Johannes Schunter

What on earth do you mean "can't explain it". It's obvious. The publishing industry is serving them garbage, and ninety-five out of one hundred indie authors is a hack with delusions of adequacy. New authors can't find readers. Readers can't find new authors. They are hidden from each other by a whirling snowstorm of trash. This problem is fixable. Hell, I could fix this problem in two hours if I could get @JeffBezos to lock Kindle's senior management in a room with me and not let them out until they really listened. It works like this. Tradpub, in the 20th century, provided a variety of services for authors. For readers, it provided one important service. Curation. Publishers were investors. They invested money in manuscripts and authors, so they vetted those authors and manuscripts for what they thought people would actually read. The system worked. If you picked up a science fiction novel from a major science fiction imprint like DAW or ACE or Baen, there was a risk it might not be to your taste, but it wouldn't totally suck. Well, the tradpub business model is dead. There's no longer any point for new authors in getting a publishing deal, because they can't do anything for you that Amazon doesn't. They add no value. As an author, Amazon provides me what publishers once did.... a store to sell my books, and a press to print them. What they don't do is provide readers the service that publishers once did. They don't vet or curate novels. Why would they? They sell every novel ever, and putting up an Amazon page for your novel costs them less than a dime. You're the one doing most of the work. They have no financial interest in rejecting anyone ever, because unlike tradpub houses, Amazon is not an investor. They have no skin in the game. Now, there's a tradpub word for the unfiltered, unvetted, uncurated output of everyone who wants to be a novelist. They call it the slushpile. And their job, back when they actually used to do it, was to strain the slush for tiny flakes of gold. Now, Amazon, who replaced them, doesn't do that. Amazon just exposes readers directly to the slushpile. Now, about this moment, you might think I am complaining how a large corporation, or a billionaire, isn't fixing my problem for zero dollars. After all, we just got done spelling out that Amazon sells all the books. Why would they have any interest in promoting one over another? Well, that's where they're wrong. This occurred to me when I was writing my first novel. I happened to mention this project to a acquaintance who was highly placed at Amazon, and he told me it was a bad idea. "The book market is saturated," he said. "We get a lot of data from people's Kindle devices, and we know what they have on them, what they have read, what they haven't and how fast they read. These people have an average of five years of unread fiction on their devices. The demand simply isn't big enough to support more supply." I nodded, and smiled, and said "how interesting", and didn't argue. But in my head, at that moment, I invented an idea I call the Thiel Fallacy: the belief that because monopolies have no competitors, they have no competition. Amazon, as an entity, was effectively bored with books, because they were convinced that, as the only game in town, they were selling all the books that could be sold. That they were selling as many books as people would read. A monopoly does not automatically have 100% of the potential market. It must compete with all of its opportunity costs, including non-consumption. In plain English, if you sell all the books there are, but people can't find a book they like, they're going to just chose not to read a book. Amazon was competing with nothing, and losing. If people have five years' untold novels on their ebook reader, this not mean they have too much stuff to read. It means they have not enough stuff that they want to read. If people could go on Amazon anytime they wanted, and easily buy a book they liked, they would have zero unread material on their Kindles, not five years' worth. So why can't people find a book they like? Because no one is sorting through the slushpile. It's just all being dumped on the readers. Imagine a store that sells everything in the universe, but it's all just dumped in a big pile. No one will buy anything, because no one can find anything. Amazon is like this. It gets an A+ at delivering you the book you ordered, but a D- at helping you find the book you want. So am I saying that Amazon should start rejecting books? Should build a whole publishing department and start reading the slushpile and rejecting things? No. Curation was just how tradpub helped readers find books they liked. It's not the only way or even the best. Amazon has data. Amazon has data from book sales. Amazon has data from Kindle. Amazon has giant computational clusters. Amazon has transformer AI technology. Amazon could build the best and most infallible book-reader matchmaking service in the history of the known universe. Omniscient, content-aware, and personalized. Then Amazon would sell more books. If you control 70% of the total book market, you don't sell more books by increasing your market share. You sell more books by increasing the size of the market. Right now, Amazon sells advertising. To authors. How retarded is that? Authors don't have money. If we did, we wouldn't be writing books and making a couple of dollars a sale. Instead, Amazon needs to advertise books for free. Not just the books that sell the most, but individually targeted marketing of the books each person would most enjoy. You don't even have to ask people which books they liked. They don't have to rate them from 1 to 5. Just look at the Kindles. Look at how fast they read it. Did they stay up until 2am reading The Hobbit? Okay, what else did people who stayed up late reading The Hobbit stay up to read? Build a correlation database. Turn AI lose on it. It should be Amazon's job, not the just the authors', to find readers for books. Because authors may be motivated, but Amazon has the data, and the eyeballs, and the money. They could single-handedly revitalize publishing, they'd make more money, authors would make more money, and readers would love them for it. All winners, no losers. All it would require is a development team, and an understanding of what the real problem is. Amazon's enemy isn't Kobo or Barnes and Noble. It's apathy.

The words "AI slop" were coined to describe an instinctive revulsion. But the people using don't know exactly what it is that's bothering them. It's not the quality of the visuals. It's not the ease of creation. And it's not any sort of style cue. It's the lack of artistic vision. What's artistic vision? It's the combination of two things: - An idea in my mind that I want you to see in the way that I see it. - A plan for how I'm going to get it across to you. Right now, an AI can't have artistic vision, because it has no object model, so it can't have an idea, and it has no theory of mind for you or anyone else, so it can't make a plan for how it's going to show you the idea that it doesn't have. So when you tell an AI "make me a short film about a cat playing basketball" you get slop. Lots of pictures and sound, no artistic vision. But when spend thirty-seven hours alternately banging your head against the desk in frustration, and wrangling the AI until it executes your exact idea and your exact plan, then what comes out will have artistic vision, because you supplied it. Someday, AI will be able to have a true artistic vision of its own. On that day, it will be a person. Until that day, AI is a terrible artist, but a wonderful brush.