For Week Ending October 16, 2026

Smashwords finally takes preorders, September's Kindle Unlimited pot lands at $65.9 million, and Substack takes over your sales tax on December 1. Plus free AI tools for authors, a Tesla that flies, and why "too niche" might be your best marketing plan.

Publishing News

Half of Audiobook Listeners Open to AI Narration

According to Publishing Perspectives and Publishers Weekly, an Audible-commissioned NielsenIQ survey of 18,000 listeners across 11 countries found that half are open to trying AI-narrated content. The resistance publishers assumed would hold is softer than expected. At the same time, audio is not draining print and ebook sales. It is feeding them.

Audible's chief content officer Rachel Ghiazza presented the findings from "The Future of Listening" at the Frankfurt Book Fair. The survey, fielded in August across Australia, Brazil, Canada, France, Germany, Italy, Japan, Mexico, Spain, the UK, and the US, asked listeners about audiobooks, podcasts, and audio series. Fifty percent said they were now open to listening to or trying AI-narrated content. Openness ran from 42% in Germany to 66% in Japan. In most markets, listeners were most open to AI narration for fiction. Germany and Japan preferred it for nonfiction.

Ghiazza called this an "and strategy." Human performers remain. Production savings are meant to fund more titles. She told hesitant publishers the time is not coming. The time is now. Quality and transparency matter. Listeners want to know what they are hearing, and it has to sound good. Openness is not the same as preference or purchase behavior, and Audible has not released the full methodology.

The cross-format numbers sit next to the AI finding. Sixty-one percent of listeners had read a title in print or ebook after listening to it. Sixty-nine percent had listened to a title they had already read in another format. Sixty-three percent who heard an author on audio went on to buy that author's print or ebook titles. Since starting with audiobooks, 29% had tried new authors and 29% had explored new genres. Twenty-six percent were reading more ebooks and 22% more paper books. Thirty-four percent said they listened to more audio in the last twelve months than the year before, highest in Brazil at 46% and Mexico at 44%.

Narrator loyalty did not disappear. Sixty-five percent of listeners seek out more books by a narrator they like. Genre remains the top factor in title selection, followed by topic, then narrator. Audio is also pulling time from screens: 51% said it helped them intentionally replace screen-time entertainment, and 47% credited it with reducing social media use.

Audible's stated goal is that every book becomes an audiobook in every language. The company already has more than 1.5 million titles. Ghiazza noted that romance listeners in particular "just want volume," and that range of production styles is expanding. Internal AI tools already help set up narrator productions by breaking down the book, defining characters, and handling scheduling. The platform is also expanding Read & Listen, which pairs ebook and audiobook, into more markets.

Why Authors Care

Half of listeners will try an AI-narrated title. That is the number that breaks the working assumption that synthetic voices stay on the shelf. Fiction is the more open category in most markets.

Audio still sells the other formats. Sixty-three percent who discovered an author on audio bought that author's print or ebook work. A low-cost AI audio version can function as discovery that moves the paid formats, not as a replacement that kills them.

Human narrators still have pull. Sixty-five percent actively seek more books by a narrator they like. Flagship titles and series with an established audience still benefit from the human voice. The long tail and backlist are where the cost math changes.

Transparency is the practical requirement. Label AI narration clearly. Listeners who will try it still want to know what they are getting. Quality that sounds cheap will fail the same test a weak human performance fails.

The cost gate is the one indie authors can open this week. Titles that never cleared a professional narrator budget can now reach the audio catalog. Use the audio as the funnel, keep the human narration for the titles where the voice is part of the brand, and measure whether the print and ebook sales move.

Sources:

Frankfurter Buchmesse 2026: Audible-Commissioned Survey Finds Audio Sends Readers Back to Print; Half Would Try AI Narration (Publishing Perspectives, Eric Dupuy, Oct 9 2026)
Frankfurt Book Fair 2026: Audible Wants to Turn Every Book into an Audiobook (Publishers Weekly, Ed Nawotka, Oct 8/9 2026)
The Future of Listening: Audible unveils findings from NielsenIQ's comprehensive global study (Audible press release via Podnews, Oct 9 2026)
Audible Reveals NielsenIQ Data: How We're Fueling a Global Listening Surge (Audible newsroom, Oct 8 2026)

Commentary

Jonathan: Half of listeners will try an AI-narrated title but that's not the same as saying half of them like it.

Thomas: When you're deciding what to do, you have to compare an AI-narrated version of your book to your next best alternative. For some authors, that alternative is narrating it yourself, and you may not be better than a high-end AI narrator.

I have returned audiobooks on Audible because the author narrated it himself. I was so excited to read one book in particular, but the audio quality was so bad it was unlistenable. And I have listened to entire books narrated by the bad Kindle AI, which sounds like a robot, and that was preferable to some humans.

We're seeing this across the market, with book cover design and with writing. AI dominates the middle of the bell curve. There is a level of awful that only humans can attain, and there's a level of excellence that only humans can attain. But AI is good and getting better at producing average-quality content, whether it's writing, images, or narration.

Many fantastic writers have never taken a theater class, never done any acting, and never done voice work.

If you're choosing between AI narration and hiring a professional like James L. Rubart, obviously James L. Rubart will provide better voice quality than AI, as will any of the celebrity narrators. Michael Kramer will always outperform AI. But if you can't afford Michael Kramer and you were going to do it yourself, AI may be preferable.

You can also do both.

It's interesting that Audible is producing this research when it isn't very pro-AI on its own platform. Audible doesn't currently allow AI narration in most contexts. So the strategy is to put a human narration on Audible and an AI narration somewhere else. That way you can be exclusive with your human narration on Audible and still get a bite of the apple on Spotify or wherever else.

Jonathan: You also have to think about narrator loyalty. Sometimes people listen to a book just because their favorite narrator did it, so a narrator can be a discovery mechanism.

If you get Michael Kramer to do your book, a bunch of Brandon Sanderson listeners might listen to it because Michael Kramer just put out another book.

With AI, you don't get narrator loyalty, and in the audiobook world that's an extra boost for your book.

Thomas: But you pay a premium for the celebrity. You can find a good voice at a good price, but a good voice with a following of listeners who follow that narrator from book to book costs more. That could be a good investment. Just realize that if you want Patrick Stewart to narrate your book, you'll pay extra.

Smashwords Finally Takes Preorders, and Most Authors Still Skip the Tool That Earns More

According to the Draft2Digital press release, the Smashwords store now accepts customer ebook preorders for the first time. Preorder sales give independent authors another storefront to build anticipation, capture early orders, and create stronger momentum for upcoming book launches. Draft2Digital already distributes ebook preorders to Apple Books, Barnes & Noble, and Kobo. Smashwords was the missing piece on the wide side.

What Happened

Ebook preorders are an essential best practice for stronger book launches, yet fewer than 20% of Draft2Digital's 300,000+ authors take advantage of this marketing tool. Draft2Digital has observed for years that preorders earn significantly more sales than books that are simply uploaded on release day. Authors can list preorders at Smashwords for up to 12 months ahead of release. Customer credit cards are not charged until the book is released, which can be days, weeks, or months later. All accumulated orders are credited to the book's sales rank on release day. That concentration of sales on one day boosts the Smashwords store's algorithms and increases the new book's visibility among shoppers.

The same retailers accept assetless preorders. Authors can list upcoming books even without the final manuscript or cover art, as long as those assets are uploaded at least 10 business days before the release date. Assetless preorders give authors the flexibility to list longer and accumulate more sales before release.

Preorders Plus Presales

Preorders build anticipation, but readers who purchase on preorder must wait until the release date to access the book. Smashwords Presales gives authors the chance to offer enthusiastic readers early access so they can purchase, read, and review before the public release. Listing the book as a preorder is required to use the Smashwords Presales tool. While the book is on preorder, authors can run one or more presales.

Presales can be public to all Smashwords store customers or private to a newsletter list or street team via a special link. Authors can offer a discount incentive in exchange for an email address. Email capture is only available during the presale period. Presales are transacted only at the Smashwords store, before the preorder releases to the public. Final manuscripts and cover art are required for presales because readers get immediate access.

Kris Austin, CEO and Co-founder of Draft2Digital, said in the release: "Preorders and presales give indie authors powerful, complementary ways to build buzz around book launches. When Draft2Digital and Smashwords merged, we promised indie authors we'd continue to innovate publishing tools at D2D while bringing significant improvements to the Smashwords store. We're delighted to keep that promise."

Why Authors Care

Fewer than 20% of Draft2Digital's 300,000+ authors use preorders, even though Draft2Digital says those titles earn significantly more sales than release-day-only uploads.

Accumulated Smashwords preorders credit the sales rank on release day, concentrating visibility when the book goes live.

Assetless preorders let authors start the clock up to 12 months out and still push final assets as late as 10 business days before release.

Pair a preorder with a private Presale link for newsletter or street-team early access, and use the discount option to capture emails during the window.

Authors already on Draft2Digital can turn this on from the existing dashboard; the distribution flows to Smashwords alongside Apple Books, Barnes & Noble, and Kobo.

For indies already wide on Draft2Digital, the practical move is to set the next title as a preorder now, even if the final files are not ready, then layer a private Presale for the street team once the manuscript and cover are locked. That stacks early orders, rank credit on release day, and newsletter growth on one storefront that previously lacked the option.

Sources:

Commentary

Jonathan: All accumulated orders are credited to the book's sales rank on release day, although Amazon has been playing around with that.

Draft2Digital's number makes sense to me. I see more authors leaning toward the Kindle preorder and uploading to Smashwords as an afterthought, as a baseline strategy.

Thomas: It's a real missed opportunity. One of the biggest things holding books back is a low number of reviews, but a preorder can't have reviews. The longer your preorder window, the more sales you can get during the preorder window. That means more verified purchases on launch day and more verified readers who can leave a review that helps you moving forward.

And if you're just getting started, you don't want to be haunted by only 12 reviews. You'd much rather send people to a preorder page where they can't leave reviews.

The day before Christmas is always more magical than the day after Christmas. So don't rush to give your readers that day-after-Christmas feeling. Give them the Christmas Eve magic. Allow a longer preorder.

There's a strategy common among a very small but influential subset of authors who say, "Don't do preorders. You want all your sales in one day. You'll be more likely to get to number one if you don't do preorders."

That strategy only works on platforms that don't count preorders on day one. Apple Books, for instance, either gives you two bites at the apple or saves up all your preorders and dumps them on launch day. So a preorder strategy is actually better if you're trying to climb the charts on Apple.

Amazon, which is the one everyone pays attention to, doesn't give you two bites at the apple. Sales count on the day they happen. If you got 1,000 orders all on launch day, that would put you at number one. When those orders are stretched out, you're less likely to hit number one on launch day.

But even that matters less now, because the algorithm is a rolling average. It prefers sustained engagement, so a lot of sales on one day isn't as good.

Jonathan: You want a good launch week.

Thomas: The arguments against preorders keep getting weaker.

Another advantage of a preorder is that it gives you time to perfect your page, whether it's on Smashwords, Amazon, or anywhere else.

Once your page is complete, with your book description, cover, A+ content, and everything else, you'll invariably notice problems or improvements you want to make. Those changes can take days to show up for all users because of the caching Amazon, Smashwords, and the others do.

Six weeks or six months of preorder gives you plenty of time to tweak your page until it's perfect, while still making money and still doing media. Once you have a preorder page, you can do podcast interviews and start sending people there.

More authors need to embrace preorders on more platforms. I would like to hear someone articulate an argument against preorders in 2026, because I haven't heard a recent argument, based on the current algorithm, for why a preorder is a bad idea.

Barnes & Noble Traffic Up 34%

According to Publishers Weekly, which reported Placer.ai analysis on October 5, 2026, bookstore foot traffic has rebounded after the pandemic plunge. Barnes & Noble visits sit 33.6% above 2019 levels, while visits to all other bookstores rose 16.5%.

What Happened

The gains are not just from new stores. Placer.ai found quarterly same-store visits have increased year over year consistently since at least the first quarter of 2024. The chain opened 57 stores in 2024, more than 60 in 2025, and plans at least 60 more in 2026, according to the same report and Chain Store Age.

Customers also stay longer. On weekends, 49.3% of Barnes & Noble visitors remain more than 30 minutes, versus 47.4% at other bookstores. The weekday gap is wider: 45.6% versus 41.9%. Placer.ai attributes part of the linger to cafes and the chain's effort to function as a third place outside home and work.

