This is part of a series of SA Forum essays produced with the World Economic Forum that runs during its annual meeting, the Summit on the Global Agenda, held in Abu Dhabi November 18-20, 2013.

The human race has never needed innovation more urgently than it does right now. We are causing and facing unprecedented problems, and those problems require unprecedented solutions—in a word, innovation. Much of the knowledge essential to innovation lies hidden within existing patent literature. Yet in its current formulation the patent system—the source of the largest body of open technical knowledge in our species’s history—has swollen to an almost inconceivable complexity and opacity.

It is time to reclaim the knowledge buried in this morass.

Patents are teachings, true recipes for enterprise. By law they are fully open documents that exist for the purpose of enabling innovation. The idea—historically, at least—is to tell the world exactly how to create or implement whatever new, nonobvious and useful tool or process the patent covers. In exchange for this full disclosure the patent’s owner acquires a right to control its use for a specified time and within a set jurisdiction.

Millions of patents are now in force in dozens of countries, with thousands more being issued every day. China, Japan and the U.S. each issued about a quarter of a million patents last year. Every one of those documents differs in quality, reach and implications; reading any one of them is as much fun as root canal surgery.

There are several reasons for this. One is that lawyers have learned to hide the ball inside intentionally opaque patents. The standards of patent-granting agencies tend to range from mediocre to execrable, and from incomprehensible to inconsistent. The whole process is painfully contentious, litigious, expensive and fraught.

Another problem is that the terrain is always changing: each new field of technology gives rise to new regulations and standards. Patterns of influence and control, of ownership and rights shift with the winds. These new patterns rarely become explicit in the patent literature, and yet inventors and investors must identify and understand them in order to conduct business—especially in the most exciting sectors, where new science prevails.

An even more formidable obstacle for innovators is that while a patent may allow you to keep other people from following its instructions, it doesn’t necessarily give you the right to follow those instructions yourself. As the saying goes, a patent is the right to “sue not to do.” Someone else may hold a patent that’s essential to making your invention work. The result: so-called patent trolls are carrying out wholesale aggregation and assertion of patent rights, and warring global firms have adopted a strategy of mutually assured destruction via patent litigation.

These problems are compounded for true innovations—economically impactful products and services built on new science and technology. Real innovations seldom, if ever, depend on a single invention or even a small number of them. On the contrary, modern innovations generally spring from hundreds of discrete items of intellectual property (IP).

You could think of individual pieces of IP as isolated features from a maritime chart: a submerged rock, a coastal inlet, a deep channel, a current, a reef, a shoal, an expanse of open water. Knowing the locations of just some of these won’t help you set an intelligent course. You need the whole chart, so you can minimize your risk and get where you want to be—even if you aren’t sure exactly where that is. The most valuable map doesn’t choose your route for you—it enables you to go your own way, informed of the hazards you might encounter. We can’t know all the opportunities that lie ahead, but it’s good to know where there’s danger.

Making such a map for the global body of some 100 million patents and associated scientific, business and technical literature requires a new, open practice my colleagues and I call “innovation cartography,” operating with a shared evidence platform—a sort of Wikipedia for IP. With machine learning and computational linguistics we can create charts for scientific and technological innovators. Much of the information they need exists already in the patent literature. It’s just a matter of making it publicly available in intelligible form.