind of Logan's Run world: If you're ending a third decade, you're obsolete.

But hold on. A recent study has finally collected some data on age and high tech innovation and found that older geeks are just as successful as young Turks. What's more, the chronologically advanced are especially successful at solving problems we increasingly — and desperately — need solved.

In other words, the high tech future may belong to the over-30 set. The Ewing Marion Kauffman Foundation surveyed 652 US-born CEOs and heads of product development who founded high tech firms in the boom (and bust) years of 1995 to 2005. Both the average and median ages were 39 — far older than the mythic dorm-room visionary. Turns out those youthquake pioneers don't really represent the pack. They're outliers.

So why is our intuition wrong about this? Because young and old founders create different types of startups.

Mature entrepreneurs tend to launch startups that require huge amounts of capital — biotech companies, energy firms, outfits that make expensive hardware. Startup costs in these areas include tens of millions for research resources, large staffs, maybe a laboratory. Then, to take their invention to market, they have to navigate complex, entrenched industries, which requires connections. "You need to know how to run a company right off the bat and inspire confidence in investors," says Vivek Wadhwa, a Harvard Fellow who coauthored the Kauffman Foundation report.

In contrast, those sexy Web-service firms that have dominated headlines on and off for the past decade require almost no capital. The "social software" market also rewards people who intuitively understand new media experiences. "There's been social change, too," says Paul Graham, cofounder of Y Combinator, a seed-funding firm. "Ten years ago, it was bizarrely unusual for someone graduating college to launch a startup. Now almost everyone who gets a computer science degree at least thinks about doing it."

In essence, the high tech world divides itself: Young people create the way-kewl consumer software — the Twitters and the Loopts — and older folks tackle the heavy-industry stuff. Young founders hack information; old founders hack atoms.

But we're moving to a world where we need more and more of the latter. Think of some of the thorniest high tech challenges — solar energy, battery systems, plug-in cars. These all reside in the world of atoms. Whoever cracks the problem of carbon sequestration is going to reap a multibillion-dollar reward. But they'll have to solve some hellishly complex physics puzzles and then introduce the solution to an energy industry riddled with byzantine state-by-state regulations and run by an old-boy network of cigar-chewing gazillionaires. Not something easily accomplished in sweatpants.

When you look at it this way, the constant hype over social applications like Facebook or Tumblr can seem a bit misplaced. I'm not saying that Web 3.0 or 4.0 apps are going away (or that they'll stop being fun). But here's my bet: When we finally start solving our global energy and resource dilemmas, the next generation of media-feted tech CEOs will look more like your parents than your kids. Or, to put it another way: Don't trust anyone who wants to put an age limit on innovation.

Email clive@clivethompson.net.