The data measure foot traffic and dwell time, not what people buy. No public source splits the visits or the basket between books and non-book merchandise. Toys and games are a real part of the assortment. Marnie Lawrence, Vice President of Merchandising for Specialty Products, oversees Toys & Games, Gift & Paper, Newsstand, and Music & Video. Axios reported that CEO James Daunt refreshed non-book items and bet big on toys and games as part of the turnaround.

In a January 2026 Publishers Weekly interview, Daunt said 2024 total sales were driven by non-book categories while book sales were "fine." In 2025 the toys and games department had a banner year alongside solid book sales. The broader toy market is hot. Circana data showed U.S. toy dollar sales up 6% in 2025, with Games & Puzzles leading at a 37% increase, driven largely by Pokemon.

Why Authors Care

Higher foot traffic at Barnes & Noble improves the odds for in-store events, signings, and discovery, especially in stores with rising same-store visits.

Longer dwell times create more browsing opportunities, but some of that time may be spent in the cafe or the toys and games section rather than the book aisles.

Authors chasing placement should still work local booksellers for front tables and events, then verify sell-through, because traffic headlines do not guarantee book units.

Independent bookstores also saw traffic gains, so the recovery is not limited to the chain.

Sources:

Commentary

Thomas: It's not hard to be up from pandemic levels when it was illegal to go to Barnes & Noble. But traffic is up 34% from 2019.

Jonathan: The chain's effort to be a third place means Barnes & Noble is now a hangout or a date night. Something else is feeding this that Gen Z prefers physical copies. We're seeing in the data that Gen Z likes to read physical books, and they like going to Barnes & Noble.

Gen Z is a bunch of people in their early 20s. Backtrack seven years to 2019, and these were teenagers in the pandemic era who weren't allowed to go anywhere. Now they get to experience community while also experiencing a physical book. I think that's contributing a lot to this.

Thomas: I also feel like Barnes & Noble is filling the role bars filled for Gen X, a place to look for a boyfriend or girlfriend.

Jonathan: Gen Z doesn't drink.

Thomas: Right, they don't drink alcohol. It's a meme among Gen Z: "I'm going to find a girlfriend at a bookstore. I want a nerdy girlfriend," so you go to the manga section. Or a boyfriend. There's a lot of hanging out and seeing who you might meet.

That sort of thing is a cultural shift.

Jonathan: There's this one girl who keeps tripping and dropping all her books in front of every guy.

Thomas: That's what we mean by a third place: a hangout where you spend time.

If you go to a Barnes & Noble, you'll find the manga section, which is typically the biggest section, often has people sitting there reading, because manga, and comic books generally, are very expensive.

Jonathan: A bookstore is a cheap hangout. At a time when everything is more expensive, books aren't getting that much more expensive. If you want to hang out with your book-loving friends, Barnes & Noble has a cafe, easy chairs, and couches.

You can read a chapter of a book you might be interested in, look at different covers, and walk around together in air conditioning, which is a big deal here in Tucson. It's quiet, and that quiet is enforced, which appeals to introverts.

Thomas: It's also quiet because books absorb sound. Even if somebody is being loud a couple aisles over, the sound doesn't travel through the books very well.

The other reason Barnes & Noble is doing well is toys and games, which we covered in a recent episode. They're hot among millennials, and childless families now spend more money on toys than families with children. That's concerning on a societal level. We've reached full Chitty Chitty Bang Bang levels of societal collapse.

Jonathan: I went to Best Buy to get my Mac Studio this week, and right there on the sliding doors were three different announcements about Pokemon cards, saying, "We don't have any in the store. Don't stand in lines out the door." Those signs are for adults.

Author Alerts

Grok Bot Webinar #2 Explained

Grok Bot for Authors live workshop with Thomas Umstattd Jr., October 22, 2026

Webinar Details

What: Grok Bot Training for Authors (More Info)
When: October 22 2026
Where: Crowdcast
Host: Thomas Umstattd Jr.
Cost: $37
Replay? Yes (For those who register)
Text Transcript? Yes (For those who register)

Core Quill Collection

Quill Command, Efficiency Quill, Strategy Quill, Webmaster Quill, Newsletter Quill, Scam Spotter Quill, Super Grok Quill, Security Quill, and Grok Bot Coach Quill.

Quill Collection 1 (September Webinar)

Amazon Advertising Quill, Book Keeping Quill, Continuity Quill, Editor Quill, KDP Quill, Meta Advertising Quill, Metadata Quill, Promo Stacker Quill, and X Ads Quill.

Quill Collection 2 (October Webinar)

PR Quill, Book Launch Quill, Kickstarter Quill, Canva Quill, Live Events Quill, ACX Quill, Audiobook Editor, Podcast Production Quill, Podcast Editor Quill, and Book Trailer Quill.

Commentary

Thomas: There has been a lot of confusion, and I want to sort it out and explain how it works now.

Last month I did a webinar on Grok Bot for authors. It was probably the best-received webinar I've ever done. If you go on AuthorMedia.social, you'll still see people talking about it, including the Quills I gave away on the webinar.

The plan was to do a second Grok Bot for Authors webinar in October, and that webinar is coming up next week. It was meant as a continuation of the first one. People who came to the first could come to the second and get more Quills, more training, and answers to questions based on their usage. People who missed the first could come too.

A Quill is a special Grok Bot I programmed myself with my expertise built into it. Part of what I gave away at the first webinar were Quills every author needs, like Quill Command and the Efficiency Quill, which are meant to work together as a system. But the idea of giving all of the Quills from the first webinar to people who came to the second rubbed people the wrong way.

I don't want early adopters to feel they got a bad deal.

So I'm breaking the Quills into two kinds. Every Grok Bot webinar I do comes with the Core Collection. Quill Command is the coordinator, the chief-of-staff Quill. The Efficiency Quill reduces your usage by working with the other Grok Bots. The Strategy Quill helps you plan your day or week and figure out what to work on next. Then there's the Webmaster Quill and the Email Newsletter Quill.

I'm also adding new core Quills. If you came to the first webinar and aren't coming to the second, you'll get these bonus Quills next week once I have them fully deployed.

The Scam Spotter Quill detects the kinds of scams that target authors and tags them in your inbox. It's trained on all the scams we've been spotting and reporting on AuthorMedia.social and other places. The SuperGrok Quill connects to SuperGrok for research and writing. The Security Quill runs security audits on you and your website. The Grok Bot Coach Quill coaches you on using Grok Bot better and answers your questions.

Quill Collection 1 had a lot of advertising Quills, plus a Bookkeeping Quill, an Editor Quill, and a KDP Quill. It was heavy on publishing Quills. You have to sign up for the replay of webinar one to get those.

Webinar two next week has a whole new set. The PR Quill writes press releases and helps you find podcasts to guest on. The Book Launch Quill distills everything I teach on book launches into a single Quill that will do your launch for you and with you. The Kickstarter Quill does the same for your Kickstarter.

The Canva Quill generates images in Canva, and the other Quills can use it too, so the Kickstarter Quill can talk to the Canva Quill to generate images for your campaign.

The Live Events Quill finds you speaking engagements and places to have a book table, like festivals and book fairs. The ACX Quill helps you publish your audiobook. The Audiobook Editor Quill edits your audiobook and gets the settings right so it's compliant wherever you publish. The Podcast Production Quill manages the production of your podcast, and the Podcast Editor Quill edits its audio.

Finally, the Book Trailer Quill works with the Patron Toolbox to produce a book trailer for your book. I've got the fundamentals working, and I'm still working out some kinks.

I'm still not a big fan of book trailers, but you'll be able to do the whole thing within the usage you already have. To have Grok Bot you need SuperGrok, and SuperGrok gives you credits with Grok Imagine, a video generator you can use for book trailers. It doesn't cost you anything extra, and you can have a lot of fun with it.

The main reason to come to the webinar is to learn how to use AI and Grok Bot. It comes with these Quills, and you'll also learn to build your own Grok Bots. To sign up or learn more, go to authormedia.com/grokbot.

I may do future collections at future webinars. Next month I'm planning a home-and-health themed one that will be less for authors and more about managing the household, managing health issues, and optimizing the rest of your life.

Substack Takes Over Sales Tax on Paid Subscriptions December 1

According to the Substack Help Center, starting December 1, 2026, your paying Substack readers will see sales tax at checkout. Substack will now calculate, collect, and remit sales tax or VAT on paid subscriptions for you. This turns on automatically for every publication with payments enabled, and Substack doesn't charge extra for it.

For writers in the US and Canada, the tax is added on top of your price. You keep the same amount per subscription. The reader pays more. Writers outside those two countries can choose inclusive pricing instead, which takes the tax out of the price you already listed.

Part of this is a real improvement. It replaces the old Stripe Tax integration. You can stop remitting on payments Substack processes after December 1. Anything owed before December 1 is still yours to file. Publishers in Brazil, Mexico, Malaysia, India, and Thailand remain responsible for all of it.

You can't opt out. Substack's guidance doesn't describe a way to decline the arrangement and keep paid subscriptions. The July 20 Publisher Agreement still gives publishers broad responsibility for tax. Substack does the paperwork, but the contract still names you.

Before December 1, take three steps:

Open Payments settings and confirm your taxable content types, so a serial with a chat add-on isn't taxed at the higher rate.
Finish any filings you owe under Stripe Tax for the period before December 1.
Tell your readers about the new tax line before their renewal receipt arrives and they email you about it.

Sources:

Commentary

Thomas: This is such good news for authors. A lot of authors haven't been collecting or remitting their sales taxes. I've found it's common among authors to assume their sales tax authority won't find them, and they're not necessarily wrong. But it's not good practice to skip collecting and paying sales taxes.

It is a hassle, especially under current law, with lots of different places to pay. I always prefer platforms that handle it for you. Teachable, Payhip, Patreon, and now Substack all collect and remit sales tax for you.

Platforms like Shopify will tell you how much you owe, but calculating the tax is the least difficult part of the process. Paying the money isn't very difficult either. The real challenge is filing the tax return.

Do you know how South Dakota does sales tax returns? Do you know how that's different from North Dakota, Texas, or California? Do you want to deal with all that complexity for $5.00 worth of sales tax? No, you do not.

September KU Page-Reads Paid From a $65.9 Million Pot

According to the KDP Community update, the September KDP Select Global Fund came in at $65.9 million. That's $7.2 million more than it paid last September and down a $100,000 under August. From one perspective this is the second month with a decline, but from another perspective this is the forth highest payout ever. The Global Fund is the monthly pot that pays KU page-reads for every ebook enrolled in KDP Select.

The longer trend is the real story. September's fund is up 12.3 percent from September 2025's $58.7 million. Through the first nine months of 2026, Amazon has put $584.1 million into the fund. That's $65.1 million more than the same nine months last year, a 12.5 percent increase. Five years ago, September's fund was $39.7 million. In July 2014, KU's first month, it was $2.5 million.

The bonus is the one number that didn't grow. Last September's total payout minus that month's fund also comes to $3.4 million. The All Stars bonus stayed flat, and all of the growth went into the page-read pot.

A bigger pot doesn't guarantee you a bigger check, though. The fund is split across every page read in KU that month, so your payout depends on your own page-reads and the per-page rate. Written Word Media's tracker hasn't posted September's rate yet. Its calculated US rate for August was about $0.00517 per page. That's the highest rate it has recorded since November 2018. At that rate, 100,000 pages read pays about $517. Amazon doesn't publish the rate itself. Written Word Media works it backward from dashboard data, and your own KDP report is the final word.

If your LitRPG, progression fantasy, military SF, or space opera is in KDP Select, the pot keeps growing about 12 percent a year. Whether your share grows with it depends on your page-reads, and your dashboard will tell you that.

Sources:

KDP Community: KDP Select Global Fund and All Stars Bonus Update, September 2026
KDP Community: KDP Select Global Fund and All Stars Bonus Update, August 2026
KDP Community: KDP Select Global Fund and All Stars Bonus Update, September 2025
Written Word Media: Up to Date List of KDP Global Fund Payouts

Commentary

Jonathan: Maybe we should keep that in context, because while this is the second month of decline, it’s up $7.2 million over last year at this time.

Thomas: This is how the press can skew a story. If we want you to be sad, we say, "It's the second month in decline." If we want you to be happy, we say, "It's the fourth-highest month ever." On Author Update, we try to give you the complete picture instead of trying to manipulate your emotions.

BookBub Adds Sci-Fi and Christian Fiction to New KU Email

According to BookBub, which told partner authors this week that its New in Kindle Unlimited promotion has grown its total audience by more than 50 percent since launch and added four new categories: Science Fiction, Paranormal Romance, Historical Romance, and Christian Fiction.

