"Writing? In 2024?"
Monday, April 29, 2024
(~2,400 words, 12 minutes)
@northshorewave Re: this publishing post:
I've read through the post that was linked, and an earlier related post by the same author that preceded it.
Her position is that the traditional publishing industry is essentially buying books as lottery tickets, paying for most of them using the few big winners they can't predict.
NorthShoreWave - The personal qualities of NSW specifically.
LLMs - Is AI a threat right now? Mostly as spam.
How Many Readers? - One famous book had 3,000 readers on an email list before its Amazon e-book debut, and went on to traditional publishing.
Funding Options - Many authors and artists are currently using subscription services. Some reasoning and numbers are provided.
Illustrations - Should you use illustrations? This lengthy section does a bit of fundamentals analysis of posting to suggest that maybe, you should.
Interaction - Reader replies are one method by which a post will spread.
Search - The people who want to read your story can't read it if they don't know about it. Writing a good book is essential, but only half the battle.
Some thoughts for you:
1 - NorthShoreWave
You implicitly asked if we had discussed your story in detail before, but the answer is that we hadn't. I have a sense of what you're trying to accomplish based on what I've observed of your character. While you think of yourself as seething, I think you're actually wise, compassionate, self-aware, and able to view things from multiple perspectives. A significant number of people are much worse at practicing at least one of these virtues. On its own, that's not enough to write a best-seller, but I think it does provide you with an advantage.
2 - LLMs
Based on my experiments (see @mitigatedai for some logs), I wouldn't worry about competition from AI. For you, the chief issue caused by AI will be spam. AI moves sideways (different text) and downwards (less meaning). I may tell LLMs to "combine Inspector Gadget and Death Note," but...
Do I actually use the information provided? No.
3 - How many readers do you need?
From one of those publishing posts, to get a sense of the number of readers you need...
Andy Weir first published The Martian as a serial for his own blog, then as a self-published novel on Amazon, then as a traditionally published novel with Random House.
“I had an email list with about 3,000 people on it, so, initially, the audience was roughly that much,” he tells me. “When I first posted it to Amazon, I didn’t do anything to market or publicize it. All I did was tell my readers it was available there.”
The book was on Amazon for five months, at a price point of 99 cents, and he sold 35,000 copies before Random House bought the rights in February of 2014.
Note that being a provocative firebrand doesn't necessarily mean you'll sell copies. Some politicians with tremendous name recognition failed to move copies of their books.
4 - Funding Options
I don't recommend using a Kickstarter to publish your book at this time or in the near future. You just don't have the name recognition, but also, Slashdotter Caimlas (who I don't know, so I don't know how trustworthy he is) wrote:
I'm personal friends with a number of authors who publish books in one of several subgenres. Mostly, they rely on Amazon's Kindle Unlimited: some of them are prolific enough that their book sales account for most of their income, simply based on peoples' reading of their works.
Mostly, unless people want a piece of history or something they can reference, folks seem to hate having clutter. Fiction that sells isn't usually, primarily sold as a hardcopy book anymore, I don't believe - short of the kinds of books that end up at the end of the grocery store isle or in an airport novelty store.
A lot of publishing is done online these days, often through subscription services such as Substack (for essays) or Patreon. (Kindle Unlimited is also a subscription service, costing $12/mo.) As an example, the webcomic Spinnerette has a Patreon (bringing down $3.3k/mo), and then runs Kickstarter campaigns for print runs (volume 8 raised $27k).
To give you an estimate, Spinnerette's Patreon has only 536 subscribers, and pulls down $3.3k/mo, but you probably haven't heard of it. El Goonish Shive, which I'm confident you have heard of, brings in $3.6k/mo on 2.4k subscribers. The famous Kill Six Billion Demons has ~5.4k subscribers, bringing down ~$8.4k.
In Patreon terms, a good foothold to try for might be 100 subscribers at $3/mo each, with an initial focus on getting to 50.
5 - Illustrations
You've posted some drawings. They have some character, showing that you have the basic aptitude to develop the skill if you applied yourself to regular practice. However, the proportions are too far off to attract much attention (except as a stylistic choice, which, I can tell, it is not).
This blog tends to break things down into their abstract fundamentals for analysis. I promised myself I wasn't going to do that here, but eh, we'll do just a bit.
To quote one of the publishing articles...
“People tend to buy the books that are already really popular,” Deahl says. “They look at the bestseller list to see what they want to buy and that reinforces this tiny amount of books at the top. It’s a very top-heavy system. The tricky thing in publishing is success begets success. But it’s really hard to create that spark.”
Let's stop to think about this.
a. Banter - Fame
There is one layer to this that you can't do much of anything about, which is that people will watch the same shows their friends watch in order to have something to talk about with their friends.
b. Investment - Background
However, there is another layer over which you have more influence. It's very easy to make a quick judgment of a movie based on its visuals, or a short trailer. It's also relatively easy to judge short songs, since they're only a few minutes long (but I don't find myself doing this often).
In order to judge a book, you have to read the text and process it. You can't make a snap judgment off a single picture, because you have to read the text first to produce the mental picture.
This website does have viral text posts, but they're like...
You seem to have fundamentally misunderstood me, Anon. Go read all 5,640 posts again.
Some of these posts can get a bit long, but it's usually a back-and-forth where each individual post is short. Often, they'll mix in images, or memes.
People supposedly read at 200 words per minute. Based on that estimate, this blog's most viral post of all time can be read in 5 seconds. That's about the same amount of time someone would spend looking at a jpeg.
That doesn't mean people don't enjoy effortposts. They will follow a blog upon encountering a good effortpost! They just don't like or reblog them.
