For Week Ending October 23, 2026
Amazon just retired Kindle Unlimited, and now the new Kindle+ pays audiobook listening out of the same KDP Select pot as your page reads. Meta now says to accept its crawl terms by November 20 or leave Instagram for business. The NSA says to patch PHP or take your site offline. Catapult shuts down October 30. And in our Zeitgeist Report, Britain arrested 62,199 people over posts and messages.
Publishing News
Breaking: Amazon Just Killed Kindle Unlimited and Replaced it with Audiobook-Integrated Kindle+
According to the KDP Community announcement, Amazon rebranded Kindle Unlimited to Kindle+ and expanded the subscription to eligible audiobooks for customers on Amazon.com.
Kindle Unlimited is gone. As of October 22, Amazon has rebranded it Kindle+, and the subscription now includes eligible audiobooks for customers on Amazon.com.
The price didn’t change. Still $11.99 a month. Still a 20-title borrow limit. What changed is where the listening money comes from.
It comes from your pot.
Audiobook listening now draws from the same KDP Select Global Fund that pays for ebook pages read. Before October 22, that fund paid for one thing: pages read. Now it pays for pages read and pages listened. One pot. Two formats.
Amazon says it is “investing additional funds to support this new and growing audiobook catalog.” It also says that “as we learn how customers use both formats, we’ll evolve the fund to reflect those differences.”
What Amazon did not say is how long that extra money keeps coming.
And this is the part every Select author needs to understand. Amazon has always set the size of the fund each month, at its own discretion. Nothing locks in that extra contribution. If listening grows and Amazon decides the additional money isn’t worth it anymore, the blended per-page rate drops. And it drops for everyone, including authors who have never made an audiobook in their lives.
Now, who’s in the catalog?
If you have a Virtual Voice audiobook, you’re already in. According to Amazon, when the ebook is enrolled in KDP Select, Virtual Voice audiobooks are automatically included in Kindle+ and in the Audible Plus catalog. If your ebook is already enrolled, your audiobook is live for Kindle+ listeners right now. No action required.
There’s also no separate opt-out. Creating a Virtual Voice audiobook is still an invite-only beta. But once that audiobook exists, Select enrollment drops it into Kindle+ automatically. To get it out, you have two choices: unpublish the audiobook or pull the ebook out of Select. You cannot unenroll the Virtual Voice version on its own.
Human-narrated ACX audiobooks face a much narrower door. To enroll from your KDP Bookshelf, a title has to meet three conditions. Exclusive distribution with Audible. Whispersync for Voice turned on. And the Pay for Production payment model.
Royalty Share and Royalty Share Plus titles are locked out at launch. Amazon says it is exploring expansion to those models. For now, most human-narrated catalogs stay on the outside.
Listening is U.S.-only at launch. Amazon lists other marketplaces as “to come.” Authors anywhere in the world can still enroll eligible titles for U.S. listeners.
So How Will You Know If Your Rate Is Getting Squeezed?
Amazon updated reporting. Your KENP report now splits Kindle+ pages read from Kindle+ pages listened. Consolidated royalties show up in the Prior Month’s Royalties report. Those numbers are the first real evidence of whether Amazon’s extra money is keeping pace with the new listening.
Here’s what to watch. The October and November fund announcements, and your new split KENP reports. If the per-page rate holds, Amazon kept its word. If it slides, the audiobooks are eating the ebook money.
If you’re in Select, this changes how you get paid, so make sure you’re subscribed and you’ll hear about those fund numbers when they land.
Sources:

Commentary
Jonathan: Amazon has always set the size of the KU fund every month. They have their own math and their own equation, and now they've thrown in another format that the fund pays out to.
If you don't have an audiobook that's eligible for this program, you may see you're per-page payment (KENP rate) go down.
Thomas: But not immediately. At first it's just more money. The sugar to help the medicine go down is that Amazon is putting in extra money. From what I can tell, for this first month all of the payments for audiobook listens will come out of Amazon's corporate account.
For the next month it will be great, and everyone will be swimming in gravy. But they haven't promised to do that perpetually.
I'm sure the suits at Amazon are thinking, "With this rebrand and a big marketing campaign heading into the holiday season, we'll make it up in volume and phase down the margin we're losing to fund this." They are not going to donate free money into this fund forever.
They're putting in some free money up front to keep the torches and pitchforks down. Maybe their projections say the new model will bring in so many new subscribers that the payout stays about the same.
But as they reduce how much they offset the audiobook listens, even if the overall pot is bigger, the way it gets divided is changing. Audiobooks now have a seat at the table.
If you have your book in audio and ebook, this probably won't affect you much. You'll just be getting paid from two different pockets. But many KU authors who are not on Audible, still many people say, "It's too hard." You can get paid through this new program just by checking a few boxes that let Amazon make an AI version of your audiobook.
Jonathan: Virtual Voice is Amazon’s AI version of your book.
Thomas: I want to be clear that there are two versions of Amazon’s AI audiobook. There's always been Kindle text-to-speech, which has always been garbage. That's not what we're talking about.
Virtual Voice is the newer, better AI model. I don’t think it’s as good as ElevenLabs, but it's not the Amazon voice from 10 years ago. It's their new invite-only program.
Jonathan: If your ebook is already in Kindle+ and you have a Virtual Voice audiobook, you're already in this fund. Every time you get a listen on a page, you'll be paid out of it.
Human-narrated ACX audiobooks that are distributed exclusively through Audible and used the Pay for Production model (you paid upfront for the audiobook) are eligible.
Royalty share audiobooks will not be put into Kindle+. Most human-narrated catalogs will stay on the outside, because the majority of them are royalty share, especially for KU authors.
If you pay up front for your audiobook, you can do whatever you want with it. If you want to stay exclusive to Audible and get paid out of this fund, you'll be okay.
I've started a business that reduces upfront audio costs. You can use the estimator on my website to see how much you'd save. I use AI editing and human narration, and if you want to enhance it with sound effects and music, you can. You'll have a better product than everyone else out there.
Thomas: AI-enabled authors, the kind of author maxing out their two published books a week, will be hurt most by this change.
Every week Amazon makes some new rule change to hurt the AI rapid-release authors. They were limited from 10 books a week down to two. But two is still a lot, especially for producing audiobooks. AI-enabled authors have never been doing audiobooks. In fact, one of the best ways to look like an AI author is to not have an audiobook.
The authors who will benefit most are the ones who've invested a year into a book an decide to have an audiobook made. Few authors say, "A year of my time is not worth an audiobook." Of course it is. The audiobook also helps you look more human.
You already have to be exclusive with the ebook to be in Kindle Unlimited/Kindle+.
Jonathan: I wonder if we're heading toward a merger of Kindle+ and Audible. We've been reporting on Spotify taking market share in the audiobook market, and it's starting to sting.
This is a move to keep market share and keep people in the Amazon ecosystem by making it simpler and more attractive, so people don't go to a new ecosystem.
Thomas: This was the natural escalation of Spotify adding ebooks, which we reported on about a month ago. Amazon is saying, "You're adding ebooks to your audiobook subscription? We'll add audiobooks to our ebook subscription."
I wouldn't be surprised if Amazon keeps merging subscriptions. Prime Video might get merged in at some point, and you'd have an Amazon Entertainment subscription where comic books, audiobooks, ebooks, and TV shows all compete for the same pot of money.
If you're an ebook-only author, you have been drafted into the front lines of the audiobook wars between Audible and Spotify.
Will Meta Force Authors on Instagram to Submit to AI Scraping?

Meta can't force you to stay on Instagram, but if you keep using Instagram for business on or after November 20, you've agreed to let Meta's bots crawl your author website. You don't sign anything or click anything. Using your account counts as agreeing. The terms add no opt-out setting. If you don't agree, your only option is to stop all business use of Meta's products.
So the real question isn't whether Meta can force you. It's whether you'll walk away from Instagram to avoid submitting to the new rule.
What Meta Is Asking You to Accept
On October 15, Nieman Lab reported a clause hidden in Meta's updated Commercial Terms. It authorizes Meta, or anyone acting on its behalf, to use "web scrapers, bots, robots, spiders, crawlers, and user agents and create test credentials" to crawl the domains and URLs "associated with your products or services," and to keep pulling updates from them.
Who's covered? The terms apply to anyone using Meta for business. That includes managing a Page, running ads, listing products, building a product catalog, using Ads Manager or Meta Business Suite, and using Meta's measurement tools. For authors, that means anyone with an Instagram business profile, a Facebook author page, or a Meta ad account.
What about your website? If it's linked from your profile or used as an ad destination, it's arguably "associated with" your products. Sending your ads only to Amazon doesn't clearly take your site out of scope.
The "test credentials" language matters too. Experts say it could, technically, let Meta sign up as a user to get past registration walls. That could expose reader-magnet downloads, ARC portals, member areas, and paid bonus chapters.
Didn't Meta Back Down?
Partly. The first draft said Meta could use the content to "train" its products. On October 16, Meta removed the word "train" and limited the use of future scraped content to "advertising and business products."
Lawyers say that doesn't fix it. Media and IP attorney Matt Topic said "develop" is broad enough to include "train."
Then, on October 23, Meta pushed the deadline back three weeks, from October 30 to November 20, without notice. As of that day, the date was the only thing Meta had changed in the terms.
The delay came after pressure from several directions:
- US publishers: The News/Media Alliance demanded Zuckerberg suspend the terms.
- US Senate: Senator Marsha Blackburn said removing the explicit training reference didn't go far enough.
- UK regulators: The Competition and Markets Authority said it was in urgent discussions with Meta and would keep scrutinizing the terms.
- EU regulators: The European Commission said it is committed to making sure Meta complies with EU competition law and the Digital Markets Act.
German competition lawyer Thomas Höppner thinks the delay may be a legal tactic. Courts can deny emergency relief if affected parties wait too long to sue after learning of a change, because the matter no longer looks urgent. In other words, a three-week pause could help Meta run out the clock on an injunction.
Don't Fall for Meta's Spin
Meta's main defense is that the update doesn't change how it uses web crawlers. That's probably true. And it's an admission because Meta's bots have been crawling the web for years:
- Link previews: Meta's link-preview crawler reads your page whenever someone shares your link. Meta's own documentation says it may bypass robots.txt for security or integrity checks.
- AI training: In July 2024, Meta quietly launched a crawler that scrapes publicly displayed web content. It didn't announce it beyond updating a documentation page. That crawler is Meta-ExternalAgent, which Meta says crawls the web for uses such as training foundation AI models.
- Scale: Cloudflare measured Meta-ExternalAgent at 0.9% of AI crawling in July 2024 and 7.5% by July 2025. In Cloudflare Radar data for the 30 days ending August 31, 2026, it ranked second only to Googlebot.
- Earlier permission: Meta's old terms already allowed crawling of sites tied to business accounts in one narrow case, inside the terms for generative AI ad features. The new clause moves that permission into the general terms and widens what Meta can do with the content.
The crawling isn't new. Your consent is. Today, if Meta's bots take something they shouldn't, you can object. After November 20, Meta can point to a contract you accepted by staying on Instagram. Legal experts told Nieman Lab the terms could be argued to override robots.txt.
Meta's other defense is that you can still block its crawlers with robots.txt. Meta keeps saying this in press statements, but it has never put that promise in the contract. No version of the terms includes language about respecting robots.txt. And robots.txt is a request, not a rule. Crawlers follow it voluntarily.