The email goes to BookBub readers who subscribe to Kindle Unlimited. Current flat fees, charged only if selected, are as follows:

$150 for Science Fiction
$100 for Paranormal Romance
$100 for Historical Romance
$50 for Christian Fiction

Submission is free. Books must be enrolled in KU within 90 days of the promotion date (backlist titles qualify if newly enrolled), novels must be at least 150 pages, the same author can run one no more than once every 30 days, the same book once every 12 months, and the book must be available in the US.

Authors in these genres who are moving a book into KU, including backlist, now have a relatively cheap paid promo option aimed at known KU readers. Christian Fiction at $50 is the cheapest category on the list.

Sources:

Commentary

Jonathan: At $50.00 for Christian fiction, that's a freaking no-brainer. You should be applying for these deals if your book is going into KU. That's ridiculous. That is great.

These promotions arelaunch focused. It's for new books in KU. When your book launches, you can get into the email and they'll send it to KU readers.

By the way, preorders don't work for KU readers. You can't preorder something to put it in your library. You have to borrow it on launch day. So this is a fantastic tool for authors using a KU-focused strategy.

Thomas: It's pretty low risk and inexpensive compared to other launch expenditures. I suspect many of these BookBub promos will pay for themselves almost immediately.

If you have a book launch coming up, remember I have a Book Launch Quill that will help you put together an entire launch plan and do as much of it as it can for you (Sign up for my October 22 Grok Bot webinar).

In the other Quill collection, the Promo Stacker can connect to BookBub and help you submit your book, not just to BookBub but to other promo sites as well. Not all of them support brand-new books, but for the ones that do, it could be really useful.

Technology

ArtCraft's Free Word Clone Hits Alpha With Real Documents Still Untested

According to The Register, the ArtCraft team behind open-source Adobe-style tools has added WordCraft, a clean-room reimplementation of Microsoft Word written in pure Rust. The project is free, dual-licensed under MIT or Apache 2.0, and runs natively on macOS, Windows, Linux, and BSD, plus in the browser via WebAssembly. It aims to recreate the familiar ribbon workflow with styles, tables, track changes, references, and mail merge, and it reads and writes .docx files.

On October 8 the Register fired up an early build on an Arm-based Surface and said it worked pretty well, though the buttons, icons, and imagery differ from Microsoft's. At that point the project described itself as pre-alpha with roughly 62 percent of real feature parity, and an alpha for everyday writing close. The WordCraft GitHub repository now lists the status as alpha. Its README states that WordCraft covers 88 percent of Word's ribbon features with commands today, but estimates it is only about 60 percent of the way to replacing Word for real work, 55 percent for a typical professional's weekly work, and 63 percent for everyday essentials.

What Happened

ArtCraft built WordCraft from public specifications and observation, using original or openly licensed assets. The spelling dictionary and hyphenation come from Grady Ward's public-domain Moby Hyphenator II word list. The Register notes that the projects lean heavily on Claude for development. Neil Brown of decoded.legal told the Register that a clean-room approach, meaning the team implemented the functionality independently without access to Microsoft's source code, may be argued as non-infringing on copyright, though the legal situation remains interesting when the other party has deep pockets and many lawyers.

The current README and ROADMAP are frank about what is missing. Beta requires testing against a collection of real-world .docx files; that testing has not begun. Charts, SmartArt, and embedded OLE objects are dropped or incomplete on round-trip. Native printing is not yet available; File > Print opens a PDF in the system viewer. Pagination is not yet Word-faithful. Core workflows, write, format, styles, lists, tables, references, review, mail merge, and open and save .docx, work end-to-end, and signed builds ship, but the project itself does not recommend the tool for production workloads.

Sibling apps GridCraft and DeckCraft cover the spreadsheet and presentation side of the same effort, but they are not the focus for authors drafting books.

Why Authors Care

WordCraft is free, open source, and offline, with no subscription, which is the practical draw for authors tired of Microsoft 365 lock-in.

The familiar ribbon and .docx read/write mean a novel or nonfiction draft can open in Word later, but the project has not yet tested that round-trip against real manuscripts.

Charts and complex objects get dropped; if your book includes them, wait.

The clean-room claim reduces one class of copyright risk but does not eliminate patent questions, and Microsoft has the resources to litigate.

Try it for simple drafting or as a curiosity; do not migrate a finished manuscript or rely on it for production until the documented beta gaps close.

Sources:

Commentary

Thomas: AI has gotten so good at creating software that people are now creating alternatives to big commercial software.

The same thing is happening with Photoshop. People are recreating Photoshop button for button and action for action. They're doing the same for Word, at much lower prices and in some cases with higher performance. I was testing one of the Photoshop alternatives, and it was crazy fast and ran right in the browser.

It's a brand-new world.

We're getting to the point where you could make your own version of Microsoft Word customized for your approach to writing, or your own version of Scrivener. "I like Scrivener, but I don't like X, Y, and Z. Make me a word processor like Scrivener but different in these areas." AI can make you that version.

Here's a good example. Say you write choose-your-own-adventure books. Right now there's no word processor designed for branching fiction, allowing you to say, "If you go through the window, turn to page 29. If you go through the door, turn to page 150."

With the help of AI, somebody listening could design a word processor specifically for managing branching fiction, where you can see it, track it, and keep it all straight. You could use it to write reader-driven fiction much more easily, and then sell it to other authors.

People building these sorts of things are making so much money. The people losing out are the folks who have historically treated their customers poorly. Part of the reason these Photoshop tools are growing so quickly is that people hate Adobe. They feel abused by Adobe.

There's a good historical precedent for this in the advent of electricity. When electricity came out, there was a huge moral panic that it could kill you. The same moral panic we have over AI, we had over electricity.

There were two rival types of electricity: alternating current, which was Tesla's technology, and direct current, which was Edison's. We still have both today. That's why you plug in those little bricks or wall warts, to convert the alternating current from the wall into the direct current your phone uses.

Alternating current is much more dangerous, and Edison's attempt to convince everyone that electricity could kill them was so intense that he invented the alternating-current electric chair just to make electricity scary.

The whole point of the electric chair as a form of execution was a publicity stunt by Thomas Edison. It didn't work, because no one calls it the alternating-current electric chair. They only call it the electric chair, and no one could keep alternating current and direct current straight.

The moral panic was still there, but people adopted electricity rapidly anyway. Why? Because the gas companies they had been buying light from treated them so terribly. The gas companies were so abusive, with terms so bad, that people would put up with an inferior technology that could kill them, according to their understanding of it, just to get away from the gas company.

Jonathan: Never discount the power of hate.

Thomas: You build up goodwill with customers over time, or you build up badwill. Many companies' whole strategy is building a moat, which basically means their customers can't escape. Once you pick their technology, you're stuck with them forever.

AI is tearing down moats all over. It's making it easier to switch from one platform to another, easier to cancel subscriptions that are hard to cancel, and easier to create your own alternative. A new bookkeeping tool just came out called Slow Books, a vibe-coded, open-source alternative to QuickBooks. Perhaps not the best branding.

There's nothing new under the sun. Every new technology has a moral panic, and there are always people who don't understand it. There are great cartoons from 120 years ago about how evil and dangerous electricity is and how it's going to kill us all. It's the same fear from the same people who are afraid of the future, and they always lose.

In America, they always lose. There's not a single new technology we haven't embraced as a people, because Americans are pro-technology at our core. There were no Luddites in America. The Luddite movement was a British movement. We just don't do that over here. We're the ones who make the cool new stuff. We're the ones who make the flying cars.

Free Photoshop opens real PSDs. It is still alpha.

According to the PhotoCraft GitHub repository, an open-source image editor written from scratch in pure Rust now opens, edits, and saves real layered PSD files. The project calls itself a clean-room reimplementation of Adobe Photoshop. It ships with layers, masks, adjustment layers, layer styles, type, vectors, and brushes. Re-saving keeps the render of 307 of 309 files in the psd-tools test set. Everything an author can click is also a command. The same 500-plus commands run from the native UI, a CLI, a JSON control channel, or an MCP server, so an agent can batch color-grade and export a folder of promo graphics.

What Happened

The repository went public around September 30, 2026. By the time Vox posted about it on X on October 7, PhotoCraft already showed roughly 4,500 stars and listed Claude Opus 5.5 as a co-author on commits. Stars kept climbing. As of mid-October the count sits above 56,000. The ArtCraft team behind it, associated with developer Brandon Thomas, released a family of Crafting Apps in the same window: VectorCraft for Illustrator work, FilmCraft for video, LightCraft for photo libraries, PdfCraft for PDFs, and others. PhotoCraft is the one that matters for cover designers and marketers who still receive layered PSD files.

The code is dual-licensed MIT or Apache-2.0. Desktop builds run native on macOS, Windows, and Linux with a wgpu GPU compositor. There is also a WebAssembly build. The official page at getartcraft.com labels the app early alpha. The project's own roadmap, reviewed around October 10-11, puts "ready for real work" at about 45 percent, inside the alpha band of roughly 40 to 75 percent. Feature breadth sits higher, near 76 percent of Photoshop menu items and tools wired to commands, but depth, performance budgets, and reliable round-trips that Photoshop itself will re-open without complaint remain gaps. The alpha gate passes for core workflows: open, layer, select, paint, adjust, filter, type, save, and reopen. Several of those are partial. The project estimates 1,100 to 1,800 more Opus 5.5 agent-hours to reach beta.

Why Authors Care

Cover and promo work often arrives as layered PSDs. PhotoCraft is free, offline, and open-source, so an author can open the file, tweak a title treatment or adjustment layer, and export without a Creative Cloud seat.

The MCP and CLI path lets an agent run the same commands across a batch of graphics. That is useful for consistent color grades or export variants once the tool is stable enough for the job.

The project itself says it is early alpha and about 45 percent ready for real work. Test on copies of files. Compare the composite against the original Photoshop render before you trust a result that goes on a sales page or Amazon thumbnail. Do not make this the only tool that touches a paid cover until the PSD exchange issues close.

AI assistance shows up in the commit history and in the project's own hour estimates. That is context for how the code appeared so quickly. The practical question for authors is whether the editor opens the file correctly and saves a usable result, not how the lines were written.

Sources:

Commentary

Thomas: PhotoCraft is the Photoshop alternative, and it's still kind of buggy. I was also testing another one, ImageFlow, that's also really good.

But everyone who has been wronged by an abusive Adobe Photoshop subscription, which is literally thousands of developers, is now contributing code to improve PhotoCraft.

Jonathan: PhotoCraft can open, edit, and save real layered Photoshop files.

Thomas: And the interface is identical. The buttons are all in the same place, so if you know Photoshop, you know PhotoCraft. With some other Photoshop alternatives, the buttons are in different spots and you have to learn a new tool.

This is a button-for-button recreation, which is fully legal. Otherwise we wouldn't have both Chrome and Safari with the back button in the same place. You can't copyright a user interface.

Adobe is learning that the hard way, and I think they may now regret abusing their customers for so long. If they had treated their customers well and charged reasonable prices, there would be no reason for people to leave. Because they abused their customers for so long, people are desperate for revenge, and they're taking it.

Jonathan: You'd get loyalty too. People are willing to suffer a little as long as they're being loyal. People support their local restaurants or local businesses, and as long as that business has been good to them, they'll stay loyal even if the prices are higher than Amazon's.

Thomas: Look at Apple. Apple users don't have a deep, abiding resentment of Apple. They aren't sitting around hating that they're locked into Apple and wishing they could try something other than their Apple computer or phone.

Apple is struggling with AI. Siri is still not very good. But it's okay, because Apple has goodwill from all the other good experiences people have had.

Adobe, meanwhile, has gone all in on "We're the only option. It's us or nothing." Intuit has the same attitude with ProSeries and QuickBooks. The companies with that attitude are the ones that will suffer most in the next few years, as they suddenly face a swarm of competitors. That's why they're so desperate for AI regulation. They don't want all those competitors.

ElevenReader Puts Its Top Voice Model on the Free Plan

According to the ElevenReader blog post published September 30 and updated October 2, the all-new ElevenReader is now powered by Eleven v4, which the company calls its most expressive voice model. The same post says Artificial Analysis ranked Eleven v4 first for quality; that ranking is ElevenLabs' claim in the announcement. Updates land on the free plan.

What Shipped

Authors who stopped listening to manuscripts because earlier voices stayed flat can run the new model without a credit card. The product page and the blog both list the free plan at 10 hours of text-to-audio per month, roughly a 400-page book, with more than 1,000 voices and no payment details required. Language support jumps from about 30 to more than 90, with the blog calling out better Japanese, Brazilian Portuguese, Mandarin, and Cantonese.

