Book publishing is a storied industry. According to industry legends, Bennett Cerf and Donald Klopfer founded Random House in the 1920s to publish "a few books... at random." Sir Allen Lane set up England's Penguin Books in the 1930s after he couldn't find a good paperback to read while waiting for a train. When I covered the London stock market in the late 1990s, I occasionally wrote about a small-cap publishing company called Bloomsbury: No one had heard of them -- until they discovered a new children's author and some books she'd written about a boy wizard.

Today there are 70,000 Americans working for book publishers, about 15% less than a decade ago. Many thousands more work as agents, editors, designers and printers. I wish them all well. But I suspect they all need to start typing their resumes. I can't see how this industry survives. Indeed, I fear the end is nigh.

The good news is that it's never been easier to become an author. That novel -- or novel idea -- you've been nursing all these years? You can get it out there, in a matter of days.

I've just written a biography of Mitt Romney, based (in part) on my notes from my days as a columnist on the Boston Herald when Romney was governor of Massachusetts. I'm self-publishing it, through Amazon. The process is so fast and easy it makes me wonder why anyone would go to a traditional publisher ever again.

(The main downside is that you do not get an advance. In this instance, my biggest risk is that Romney will blow the election with his gaffes before I can reach the shelves.)

Yes, there's been a learning process. I've had to pick up a bit about formats and templates. But I'm a technological Luddite -- and even I've mastered it. In simple terms, you copy your text into a Microsoft Word template and upload it. You can then design your own cover in Photoshop and upload that too, or use one of their templates. That's about it. It takes about two days to get your proof back, and a few more days to launch the book. Then it's available, like any other, at Amazon.com.

I'm also publishing the book on the Kindle, and that's even easier -- about twelve hours from uploading your document to publication.

This is an amazing contrast to my last experience publishing a book. Then I went through a traditional publisher. The less said about the experience the better, but the bottom line is that it took eighteen months from idea to print. Just after Lehman Brothers collapsed I went to an agent and said, "This crisis is going to be much deeper, longer and worse than people realize. I want to write a personal finance survival guide." But it took months to get a publisher to sign up, and then the turnaround time was nearly a year. By then the public mood had then changed. (It has since changed again, of course.)

A few days ago Penguin, which is now owned by Pearson, announced that its first half operating profits had fallen by nearly 50 percent. It looks worse than it is: The profits are merely back to where they were a few years ago. But what I suspect is happening to the industry is a two-stage process. First, the arrival of ebooks benefited publishers. They've been able to publish their books in electronic format and collect a bigger net margin. But then there's a second stage: Authors cut out the publisher completely and publish themselves. That's what's happening now.

Under the old system, the author was lucky to get 15% of the cover price. Today they can keep much more -- as much as 70% for electronic books.

Once upon a time we bought books in stores. The publishers had privileged relationships with the store owners. So you couldn't reach the market without them. Today, that whole structure has gone. This doesn't mean the end of all those jobs in publishing. If you are writing a serious book, you still need a development editor, a copy editor, a proof reader, maybe a cover designer, and a marketer. But in future such people will be freelancers. You'll hire them on a project basis. The best agents will restructure their roles to be managers, or "producers". They'll bring the various people together for the project.

Even under the current system, you have to do your own marketing anyway. Naturally, I'm having to deal much more with Facebook and Twitter than I have done in the past. It's a new world, brave or not.

Forrester Research predicts that ebook sales will rise from $2.2 billion last year to $10.5 billion by 2016. But even if they do, where will the money go?

In the old days, publishers could use the massive profits from the likes of John Grisham or Patricia Cornwell, and use them to help subsidize other, less profitable books. The format of the bookstore encouraged browsing: I'd go into to buy one book and pick up five others that caught my eye.

In the future there won't be the same browsing, and there won't be any subsidies. At some point, people like John Grisham or Patricia Cornwell will be able to self-publish and keep close to 100% of their output: The only constraint will be Amazon, which controls access to the Kindle.

Self-publishing will produce a tsunami of books. Very few will be financial successes. Very few will be any good. Sifting your way through the chaff for the wheat will be much harder than browsing through a bookstore. I suspect the best-selling few, like Fifty Shades of Rubbish, will "crowd out" the rest. Once upon a time Internet cheerleaders talked about the so-called "fat tail," the idea that the Internet would make marginal products profitable. The reality seems to be the reverse: The overwhelming dominance of the few.

What does this mean for the economics of writing? We'll see. Last week I had lunch in Miami with the writer Gerald Posner. He's working on a big history of the Vatican and its wealth over the centuries. It's a fascinating and important topic. The book will probably take five years, from start to finish, to research, write and publish. Will these kinds of "big books" even be economic in the new electronic age? No advances, and no cross-subsidies. The writer takes a massive risk, with pretty minimal upside too. I have my doubts.