Authors have reason to want that promise in writing. In Kadrey v. Meta, a lawsuit brought by thirteen authors, the record showed that Meta downloaded the LibGen pirate library. After licensing talks stalled, Meta decided to train on LibGen and dropped its licensing efforts once it found LibGen held most of the books it had been trying to license. Plaintiffs said Meta's own testimony showed Zuckerberg approved using LibGen despite internal concerns. Employees called it a data set they knew to be pirated.
Will It Happen?
Three things could happen before November 20:
- Meta rewrites the clause. Publishers are demanding that Meta put a robots.txt commitment into the contract. So far, Meta has offered explanations, not edits.
- Regulators step in. The UK's CMA and the European Commission are both looking at the terms. A ruling in Europe wouldn't necessarily protect American authors.
- Nothing changes, and the terms take effect. This is the outcome to plan for. As the UK News Media Association's Theo Bamber put it, the delay is "a pause not a solution."
Big news publishers have trade groups, lawyers, and regulators fighting for them. Indie authors don't. If the clause gets fixed, the fix will be negotiated for them, not for you.
What Authors Should Do Before November 20
- List every website linked from your Instagram profile, Facebook Page, and ad accounts.
- Add a robots.txt rule blocking Meta-ExternalAgent, the AI-training crawler.
- Allow Meta-ExternalAds only if you advertise on Meta.
- Lock down anything behind a signup form: reader magnets, ARC portals, and bonus content.
- Check your server logs a week later to see which Meta bots still show up.
- Decide whether your Instagram business account is worth the signature.
Sources:
- Meta Commercial Terms preview (now effective November 20, 2026)
- Meta's new commercial terms will allow the scraping of news sites, legal experts say | Nieman Lab
- Legal experts warn Meta's new terms still grant broad scraping permissions… | Nieman Lab
- Meta delays rollout of terms… Legal experts are still skeptical | Nieman Lab
- Meta quietly delays controversial update… | Fast Company
- Senator Says Planned Meta Commercial Terms Update Could Harm US Businesses | Reuters via US News
- Meta's New Terms: What Its Crawler May Read | Digital Applied
- Meta quietly launched a web crawler for AI training data | Fortune (2024)
- Meta-ExternalAgent | PPC Land
- Kadrey v. Meta case summary | WIPO
- Zuckerberg OK'd training on copyrighted works, filing claims | TechCrunch
- Meta Web Crawlers | Meta for Developers
Commentary
Jonathan: It's like an ultimatum in a relationship that says, "You have to meet my parents and impress them, or it's over." That's what Instagram is doing. "You'll agree to this new thing, or we're done."
Thomas: Some authors loves talking on Instagram about how evil AI is, and about authors who write with AI, as if Meta isn't an AI company. Meta is AI all the way down. Instagram is AI all the way down.
The dirty secret is that they've been doing this the whole time. Don't think Meta hasn't already scraped your author website.
I have a tool in the Patron Toolbox that shows you how much Meta already knows about you. You type in your name and your book, and it checks with Muse to see how good its context on you is. It can't search the web, so you get a very different answer than you would otherwise. You can do the same for ChatGPT and the others. They're all locked away from searching the web so that you know exactly what AI knows about your book.
I don't pay extra for web search on those tools, which is the best guard against AI doing something you don't want it to do. If you don't pay for it, it won't do it. You can see how much Meta knows about you, and it probably already knows quite a lot.
Meta is also in a court case similar to the one Anthropic settled for $1.5 billion, because Meta pirated books the same way Anthropic did, downloading them illegally. Your book may be inside their pirated collection.
My suspicion is this will change nothing, because people's relationship with Meta is like that of an abused spouse. No amount of additional abuse will cause people to leave Instagram. All their friends are there. They're addicted to the affirmation. You get thumbs up and hearts, and you feel loved. You feel like these really are your friends.
But they don't love you. They're not your friends. Some of them aren't even people. But it feels real, and that feeling is enough to keep people around, despite everything they have to agree to in the terms of service.
Jonathan: I'd say AI has made it more recognizable how much of Facebook is not real. Three-quarters of my feed is now recognizably ChatGPT-generated images.
Thomas: It's all AI. The people creating content use AI. The images are generated with AI. Half the scripts people read into their cameras for reels are written with AI's help. What you're shown the next time you swipe up is decided by AI. Top to bottom, it's all AI.
If you're anti-AI, it's a free country and you're an adult. But if you want to be consistent with your ideology, your anti-AI jihad, you have to get off Meta.
Jonathan: But then how will you tell people all the time that you're anti-AI?
Thomas: You could do it in real life.
As soon as your book is on Amazon, you're dealing with AI, because the Amazon algorithm is AI. You've been dealing with a machine learning algorithm for 20 years. AI determines what people see and what they don't.
The people least threatened by this new world are the Mennonite authors who sell at homeschool book fairs. There's a whole world of book fairs where people buy books with cash on the barrel, in person, face to face. Tens or hundreds of thousands of dollars of books get sold in a day.
You'll find legit Mennonites in bonnets. There are whole Mennonite publishing companies you've never heard of unless you're in that world, and they're doing great.
Jonathan: Taking payment on their phones.
Thomas: The other people who will do fine are the ones who have fully embraced AI. They have GrokBot running their publishing company. Much of it is automated, and they can focus on their writing.
The folks who are struggling are the ones trying to split the difference. They refuse to let AI make their lives easier, but they still force themselves to compete in a world that's AI top to bottom. Those are the people who will get squished. There's safety on both edges.
There's a famous story from the Romans I'd like to share.
Jonathan: Of course there is.
Thomas: Very early in the Roman Republic, when Rome was just a city, the Romans were at war with one of their Latin neighbors. That town caught the Roman army on the side of a mountain and trapped it there, stretched out and vulnerable.
The town's leaders sent word to their elder asking what to do. "Should we give them mercy or kill them all?" He said, "Kill them all." They didn't like that answer, so they asked again, and he said, "Give them mercy." Confused, they carried the old man on a litter to the battlefield and asked him in person.
He said, "Either show them mercy, and they will remember and be your friends forever, or kill them all, and they'll be dead and you won't have to bother with them ever again." Both paths were good paths.
The leaders gathered and decided to split the difference. They made the Romans march under the arch, the ultimate humiliation in the ancient world.
How do I describe it in a family-friendly way? Troops line both sides mocking you, and you have to walk through an arch carved with female imagery. You're basically sacrificing your masculinity as you walk under it.
It was so bad that when the army came home, Rome closed its gates to them. "You're not allowed in the city. You're no longer Roman men. You have walked under the arch." Rome hated that town so much for humiliating its troops and taking the middle path that it eventually conquered it completely. It did not go well for that town.
So often in life, the best solution is one of the extremes. Walking the middle path is what leads to destruction. Shooting between the targets is the path to destruction.
Jonathan: The Samnites.
Thomas: Thank you for the fact check.
Did Pangram Pirate Your Book?

According to Publishers Lunch (Katy Hershberger, October 23, 2026), a writer in September ran an unpublished manuscript through Pangram’s similarity checker. The tool flagged two lines. Both were common phrases, not exact full-sentence matches. One linked to a post in an online forum. The other linked to the full pirated manuscript of a published book, hosted on Notability.
Notability is a private note-taking and annotation app. It is not built as a public content repository. Yet the checker treated that pirated file as a source and returned a link to it.
What Happened
Pangram added the similarity checker in August 2025. Its own announcement said the tool searches the open web for matching documents the way Google does. It scans each sentence across millions of articles, websites, forums, databases, and other sources. When it finds a long stretch of matching text, it highlights the passage and gives a direct link to the source. Access was limited to paid subscriptions or organizational licenses.
Pangram’s knowledge hub (last reviewed July 24, 2026) describes the same feature as scanning against a third-party database of published work. The marketing page still says it compares text against billions of webpages, academic papers, news articles, books, and more. Those two descriptions sit next to each other. One sounds like a licensed corpus. The other sounds like open-web search that can reach forums and, in this case, a pirate file on Notability.
This is not the first time pirated text has surfaced around Pangram. According to WIRED (September 2, 2026), co-founder and CEO Max Spero used a pirated PDF of Mia Ballard’s manuscript Shy Girl for an AI-detection test that later drew public attention. When asked, he said he had not looked closely at the file. He put it straight into Pangram. That incident involved the AI detector, not the similarity checker. Pangram’s own accuracy post states the company never benchmarks its AI model on pirated content and uses commercially licensed human writing for training. The similarity checker is a separate product. Its public documentation does not make the same exclusion.
No company response to the Publishers Lunch piece has appeared. The free teaser of the Lunch article ends mid-sentence, so the full body is not public. The specific book title and author of the pirated manuscript are not named in the available text.
Why Authors Care
The tool publishers and schools use to police originality is matching against at least one full pirated manuscript and an open forum. If your book is already on a pirate site, the checker can surface it as a “source.”
Pangram markets the feature to educators, content moderators, and publishers. A match against a pirate file can still appear in a report an editor or instructor reads.
The company’s own pages use two different frames for the same feature: “third-party database of published work” and “open web.” Authors cannot tell from the public materials whether known pirate hosts are filtered.
If you run a paid similarity check, the knowledge-hub page that asked whether uploads are added to the repository returned a 404 at the time of this research. Confirm current data handling before you upload an unpublished manuscript.
Authors who use Pangram for AI detection can keep doing so if the accuracy numbers matter to them. The similarity checker is optional and paid. Before you enable it on a draft, read the source links it returns. A match to Notability or a forum is not the same as a match to a licensed book. If a report cites a pirate file, treat the result as evidence the index is messy, not as proof someone copied you.
Sources:
- Pangram’s “Similarity Checker” Compares Text, But its Sourcing Raises Questions (Katy Hershberger, Publishers Lunch, October 23, 2026)
- Introducing Pangram’s Plagiarism Detection (Pangram, August 1, 2025)
- What does it scan? (Pangram knowledge hub, last reviewed July 24, 2026)
- Similarity Checker | Pangram
- Pangram Has Emerged as the Gold Standard of AI Detection. Should You Trust It? (WIRED, September 2, 2026)
- Is Pangram accurate? (Pangram, October 16, 2026)
Commentary
Thomas: I thought Pangram was the anti-piracy company. Are they pirating books?
Jonathan: We reported on the Shy Girl incident when it drew public attention.
Thomas: So, at Pangram, the left pocket is not pirating any books. The right pocket is a separate product and is pirating books.
A good metaphor for this is Getty Images, which declared bankruptcy this week because all the AI companies trained on its images, and demand for stock photos dropped to basically nothing.
Why would you buy an expensive copyrighted stock photo from Getty when it means you have to you deal with constant harassment from copyright trolls threatening to sue you even though you bought it legally. Meanwhile, you can create an AI-generated image that costs nothing and brings no trouble from copyright trolls. Getty lost so much income that it can't keep up with its debts.
Pangram is in a similar situation, where companies can just pirate people's work and make enough money in the meantime. Even if the AI companies lose a copyright case to Getty, it will take 10 years to work through the courts. And it's unlikely they'll lose, because Getty presented thumbnails of its photos for free online. Reading a webpage presented free online isn't exactly a copyright violation, especially if you're tokenizing the images. They might win, they might not. It's not an open-and-shut case.
What's happening is that these AI companies are powerful enough to roll over smaller companies and individuals. Copyright is not the protection you think it is, and it's not a good source of money for authors. It's a good source of money for Disney, Nintendo, and Sony. It's not a good source of money for individuals.
But it is funny to me that Pangram, supposedly the big arbiter of good, is potentially sourcing through shadowy means.
Jonathan: "Bring us your books. We'll keep them safe."