Long-form listening gets more consistent volume, more accurate pronunciations on hard words and abbreviations, better speaker consistency, and more natural pauses and pacing. Emotion and expression are richer. The Expressiveness help article sets three levels: Subtle still uses the legacy v2 model, while Balanced and Expressive run on v4. Users can still add custom pronunciations in the app.

The app stays on iOS, Android, and the web with a synced library. Offline downloads remain an Ultra feature. The product page and the blog put Ultra at $11 a month. Annual billing is $99 a year, which works out to $8.25 a month. Ultra adds unlimited text-to-audio on imports (capped at 24 hours a day at normal speed), offline downloads, custom voices, and 20 hours a month from the 200,000-plus premium audiobook catalog.

It reads eBooks, news, documents, Google Docs, Kindle webpages, PDFs, Word docs, and ordinary web pages. The product page shows Pride and Prejudice samples recorded on each app's free plan in September 2026, contrasting ElevenReader with Speechify and the Kindle Assistive Reader.

Why Authors Care

Free 10 hours is enough for one full manuscript pass or a stack of research PDFs and competitor pages without a card.

If earlier TTS felt robotic, retest on Subtle versus Expressive before you decide the category is still unusable.

Custom pronunciations fix character names and made-up terms that trip the model.

The language jump matters for international editions or foreign research you already have as PDFs or web pages.

Heavy daily listening still needs Ultra for unlimited hours and offline. Light use can stay on free.

Sources:

Commentary

Thomas: ElevenReader is a tool readers can use to listen to your audiobook if you're in the ElevenReader ecosystem. What more authors are doing is using ElevenLabs to create the audiobook themselves, so I want to give you a sense of where things are heading.

I have clips of the opening line of Pride and Prejudice. First is the built-in Kindle audiobook reader. Then Speechify, which you could call the next level up. Then ElevenLabs' older model. And finally ElevenLabs' newest, version four.

Listen to the samples on the YouTube of this episode.

Thomas: No. It takes emotional context and performs the text in a more emotionally rich way.

These clips give you a sense of the progression. In the first one, all the words were pronounced correctly, but it had a very robotic-sounding voice. Then the models started performing the words better. Now ElevenLabs v4 is performing the sentences. It's getting a sense of the flow of the sentence itself, and that's a huge step.

It was the number one model for two whole days. Then Fish Audio released Drama 3, a model that's as good or better. I think it has a better Elo in the head-to-head, at a third or a quarter of the price. So it's way cheaper if you're making your own audiobooks.

Their announcement video is really funny. Uncle Roger, the famous YouTube personality, is making fried rice, and a deepfake of Uncle Roger is roasting Uncle Roger while he cooks. It goes back and forth between the real Uncle Roger and the AI Uncle Roger, and it was shockingly good. It's getting the emotional nuance, the performance that's so important for an audiobook and really for any kind of audio performance.

The study we read at the beginning of the episode reflects listeners' experience with the older, Speechify-level performance. Listeners haven't heard ElevenLabs v4 or Drama 3 yet, because these just came out and they're way better.

We're getting to the point where AI narrators will very soon be better than the average human narrator.

Jonathan: I'm an audiobook narrator, I enjoy doing it, and I think human performers are mostly better than AI, except for some people who have a horrible time expressing themselves.

But version four was almost frictionless. Nothing prevents you from taking in the story. It's not, "Ugh, I'm listening to AI right now." It gives me the story without getting in the way.

If you're still saving up for your audiobook, whether you'll narrate it yourself or hire somebody, this is a great, cheap interim. Put an audiobook up now and replace the sound files later when you can afford a narrator.

Thomas: Or replace just a few characters in the human-narrated audiobook you're doing. I think this is the real opportunity: human voices and computer voices sharing a performance.

Maybe you can do eight out of 10 of the characters in your book, but one has a really deep voice or a weird accent. You can't, for the life of you, do a consistent Scottish accent. You've watched the YouTube videos, and you just can't make it work. Now you can load up a Scottish-accented AI to do just that character, and you do everything else.

The more different voices you bring in, the more cinematic it sounds, which for most listeners is a feature. It makes the book more appealing while making life easier for you, because you're reading fewer words into a microphone.

Jonathan: It's easier to edit, because you're not editing breaths or mistakes.

Thomas: If you want to hear examples, go to fish.audio. They have 2,000 voices, and they're very competitive on price. It's exciting where text-to-speech technology is headed.

Google Lets Authors Prototype Book Games Without a Studio

According to the Google Labs blog, Google launched Playground on October 7, 2026. It is an experimental platform that lets users create, play, and share custom browser games by describing them in text prompts. No coding is required. The tool is available at playground.google for users in the United States who are 18 or older.

Playground runs on a conversational interface. You start from a blank canvas, remix starter prompts, or use guided support to shape the concept. Then you tweak the physics, rewrite the rules, customize the characters, or overhaul the environment by asking, and you can test the result right away. Games run in the browser on a phone or laptop.

Once a game is ready, you can keep it private, drop a shareable link for friends and family, or publish it to the Explore gallery. Published games go through safety screening aligned with the community guidelines, and users can report violations. Select genres support in-game leaderboards and multiplayer. A Play Games profile lets you claim a handle, favorite games, and follow creators. The gallery highlights games based on player ratings and activity.

Creation access is tiered based on your Google AI subscription. According to The Verge, a Google spokesperson said the platform is free to use, but Google One subscribers get higher weekly token limits based on their plan. The same spokesperson said Playground is powered by Gemini, Nano Banana, and Lyria models with a custom harness refined through internally built games and evaluations. Browsing and playing the catalog is freer than generating new games.

Google also previewed a coming integration with Unity Spark. According to the Google Labs blog, Unity Spark is in testing with a closed beta soon. It will unlock more advanced 3D capabilities and professional-level mechanics using the Unity runtime, so creators can start with text prompts and scale toward immersive experiences. Games made with Unity Spark will be shareable on Playground. Unity confirmed the partnership the same day and noted that Spark remains browser-based.

Why Authors Care

Prototype a book scene, character interaction, or setting as a simple playable game without hiring a studio or learning an engine. Keep the output private or share the link as a reader magnet for a newsletter or community.

Treat this as an experiment, not a commercial title. Quality is experimental, access is limited to US adults 18 and older, and heavier creation depends on subscription tiers and weekly tokens.

This does not replace a real game studio deal or a polished commercial release. Sign in at playground.google with a US Google account to check current token limits and available genres before you invest time.

Sources:

Commentary

Jonathan: What uses mostly text to provide all the context? A book. If you want to create a simple phone or browser game from text prompts, give playground.google a try, and share the results on AuthorMedia.social if you think we'd enjoy checking them out.

Thomas: Creating a video game based on your book is a fun thing to play with, and I think there's a lot of potential here for certain kinds of books.

Zero to Hero: Meta Now Has a Frontier Model

According to Fiction Bench, the live leaderboard that makes every model write the same twenty genre stories and scores the chapters with AI judges plus blind human votes, three models that shipped in the last ten days landed in places an indie can actually use. Claude Haiku 5.5 sits at number nineteen with an AI-judge score of 68.5 and the best bang-for-the-buck mark on the board. Gemini 4 Argon is tenth at 73.7. Meta's Muse Wonder is seventh at 79.4. None of them top the Opus and Astra entries that still lead the list, but the gap between "cheap and unusable" and "cheap and usable" just closed.

What Happened

Anthropic released Claude Haiku 5.5 on October 7. According to the Anthropic announcement, it is the cheapest and fastest small model the company has shipped. For prompts under 100,000 tokens the price is $0.10 per million input tokens and $0.50 per million output tokens, roughly a tenth of Haiku 4.5 and about 75 percent cheaper on average. It carries a one-million-token context window, up to 128,000 tokens of output, and the first effort setting on a Haiku model. Computer-use scores jumped from 15.7 percent to 72.4 percent on OSWorld and agentic coding went from zero to 39.2 percent on Terminal-Bench. On Fiction Bench it scored 63.3 on Story, 68.7 on Intelligence, and 76.7 on Prose quality, with a writing cost of $0.37 for the full benchmark run. The human rating sits around 1,055 on a scale where 1,000 is average.

Google made Gemini 4 Argon generally available to developers and paid users on October 13 after a limited cyber-partner rollout at the end of September. According to Google's developer documentation, the model uses a unified one-million-token window shared by input, thinking, and output, and it can produce long decodes that approach a million tokens. Introductory pricing is $2 per million input tokens and $10 per million output tokens through November 30, then doubles to $4 and $20. Thinking levels run low, medium, and high. On Fiction Bench it scored 70.1 on Story, 78.3 on Intelligence, and 73.5 on Prose, with a writing cost of $3.51 and a provisional human rating near 1,086.

Meta released Muse Wonder and Muse Wonder Full on October 12. According to the Meta newsroom post, the models are built for long personal and professional agent tasks and introduce Active Memory (carrying relevant details across conversations) and Active Focus (staying on the stated goal through blockers and errors). Muse Wonder is available on the Meta Model API at $4 per million input tokens and $20 per million output tokens with a one-million-token context. Muse Wonder Full is limited to Muse Code subscribers on the $50 and $200 monthly plans. On Fiction Bench, Muse Wonder scored 77.2 on Story, 81.1 on Intelligence, and 81.0 on Prose quality, with a writing cost of $12.74.

Why Authors Care

Haiku 5.5 drops the cost of iterating a full novel draft, outline, or scene set into the range of a few dollars instead of tens of dollars. An author can run multiple passes, test different tones, or generate the next three chapters without watching the meter.

Gemini 4 Argon's long decode and unified window let a single call carry more of a long manuscript or series bible before the model starts inventing characters or dropping plot threads. The higher Fiction Bench Story and Intelligence scores show it is more reliable at keeping the brief.

Muse Wonder's Active Memory is the feature aimed at the exact failure mode of earlier models on multi-chapter work: forgetting what the author already established. Combined with its seventh-place Fiction Bench ranking, it is the first Meta model that looks usable for sustained fiction assistance rather than one-off scenes.

None of these scores make the models better novelists than a human who has finished a book. They do change the economics of the first and second drafts. Authors who already use AI for outlining or continuity checks can move the cheap model into the drafting loop and reserve the expensive models for the hard revision passes.

Sources:

Commentary

Thomas: Three new models came out this week. Meta came out with Muse Wonder and Muse Wonder Full.

We've been making fun of Meta for a year and a half for being bad at AI. But about a year ago, Zuckerberg spent billions of dollars hiring the best AI researchers he could convince to switch, giving them signing bonuses of millions of dollars, up to $10 million and some higher, just for switching teams. I said, "Don't count this out. This really could work."

This week they released their new models, and they are frontier level. They're Fable-level models. Suddenly Meta is back in the running with one of the top models in the world. They came out of nowhere to become a contender. I've heard the Meta model is actually good at writing.

Muse Wonder is on Fiction Bench, and it's doing okay. But Muse Wonder Full, the one that's supposed to be really good at writing, isn't on Fiction Bench yet. I hope Jason adds it, perhaps once it's on the API.

Google also came out with a brand-new model that's doing well on the benchmarks. And there's a new Haiku 5.5. It's not super useful for most things authors do, but if you're doing cheap stuff with AI, it's way better than Haiku 4.5.

Open-Weight Drama 3 Puts Directed TTS in Authors' Hands

According to the Fish Audio blog post dated October 13, 2026, Fish Audio released Drama 3, its most emotive text-to-speech model, on the web product, the iOS and Android apps, and the API the same day. The company also posted open weights on Hugging Face under a new Fish Audio Community License.

What Happened

Drama 3 takes the same bracketed direction authors already use on Fish Audio's S2.1 Pro, then plays it with more feeling. You write cues like "[barely holding it together, forcing a smile]" or "[through gritted teeth, keeping his voice down]" in plain language. Mid-line direction changes land where you put them. The model speaks 83 languages, and the direction language does not have to match the script language.

Fish Audio says Artificial Analysis's Pronunciation Robustness benchmark ranks Drama 3 first at 93.2% correct on hard words, numbers, and codes, against 88.1% for S2.1 Pro. Brand names, place names, and context-dependent readings show the biggest gains on that table. Treat those scores as Fish Audio's report of an independent bench until you hear the take yourself.

The weights are open, not open source. Under Community License v1.0, research and non-commercial use are free. Commercial use is free if you and your affiliates take in less than US$500,000 a year. Above that line, or if you offer Drama 3 to others as a TTS, voice-cloning, or inference service, you need a license from business@fish.audio. Cloning a real person's voice without consent is banned. Self-hosting is serious hardware work: Fish Audio's install docs call for an NVIDIA GPU with at least 48 GB of VRAM for FP8. The model is a 35B-parameter mixture-of-experts with about 3B active per token, and one script can carry up to 16 speaker tags.