I think you already know this part, but for "acceptable" length for reblogging, it's usually best to keep it under one "Tumblr page," meaning around one screen length on desktop, or around 200-300 words. I've talked about this part before, but if the reader can see the end of the post, it feels like less of an investment to read the post, and reblogging it won't fill up a friend's Tumblr dash.
Obviously that's tough for long-form fiction, because it has to load more context about the characters in order to establish the stakes. (Unless it's fan fiction, where the audience already knows the characters.)
c. Investment - Strategy
As you know, this blog will sometimes post political cartoons and other illustrations as part of its general stream of content.
The obvious strategy is just to have some nice-looking character images or images of scenes from the story. It can be "read" faster, so it's more shareable.
I think that strategy suffers from a weakness in that it's easy to just look at the image and disregard the text. This would reduce your fiction blog to an art blog - and it is not an art blog.
Therefore, I would like to gently suggest - and keep in mind, I do not have any published novels - a different potential approach. This proposal is speculative, and this technique is not widely used.
Do you know that famous Rockwell painting, Breaking Home Ties? Rockwell is a master of telling a story with just a single still-frame painting.
Rockwell has to tell the whole story in one picture, because that's the medium he's got to work with. This limits how much story he can tell. As an author, you don't have to limit yourself to what can be told in just one image, because you have the text.
This strategy would involve a two-step maneuver.
First, the image at the top of the post communicates the essentials that the reader needs to know about the characters through the composition of the scene (so that they don't need to read background material), as well as various subtle details, while raising questions, also through the use of details/etc, to increase the viewer's curiosity.
Fortunately for the viewer, second, the questions raised by the image are answered in the text right below it.
The post would form an entry point into a network of related posts; tags for particular characters could be linked at the bottom, or links to other posts in the sequence.
Secondary characters would be ideal for this, because you can manipulate their scenarios/context/character to fit the short format, while your overall project will focus on the main characters and thus have a greater, long-term narrative investment for appropriately larger payoff.
As I wrote in my post on 'text wall memes,' people will read text in an image, and they'll even reblog it, but it's contextual. So again, this is speculative, but it should be feasible. It's a matter of creating the appropriate context.
d. Investment - AI Art
I don't think you should use AI-generated art. Yes, people will be able to tell, but the even bigger problem...
Compare this AI knockoff to Norman Rockwell's original Girl with Black Eye.
The expression is wrong. The pose is different. This is a completely different story from the one Rockwell was telling! The prompteur 'borrows' the right 25% of the image from Norman's original because he can't reproduce it. And what is that random white cloth on the left side of the image?
There is a significant reduction in the amount of intention in the image. Putting it back in involves working over the image, repeatedly, usually with inpainting, and often working against what's in the AI's training data, forcing it to pull from more and more improbable parts of the distribution (until eventually, there's no matching data in the training at all; you have to get out and draw it yourself).
I'm going to borrow a post of my own here from 2019.
This isn't oriented towards the strategy I've described, and it only got 21 notes, but note the teacup with steam and tea bag tag, the obscured flag in the background, and the Youtube-style video tracker on the bottom. The combination of the special effect, text that looks like a subtitle, and video tracker imply that the image is a screenshot from a streaming anime.
The character is casually (as indicated by the cup of tea) sitting at a computer desk (as indicated by the faintly sketched keyboard and hand position for a mouse). What's that flag in the background? It certainly doesn't belong to any extant country. (In fact, as the artist, I'll tell you - it's based on an O'Neill Cylinder.)
Obviously this art is very much just a sketch in quality terms. An AI rendering usually looks much fancier. However, an AI would not put that detail in.
e. Investment - Technical Skills
However, I will suggest the use of software if you go this route. (Or the hiring of an artist, but that could get expensive.)
Your issue is with proportions. Lots of people have trouble with proportions. (You also have trouble with hands. Lots of people have trouble with hands.)
One way to deal with this is to just train. You'd be surprised at how fast you improve if you draw from realistic sources such as photographs an hour a day for a year, even if it's just a quick sketch. You probably aren't willing to do so. You have other things to worry about, including writing.
However, you could use posing software. You could save the proportions of several characters and position them throughout the scene, as well as having a grid for the ground and potentially other props to help with positioning of items like lamp posts or the edges of buildings. (I've experimented with posing software a bit myself.)
Dan Shive (of El Goonish Shive) does not use posing software as far as I know, but he has used 3D software. Although his style is cartoonish, one thing people like about him is that he does put effort in at improvement, and the quality of his work has improved substantially. (That was actually the inspiration for the second part of the "in 2028, Hollywood runs out of ideas and adapts El Goonish Shive" post.)
6 - Interaction
Though shorter posts tend to go more viral, I find that posts which someone can reblog and share their opinion tend to show up a lot in my top posts (as long as they're only about one tumblr page long). The MOON PRISON poll is a good example of something that's approachable and neutral, but fits heavily with the themes of my blog, but other posts may take a political position that invites disagreement, resulting in discourse, and get reblogged that way. (You may also remember the silly Swift Pill poll.)
I don't recommend courting disagreement on purpose. Not only is this bad for the social environment, but it tends to make people go crazy.
7 - Search
I think you've probably noticed some of this already and are working with it (posting short excerpts, initial art). Most of this is, again, speculative. This is all just information for your consideration.
Writing a good book is the first problem. Getting the readers who would enjoy the book to find it in such a noisy environment is the second problem. I think you can do it, but if your trajectory isn't currently looking as good as you want (e.g. # followers on your story's sideblog), I would recommend expanding your strategy so that you're in a good position when the book itself is ready to launch.
10 notes
·
View notes