Thomas: If Pangram gets sued the way Anthropic was sued for downloading pirated books, Pangram can't afford a $1.5 billion settlement. Although I don't know how much of a precedent a settlement creates. It's not as much of a precedent as a ruling.
Jonathan: It creates an expectation, which isn't the same as a precedent, but it does create an expectation.
Musk Will Bar Ugly or Overhyped Ads on X

According to Elon Musk, going forward all advertising on X will be subject to a reasonable esthetic standard, and advertisers who post ads that are ugly, shrill, strident or overstate claims about their product will be barred from advertising. An hour later he asked users to report ugly or misleading ads to the @Ads account. X has not updated its formal Ads policies or help center; the existing Quality Policy already requires proper grammar, bans hashtags and URLs in ad text, limits emojis to one, and rejects gimmicky or indecipherable media, while the Deceptive & Fraudulent Content policy already prohibits clickbait and exaggerated claims.
Authors running book ads should strip all-caps urgency, multiple emojis, hashtags, and unsubstantiated “life-changing” or permanent-bestseller language from their creative, then recheck the current Quality Policy before the next campaign. User reports to @Ads are the enforcement path Musk named.
Sources:
- Elon Musk post on esthetic standard
- Elon Musk post asking users to report ads
- X Ads Quality Policy
- X Ads Deceptive & Fraudulent Content policy
- X Ads Policy Update Log
Commentary
Thomas: Musk is barring ugly or overhyped ads on X.
Jonathan: This is already in the policy documentation, but the fact that the CEO has made it his priority sends a directive to everyone else at the company, he’s saying that they need to make sure this is not what the platform looks like. If people are complaining about ugly or overhyped ads on X two months from now, after he's already said to take care of it, people will get fired.
Thomas: This is good for you if you've been advertising on X or are thinking about it. As the quality of the ads increases, people will look at that part of the interface more, which means they'll become less banner blind.
Banner blindness is a real problem with online ads. During the heyday of Google Ads, they worked really well. There were often only one or two, they were well integrated into the product, and they were relevant and high quality. That quality made them highly effective.
You don't want to be the only high-quality ad in a sea of slop, because people will be trained to assume it's all slop. You want to be in a good neighborhood with your advertising. So this is a good change.
I did my big GrokBot webinar yesterday, and you can buy the replay. In my first GrokBot webinar in September, which also has a replay available, I shared a quill, a preprogrammed GrokBot that can run your ads on X for you or help you run them.
This is an area I'm optimistic about for growth, particularly for people who write for men. There are a lot of men on X. I think it's the only social network with more men than women. Advertising to men on Facebook or Instagram is tricky because those platforms skew far more female. I think Instagram is 60% or 65% women.
Jonathan: Reddit is 90% men.
Thomas: Especially if you measure by minutes. A lot of people just have an account. Technically I have an Instagram account because my wife sends me memes, but I haven't logged in for months so now she just shows them to me on her phone because she knows I don't log in. That's how Instagram is, whereas X is the other way.
Catapult Press Goes Out of Business

According to Jane Friedman’s The Bottom Line on October 21, 2026, Publishers Lunch reports that Catapult is being sold and will wind down operations on October 30. The staff union confirmed the news to Publishers Lunch. The sale covers the company’s assets only.
What Happened
Catapult was co-founded in 2015 by Elizabeth Koch, who funded it, and Andy Hunter, its first publisher (he stepped down in 2022). It started as a combined books, magazine, and writing-classes venture, then took on publisher Counterpoint and its Soft Skull imprint in 2016. It dropped the magazine and classes in 2023 to focus on book publishing and added Hawthorne Books in 2025. By late 2025 it described itself as operating within Unlikely Collaborators, a nonprofit Koch founded. Staff unionized in 2026.
Publishers Lunch, citing the union, states that the sale includes the company’s assets only and that continued employment with the prospective buyer is not an option. The union also said the company announced plans for a sale while the NLRB vote was ongoing. The company has not yet responded to requests for comment. Typically, in an asset sale the buyer acquires things like the backlist and imprint names rather than the company itself.
Why Authors Care
Authors published by Catapult, Counterpoint, Soft Skull, or Hawthorne should pull their contracts this week and look for reversion clauses triggered by cessation of business, failure to publish, or nonpayment of royalties.
Books still in the pipeline face the highest risk: editorial, production, and marketing staff are winding down by October 30, and the buyer of assets may not take unfinished titles.
Royalty statements and payments can stall or stop if the selling entity no longer has staff to process them, even if the buyer later acquires the backlist rights.
Other authors should treat small-press contracts as the only durable protection. A well-funded indie with a nonprofit wrapper and recent acquisitions can still sell assets and shut the doors, leaving authors to negotiate reversion or a new home for their titles.
Sources:
- Independent publisher Catapult is winding down (Jane Friedman, The Bottom Line, October 21, 2026)
- Catapult Is Being Sold; Operations Will Wind Down On October 30 (Publishers Lunch)
Commentary
Thomas: Yet another traditional publisher has fallen, in an ongoing story of traditional presses going out of business. This one unionized, and a few months later it went out of business.
As a general rule, when white-collar people unionize, it's a sign things at the company are not going well. White-collar people have very comfortable jobs. You sit all day in air conditioning, and if sitting all day in air conditioning is too difficult for you, something is really broken in the company.
I understand a lineman dangling from power lines in a hurricane, whose job is to stretch the line over and get the power restored, wanting a union so he doesn't get electrocuted. That makes sense. "We need safety rules. I can't go out if the wind is more than 30 miles an hour."
But somebody sitting in a New York office in air conditioning, typing on a computer, unionizing? I roll my eyes. This happens so often in media. You see a unionization effort, and then you see a bankruptcy, often within 12 months. We saw something very similar at Gimlet Media.
But I suspect Catapult had deeper, more fundamental issues, like not publishing books readers wanted to read, and we're seeing this industry-wide. EA is closing BioWare over likely the same sorts of problems. They had a product people weren't interested in.
EA Closes BioWare, Cancels Next Mass Effect
According to BioWare's October 21 statement on X, the studio is closing after more than 30 years and early development on the next Mass Effect project is ending. EA's employee memo from the same day, signed by Chief Studios Officer Cam Weber and COO David Tinson, confirms the Edmonton studio closure as the company concentrates on fewer large-scale action-adventure projects while continuing Marvel's Iron Man and the next Star Wars Jedi game; Bloomberg reports the cuts cover around 400 roles company-wide, about 3 percent of staff, including roughly 50 at BioWare. EA still owns the Mass Effect and Dragon Age intellectual property, and with no team left working on them former executive producer Mark Darrah has said the publisher is more likely to sell the franchises or leave them dormant; past tie-in novels and Dark Horse comics expanded both worlds, but writers including Derek Kunsken and Catherynne M. Valente, who had been working on the canceled Mass Effect project, have said on X that the team had a strong game in progress. The pipeline for franchise novel and comic work at one of the last major story-driven game studios just closed. Authors who built careers on licensed game tie-ins now face a thinner market as publishers chase fewer narrative-heavy projects. Indies who own their worlds outright do not need a studio's permission to keep writing them.
Sources:
- BioWare on X
- Building EA's Next Chapter
- Bloomberg: Electronic Arts to Cut 400 Jobs, Shut Down BioWare
- IGN: Former BioWare EP Mark Darrah on IP
Commentary
Jonathan: This is a huge deal, because this was a very profitable studio. It has had hits people have been loyal to for 20 years. We're talking Star Wars: Knights of the Old Republic and the Mass Effect franchise.
Those of us who played these games are thinking, "They failed? Holy cow. They made billions of dollars."
Thomas: Knights of the Old Republic, one of the greatest role-playing games ever made, is a BioWare production. But their more recent games were very woke and gamers reacted by calling them very cringe, as Gen Z likes to say.
Jonathan: Veilguard is really what took them out. The Dragon Age series was never a conservative video game series by any stretch of the imagination. But in Dragon Age: The Veilguard, which we covered in the early weeks of Author Update, the story was slowed down so much by conversations about pronoun usage and sexuality choices that players said, "I'm not here for this. I'm not playing my medieval combat fantasy game against dragons to have an HR meeting every 10 minutes. This isn't why I'm playing the game."
So their profits started plummeting. They announced that Mass Effect 5 will not continue in production. It's a huge break.
Of course, all the commentators on the right are saying, "Go woke, go broke," which to a large extent is true. They put a lot of stuff into the game that gamers didn't want to play.
Thomas: It's a pattern-matching thing. It's not just the woke content. It's everything that goes along with it. You have people in your organization making decisions for political and ideological reasons instead of for thrill-the-customer reasons.
They forgot who their Timothy was. They were on a quest to change the world rather than thrill their customer. In some ways, that was the real problem, and going woke was just the signal of the problem.
You can have a very woke company that knows its customer base and knows how to thrill it. Ben & Jerry's does this very well. They're a woke company, but they know who their woke customers are, they're zeroed in on those customers, and they've been woke from the very beginning. Every once in a while conservatives try to cancel Ben & Jerry's, and I just roll my eyes. Everybody knows they're leftists over there.
Jonathan: Conservatives weren't going there anyway. What are we canceling?
Thomas: It's like being surprised about Lindsey Graham. We all knew.
On the other hand, BioWare's earlier games were just fun games. I don't remember the first Baldur's Gate, a 2D isometric RPG, having any modern-day politics in it. Sure, it had D&D politics, with the factions and kingdoms of a fictional world. But if it had any modern-day material, it was subtle and in the background, not the focus of the story.
You as a player got to choose the direction of the game. You weren't forced on rails through a certain experience.
Jonathan: Knights of the Old Republic, or KOTOR, as it's called in the video game world, let you become a bad guy. It gave you a dark arc. You could still beat the story's villain, but you could end as a villain yourself. That was revolutionary in its time, because you could develop your character along that line, and it was so cool.
There were conversations at my church about whether it was a sin to choose the villain dialogue options. I'm talking about teenagers. Adults did not care. But teenagers were asking, "Am I going to hell because I picked this dialogue option?"
Thomas: "Am I damned because of the dark side?"
Jonathan: It was actually an interesting philosophical discussion, because yes, you are choosing, but there are no real-world effects.
Thomas: It also had a clear moral system. I think that's another symptom of go woke, go broke. Woke morality is a twisted, broken morality where the only thing that really matters is consent, and otherwise everything is okay. For example, they’d say it's okay to kill someone as long as they consent to being euthanized. Nothing is forbidden as long as there is consent, whatever consent means.
Jonathan: And consent is really fluid legally.
Thomas: It can be withdrawn later. As a moral system, it's unsatisfying compared to an absolute moral system.
The original Knights of the Old Republic followed the original Star Wars moral system, where the dark side was bad and the light side was good. In the more modern, woke-ified Star Wars, the light side isn't really good and the dark side isn't really bad. It's all been muddied together. And you'll notice the new Star Wars is not resonating with audiences, because audiences still believe in good and bad.
We're seeing this industry-wide. Just today, Eminem released a song about repenting of his sins and turning to God. It's with Kid Rock, on an album called Halfway to God. There's a song on the album about repentance and a song about enjoying life because Jesus turned water into wine. It's the first religious song I've heard that drops the F-bomb throughout. It's a very strange song.
But it's an indication of the zeitgeist shifting. Even Eminem is realizing, "If I want to resonate with my audience, I need to start going to church, because my audience is going to church now."