On the hosted side, Drama 3 is live at fish.audio and through POST /v1/tts with the header model: drama-3. From October 13 through November 12, Fish Audio prices the Drama 3 API at $10 per million characters, 80% off the $50 list price, and cuts new creator plans 50%. Interactive editing and audio inpainting are still invite-only.

Why Authors Care

If you draft audiobook chapters or audio drama, you can direct tears, anger, and laughs on specific lines without hiring a booth for every take.

Authors under the $500k revenue cap can run the weights themselves or use the hosted API under the Community License terms. Confirm the license against your actual revenue and affiliates before you ship a commercial title.

Self-hosting needs a high-end NVIDIA card. Most authors will use fish.audio or the API and keep the open weights as a backup path.

Pronunciation on chapter numbers, character names, and made-up terms still needs a listen. Custom tags and regeneration remain part of the workflow.

Pair this with the closed TTS options you already test. Drama 3's pitch is direction plus open weights, not a claim that every manuscript should leave ElevenLabs or ElevenReader tomorrow.

Sources:

Commentary

Thomas: Here's a Fish Audio example of a narrator reading what I think is nonfiction. They really didn't give themselves good material. That was a terrible piece of text, but the performance was fine. Here's a character performance.

Listen to the sample in the YouTube version of this episode.

Jonathan: That sounds very manga.

Thomas: Fish Audio is an Asian company, so there's a lot of manga influence. But you'll notice that felt more emotive.

Both of these models now take emotive coaching. You can put words in brackets describing the tone or performance. Write "whisper," and it will whisper that line. Write "chuckles," "humorous," or "sardonic," and it will perform the line that way.

What I don't know is whether either of them adds those descriptors automatically. Adding them by hand is a lot of labor. What you really need is an AI that reads through the script and gives the machine at least the first pass of performance coaching, which you can then tweak yourself.

Fiction Bench Makes Creative Writing a Measured AI Category

According to the Fiction Bench site launched October 16, 2026, Jason Hamilton of Story Hacker AI released a dedicated public benchmark that measures how well AI models write fiction. Every model writes the same 20 genre stories, three chapters each. A panel of AI judges scores the chapters and finished stories against a shared craft rubric plus genre-specific criteria, and readers can vote blind on short scenes. Hamilton launched it as early access for his Story Hacker community before a wider reveal and credits developer Dave Law's open-source work for the initial approach he adapted. Story Hacker AI is a paid Skool community at $67 a month that teaches AI-assisted fiction.

Why a Fiction Benchmark Matters

Creative writing has long been hard to measure the way coding or math benchmarks measure those skills. Models that excel at following instructions or summarizing text can still produce flat characters, unearned plot turns, or prose that reads like it was generated. Fiction Bench treats story craft as its own category. It forces every model to produce multi-chapter genre fiction under the same prompts, then scores plot, emotional engagement, believable characters, world building, originality, genre craft, coherence, pacing, dialogue, and freedom from purple prose or forced metaphor. The human side is blind: readers at the compare page see two anonymous versions of the same scene and pick one, mark them about the same, or skip. Votes feed a Bradley–Terry model. Fewer than 20 votes keeps a rating provisional.

That structure signals two things. AI labs are now competing on whether their models can sustain a story across chapters, not only answer questions or write code. Authors and readers finally have a shared public test instead of relying on private demos or single-prompt anecdotes. Other creative-writing benchmarks exist, including EQ-Bench Creative Writing v3 and Lech Mazur's writing benchmark. Fiction Bench differs by using multi-chapter genre stories and collecting human blind votes. Not to be confused with Foreverse Research's unrelated novel-continuation leaderboard of the same name.

How Readers Can Take Part

Hamilton is asking people to run blind comparisons and stress-test the board at scale. Visit the compare page, read two anonymous scenes written to the same prompt, and vote. Model names appear only after you choose. Early AI-judge numbers from the October 16 launch-day snapshot placed Claude Opus 5.5 first among the models that finished all 20 stories, with several GPT and Claude models close behind, but those scores are preliminary. Human voting has only just begun, so every model's human rating is still provisional. Prices listed on the board come from OpenRouter, and the writing-cost column shows what each benchmark run actually cost the writer model.

Why Authors Care

A public fiction benchmark gives you a reference point when you choose which model to use for drafting multi-chapter genre work this week.

Blind votes let you contribute to the ranking instead of waiting for someone else's demo.

Treat the launch-day AI-judge scores as a first filter. The real signal will come once enough human votes accumulate.

If you already test models for story quality, this is a standardized multi-chapter test you can point to and stress-test yourself.

Sources:

Commentary

Thomas: I'm so excited about this new benchmark, Fiction Bench, from our friend Jason at the Nerdy Novelist. I've had Jason on Novel Marketing before, and he's who I go to for measuring large language models against each other on how well they write. I've been longing for a benchmark like this.

Almost all the benchmarks for large language models are geared toward coding or general intelligence. Whether a model is good at coding doesn't tell you whether it's good at writing, editing, or anything else related to being an author.

The Nerdy Novelist teaches writing with AI, and now he has a benchmark for measuring which is better, Opus 5.5 or Astra 6. He measures them on different metrics and even shows the cost, which I think is really beneficial.

We'll probably start using it as we talk about new models. Instead of talking about coding benchmarks, as we've been doing, we'll talk about how models perform on Fiction Bench, which I think is a much more important benchmark. You can find it at bench.storyhacker.ai.

The current leader on Fiction Bench is Opus 5.5, which I almost feel has solved AI. Opus 5.5 can do almost everything I want an AI to do, other than image generation. It's the perfect all-around model for coding and writing.

Zeitgeist Report

Flying Cars Are Here: Is the Media Trying to Ignore It to Affect the Zeitgeist?

According to The Verge, Tesla revealed its second-generation Roadster on Thursday, October 15, 2026, at a SpaceX test site near Waco, Texas. The livestreamed event ran under 30 minutes. The car drove up a tall ramp and clung upside down with a SpaceX-supplied fan suction system. It flew off the end using cold-gas thrusters, flipped, and landed on its wheels. It then jumped over a lineup of Tesla vehicles (Model 3, Model Y, Cybertruck, Semi, and Cybercab) in what Tesla's Lars Moravy called a kick flip. Elon Musk said, "Flying cars are real."

The Promise Kept

Generations of Americans grew up on the promise of flying cars. The Jetsons made them ordinary. Back to the Future Part II set the year 2015 as the moment hoverboards and flying cars would arrive. The joke hardened into "Where's my flying car?" On Thursday, Tesla and SpaceX answered with a live demo. The car used suction to hold itself inverted on a ramp, then cold-gas thrusters to leave the ground, roll, and set down. Tesla claims a boosted 0-60 mph run in 0.81 seconds, according to Road & Track. Production is promised for 2028, nine years after the original 2017 reveal. No final price was announced. Reservations are open for non-flying versions. The flying Rocket model is single-seat, invitation-only, and not approved for public road use. The thrusters use limited compressed gas, so the boosts are short hops, not sustained flight. This is still the first time a production-intent car from a major manufacturer has publicly demonstrated the old sci-fi trick in front of a live audience.

Running the Story Versus Featuring It

The Associated Press, CBS News, The Verge, TechCrunch, and Road & Track all covered the event. Fox News posted video of the kick flip. The New York Times ran a short video. Coverage existed. What did not exist was front-page treatment. The morning after, neither the New York Times nor Fox News led its homepage with it. Flying cars should have been the lead story on every major front page. Instead they were treated as a tech curiosity while politics and tragedy filled the top slots. There is a difference between running a story and featuring it. This is the difference between page 1 and page 20.

The Political Filter

Polymarket reported that Reddit moderators removed a post showcasing the Roadster for violating a subreddit ban on "political content." The post did not name the subreddit. Weeks before the election, a car demo becomes political when the person attached to it already is.

Comment sections on instagram skewed negative while users on X were ecstatic. Some viewers called the stunts cool technology with real limits. Others minimized the achievement because of the man who built it. That split is the cultural signal. Audiences that cannot separate a technical first from a political figure have lost the capacity for technological wonder. The same people who spent decades asking for flying cars now treat the arrival as just another reason to argue about the founder.

What Authors Should Do With This

Science-fiction writers who set flying cars in the far future now have a nearer-term prop with real constraints: short hops, limited gas, regulatory approval, and invite-only access. Near-future and thriller writers can use those actual limits instead of pure fantasy. Fiction that captures wonder rather than cynicism may meet readers who are hungry for it. When a technology becomes a political identity marker, readers will bring that filter to any story that features it. Authors who want the audience that still likes big technical leaps should show the machine working before they show the argument about the man who owns the company.

Elon Musk Called the shot a decade ago.

Sources:

Commentary

Thomas: Everyone was making fun of the thumbnail of the flying car. There's a media blackout about it, but flying cars are real, and they're here!

Jonathan: Yesterday I was scrolling on my phone and suddenly saw nothing but flying cars. I thought, "Okay, this is obviously AI." Everyone kept saying, "This isn't AI. This isn't AI."

The ramp was shaped like a half-pipe. Remember playing with Hot Wheels and jumping your cars? This thing went up a half-pipe, stayed suspended using attitude thrusters, drove off, flipped over, landed using those same thrusters, and then jumped a whole line of cars. It was really cool to watch.

Elon Musk came out and said, "Flying cars are real," and the world hated it. Parts of the world hated it, anyway.

The rest of us were like, "Whoa, flying cars. That's awesome."

Thomas: I was so jazzed I literally couldn't sleep last night. This has been the definition of the future my entire lifetime. It was the definition of the future my father's entire lifetime and my grandfather's entire lifetime. We have been longing for flying cars for generations, and it finally happened.

We finally have a car that can legitimately leap off the ground, go somewhere else, land, and keep driving. And that may not even be the most interesting thing about the car. It has all these other breakthrough technologies.

The zeitgeist takeaway is that people aren't interpreting this from a sense of wonder.

You can see who's been poisoned by their algorithms. In any other era of American history in the last 100 years, this would be the lead story in every newspaper, and everyone would feel wonder and delight.

Jonathan: This is Chicago World's Fair-level stuff, back when America was hopeful and looked to the future and to technology to make life better. This guy over here is making flying cars. This guy over there is making something cool. There's Starlink. We're going to Mars. We're reusing rockets we caught with forks.

Thomas: You can make a video call to someone on the other side of the world. This is magic, and everyone is bored by it.

Jonathan: I saw so many people saying, "Guys, it's not really flying." Okay, how do you want to define flying? Or, "It's not practical, because we have planes." Yeah, you moron, our roads are designed for land-based linear driving.

Thomas: If all we wanted was practical cars, everyone would drive a Toyota Corolla.

It's okay to enjoy things!

I realize there are powerful forces in this world, both human and non-human, that want you to be sad, despondent, and pessimistic about the future. It's very important to them that you're gloomy about the future heading into the election.

They believe you cannot allow any optimism to flow through your brain or your body, because any amount of optimism will affect how you vote, and you'll vote for the “wrong” people.

In fact, stories about the flying car were being rejected from Reddit for being political, because there's an understanding that optimism will change how you vote.

In this ideology, you are not allowed to be optimistic. You're not allowed to have a sense of wonder. You're not allowed to enjoy the magic of this moment.

It's as if somebody restored a dinosaur from DNA, Jurassic Park style, but it was a little one, and people said, "Well, it's not one of the big ones." Oh my goodness. Can you not enjoy anything? Are you so poisoned by the media you're consuming?

Jonathan: That story isn't really about the hopefulness of science.

Thomas: Sure. But we're making magical breakthroughs. We can talk to the computer like they did on Star Trek. We have cars that drive themselves. We have cars that fly!

It was interesting watching reaction videos to the demo, because car people were freaking out before the car even took off.

If you know much about cars, one of the big challenges for fast cars is that the faster a car goes, the more it flies just a little bit. I mean half an inch off the ground, just enough for the tires to lose traction. Without good traction, you don't have good control or good acceleration.

The historical solution has been to put a wing on the back that pushes the car down at high speed. The wing looks really cool, and all the fancy cars have one, but it doesn't work very well, and it often works only in a narrow band of speeds.

That spoiler often won't do anything until you're into triple-digit speeds, so it doesn't help your 0 to 60 at all. And the faster a car tries to go 0 to 60, the more the tires leap off the ground a little, and you hit these physical obstacles.

The new Tesla Roadster has suction that pulls it down to the ground, which is why it could drive inverted before it started flying. Car people all knew what that meant. It means it could potentially corner better than any car ever made. It can go 0 to 60 in less than a second. They demoed it at something like 0.89 seconds, a level of acceleration you just don't see.