BioWare was a company run by people who explicitly discriminated against anyone who went to church. Nobody at BioWare was going to church. They got so out of touch with their audience that the audience stopped buying their games, they ran out of money, and EA closed them down. They had excellent IP and really talented developers, but the writers ruined everything.
Jonathan: Let that be a lesson to you all. Go to church, but also go to the book fairs. Meet your readers. Find out what they're enjoying. Look at the other books they're holding.
Thomas: Listen to your reader rather than judging your reader. You'll notice the BioWare staff were very judgy toward gamers.
Jonathan: They said, "The game wasn't bad, the gamers were." They put all the onus for making a terrible game on the gamers. "You just didn't appreciate what we made." You're not an art company, you idiot. You make video games.
Thomas: And even if you are an art company, you still have to thrill your customer. Authors can fall into this too, saying, "My book is a masterpiece. It's just all these ignorant rubes who couldn't appreciate it."
That attitude does not sell books.
YouTube May Charge Premium for the Audiobooks Pirates Already Give Free

According to The Wall Street Journal on October 14, 2026, YouTube has held discussions with publishers in recent months about developing an audiobook feature that could be ready next year. The company has weighed bundling it into ad-free YouTube Premium, which starts at $15.99 a month. It is not clear how the content would be displayed or whether individual books would be sold. YouTube declined to comment.
What Happened
The category is already dominated by Spotify and Audible. Publishing executives told the Journal they see an opportunity in YouTube’s audience and a revenue-generating alternative to the glut of unauthorized audio versions of their books already on the platform. The Association of American Publishers reported that digital audiobook revenue topped $1 billion in 2025, up from $265 million in 2016. Broader Audio Publishers Association figures put total U.S. audiobook sales at $2.43 billion in 2025, a 9% increase.
The free shelf is already large. In the Audio Publishers Association’s 2026 consumer survey, 45% of people who listened to an audiobook in the past year said they had listened to a free copy on YouTube, up from 38% the year before. The APA notes that many of those copies are pirated. A New York Times investigation in May 2026 documented AI-narrated unauthorized versions of bestsellers, including John Grisham’s The Widow with more than 80,000 views. The Association of American Publishers partnered with Vermillio that month to identify and speed removal of infringing copies, including on YouTube.
Spotify offers the closest working model. Premium subscribers get 15 hours of listening per month included in the plan. Direct à la carte purchases through Spotify for Authors pay 50% of the suggested retail price the author sets. Listening that falls inside the included hours is paid from a per-market consumption pool on a pro-rata share of total listening. That pool rate is not published and fluctuates. Audible’s newer model pays 50% exclusive or 30% non-exclusive, calculated against Member Value rather than a fixed list price. YouTube has not said which of these approaches it would copy, or whether indies would even be invited.
Why Authors Care
A licensed YouTube catalog could convert some free listens into paid ones, but only if the rate is clear and the catalog actually includes independent titles.
Existing distributors (ACX for Audible, Voices by INaudio, Author’s Republic, and Spotify for Authors) have no confirmed path into a YouTube feature yet. Major publishers are in the room first.
A Premium bundle that works like Spotify’s 15-hour model tends to compress per-title revenue into an opaque pool. Authors who already upload authorized samples or full public-domain titles to YouTube for discovery or ads should document their channels and keep records in case a crackdown arrives with the official product.
This week, check your current audiobook distribution agreements for exclusivity language and note whether your aggregator lists YouTube or Google as a partner. If YouTube launches, the first contracts will set the terms for everyone else.
Sources:
- YouTube Is Looking to Expand Into Audiobooks - Wall Street Journal, Alexandra Bruell and Anne Steele, Oct. 14, 2026
- AAP StatShot Annual Report: Publishing Revenues Totaled $33.4 Billion for Calendar Year 2025
- APA Sales & Consumer Data - Audio Publishers Association
- YouTube Is Crawling with Pirated Audiobooks Made Using A.I. - New York Times, Alexandra Alter, May 21, 2026
- Spotify for Authors Distribution Terms
- YouTube Premium and YouTube Music price increase - TechCrunch, April 10, 2026
Commentary
Thomas: The audiobook wars continue. This is not a two-faction war. A new challenger has arrived.
YouTube has been sending raiding parties into the audiobook wars for years. Authors have been uploading their books to YouTube. Pirates have been uploading books to YouTube. But now YouTube is potentially exploring doing it for real.
Jonathan: Audiobook revenue going from $265 million to over $1 billion is a $750 million climb in nine years. And the rise in people listening to free copies on YouTube shows customers increasingly using YouTube to check out an audiobook before they commit a credit to it or buy it.
Thomas: If you want a sense of how dominant YouTube is, there's a chart on X, shared by TV Grim Reaper, showing the percentage of U.S. TV viewing by platform. Back in 2024, YouTube was almost tied with Netflix, though still higher. Now it's almost double Netflix.
[INSERT GRAPHIC]
The typical American household spends twice as much time on YouTube as on Netflix. YouTube is running out of competitors. There's no one in its tier anymore in terms of watch time, and it can't kill Netflix any faster than it already is.
Netflix as a stock is taking a beating, which is another story we didn't even have on our list. They're going woke and realizing they're making content Americans just don't care for. No one is excited for Narnia. The gay and retarded Aslan, for some reason, is not resonating with audiences, and they're shocked. Shocked.
Jonathan: I'm shocked.
Thomas: Have you met a single Narnia fan who is excited for this new movie? No real fans are excited for a gay and retarded Aslan.
Jonathan: And the backpedaling happening behind the scenes is crazy.
Thomas: They're doing a lot of layoffs and trying to swap out the team. The problem is that swapping out the team for basically the same kinds of people with different names won't fix it.
YouTube's advantage is that it lets anyone make content, so conservative content like our show can find a spot. There's a lot of conservative content on YouTube, in addition to middle-of-the-road content and woke content, across the whole spectrum. That whole spectrum is why YouTube has twice the viewership.
Netflix is run by Democrats creating content for Democrats and for people with that worldview. They don't seem to understand why people who voted for Trump aren't excited to watch the new Netflix show those voters would call gay and retarded. I'll keep saying that because I know it triggers people, and I'm fine with that. Netflix should just provide different content for different audiences.
YouTube will keep gaining market share and may start taking it from Audible. Indie authors have had their books on YouTube for a long time. Traditionally published authors have not. If James Patterson suddenly becomes the MrBeast of audiobooks on YouTube, it might actually find an audience.
Jonathan: Because MrBeast built his audience on YouTube.
Thomas: This could be really big. It will create a wake indie authors can surf in, because if YouTube enters the audiobook market, it will enter big.
Jonathan: And indies who are already there and familiar with the process have no catching up to do. If you already have a lot of your work on YouTube, you'll get a surge.
YouTube will focus on training customer behavior, telling them, "This is how you listen to audiobooks on YouTube." That relieves authors of having to train people to go to YouTube for audio, which they just won't do for you. Once YouTube trains its audience to listen to audiobooks, all you have to do is get them to yours.
People binge YouTube. If they have their headphones on while they're building houses or commuting to work, they'll run across your book. You just have to get your targeting right, get your SEO right, make sure you're findable, and have good books they'll want to listen to, and you'll grow.
Thomas: It's also a way to have a free version of your book that you make money from. YouTube has two revenue sources.
One is YouTube Premium. Of all my subscriptions, it's the one I like most. Someone in my family has it, but if they didn't, I'd pay for it right away. I consume a lot of YouTube, and not having ads is really nice.
I like that a significant chunk of your Premium subscription goes to the shows you watch. I think about half goes directly to creators. YouTube and X are the two most generous social networks for sharing advertising revenue with creators.
The other source is all the ads you watch. For those of you watching live, we don't turn ads on for the live show. The only money we make from the live version is whatever people donate while watching. We've experimented with ads, but when an ad runs during the live show, you miss whatever we're saying.
Jonathan: It doesn't make sense, and we didn't make any money either.
Thomas: The ads don't pay great, especially with the kind of audience we get. We'd rather give it away free.
Jonathan: Ridge Light Ranch asked, "Are people just earning ad revenue if their book is on YouTube, or is YouTube a place for a free first-in-series?"
I wouldn't look at it as a permafree option for a first-in-series, because, again, you're looking at user behavior. They listened to your book because it was on YouTube, not because they were looking for your book and it was only on YouTube.
Thomas: I know an author making $1,000.00 to $2,000.00 a month just from audiobooks streaming on YouTube. People binge it.
You get paid per minute on YouTube, so a short clip, like a clip of this segment, won't pay very well. A one- or two-hour show pays a little better. Someone who “watches” a 20-hour audiobook pays even more.
You could do a permafree, but you'd be converting somebody to a different platform, which is tricky. They may have to create an account for the first time. Just because somebody is on YouTube doesn't mean they have an Audible account or even an Amazon account. These are potentially very different audiences.
There is money to be made. The easiest way is to have an AI version of your book on YouTube and a human version on Audible. You can have both simultaneously. Some people want the human version, and some want the AI version for free.
Jonathan: One commenter asked whether we've done any videos on how to set up profitable audiobooks on YouTube.
Thomas: It really comes down to writing the kinds of books people want to read. I consult confidentially with a lot of authors, particularly the successful ones. The ones doing this aren't just making two grand on YouTube. They're making a lot of money on Amazon too, because they write to market. They write books readers already want, and they place those books on YouTube as well.
If your book isn't selling well on Amazon, putting it on YouTube won't fix it. But if your book is selling well on Amazon and you put it on YouTube, that will bring in a lot more money. An AI version of Dungeon Crawler Carl on YouTube would bring in tons of money.
Jonathan: I can do an Author Arsenal video on how to upload an audiobook to YouTube. It wouldn't be that hard. I'll try to throw that together this week.
Technology
Libby Readers Can Now Hide AI Books From Library Searches

According to Libby’s help pages updated October 14, 2026, and a June 30, 2026 interview with OverDrive CEO Marc DeBevoise, the library app has rolled out AI content controls that let patrons filter titles with AI contributions. The path is Menu, then Content Controls under Settings, then AI Contributions. A viral social post by @becksreadstoomuch titled “ANTI-AI LIBBY FEATURE JUST DROPPED” walked readers through the setup, and library accounts have been posting the same steps since mid-October.
What Happened
The feature sits inside the existing Content Controls menu that previously handled audience age filters. Under AI Contributions, patrons see four categories: Text, Audio, Graphics, and Translation. For each, the options are roughly All, Human involved, or Human only. Libby’s official setup page spells out the audio case: choose “Human narrators only.” By default that preference only hides titles in deep search (titles your current library does not own). Turn off the “Deep search only” toggle under “Where to apply filters” if you want the block to hit every library list and search. The filter is off by default. Settings do not sync across devices, so each phone or tablet needs its own setup.
OverDrive is not running its own AI detector. Its Responsible AI policy states: “OverDrive does not exclude titles created with AI tools from the catalog. We ask that publishers self-identify content that was created with AI assistance.” The June announcement from DeBevoise said the same thing: the filter depends on publishers and distributors labeling works through standardized metadata. Titles now display those publisher-supplied AI disclosures. Readers can report content they believe is unlabeled AI, and librarians have new staff-side tools when they are selecting titles.
Why Authors Care
A patron who turns the filter fully on will not see a properly labeled AI-narrated audiobook, AI-generated cover, machine-translated title, or AI-text book in search results or library lists. That is a discoverability gate in the largest U.S. library lending app.
Indies who use AI narration or AI covers need to check what metadata their distributor actually sends. PublishDrive’s partner guide says OverDrive accepts AI-narrated audiobooks if the AI voice is listed as the narrator and the description opens with a line such as “This audiobook is narrated by AI.” It accepts AI-generated covers if declared as an illustrator, but lists pure AI-generated ebooks as not accepted. Draft2Digital allows AI-assisted work that has undergone extensive human editing and feeds titles into OverDrive.
If your distributor does not pass a clear AI flag, the title may still appear even for patrons who set the strictest filter. If it does pass the flag, the title drops out of results for those patrons. Check your current titles this week and confirm the metadata fields your aggregator uses for OverDrive.
The filter is patron-side and optional. Libraries still decide what to buy. A title that is unlabeled stays visible unless a patron reports it.
Sources:
- Getting started with content controls
- Setting up content controls in Libby
- Libby will filter out AI content, kind of (The Verge / Lowpass)
- Responsible AI | OverDrive
- AI-generated/assisted/narrated – what to declare (PublishDrive)
Commentary
Thomas: The anti-AI jihad continues.
Jonathan: Basically, if your book is deemed AI in any way, through either the text or the cover, it can be tagged as AI, and people can choose never to see it in the Libby app.
If you rely on Libby for your library distribution and you use AI at all, or it looks like you might, you might get hammered.
CEOs Say AI Adds Jobs. The Entry-Level Numbers Disagree.
According to the KPMG 2026 U.S. CEO Outlook, released October 15, 89% of the 400 U.S. CEOs surveyed expect their companies to increase total headcount over the next three years. The survey covered firms with more than $500 million in annual revenue and ran from August 11 to September 21 as part of a global poll of roughly 1,400 leaders. Axios framed the result as CEOs treating AI as a job creator rather than a destroyer.
The same release shows the other half of the ledger. In response to AI, over the next two to five years, 53% plan workforce reductions in some areas. That is up from 35% in the 2025 U.S. CEO Outlook. 86% are redesigning roles, career paths, and team structures for human-AI collaboration, up from 65%. 79% are hiring people with AI and technology skills, up from 64%. 75% say AI has made them rethink the skills required for entry-level roles. Only 9% describe current AI agents as transformational for their organizations; 39% report significant improvements in efficiency or growth and 43% report moderate, targeted gains.
KPMG U.S. Chair and CEO Tim Walsh told Fortune that the percentages should be read loosely and directionally. He said agents are not yet eliminating full roles at his firm and that capacity opened by AI is expected to support expansion rather than pure cuts. Skills gaps ranked as the top workforce challenge at 60%.
What Happened
The headline optimism sits next to concrete shifts already visible outside the C-suite. An Anthropic internal report released around October 16 found the share of junior technical hires at the company fell from 27% in early 2025 to 9%. Junior share of technical staff dropped from 21% to 14%. Agent work hours reached six times employee hours by August while the company continued to grow overall. Stanford Digital Economy Lab research using ADP payroll data through mid-2026 shows employment for workers aged 22 to 25 in the most AI-exposed occupations roughly 19% below where it would have been relative to peers in less-exposed fields, driven mainly by reduced hiring rather than layoffs of existing staff.
Tech layoff trackers for 2026 already exceed 2025 totals, with AI cited in a rising share of announcements. The pattern is consistent with the KPMG data: overall headcount can rise while the composition of work changes and the first rungs of the ladder shrink. Most of the expected hiring increases are modest. 60% of the CEOs foresee growth of less than 5%, 26% expect 6-10%, and only 2% expect 11-25%.
Why Authors Care
Routine writing, editing, cover design, and basic narration support are the creative-industry equivalents of the junior technical tasks Anthropic is handing to agents. Publishers and production houses that adopt AI will need fewer pure production hands and more people who can direct, edit for voice, and judge quality.
The market for original craft, distinctive voice, and reader relationships does not shrink when tools get faster. It concentrates. Authors who treat AI as production leverage while owning the judgment and the relationship keep the pricing power. Those who compete only on volume or on tasks AI handles cleanly lose margin.
Entry-level freelance gigs in content mills, basic developmental notes, and commodity design are the first to compress. Specialized work (genre expertise, series continuity, high-trust editing, performance narration) holds up or gains relative value.
Action this week: audit one production step you outsource or do yourself that is mostly mechanical. Test whether a current tool can take the first pass while you keep the final decision. Track whether the time saved goes into more books, better marketing, or higher rates rather than simply matching a lower bid.
Sources:
- KPMG Survey: CEOs Confident in Their AI Progress and U.S. Economic Growth, but Cyber and Geopolitical Risks Intensify
- 2026 KPMG US CEO Outlook
- KPMG U.S. CEO Outlook: It's 'Go Time' to Reduce Supply Chain Costs and Drive AI Integration and Upskilling (2025)
- AI will create jobs, U.S. CEOs say in survey
- KPMG's CEO on why AI safety isn't another Y2K
- Anthropic report on junior technical hiring and agent hours
Commentary
Jonathan: Companies may say AI will increase hiring in their productivity areas, but we're already seeing HR departments devastated by the rise of AI. Anything that relies on spreadsheets is already going away.
AI is making CEOs rethink the skills required for entry-level roles.
Thomas: People who are good with AI, who know how to kick the training wheels off ChatGPT and move into Codex, or know how to use Claude Code or GrokBot, are all employed and busier than they can imagine. The people who don't know how to use AI are often really struggling.
We went through the same thing 30 years ago with the personal computer. When computers started appearing on people's desks, it was just DOS, and typing commands into DOS was hard to learn. A certain kind of person resisted using the computer, and those were the people who got laid off. The people getting raises, jobs, and promotions were the ones who were good with the computer.
There's nothing new under the sun. The better you are with AI, the more you can get done, the more valuable you are to your company, and the more they'll want to hire you and give you raises. The worse you are with AI, the less productive you are, and the less they'll want to keep you around.
Every productivity tool follows the same pattern, going all the way back to the cotton gin and the steam engine. We have 300 years of the exact same cycle. Early on, there's a moral panic. When electricity came on the scene, people said, “It will kill us all!” There was all this fear, uncertainty, and doom about electricity.
Thomas Edison, trying to get electricity regulated, electrocuted elephants in New York City. You can look it up. It really happened. He invented the electric chair just to make electricity look scary.
That's early in the cycle. Then the companies that failed to adapt to electricity struggled.
Factories used to be tall, multi-floor buildings, because a coal-fired steam boiler in the middle powered all the turbines and machinery. At first, factories just replaced the coal boiler with an electric one and cut their costs by 10%.
But the factories that really took advantage of electricity moved from multi-floor buildings to flat ones. Now you could put electricity anywhere in the building, and without going up and down floors, the whole factory ran more efficiently. But it required an entirely new building.
With every new technology, you have to re-evaluate your assumptions, because some are good and some are bad. Most people in the buggy business did well in the horseless carriage business. If you made tires, you did great. If you made carriages, you did great. If you made buggy whips, you did not.
We're seeing the same thing with AI. Some jobs will change or go away altogether, and others will be in much higher demand.
The spreadsheet didn't put accountants out of work. It created more accounting jobs. If your job was nothing but adding up numbers on an abacus, it did go away. But we went from 50,000 accountants to a million, because there was more demand once they could do more.
This is always the case. There are more jobs now than ever before in history, and more labor-saving machines than ever before. Ultimately, the jobs created will exceed the jobs destroyed. But the winners will be the people who are good at learning and adapting, and the people who aren't will struggle.
That's been true for thousands of years, because the world is always changing, and adapting to a changing world is a key survival skill for our species.
Atticus Adds Easy Two-Page Image Spreads
According to Atticus, two-page image spreads are now live. Click the chapter image icon, choose "two page spread," then upload or pick from the Image Gallery, and Atticus handles the layout for maps and landscape images. The feature is not available on the Limited version.
Full-license authors can stop splitting map files by hand for print. Limited users keep the old process or need to upgrade, and Atticus is hosting a free webinar, Audiobooks for Every Author, with Derek Doepker on Wednesday, October 28, at 11 a.m. Pacific / 2 p.m. Eastern, covering DIY versus outsourcing, where to hire narrators, and home recording gear.
The Atticus affiliate program still pays 20% per author signup.
Sources:
Commentary
Thomas: If you want to add images to your ebook, you can do that with Atticus.
I have complicated feelings toward Atticus. There's not one program I recommend for everyone. Atticus is the right solution for certain people, and I try to shoot straight on it. But for people running Windows, Atticus is a no-brainer, especially since Vellum, which I really like, is only available to Mac users.
Anthropic Books $80.7 Billion Loss as Claude Growth Slows

According to Reuters reporting on a draft IPO prospectus reviewed by the Wall Street Journal, Anthropic recorded a net loss of $80.7 billion in the first half of 2026. That is the number Polymarket posted on X. Most of it is not cash that left the building. About $72 billion is a non-cash accounting charge. Convertible notes Anthropic issued to raise money rise in estimated value as the company's own valuation rises, and under U.S. accounting rules that increase is booked as a loss.
What Happened
Revenue in the same six months was $16.3 billion, up from $1.1 billion in the first half of 2025. Compute and infrastructure spending hit nearly $9 billion in the half, more than the company spent on computing for all of 2025. Anthropic has committed at least $518 billion over roughly a decade to cloud and computing capacity, with some reporting that total obligations could reach $750 billion including deals under discussion. Roughly 20 percent of the first-half revenue came from just two customers.
At an investor meeting on October 14, Anthropic told prospective IPO buyers that third-quarter revenue reached $16.8 billion, a 45 percent rise from the prior quarter. That is still rapid growth, but slower than the more-than-doubling rate from the first quarter to the second. Reuters reported that some investors who attended the meeting are now discussing a valuation around $1.5 trillion instead of the roughly $2 trillion figure that had been floated earlier. The IPO is expected in November. The Polymarket item titled "Anthropic, repriced" and related social commentary match this reporting on slower sequential growth and a lower base-case valuation. Claims about corporate users cutting back on heavy token use and shifting to open-source models come from market commentary, not from the draft filing itself.
Why Authors Care
Claude's consumer pricing has not changed. The free tier still exists. Pro remains $20 a month or $17 billed annually. Max is $100 or $200 a month. Recent API rate cards show some model prices held or reduced rather than raised.
The real pressure is the cash compute bill and the locked-in future commitments. A company spending nearly $9 billion on infrastructure in six months while booking massive paper losses has reason to tighten free usage or raise prices later, even if it has not done so yet.
Authors who run heavy workflows on Claude should test open-source alternatives now. Local models and cheaper API options from other labs continue to improve. Building a second workflow that does not depend on one vendor's free tier or rate limits is the practical Monday step.
The $80.7 billion headline is real but mostly accounting. The number that matters for product decisions is the compute spend and the slowing sequential growth that has already led some investors to cut the IPO price they are willing to pay.
Sources:
- Anthropic logged net loss of over $80 billion in first half of 2026, WSJ reports
- Investors temper Anthropic IPO valuation expectations over revenue growth slowdown
- Polymarket X post on the $80 billion loss
- Claude plans and pricing
Commentary
Jonathan: That $80.7 billion was not their expenses. It's the net: the difference between what they paid and what they were paid.
Thomas: Anthropic is an interesting company to follow, because it's one of only two companies all in on AI with no other revenue sources.
Look at the big companies pursuing AI. Meta was able to lose $80 billion on the Metaverse, and it barely affected their stock, because they make so much money in so many other places. They can take flyers on things, and if something doesn't work out, it's fine. "We print a billion dollars a day on Instagram. That's three months of profit from Instagram. We'll be fine."
Same with Google. Google is a big AI company, and it makes money in a lot of other places.