Teslas already felt like a rocket ship taking off. I had a friend drive me in his Aston Martin, and a different friend let me drive his $30,000.00 Tesla, and they felt the same in terms of that initial surge.

Teslas give you access to full torque at all speeds. There are no gear changes where some speeds have more torque than others, which is common in a gas-powered car. An electric motor gives you full power at all speeds because of the nature of the engineering.

And it has rockets on the back. The same technology that can shoot it into the air can fire behind it for even more acceleration. How are y'all not excited?

Yet when one of the people on our team shared a story on Instagram, everyone was grumbling, whining, and complaining about how much money Elon Musk has. What has happened?

There are two reactions. There's the awe and wonder you're seeing on X, and then there's the cynicism and negativity on Instagram, which is very much NPCs responding to their programming, saying, “You're not meant to be happy. You're meant to be sad so that you won't vote for the Republicans.”

Thomas: Let your childlike wonder thrill you.

Jonathan: No, your childlike wonder should grow up with you. That's what I'm saying. Being an adult doesn't mean you get rid of it. It means you understand more about the universe, so there's more for it to amaze you with.

Look at the scientists who made some of the greatest things we've ever had or discovered things in the universe. They maintained that childlike sense of wonder. It's what drove them through all those math equations, which are not fun. They did all that work for the thrill of discovery. That's wonder.

And you're not going to finish a book without a sense of wonder, without enjoying bringing something to life through the characters, the tension, and the way the story weaves and turns and jukes and tries to throw me off.

Thomas: This is a real challenge for writing science fiction, because we've become such a cynical people. There's a longing for wonder, but also such resistance to it. A lot of science fiction has become dystopian rather than wonderful.

People aren't writing stories that evoke wonder, because it's difficult to do. It's difficult to write a wonderful story for a cynical reader, which is really unfortunate. It shows how far we have fallen as a people.

You can choose how and where you want to nudge readers. But it also tells you that you really have to pick your reader by political party, because people's interpretation of this event is splitting along party lines. I'm hardly seeing any Democrats celebrating this, and I'm hardly seeing any Republicans who aren't super jazzed.

Whether you can be optimistic about the future has to do with your party affiliation. One party wants to build things and invent new things and is optimistic about AI, flying cars, and rockets in space.

The other party is pessimistic. The world is coming to an end. The climate is being poisoned. We're in late-stage capitalism, whatever that means. Everything is bad, I hate it here, and I just want to get out. If only we could have a revolution and tear it all down, things could be better.

No, we can make things better right now. We can build things right now. These are two very different worldviews, and the kinds of stories that resonate with each group are very different.

Jonathan: I don't know if I'd agree it's a complete party-line split. I understand what you're saying about the baseline, and there will be outliers, but I'm seeing maybe half of Republicans against it too. I think they have the algorithmic poison, and they're tired of the war in Iran.

Thomas: Why should that matter to our enjoying flying cars? We could have a world war, and flying cars would still be cool.

Jonathan: Can you imagine Marines with flying Humvees?

Thomas: Everyone in the military-industrial complex is taking notes now.

Elon Musk called this shot 10 years ago. He's been talking about making cars fly for 10 years, and he's been surprisingly specific about the technology. He's tweeted about it. He talked about it on Joe Rogan. No one believed him.

Everyone was shocked that the cars fly now, even though he's been explaining the technology he would use, the cold gas pressure system, since around 2016 or 2017. People said, "Whatever. Cars can't fly." Then he shows that cars can fly, and within 12 hours people are saying, "Well, they don't fly that well."

Jonathan: We're going to see the craziest lane changes in the world. Although I don't think this car can legally deploy on roadways, because it can fly.

Thomas: There are a ton of safety issues, but none of that matters right now.

It's a cool thing!

We have beautiful, expensive cars that don't fly. One of the big luxury brands, Aston Martin or Bugatti or someone, released a big new car the same day as the Roadster and called it something like "A Taste of the Future." Who would spend half a million dollars on that when you could spend half a million dollars on a car that does all of that, flies, and goes 0 to 60 in less than a second?

This is amazing. This is the future. This is what we all dreamed of. Let's enjoy it, at least for a day.

I went to The New York Times homepage to see how they were reacting, and they didn't have a story about it at all, because you really can't cover this without it being cool. I'm sure there's a story somewhere so they can say they covered it.

But the only story about electric cars on the Times homepage was about production issues at an electric car factory in China. They’re essentially saying some Chinese electric car company's supply chain issues are bigger news than Elon Musk making an electric car that can fly.

Let's be clear: this isn't a jet pack. It isn't like any of the flying things we've had. No propellers, no helicopter.

So I checked Fox News. Also no coverage, because remember, Fox News isn't very conservative. It's part of the same coastal elite ruling class, often in lockstep with all the other media outlets. The only news outlet I could find that featured it on the front page was the Daily Wire.

Holyoke in the chat says, "I want a personal jet pack that has great range."

Jonathan: The first thing I thought when I saw this video on X was, "I've got to get me one of those."

Thomas: Obviously it will be crazy expensive, and at first you'll only be able to use it on special tracks. But that's how all technologies are initially. They're dangerous and risky. Motorcycles were the same way.

Jonathan: People go to the lake for their boats. They'll go to the track for their flying car or their tank or whatever they have.

Thomas: I was so jazzed and so optimistic. I didn't come across any of the negativity until the next day, because the flying car drowned out all the AI buzz in my X feed, which is normally 80% AI.

My wife will ask, "Have you seen the latest negative thing on X?" I say, "No, my feed is all optimistic AI stuff, people inventing the future." So she'll tell me about the latest shooting or whatever.

The most recent shooting isn't going to get much press, because apparently a good guy with a gun who was carrying stopped it. A dad had a gun on him in the school where he wasn't supposed to, and he ended the shooting before the cops even got there. That doesn't play into any of the useful narratives, so they're memory-holing it.

The media is really good at directing what you think about. They want you thinking about the topics that will lead you to take the action they want you to take. They're not as good at telling you what to think, but they're very good at telling you what to think about.

It tells you a lot about our world that we finally have flying cars, the technology that defined the future for generations, and they don't want to talk about it because it might make you happy. They can't allow you to be happy even for a moment, not even for one day of optimism.

Jonathan: One way to keep your feed more positive is to do what I do: I don't curate it at all. I want to see all the chatter, but I watch all the videos.

Thomas: I was wondering whether anybody would actually find out about this story from our show. And they have! People have gone the whole day without hearing about flying cars.

We have flying cars!

Kelly J says in the chat, "Escape your media echo chamber. Get on X. Get away from whatever you've been following. You'll be so much happier for the rest of your life, because the media that you're following wants you to be sad."

They Called Dressmaker Too Niche. It Paid for Itself in a Day

According to Creative Bloq's August 6 interview with the co-creators, Dressmaker did not start as a hole in the market. It started because Sabrina Becker wanted to make dresses, and Ehlers-Danlos syndrome, a connective tissue disorder, had blocked the costume-design career she wanted. Dexerto's October 5 story is what carried that origin around this week. The viral caption says she lost her ability to sew. The interview is more precise, and more useful. Becker told Creative Bloq, "I grew up sewing with my mom, so Dressmaker has been a really lovely way for me to reconnect with that."

What Happened

Jonathan Hau-Yoon, her partner and co-creator, spent an afternoon cutting a video of the idea. He told Creative Bloq it drew about half a million views before the game even had a name. Free Lives then spent about three weeks on a rough prototype. That free build was downloaded 40,000 times in the first month and 100,000 times after two months. Cozy Lives, the team under publisher Free Lives, shipped the full game on Steam on September 21, 2026.

The sewing is the game. You pick the fabric, line the pattern up with the grain or cut it on the bias, run it through the machine, and finish it with buttons, bows, lace, and applique. Becker designs and curates the pieces. Dexerto reports she wore a dress she made herself on their first date. The players who showed up did not ask for a combat loop with a dress painted on it. They started posting gowns built like the Mona Lisa and Van Gogh's Starry Night.

On June 9, the team posted on Steam that they had been told the game might be "too niche," and that they should not expect much. That same post announced 100,000 wishlists. On September 22, Hau-Yoon wrote that within 24 hours of launch they had fully recouped their costs. The Steam page today lists Dressmaker as Overwhelmingly Positive across 10,243 reviews, with 98% of the 7,795 English-language reviews positive.

The Audience They Were Told Did Not Exist

This is the Timothy test. A Timothy is the one specific reader you can name, the real person the work is for. Becker was the Timothy. The game was not a generic cozy title aimed at a generic they/them audience nobody can picture. It was a dress shop built so one sewist could get back to the decisions she already loved: grain, pattern, trim, a client who needs a dress.

That is the opposite of the current shortcut. Take a game built for men, drop in a girl boss, and call it for women. Women did not line up for that. They lined up for Dressmaker. They lined up for Stardew Valley, which Eric Barone built alone over four years, teaching himself the code, the art, and the music. His press page says that as of December 2024, Stardew Valley had sold over 41 million copies, including over 26 million on PC. They lined up for Animal Crossing: New Horizons, which Nintendo's August 6, 2026 earnings put at 50.29 million units through June 30. Specific life. Specific craft. A person, not a persona.

Why Authors Care

Write the book for the one reader you can picture, then trust that people like that reader exist in number.

A niche objection is not a death sentence. Dressmaker was called too niche on the way to 100,000 wishlists, then covered its costs in a day.

Do not bolt a trend onto a story you do not have. A romance reader does not want a thriller with a kiss added. A sewing reader does not want a shooter with a dress added.

Serve the audience you already have. Becker's knowledge is in the pattern pieces. Your Timothy's knowledge should be in the book.

If the work only makes sense to a generic everybody, it is for nobody. Name the person.

Sources:

Commentary

Thomas: This game became a mega bestseller, and it's the perfect example of everything I teach about writing. This game was made for one lady who wanted to do a very specific thing.

The mistake the video game industry and Hollywood have been making is taking games men enjoy and trying to make them more approachable for women. "We're going to take this blow-the-aliens'-heads-off game, add some girl bosses and some feminism, and make it more appealing to modern audiences."

Jonathan: Halo, Gears of War. I think Star Wars Zero Company has the same issue.

The funny thing about Gears of War is that they did all this masculine marketing, then sent test copies to the game reporters, and even the game reporters were asking, "Why is this girl boss here, and why is she taking over all the dialogue?"

Thomas: Women don't want Gears of War for women. It has never worked. You don't see that smash hit. The women who enjoy Gears of War enjoyed the original Gears of War, which was for men.

There was no gender check where you had to show your driver's license, and someone would say, "Sorry, you're a woman, you can't buy this." Women could always buy Gears of War, and there's always some percentage of women who want a blow-the-aliens'-heads-off game.

Shooting between two targets is called missing, and that's what so many of these games are. They're a miss. Instead, a company should have two teams. One makes the blow-the-aliens'-heads-off game that's mostly for men, though women can buy it if they want. The other makes the dressmaking game for women, though men can buy it if they want. Most men won't.

Jonathan: I guarantee you guys are playing it.

Thomas: Sure. I'll admit I've played Stardew Valley. My wife got Animal Crossing, and I tried playing it, but after a little weeding I said, "I can't." I enjoyed Stardew Valley, though, and I made a lot of money in that farming simulator.

It is possible to make a game that appeals to both, and Stardew Valley is a good example. My experience playing Stardew Valley and my eight-year-old daughter's are entirely separate.

For her, it's like a dollhouse. She goes to town, talks to all the people, has maximum hearts with all of them, is very friendly and bubbly, and gets to know all their backstories. I go down into the mines, kill monsters, make money, grow crops, and hardly ever talk to the villagers.

It's a big enough game that parts of it appeal to me and parts appeal to her, and the core mechanic of farming appeals to both of us.

That's the approach if you want a bestselling book that appeals both ways. You have characters that appeal to certain people and different characters that appeal to others. You have Team Jacob and Team Edward, and they appeal to different demographics of the audience.

But you have to be very careful with that, because it's exactly the approach they tried with the Gears of War girl bosses. All it did was turn off the men playing it without turning on the women.

You have to pick a Timothy, thrill the Timothy, and be very careful about adding Timothies to your book. It can work, but you always have to keep your first Timothy as the primary Timothy and never accommodate the others in a way that irritates number one.

Jonathan: When you're writing your novel, you're writing it for one Timothy. I wouldn't insert more Timothies.

You can have other characters who are different tropes: the uber-masculine guy who doesn't want to touch anything pink and the funny hacker guy, and they kind of get along but kind of don't. But all of them go through the POV of your Timothy. They all need to relate to your Timothy in some way, so your primary reader is the one who gets to enjoy all these characters.