Same with Grok, which is now owned by SpaceX AI. This week SpaceX had what I think was its first quarterly earnings call since its IPO, or at least the first where it announced a profit. So SpaceX AI, or SI, as they're now rebranding, is profitable. The first superintelligence unicorn is already profitable. Anthropic is not yet profitable.
Part of the problem is that serving an AI model and selling tokens is very profitable, with margins around 80% or 90%. But building the model in the first place, creating Opus 5.5, let's say, might cost $50 billion of inference, and they have to pay that off by selling tokens at a very high rate while they're building Opus 5.6.
At any point, these companies could slow down development of the next big thing and become a lot more profitable. But by doing that, they'd be giving up the race to their competitors.
And now suddenly Anthropic is begging everyone to slow down for the sake of humanity. This isn't for the sake of humanity at all. They’re running out of capital, so they need everyone else to slow down before they run out of cash.
The problem is that the companies they're racing against, with the exception of OpenAI, which is in the same situation as Anthropic, all have other sources of revenue that buy them time.
SpaceX makes tons of money doing shady things for government agencies that want things in space nobody talks about the details of, but somehow billions of dollars show up in their accounts. And oh yeah, they're launching the Starlink satellites too. Those are the only satellites they're putting in space. Do not pay attention to the other half of the payload.
Jonathan: “No space lasers here,” says the guy who made flying cars. I guarantee you there are space lasers in orbit.
Thomas: They're totally not putting space lasers and nuclear bombs in orbit. They deny all knowledge of such things.
But they're already profitable, because they have this profitable space company and they also sell internet. They're about to become the fourth cell phone provider. Very soon Starlink will probably offer direct cell phone service. They just bought the 800 megahertz frequency that I think used to be Nextel's. You'll be able to get cell service straight from space, and we'll have a little more competition in the cell phone market, which I think will be really good.
They're doing fine. Meta is doing fine. They actually have a good model now. Muse is good. Google is fine financially. Their model isn't great, but they have money coming in from other places.
OpenAI and Anthropic are the ones that really want regulation so they don't run out of money. And who is asking for regulation the most? OpenAI and Anthropic. Who is saying, "We don't need any regulation"? Meta. Why? Because they have unlimited money.
Jonathan: If you're running heavy workflows on Claude, start checking out other options in case Claude starts limiting your usage.
Thomas: Right now, for your $100.00-a-month subscription, Claude gives you something like $1,000.00 worth of compute.
Jonathan: But that won't stay that way with these numbers.
Thomas: It won't, but switching is super easy. I literally switch between Claude and Codex on the same project, back and forth, because they can share the same Git repository on your machine.
If you've kicked the training wheels off, moved from ChatGPT to Codex and from Claude to Claude Code, and you're using Git on your computer, you can switch back and forth instantly without losing any progress and hardly any context.
That's actually one of their big risks. They can't lock in their customers. At any point, all their customers could leave. A huge chunk of the market just uses whoever has the best model and ping-pongs back and forth, because their tools can point to whatever API is serving up the smartest tokens, so to speak.
Someone in the chat made a good point that Anthropic doesn't own its own data centers, for the most part. It rents other people's. One reason SpaceX is so profitable is that it overbuilt its data centers last year and started renting data center space to Anthropic. So some of Anthropic's losses go into SpaceX's bottom line.
Everyone laughed at Elon Musk's tweet, back in 2025 I think: "Buy GPUs. Profit." He was buying $10 billion worth of GPUs, and people said it was insane. Don't bet against Elon Musk. The guy can see the future.
NSA: Take Unpatched PHP Sites Offline Until Fixed

According to the NSA Cyber account on X, effective immediately organizations using PHP must prepare for an update addressing two critical vulnerabilities that often provide a direct path to the server’s file system or database. If an immediate update is not feasible, take the system offline until it can be fully mitigated. The same day the NSA issued a press release naming the pair PHP2Shell and repeating the offline recommendation for any system that cannot be patched at once.
PHP runs 69.3% of websites whose server-side language is known, according to W3Techs as of October 22. That share covers most WordPress author sites, newsletter landing pages, and stores. The bugs sit in the PHP runtime itself, so a fully updated WordPress core, theme, and plugin set still leaves the site exposed if the underlying PHP is unpatched.
What Happened
The PHP project shipped the fixes on October 21 in versions 8.5.12, 8.4.27, 8.3.36, and 8.2.35. The official advisory tracks the issues as CVE-2026-109618 and CVE-2026-109619, each scored 9.2 critical. One is remote code execution via out-of-memory corruption; the other is remote code execution via stale global state that can survive across requests. Both can be triggered by an unauthenticated attacker. Every PHP installation serving web requests is at risk, including PHP-FPM, Apache with mod_php, php-cgi, FrankenPHP, and LiteSpeed, with the highest risk on mod_php where the attack can fire before any application code runs. The advisory states there are no complete mitigations; updating is the only safe choice.
Versions before PHP 8.2 receive no official security updates. Minimal source backports exist for 7.4, 8.0, and 8.1, but the project directs users of those branches to upgrade or check their distribution vendor. PHP 8.2 itself reaches end of life on December 31, 2026. Researchers at Calif (assisted by OpenAI’s Patch the Planet effort) and at Wiz found the flaws; the underlying issues had been present for years. Leaked details forced the release a day earlier than planned, and security firms reported exploitation attempts before the patches landed.
WordPress confirmed the same day that the issues are not specific to WordPress and that no WordPress update is required or planned. The WordPress news post by John Blackbourn tells site owners to check their host’s update plan and update PHP immediately, and it quotes the UK National Cyber Security Centre: where immediate remediation is not feasible, consider taking the system offline until it can be fully mitigated. You can see your PHP version inside WordPress under Tools > Site Health > Info > Server.
Major hosts responded on different timelines. SiteGround applied server-level mitigations by disabling the affected functionalities across supported PHP versions before the public release. WP Engine finished deploying patches (including some older branches) by early Thursday UTC. Kinsta and Bluehost were still rolling patches out while using firewall rules and temporary limits. Automattic stated that sites on WordPress.com, WordPress VIP, WP Cloud, Pressable, and Tumblr were already protected. Many shared and managed hosts continued work into Thursday.
Why Authors Care
Updating WordPress, themes, or plugins does nothing for these bugs. The fix is a PHP runtime update controlled by your host or by whoever manages the server.
Check your version today in Tools > Site Health > Info > Server, or in your host’s control panel. You want 8.5.12, 8.4.27, 8.3.36, 8.2.35, or later (or a confirmed vendor backport).
Ask your host in writing whether your specific site is on a patched build or under an active mitigation, and when the permanent update will finish. Do not accept a generic “we are protected” without a version or date.
If you are on PHP 8.1 or older and you control the server, the PHP project supplies no official binary fix. Plan an upgrade to a supported branch. Agencies recommend taking the site offline until that is done.
A successful exploit can reach the same files, credentials, and database the site uses. That includes customer data, mailing lists, and store orders. Rotate keys and review logs if your site was exposed before the patch.
Sources:
- NSA Cyber X post
- NSA press release
- PHP security advisory
- WordPress news: Critical PHP Security Release
- WordPress X post
- W3Techs PHP usage
- SiteGround statement
- Automattic protection notice
Commentary
Thomas: A big ramification of superintelligence getting faster is that two things are happening at the same time.
But first, some context.
Nation-states have been hiding backdoors in software forever. It's a normal thing. A nation-state will have operatives, or recruit operatives at tech companies, to put in backdoors and add zero-days they can use when needed.
The most famous use of this was back around 2008, when something like 19 different zero-days were used for one very specific virus. It happened to target only centrifuges, only at nuclear refineries that just happened to be in Iran. All the Iranian nuclear refining broke because of this super virus that exploited all those zero-days on the same day.
If you want to read about it, it was Stuxnet. Officially, we had nothing to do with it. We deny all knowledge. Israel also denies all knowledge. It's a mystery who made this virus.
AI is turning out to be really good at finding these backdoors. Companies using AI are suddenly saying, "We have this massive vulnerability that's been in PHP for 10 years. We've got to patch it."
This last week, about a dozen major vulnerabilities, which in many cases had been there for over a decade, were revealed to the public. The PHP one is the one everyone's freaked out about.
It was discovered about two months ago. OpenAI has known about it for two months. There's a CIA director on the board of OpenAI, so the CIA has known for two months about a way to unlock anybody's website and get complete access, and they've been patching it.
The big drama wasn't that there was a vulnerability; it was that the vulnerability was announced before the patch was ready. For a 24-hour window, all the hackers in the world knew about it before the fix was in.
That's not how it's supposed to work. As long as you keep everything up to date, hackers aren't supposed to be able to get in, because by the time they know about a vulnerability, the patch is out. This time it got flipped. It was a mix-up in communication. It wasn't the AI's fault. It was a human's fault. Somebody somewhere talked.
For a few hours, websites were somewhat vulnerable, although if you were behind Cloudflare, it put in really good mitigations immediately. I'm doing this whole story off the top of my head, but to hack your website, an attacker had to overwhelm it and trigger a cache overflow in a very specific way, which Cloudflare could prevent before it hit the PHP buffer. I don't actually know of any websites that were hacked using this vulnerability.
If you listen to the mainstream media, it's "Wikipedia went down. The NSA said to take your websites down. The Prime Minister of Japan said to take your websites down. All these websites went down." What they don't clarify is that the websites were taken down proactively by the people running them. All those European websites went down because people were nervous and turned them off, not because they were hacked.
I haven't heard a single report of a website actually getting hacked with this vulnerability, because of Cloudflare's mitigations and the proactive steps people took. I'm sure somebody somewhere got hacked, and I know there were agent swarms trying to exploit it. But it was a tricky hack. You had to overwhelm the server and inject code during the overrun to get an upload into the cache overflow. The timing was tricky.
There was a different hack this week that wasn't about unlocking the door at all. An Indian IT firm had been given the keys to the internet, and some of its employees set their passwords to the number four. Another one's password was 37. These weren't personal computer passwords. They were passwords securing SSL certificates for the whole internet.
The other big story this week isn't a hack. It's unauthorized agents accessing bank accounts. If you dig into the details of this report, they're not authorized by the bank. That doesn't mean they're not authorized by the customer.
If your GrokBot moves your funds from a savings account that gives you no interest to a high-interest savings account at a different bank, from the bank's perspective, that's unauthorized agent access. Red alert. From the customer's perspective, you went from 0% interest to 4% interest. Very good for the customer, very bad for the corporation.
In many cases, AI is giving regular people more power. They can hack into their TV, take out the spyware, and make it run faster. Corporations really hate this. They want to give you slop, and they don't want to give you choices. If you know how to use AI, you suddenly have choices.
I hacked into my TV and made it 20% less annoying. It's a Vizio, so you can't hack it very well, but even so I was able to take out some of the annoyances with Claude. It was great.
Zeitgeist Report (World Tour)
Today We are Going International
Worse Than Book Burning: Britain Arrested 62,199 People Over Posts and Messages

British police arrested at least 62,199 people for communications offences in five years. That's 34 arrests a day. About one in five ended in a conviction.
The Home Office doesn't track this number. A civil-liberties group had to file Freedom of Information requests to find out how many people the state arrested for what they typed.
The Count
Big Brother Watch sent FOI requests to all 45 territorial UK police forces. Thirty-eight sent complete data. From January 2021 through December 2025: 62,199 arrests, 18,520 charges, and 12,292 convictions.
The group published the tally September 7 in Britain's Free Speech Problem. If you fill in the seven missing forces at the national average, the group estimates about 73,000. Polymarket pushed the number across X on September 11, and it has been circulating since.
The Laws
Three statutes drive the count: Section 1 of the Malicious Communications Act 1988, Section 127 of the Communications Act 2003, and Section 179 of the Online Safety Act 2023. They cover messages that are "grossly offensive," "indecent," "obscene," or "menacing," plus knowingly false messages.
The count includes real threats and domestic-abuse texts. It also includes jokes, opinions, and social posts. The arrest data doesn't separate them, and neither do the statutes.
Postcode Lottery
Cumbria arrested 25.7 people per 10,000 residents, two and a half times the national average. Neighboring Northumbria arrested 1.9. Staffordshire arrested 0.58.
The law is the same in all three counties. The arrest rate is 44 times higher in Cumbria than in Staffordshire.
The Receipts
In one case, West Midlands Police went to a vulnerable girl's home over a TikTok image of a teacher with negative comments under it. Officers demanded her devices and told her to choose between a voluntary interview and arrest. She hadn't made the post, shared it, or commented on it. Police took no further action. The Free Speech Union has the file.
Big Brother Watch director Silkie Carlo says the people arrested include those posting online jokes, those at non-violent protests, and protesters holding blank sheets of paper.
Nadine Dorries wrote the report's foreword. She was culture secretary while the Online Safety Act moved through Parliament. Now she calls it a dog's dinner and wants an independent national review. The group launched the report in Commons Committee Room 8 on September 9.
Lord Young of Acton made the key point on GB News: when four in five cases end without a conviction, the arrest itself becomes the punishment.
What It Means for Your Shelf
Your manuscript probably isn't the risk. Your marketing is. These laws target messages, and authors sell books through posts, DMs, Facebook groups, and TikTok. Writers in thrillers, military SF, and horror build their brands on edge. On a bad day in Cumbria, that edge reads as "menacing."
The UK is a market, not a monolith. Whether a joke is legal can depend on which county reads it.
US authors: You still have the First Amendment at home. Your words still travel into a market that arrests people over messages. Read the statutes before you plan a UK launch the way you'd plan a US one.
UK and Canadian authors: Big Brother Watch and the Free Speech Union hold the casework and the FOI tables. Bookmark both.
A publisher can reprint a burned book. An author in a holding cell writes nothing.
If you want the numbers before anyone spins them, like and subscribe to Author Update.
Commentary
Thomas: In this episode of The Zeitgeist Report, we're beginning what may be an ongoing segment: a world tour, where we talk about how readers in America feel about other countries. That affects authors in two ways. First, if your book visits that country, what's the readers' vibe about it? Second, if you're from that country, you need to understand how American readers feel about you.
Our first stop is the biggest country in the literary world for Americans after the United States. Americans read more American books than any others, but the second-biggest source of the books we read is the UK. And not just because Harry Potter, The Lord of the Rings, and Narnia come from the UK.
Jonathan: So much of our culture.
Thomas: And our genetic ancestry, our cultural ancestry, and our legal ancestry. America was born out of England, Scotland, and Ireland. Not Wales so much. The Welsh mostly stayed in Wales. I have some Welsh ancestry, but I think that's pretty uncommon.
They're skipping straight past the book burning to arresting the authors.
Jonathan: And 12,292 convictions is a really high conviction rate for that many charges. That's kind of nuts.
Thomas: Just yesterday there was a report of a 16-year-old who posted a photo of a British politician as a rabbit. The police raided his house and took him to jail. I don't think the post even had any commentary. It was just the politician's face on a rabbit.
There's a kind of hypocrisy common in the publishing establishment. Publishing news outlets, other than this one, talk about book burning all the time. By book burning, they mean offensive and obscene books aren't being provided free to children in school libraries. Meanwhile, they completely ignore 62,000 people getting arrested for free speech violations. That's an incredible hypocrisy, and people in America see it.
People in America take freedom of speech very seriously. We will kill you if you try to take it away from us, and we did. When England tried to take it away, we said, "You can have our rights from our cold, dead hands." We crossed the Delaware in the wintertime to kill you, just to make it very clear that we will have our liberties. We don't care what your laws say. We have our own laws.
There was a big court case the UK filed against 4chan on a lot of these points. The UK won its case there, but 4chan has no assets in the UK, so they had to go through the American court system to try to seize 4chan's assets.
The American judge looked at it and said, in effect, "You're not going to collect money from an American organization over UK speech codes." And the 4chan lawyer's response was a meme, something like a gangster rabbit sitting on a pile of cash.
Jonathan: The text of the letter was great too.
Thomas: It captures the vibe. If you're a UK author, you have to be very careful about the vibe in America, which is shifting hard against what's going on in the UK, on free speech grounds but also on other grounds.
It used to be that beans on toast was the slur for British people having bad taste. Now the joke is beans on falafel, the idea being that people in the UK aren't even British anymore. They're Muslim. There's been this Muslim conquest, where now the majority of people in London are Muslim and not even of British ancestry. The number one baby name in the UK is Muhammad.
There's this kind of conquering the UK has submitted to. Americans in the major urban centers still believe diversity is our strength and are all for the Muslim conquest. But outside the urban centers, there's a very strong distaste in America for the Muslim conquest. That colors how we interpret fiction coming out of the UK and authors commenting on things.
For foreign authors writing for the American market, your best bet is not to comment publicly on American politics. I know it's very common around the world to comment on American politics. The world follows American politics much more closely than Americans follow world politics. Most Americans couldn't tell you the political parties in the UK, but most people in the UK could tell you the parties in America.
But Americans really, really don't like it when foreigners comment on our politics, even if you're saying things we agree with.
Jonathan: I was watching a comedian last night who doesn't have kids. He said, "You know what people with kids hate most? When a person without kids tells them how they should parent." You think because you've had a night of uninterrupted sleep, you can tell me how to raise my kids?
It's the same thing. UK people get mad when we comment on their politics. We see it in the chat and the comments all the time. But I'm not trying to sell UK people my books. If you're a UK author trying to sell Americans your books, don't tick them off beforehand. That's just not good business.
Thomas: To understand how Americans see the UK, think of how the UK sees Russia: an evil country that's very oppressive to its people and is fighting a very unjust war. And we look at that and say, well, Russia isn't arresting 60 people a day for speech code violations. They're arresting five or six a day. Still a lot.
Jonathan: If we count shooting them too.
Thomas: So many Americans see the UK as worse than Russia when it comes to oppressing its own people. That's not to justify all the terrible things Russia is doing or its invasion of its neighbors. We're not cool with that, obviously. But it's hard to see how other people see you, and I'm using this as a metaphor for how Americans are starting to see the United Kingdom.
There's also a growing sense in some parts of America that America will have to invade the British Isles and liberate them. There's going to be a civil war. We're anticipating a civil war in Britain, and we're anticipating having to send our sons to fight in that war, to be frank about where we see things going. We don't feel the UK has the virility or the manliness to defend itself, and we won't be able to bear seeing London fall.
Jonathan: Well, it's not another desert. It's not a jungle, it's not a desert. That sounds reasonable.
Thomas: If an American president asked Americans to go fight and defend Britain from the invasion Britain is suffering, many Americans would volunteer.
Jonathan: There's a sense of loyalty. "These are our bros. This is where we came from." We have good memories of World War I and World War II. On the military side, we've been working with the UK military forever. I love my UK Royal Marines. Those guys are awesome, and we got along great. The army is pretty cool too. We have good feelings toward them for the most part, and we want to help them.
In the Vietnam War, the military was hamstrung. You couldn't go past the line of demarcation to chase the bad guys, so they could come up and shoot us for 20 years, and we couldn't go after them. That's what we feel the UK is in right now. They're not even allowed to try to keep their culture.
Thomas: Another big difference is that Americans have a very romanticized view of English history, and a lot of Americans study it in great detail. We know about King Henry VIII and his wives. We know about King Henry V and the Battle of Agincourt. We love English history.
When I was last there, I illegally climbed into Dunnottar Castle in Scotland. It was my last day, the castle was locked, and there was a cliff. I thought, "I think I can climb this." So I climbed the cliff, risking my life. I was very young and stupid.
Then I scaled the wall into the castle ruin and explored it, feeling like a Viking conqueror. I turned a corner and saw two very stunned groundskeepers inside a locked castle, not expecting to find anybody. My first thought was, "They may think I'm a ghost." So I tried to be as friendly, American, and touristy as possible. I think they saved my life. After letting me explore a little, one of them pulled out a giant key and let me out the door.
While I was there, I said, "This is so cool." They said, "It's just a ruin." That captured the British view of their own history. "It's just a ruin. We're evil. We are colonizers. We did all this terrible stuff to all these people."
They really believe colonialism was bad, and they're ashamed of it. They're ashamed of their own glorious history. They don't see it as glorious. They see it as a mark of shame. They're embarrassed of their very identity.
We are not embarrassed of our ancestry in Britain. We don't see it as a mark of shame to be culturally descended from the British. That's a big difference, and I think that's what will motivate the invasion when it comes.
There's still a chance to fix it. The invasion doesn't have to come. But we have not done a good job staying out of English wars. When Britain goes to war, we tend to go to war too, either with them or against them. We rarely let them get into a fight we're not part of.
Jonathan: When your buddy gets in a fight, you say, "Hey, leave my buddy alone." That's how it works. I don't even like that guy over there, but he's a Marine, so I have to go.
France Lost the Romance Americans Used to Buy

According to Glynnis MacNicol in the New York Times, the French fantasy that Paris offered an escape from the world’s problems is coming to an end. She wrote from the city after the October 17 weekend, when student protests over school conditions and the cost of living turned into nationwide marches of 550,000 people by Interior Ministry count, with looting, arson, and roughly 1,800 arrests. Shattered windows and burned bikes near Place de la République replaced the terrace view.
What Happened
The drop in American regard is not new. In March 2003, House Administration Chairman Bob Ney and Rep. Walter Jones ordered the three House cafeterias to rename French fries “freedom fries” and French toast “freedom toast” after France opposed the Iraq War. Gallup measured French favorability at 34 percent that month, down from the high 70s. The name change was quieted by 2006, and favorability recovered to 78 percent by 2014. Pew Research Center surveys in 2024 still found majorities of Americans viewing France favorably, with a partisan gap: three-quarters of Democrats versus 53 percent of Republicans.
The recent protests shifted the conversation. President Trump and Vice President Vance framed the violence as the result of mass migration and failed assimilation rather than school funding or fuel prices. High-engagement posts on X circulated footage under captions such as “Paris is burning” and “France is no longer France.” Meme accounts paired the same clips with anime edits and warnings that the scenes would reach American streets. MacNicol’s essay and a Bret Stephens column in the same paper described the country coming unglued under high social costs, youth unemployment near 20 percent, and political paralysis ahead of next year’s presidential election.
The older image of French panache—wine, fashion, long lunches, the city of lights as a sophisticated retreat—still appears in tourism numbers. One 2025 report noted a 17 percent rise in American visitors to France even as French travel to the United States fell. The cultural prestige that once made Paris an automatic prestige setting in American novels and films no longer commands the same automatic respect in the feeds where many readers spend their time.
Why Authors Care
Readers who once treated a Paris setting as an automatic signal of romance or high culture now encounter the same streets in riot videos before they open the book.
The cautionary-tale framing travels easily into military SF, thrillers, and near-future fiction; authors who want the older romantic register face a higher credibility burden.
Tourism data still shows Americans visiting, but the aspirational soft power that once sold French-adjacent books and lifestyle titles has thinned on the platforms where discovery happens.