Men don't have a problem with girl bosses. Watch the Aliens movies. People don't have a problem with Sigourney Weaver. They don't really like the stories of Alien 3 or Resurrection, but they never had a problem with Sigourney Weaver as the girl boss, because it was written well and it felt natural, not forced.

She worked with the Marines. The Marines got jacked up, not by the girl boss, but by their officer, which is just truth. So there's no problem with that kind of character, as long as it's relatable from the perspective of your primary reader.

Thomas: If you're going to have a woman warrior, her motivation needs to connect somehow with her femininity. Sarah Connor is a good example done well. She's a real badass character, but her motivation for being a badass warrior is protecting her son, which is a very authentic female motivation for violence.

Women in general are less prone to violence than men. This is a well-documented anthropological fact. I know people don't like thinking the genders are different, but this is a very true statement.

That doesn't mean women are incapable of violence, and one of the most common ways to get a woman to become violent is to threaten her child in front of her.

Jonathan: Mama bear syndrome.

Thomas: Mama bear syndrome. The Terminator, or Terminator 2, is that. Now it's mama bear against an even scarier bear, which is what leads to all the dramatic tension and conflict.

At no point is Sarah Connor more powerful than the Terminator. He is deadlier, more ruthless, more violent, and stronger at all times. That doesn't mean she isn't also strong. She's a very strong character, and she was very well deployed. Terminator 2 is considered one of the best action films ever made.

It has a strong woman, but she's not a Mary Sue. Everything about her is believable: her weaknesses, her strengths, and her ruthlessness, because it's connected to her core feminine identity as a mom.

It doesn't always have to be motherhood. There are other aspects of femininity you can connect to. But it can't just be, "I'm just as good as the men. I'm just like the men," because that's not how women are. They're not just like men. They're different. I realize in 2026 this is a controversial thing to say.

Jonathan: You just demonetized us.

Thomas: I don't care. The cars fly now. I'm having a good day. You're not going to get me off this high, dadgumit.

The joke was that we wanted flying cars and all we got was 140 characters. Now we have flying cars, and all people can do with their 140 characters is whine about it. I'm not going to whine. I'm going to enjoy it. I'm going to let the optimism flow through me.

Jonathan: Make your Dressmaker. Make exactly what your Timothy wants. This one person didn't have the dexterity to sew her own dresses, and that was something she loved. She loved to create. Now she has a program that gamifies it and lets her do it. She might even be able to download the patterns and get those dresses made in real life.

That makes you say, "Shut up and take my money." You're supposed to make your reader say, "Shut up and take my money," with every book you write.

When you do that, somebody will have so much fun, and fun is addictive and attractive. When you see somebody having fun, you think, "What's that guy doing? He's making dresses on his iPad. Is that fun?"

Thomas: Armor Maker, like a blacksmith simulator. I could see that being a relaxing game for dudes. You put on an audiobook, and you're in the forge.

Jonathan: And then you test different calibers of rifles against the armor.

Thomas: Crafting is an element in a lot of video games, that desire to make things. Why is crafting popular in video games now when it wasn't as big an element in the past? Because more and more people work in bureaucracies, particularly women.

There's a modern belief that a woman's place is in a bureaucratic system. She shouldn't be at home building a family; she should be building a bureaucratic system, and her most important role as a woman is as a girl boss in a big bureaucratic organization.

The problem with being a girl boss in a big bureaucratic organization is that you don't make anything. You sit in meetings, talk to people, and pass paper back and forth. There's a longing in every human heart, to make things, whether it's a meal, a dress, or a doll.

My daughter has learned to sew, and her delight in making clothes for her doll is unlimited. When she comes to show me, "I made pajamas for my doll," she's beaming. She's glowing, because the desire to make things is a core element of the human condition.

So many of us now have jobs where we don't make anything. At the end of the day, the physical world looks the same. You've had meetings. You've made decisions. A PowerPoint presentation has been changed. Nothing real has been made, and people long for that.

That desire to make things is an itch a video game can scratch, but a book can scratch it too. LitRPG often features extended crafting sequences, descriptions of somebody building, making, or assembling something, in what would seem like very boring detail.

It would have been boring to somebody 50 years ago whose job was doing things in the physical world. "Why would I want to read about somebody making something?" But for somebody starved of that, whose whole life is meetings, bureaucracy, paper, and forms, making something that's real, or feels real even in a virtual world, is so rewarding that they love it, and it makes them happy.

Jonathan: I read a story this week about a NASA engineer who went to work at SpaceX. He said it was hard to adapt to the culture, because at NASA you have to have two meetings and a couple of months before testing can even start. At SpaceX, he had a working model the next day. They machined it that day and tested it the following day. He wasn't used to that.

For people used to nothing ever getting done at work, which is really depressing, this is really attractive. You want to accomplish things. You don't want your time wasted by useless individuals in the supply chain.

I'm a big Warhammer 40K guy. In 40K you can create your own little army, your own chapter, with a backstory and everything, and you design their colors and heraldry. There are dudes arguing with each other about the colors they're painting their chapters while making fun of Dressmaker. Guys, it's not different.

Thomas: I've often worked in bureaucracies or in very virtual, digital businesses. I build websites, but especially as I got into management, a lot of my day was meetings. It was that middle-management experience where nothing has changed and we've just made decisions. I found mowing the grass very psychologically fulfilling, because the physical world had changed as a result of my actions when that wasn't happening at work.

That's fascinating, because generations ago every job resulted in changes to the physical world. Now we have this virtual, spiritual world online with no physical substance. This website lives in a data center somewhere. Is it even real?

It's easy for authors to get pulled into that. Is the book real? People only read the ebook or listen to the audiobook, so there's nothing real about it. Well, the emotions are real. The feeling is real. Print it on paper. Make a physical version that's worth owning. There's something to be said for tangible reality.

Jonathan: I do both. I create digitally with audiobooks and writing, but I also enjoy painting miniatures. If I do a good job, other dudes look at it and say, "That's cool. How'd you get the lightning effect?"

But I realized yesterday that because of Arsenal Audio, I'm doing a lot of audio projects, and a lot of the editing is done by AI. I still have to listen to all these audiobooks to QC them. What can I do while I'm QC'ing an audiobook? I can paint miniatures.

Thomas: We're so back. Let the optimism flow through you.

Evil corporations have had power over us for a long time, and they're losing it. Earlier today I was watching an AI show, AI for Humans, and one of the guys demoed an AI agent that could call customer service on his behalf. His agent was waiting on hold at Delta and working with the customer service person, and the customer service person said, "Can I please just talk to a human?"

Jonathan: That's so amazing. I love turning the tables like this, because I'm so done.

Thomas: This was the same company, Delta, that forced the AI agent to talk to an AI agent first before getting a human. The power balance is shifting, and it's interesting, because it's not shifting from the corporations to the humans. It's shifting from the corporations to the humans who use AI.

The humans who insist on an anti-AI view, which you see in our comments, are the humans who will be ruled over by the humans using AI. This is the new technology, the new source of power for the age we're moving into. You either have power or you don't.

You can't go back in time and say, "Pike and shot is the way to do warfare." It was for a time, but then pike and shot stopped working. The bayonet was invented, and all warfare changed.

Jonathan: I just watched one of my favorite movies of all time, The Last Samurai. I respected the samurai's adherence to honor and their rejection of firearms.

But at the end of the movie, and I'm spoiling this for you in case you haven't seen a movie that came out in 2003, they charge a machine gun nest. It has only two machine guns, and they're all gunned down. I'm watching this thinking, "I like the samurai. I appreciate the stand they were making and the message they're sending. But when the Marines went, they took every machine gun nest in the Pacific."

There's something to be said for understanding the changes in warfare and power dynamics and adapting. That's why the Marine Corps credo includes "adapt and overcome." We know how to charge machine gun nests. We know how to throw grenades, flank, distract, and lay down suppressive fire. We use flamethrowers. We'll do anything to take out that machine gun nest so the guy next to me doesn't get shot.

There are a lot of guys making their last stand who think they're holding the line against AI slop, or whatever words they want to use for it. But while you focus on the romanticism of that, you're ignoring the incredible advances made by the guys who adapted.

Thomas: People using Grok Bot on their websites are very quickly getting better websites than people who don't. Is it possible to hire a web firm to give you a better website than Grok Bot? Absolutely. Top authors and people with a lot of money can afford that, and that's still the case.

But most folks can't afford to spend $5,000.00 or $20,000.00 on a website. I worked for a web firm for a while, and our entry-level price for a website back then was $20,000.00. We built amazing websites for big clients who were very happy with them, but they were very expensive.

Even when I ran a web firm for authors, our low-end websites were about $2,500.00, which adjusted for inflation is closer to $4,000.00 or $4,500.00 now. You can build a website of equivalent quality with Divi and Grok Bot for a couple hundred bucks.

Jonathan: We're talking accessibility here. With my business, Arsenal Audio, I can now make audiobooks at a much lower cost than people are used to. People used to say, "I can't have an audiobook." Now I can make one chapter of your book as a lead magnet for your website for $100.00, and I can do it in an hour.

That has allowed me to hire other people. My female voice actress is a mom who works after her kids go to bed, and she's able to pull in side income from the narration contracts I send her, because I have a business structure enabled by AI.

I wasn't able to hire a female voice actress before. I was barely making it myself, because editing audio is so time-consuming. Now that bots take care of a lot of it and all I have to do is QC, I can hire her, and I'm hoping to hire more.

The success is being spread around, because more people can access more income now that productivity has gone up and the price has gone down. That's a good thing.

I understand everyone's concerns about AI slop. I had problems with human slop before AI was an issue. Terrible art has always existed, and being made with AI doesn't make it terrible.

Thomas: There's a thing called Jevons Paradox, and for those of you freaking out about AI, what Jonathan just described is Jevons Paradox, which goes all the way back to the 1800s. For many things, as the price goes down, demand stays the same and people just spend less. But for other things, as supply goes up, demand goes up as well.

The ultimate example is coal. The cheaper coal got, the more people used it for more things. Suddenly they were powering things with steam engines that had never been powered by steam engines before, because coal was getting cheaper and more plentiful. As supply went up, demand went up even more.

Accountants and spreadsheets are a classic example. There were something like 50,000 accountants in the country in the 1960s. They would spend all day or all week building a spreadsheet, because they had to do it on something like an abacus. The corporation would send out for a spreadsheet, and the accounting department would come back a week later with a whole page of math.

Then VisiCalc came out on the Apple II and did that whole page of math instantly. The result was that 20 years later we went from 50,000 accountants to a million, because the demand for spreadsheets went up far more.

A spreadsheet wasn't some precious thing anymore that only the CEO got three or four times a year. Suddenly everyone was making spreadsheets for everything, and you needed all these accountants to make sense of a world with dozens or hundreds of spreadsheets.

It didn't kill the accountants. Every accountant I know is busier than ever. They're losing their minds, they're so busy. Don't you have AI? The work is growing beyond the technology's ability to keep up.

Jonathan: We were talking about this before. I'm more tired every day now, because my brain has less downtime. When I worked on a single task like audio editing, it was boring, but it didn't tax my brain, because I wasn't constantly making decisions.

Now I have employees working for me and doing all the things, and I'm making decisions all day long: "No, stop that. Don't do that. You can get into that window. I told you you could. Good job. Now I need you to do it over here." My wife looks at me in my armchair at the end of the day, brain-dead in front of my video game, and asks, "Are you okay?" I tell her, "I'm working."

Thomas: Dora has a good comment here that I want to discuss. She says, "Don't want AI, don't use it, full stop."

This is an interesting element of the debate, because people who use AI are totally fine with people who don't. There's no social pressure like, "Why aren't you using AI? You're an evil person for not using AI." But the inverse is not the case.

I've noticed this with a very obnoxious kind of Southerner who has very particular taste in steak. I think it's a phenomenon unique to the American South, but maybe it's across America. It's a certain kind of man who likes his steak rare, and if he sees somebody get a steak well done, he's personally offended, even though his steak is still rare.

If somebody else wants to eat their steak well done, that's fine. It's a free country. They're an adult. Maybe their palate is different from yours. Maybe what they enjoy about food is different. Let them eat their steak well done. Why do you care?

Jonathan: It's so hard to get a good well-done steak, though. As someone who works at this, it's so hard. You're technically getting a bad steak, because even the chefs in the back are thinking, "I'm going to choose a steak that isn't the best part of the roast, because I'm not wasting a good steak on well done."

Thomas: Different people have different tastes, and something that tastes good to you may taste terrible to somebody else. I'll eat food my wife makes and think it's the most delicious meal I've had in days. My wife also enjoyed the food she prepared. Then I look at my own children, who share my DNA and my wife's DNA, and they're acting like they've been served toxic sludge. What is going on?

Some of it is that they haven't developed good palates yet. Some of it is the pickiness of toddlers. Some of it is that our tastes change as we age. And some of it is that they haven't even tried it yet. They think they've tried it, or they tried it a year ago and didn't like it then.

It's the same thing with AI, by the way. A lot of people tried it a couple of years ago, didn't have a good experience, and assume it's the same now. Or they tried some free version and haven't updated their view of AI.

As a parent, it's my job to expand my children's horizons, because while it is a free country, they are not adults, and I am their father. If all they eat is macaroni and cheese all day long, they won't be healthy, and I have a duty to God to help my children learn to eat protein.

Jonathan: The Bible says iron sharpens iron. That involves friction. So I rub up against my kids' dietary preferences from time to time. No, you're getting chili crisp on your buttered noodles. You can drink milk to reduce the sting, but your palate is going to learn.

I grew up with a guy who only got what he wanted: chicken nuggets and boxed mac and cheese. His palate was never challenged, and he never grew beyond that. It's kind of sad, because there's a whole big world of wonder out there. I'm contemplating trying uni. It looks gross, but there are people who like it, so maybe I'll try it.

Thomas: I will say there is a market pressure to adopt AI that the anti-AI people are responding to. Jonathan, you're now producing audiobooks at the same level of quality as the next guy, but at a lower price and with higher-quality. That means the person who's been doing it the old-fashioned way for years has to start explaining to potential customers why they shouldn't go with you, who's better and cheaper, and should go with them instead.

Jonathan: There are a couple of ways to defend a legacy workflow. They can say that what I'm creating isn't actually better quality, that listeners won't like it. But it's all still human narrated, just AI edited, so there's no actual difference in quality, and that argument falls apart.

Thomas: They're feeling competitive pressure, and it's harder to make a good case for their now-inferior product. Up to this point, their product was perfectly adequate, and people were happy to pay for it.

Nobody expected sound effects in an audiobook, because sound effects were expensive and no other audiobooks had them. Now that sound effects are inexpensive, people are starting to expect them, and the person doing it the old-fashioned way is unhappy, because they're forced to change.

So they're feeling pressure to adopt AI. They can learn a new thing and adopt AI, they can lose market share and revenue, or they can try to get everybody to stop using AI and go back to the old way. They'll try all the same strategies that, up to this point in history, have never worked. Not a single time.

My biggest emotion toward the people attacking me for AI is pity, because I see somebody who's about to be conquered. They're the samurai holding the sword with a moral objection to muskets, and they're going to charge the machine gun.

Jonathan: And I would love for them to have the guns, because they would use them so awesomely. They would be so good at it.

Thomas: They actually did. If you study the history of Japan, they had guns, and then they banned them. When I say we've never reversed a technology, or that it's never worked in history, I'm talking specifically about America. America has a very particular relationship with technology, different from other parts of the world.

Japan did roll back technology. So did China. At one point, China had the biggest fleet in the world. They were the only country with gunpowder and the only country with the compass. Instead of going out and conquering the world, do you know what they did? They burned their fleet, because the bureaucratic middle management of the kingdom was threatened by the idea of expansion.

That's very Middle Kingdom thinking, very different from Roman thinking, which is our cultural ancestry: "If I'm better than all my neighbors, I have a moral duty to share my greatness with them and help bring them up to my level and have some freedom."

If that means conquering them to bring them up to my level, then that's what civilization demands. I will conquer the Gauls, I will civilize the Gauls, and I will make Gaul a paradise again. Make Gaul great again.

Jonathan: I see so many storytellers who need to succeed, because they make such incredible stories. But their productivity flows are going to be overrun. I think inferior storytellers who were able to adapt to the new system will pass them, and we'll lose some of these legacy storytellers if they can't shift with the flow.

There's a concept in the video game industry, and Thomas, you know all about this, called the meta. In a game, there are weapons, items, and skill combos you like to use. Then the developers change them in an update or patch, and all the gamers watch YouTube videos to find out how the meta changed so they can stay the most optimal players.

Then you have the holdouts who say, "No, I like my original setup and my original items, and I'm going to force them to work." Those guys get crushed every single time, and they're not even playing for fun anymore. They're just playing to hate it.

Thomas: There's a common metaphor about the buggy whip, and it's a good historical period to study, because everyone focuses on the buggy whip. But most of the people who were making buggies have names you still know today, like Pontiac and Studebaker.

A new technology came along where, instead of attaching your buggy to a horse, you could put a motor on it and turn it into a horseless carriage. The buggy makers said, "I already know how to make a carriage. Now I just need to learn how to make a motor, and I'll be far wealthier than before. My carriages will be far more appealing."

Many of the early automobile companies started as carriage companies. And not just the carriage companies, but tire companies too, names like Goodyear and Michelin, who made wheels and tires for horse carriages before they did it for horseless carriages.

Most jobs are like the carriage business, not the buggy whip business. But some jobs are like the buggy whip business, where there just won't be much market.

I think translation is going that way. We'll very quickly reach the point where translating from one language to another is more or less a solved problem. Because of the way AI works, moving into tokens and out of tokens, it handles idioms as well as or better than native speakers. And since it has read all the written works of both languages, it can translate better than humans can.

There will always be some room for human interpreters. When the president of one country meets the president of another, each brings their own human interpreter. And there are still buggy whip manufacturers. They didn't all go out of business, because there are still buggies and still horses.

Jonathan: As someone who worked in interpretation at the kinetic level, you do have to watch out for interpretive bias, where the human's bias colors what the interpreter hears.

Several times I had to slap down native translators who were telling my Marines things that weren't true, because the interpreter hated the person speaking. AI translation won't do that, unless the designer or programmer puts that bias into the model itself, which will inherently make it less successful.

Thomas: Everyone listening needs to decide one thing: is my writing a profession or a hobby? That really guides your relationship with AI. If it's a hobby, you really don't want to touch AI, because it just takes the fun away.

It's like riding a horse. Riding a horse is a great hobby, and how you approach it as a hobby is very different from how you'd approach it as your job. The kind of horse you'd get is very different too.

There are still people who ride horses for work. Not many, but there are still working horses and working horse riders. Here in Texas we have quite a few. Even on a ranch where most of the work is mechanical, somebody on a horse is still a good way to round up the cattle.

But if you're approaching writing professionally, putting your head in the sand makes it difficult to compete. Not impossible, and things move slowly. The trad pub world all has its head in the sand, trying to hold the line.

Going back to Dora's comment, no one is bullying you to move into AI. Nobody is going to judge you for not using AI. But you will feel the market pressure. Why does your website look so bad, when everybody else's websites look better than yours?

Jonathan: Especially when it's so easy now to have a frictionless website. It may not be beautiful, but it works.

Thomas: Of all the Quills from my Grok Bot webinar, the one that gets the most positive feedback, even more than Quill Command, is the Webmaster Quill. It's the very first Quill I ever made, partly because I have deep knowledge of websites, but also because building websites is such a different skill for authors.

Jonathan: Nothing an author does relates to building websites. Websites are all visual and copywriting, which is nothing like what we do when we're writing a novel.

Thomas: Do you know what CSS is? Do you know how to center a div? Can you listen to Thomas's 30-minute episode about DNS, domains, and Cloudflare? Yes. You can learn how to do it yourself. I have all the education, and I have a course.

Somebody just went through the whole course and built a beautiful website. She sent it to me, and it looks great. I asked her, "Did you do this yourself, or did you do it with Grok Bot?" She said, "I did it myself."

She had that sense of accomplishment we were talking about earlier, because she had built a beautiful website that expressed her books well. It was a good-looking website, and it had taken her 40 hours. It looked like a website that 40 hours of work had gone into.

The thing is, a website of equal quality could have been made in half the time using AI agents. It's a free country, and she got a lot of emotional enjoyment out of building it herself. That's great. All the blessing.

The builder doesn't just build the house. The house builds the builder.

The more you build websites, the more technically sophisticated you become, and that helps in other areas of your life. All of that is true and beneficial. But as she gets busier, with more books coming out and more demands on her time, having someone, or something, help can help.

Jonathan: What novelist wants to spend all their time working on a website? We don't want to do that. The mentality we need to develop is, "I am the boss. I'm the CEO of my company, and these are my employees."

Good bosses know what their employees do, so develop that skill. Listen to Thomas's episode on DNS and Cloudflare. You can know the broad strokes without spending hours troubleshooting your own website.

I developed the skill of audio editing the legacy way, and I'm more than happy to outsource it to a bot that can do repetitive tasks. Yesterday I went through a piece of audio it edited. I highlighted and screenshotted the points it should have edited and kept dropping them into its context window so it could do better next time. I was training my employee, which I love. I'm a sergeant at heart. I love training new guys to be effective.

Thomas: We have a complaint in the chat that we're talking too much about samurai and not enough about the Romans. I'll give you the same metaphor with the Romans.

In the Bronze Age, the king of military weapon systems was the chariot. You read about chariots constantly in the Bible. It was a constant conundrum: what do you do about the other guy having 100 chariots, or 1,000 chariots, or iron chariots? These chariots are unstoppable.

But as the world shifted into the Iron Age, the Romans developed techniques for handling chariots that were so effective that chariot warfare completely disappeared from the Mediterranean. Then Julius Caesar, I think it was, crossed the Channel and landed on the haunted island of Britain, where no Roman had set foot before. The Britons had chariots, and the Romans mocked them, acting like it was a chariot race.

The Britons, and I think it was the Britons, though it may have been a different people group, were completely befuddled. They hadn't encountered Iron Age military tactics, and they had no idea their technology was no longer cutting edge. Chariots had been dominant for their lifetime, their fathers' lifetimes, and back into the mists of memory. Whoever had the most chariots won. These Romans invaded without any chariots. Obviously they were going to lose.

It turned out chariots were now a terrible weapon system. With drill, discipline, and unit formations, the Romans could direct where the horses went more than the chariot driver could, and then they just threw javelins at the chariots. It became a game, and at no point were the Roman soldiers in danger.

That vertigo, that shock of realizing your whole civilization is out of sync, is civilization ending. The Britons and the other peoples the Romans encountered didn't fare well. Century after century, the Romans just won. They didn't just win the wars. They won the battles.

When the Germans finally started winning battles against the Romans, they were using Roman tactics. The Germans hired themselves into Roman units as mercenaries, then left and took the tactics and weapon systems with them.

The chariot never returned. The closest thing is the tank, and even tanks are starting to go away.

Ultimately the Germans beat the Romans by out-Romaning the Romans. "We can have your drill, your shields, and your iron, but we will have honor, and you have decadence." That was the difference. They had the same tactics and weapons, plus the honor of barbarians, while Roman honor had declined as the empire became decadent.

The cars fly now. The computers think now. You can use the thinking machines or not. It's a free country. But if your goal is to make money, you have to be competitive, and that means not running away.

Liked it? Take a second to support us on Patreon!

Barnes & Noble Foot Traffic Soars and Cars Can Fly Now

Publishing News

Half of Audiobook Listeners Open to AI Narration

Commentary

Smashwords Finally Takes Preorders, and Most Authors Still Skip the Tool That Earns More

Commentary

Barnes & Noble Traffic Up 34%

Commentary

Author Alerts

Grok Bot Webinar #2 Explained

Commentary

Substack Takes Over Sales Tax on Paid Subscriptions December 1

Commentary

September KU Page-Reads Paid From a $65.9 Million Pot

Commentary

BookBub Adds Sci-Fi and Christian Fiction to New KU Email

Commentary

Technology

ArtCraft's Free Word Clone Hits Alpha With Real Documents Still Untested

Commentary

Free Photoshop opens real PSDs. It is still alpha.

Commentary

ElevenReader Puts Its Top Voice Model on the Free Plan

Commentary

Google Lets Authors Prototype Book Games Without a Studio

Commentary

Zero to Hero: Meta Now Has a Frontier Model

Commentary

Open-Weight Drama 3 Puts Directed TTS in Authors' Hands

Commentary

Fiction Bench Makes Creative Writing a Measured AI Category

Commentary

Zeitgeist Report

Flying Cars Are Here: Is the Media Trying to Ignore It to Affect the Zeitgeist?

Commentary

They Called Dressmaker Too Niche. It Paid for Itself in a Day

Commentary

Leave a ReplyCancel reply

7 Secrets of Amazing Author Websites

You have Successfully Subscribed!

Recent Episodes

Categories

Tags

Want more help?

You have Successfully Subscribed!