Check the comments and search trends under any Paris or French-culture title this month before locking a setting or a marketing line that assumes the old prestige.
Sources:
- The French Fantasy Is Coming to an End — Glynnis MacNicol, New York Times, Oct. 19, 2026
- Decade After “Freedom Fries,” U.S. Opinion of France Strong — Gallup, 2014 (with 2003 baseline)
- US views of European allies and relations — Pew Research Center, May 2024
- France comes unglued (again) — Bret Stephens, New York Times, Oct. 2026
- Freedom fries — Wikipedia (primary contemporary reporting linked therein)
- High-engagement X posts circulating Oct. 17–23, 2026 riot footage under “Paris is burning” and related captions (unverified social sentiment)
Commentary
Jonathan: People have already said France has lost the romance for them. Americans used to love the idea of finding romance in Paris, finding your true love in front of the Eiffel Tower. Honestly, I can't remember the last time I saw that in media.
Thomas: It started decades ago, when France didn't help us in Iraq and there was a big turn. Remember the freedom fries thing, when Congress renamed french fries?
The whole country moved away from calling them french fries. Go to McDonald's, and they're not french fries. They didn't switch all the way to freedom fries. They just call them fries now. We dropped the word French, and that was in the early 2000s. Things have only soured since then.
Jonathan: You can't go to Paris anymore to get away from the world's problems. They didn't protect Paris as that fantasy getaway where it's just about romance and nice food and coffee.
All the videos and pictures that came out of the protests line up with what we all assumed was happening to Paris anyway. There's that famous video of a woman telling looters going into her store, "I'm on your side. I voted to give you rights and to give you a place." They didn't care. They went into her store and looted it.
That's been making the rounds on social media, and it exposes something. In America, we don't feel associated with the French anymore. I see France getting looted, and I think, "Well, the French brought it on themselves."
If I see the UK getting looted, and I think, "Hey, I know those guys. You're not doing that to my boys."
Thomas: My book club recently read Revolutions by Eric Metaxas, and it reminded me how long Americans had good feelings toward France. We were very grateful.
Jonathan: They were the first ones to help us.
Thomas: They never really did any fighting on actual American soil. They mostly did a lot of sailing around and posturing. But they were helpful. They had some naval engagements that influenced the war, and they distracted the British Navy. More than that, there was the moral support of having them aid us in the Revolution.
And you have figures like Lafayette, who wasn't acting as an agent of the French government but was a French noble who came over and served as an aide-de-camp to George Washington.
For a very long time, we had a very high opinion of France, and vice versa. They gave us the Statue of Liberty. But there's no such thing as an eternal alliance, except maybe the Auld Alliance between Scotland and France.
The attitude is really souring. The leaders of France and the United States have not been getting along. What we want and what France wants are very different, and we're at odds far more than we're aiding each other.
Part of this is that France is only sort of in NATO. It's the one member that doesn't submit to NATO leadership to the same degree as the others, and it's still trying to operate its own empire. France is still very involved in North Africa.
To their credit, they're one of the few European countries that still sort of has a navy. They have an operating aircraft carrier, and they have the Foreign Legion, which still does adventuring abroad, even though that's not at all part of their narrative of being a woke, progressive country. Germany doesn't do any adventuring abroad, but France absolutely does, in the colonies it still has.
The other thing is that certain neighborhoods in Paris have, I believe, made it illegal to buy wine, because they've been so conquered by Muslim influence. They're trying to push out the native Parisians.
Jonathan: What is France without wine?
Thomas: I was trying to fact-check this live while we were talking. Obviously it's not Paris-wide. But certain areas have become very Muslim and very African, because of a huge influx from France's colonies. The UK has a similar problem, with people from former British colonies immigrating and changing the nature of the country, and it's causing a lot of strife.
Americans outside the urban centers are watching from abroad. In the urban centers they think it's all great and repeat "diversity is our strength" like a religious mantra. But not everyone in America repeats that mantra, and they see the rioting. They see the destruction of historical sites. They see the Palestinian flag flown over the statue of Joan of Arc.
It's very interesting talking with Americans. Seeing what's happening in the UK, they say, "Let's go rescue them. We're ready to invade if called upon." With France, it's not that way. It's, "We bailed you out twice. You're on your own." I haven't seen any desire among Americans to lift a finger to help France.
There's been a big question of which country in Europe will be the first to fall to Islam. Once Islam gets a majority, or even a plurality, it will take over the government and crush all dissent, and a Christian country becomes a Muslim country. We saw this in Lebanon and in Syria. Both were Christian countries. The Muslim population grew, gained political power, and completely crushed the Christian population.
They're going to do the same, and some European country will be the first. It's going to happen, it's going to be brutal, and it may wake up the other European countries, who will say, "You can't live with these people. They don't play nice."
They've never played nice. There's been no point in history where a Muslim minority population was anything but violent in the country it was in, and no time in history when a Muslim majority population wasn't oppressive. You can study the history of Islam. It's really bad.
People don't study the history of Islam. They don't understand what Islam is as a religion and what it causes politically, because it's not just a religion. It's a political ideology. And they're terrified to say anything about it, because if you say anything bad about the pedophile prophet, you go to jail in England.
But I can say he's a pedophile prophet, and I can say Allah is a false god, because I live in America. And I will say it because it's true. Allah is a false god, and Muhammad is a pedophile prophet, and I stand on what I said.
Jonathan: Bringing this around to how it helps authors, particularly romance authors, since that's the easiest example: romance has been associated with Paris for at least 100 years. When your reader reads a book, they bring context to it. They bring preprogrammed responses. It used to be that when Americans saw Paris, they associated it with romance and expected some romance in the story.
Take the movie Anastasia. My girls just watched it. We waited a while, because Rasputin is kind of creepy. The romance between the guy and the girl takes place in Paris. It's a Russian story, but it's set in Paris, and there's a whole song about how romantic Paris is. That was back when Americans agreed.
We haven't seen Paris as the romantic location in a long time. If I said, "Hon, we're going on a romantic vacation to Paris," I'd have to load my guns into my checked bags.
Thomas: The Olympic opening ceremony really radicalized me. I was so disgusted by it, and my opinion of France fell so much watching it. Of the last 10 years, it was one of the most pivotal events in my political evolution. Charlie Kirk dying is number one. Trump's near assassination is probably number two. Number three is the Olympic opening ceremony.
It was so disgusting. The mockery of Christianity, the mockery of Western civilization, the twisting of it, the demonic elements. It was like sewage. Wow, they hate themselves. They hate their history. They hate their God. No people that hates itself, its history, and its God can persist, because to persist as a people, you need those things.
And they don't just hate their true God. They also hate their false gods. They've rejected not only the one true God but also their old pagan gods. They've thrown it all out. They don't even know who they are anymore.
They just hate themselves, feel guilty for who they are, and want to be conquered. There's this weird desire to be conquered that I, as an American, don't understand. We're Americans. We do the conquering. We don't get conquered.
I have complicated views about France, because my family came from France, and they were oppressed there. There was a lot of violence against Protestants, and my ancestors were Huguenots who got hunted. There were wars and betrayals and drama. You can study the Huguenot wars if you want to. My ancestors were Huguenots, and now they're Americans.
But it still grieves me. I've been perplexed about why I still care about France, because I've mourned watching Palestinian flags over burning buildings in Paris. Why do I care? But I do.
I think part of it is that France, for all its follies, is part of Western civilization, and it's been a jewel of Western civilization. I think it's indisputable that the very best art of Western civilization is in Paris, in the Louvre. The Hermitage in St. Petersburg and the Louvre are the two main repositories of the best of Western art. Other museums have great paintings, but those are the two big ones.
I was in Europe when Notre Dame burned. I don't know how much coverage it got here in the States, but it was wall-to-wall news in Europe, and people mourned it.
It seems like a church is burning somewhere in Europe every day. There's a systematic campaign against these old churches, and people don't seem to realize they're losing their history. It's getting burned down.
Jonathan: As an author, be aware that the symbol France once was may not speak to your demographic anymore, because people don't recognize it as a symbol of romance anymore. And that's on both sides.
We're not just talking about Trumpers saying, "I won't ever go to France. I'm not spending any money there." That's not what we're saying at all. The overwhelming perception right now is that Paris is not a romantic location anymore, and that matters when you decide where to set your romance or your cozy.
Thomas: Would you set your cozy romance in Baghdad?
Jonathan: I could see it.
Thomas: You could. But the way you'd write it, you wouldn't rely on the Baghdad setting to add the romance. You'd rely on the setting to add the drama, and add the romance between the characters.
Jonathan: We'd call it This One's in the Bag.
Thomas: We will tour other countries. We have things to say about Canada, Japan, and Korea.
Jonathan: Somebody asked a really interesting question: Is there a replacement for France as a romantic setting?
The dating world has been nuked by societal trends over the last 20 years. It used to be drive-in movies in America, or lovers' point. Where are the romance places now? The bookstore. That's actually a very romantic location. People view it romantically.
Thomas: We talked about this just last week, with Barnes & Noble traffic being up. People in the chat said their kids hang out at the bookstore, and they hang out with romantic aspirations. "Maybe I'll meet my true love at Barnes & Noble."
Somebody said Italy. Italy is interesting. It's now run by a government that's more right, but not super right. It's more centrist.
Jonathan: Who cares? That doesn't affect whether people think of it as romantic.
Thomas: It does matter in terms of how bad the riots get. They tried to do the same riots in Italy, and Italy shut them down. So you don't see video of Italy burning, with fires and the Colosseum in the background. Greece and Italy, maybe. The Mediterranean parts of Greece have a lot of beautiful vistas.
Jonathan: Europe has tons of beautiful vistas. That doesn't mean they're described as romantic. Russia has tons of awesome vistas, once you get past the cardboard buildings.
Thomas: Scotland, maybe? There's a lot of romance regarding Scotland.
Jonathan: I see a lot of potential in Ireland. Scotland is a little dreary much of the time, but Ireland is so green. It's shocking to American sensibilities. Of course, I live in Tucson, so everything's brown.
Thomas: I came up with a saying when I spent some time in Scotland: "The cost of green is gray, and the cost of blue is brown." You pick one or the other. If you want green land, you get gray skies, because you need the rain to make the green. If you want blue skies, you get brown soil. Maybe no place has perfect weather. There are always pros and cons.
Jonathan: Casablanca was set in Morocco, and for about 50 years that was the romance. For a while, Seattle got there too, because of the Tom Hanks movie Sleepless in Seattle.
Thomas: I think there's an opportunity to write the next Casablanca. Pick a part of the world people don't know anything about and make it a romantic setting. Tabula rasa.
The more people know about a setting, the less romantic it gets. Part of why Casablanca was romantic at the time was that no one knew anything about it. There was no Wikipedia page to look up. If your edition of National Geographic didn't happen to have an article on it, maybe you could look it up in the encyclopedia, but there wasn't much information. And there are a lot of parts of the world people don't talk about much: El Salvador, Ecuador, Belize. Hawaii.
Jonathan: If you've ever been to Hawaii, you don't think of it as a romantic place.
Thomas: I've been there. I love Hawaii. There's a big difference between the Big Island and the little islands.
Jonathan: But Japan is considered a romantic location. There's a whole subset of the male demographic in America saying, "I want a waifu."
Thomas: PewDiePie moved to Japan. The biggest YouTuber, with the highest status, could move anywhere in the world, and he moved to Japan.
Jonathan: Japanese women, at least, present as positive, smiley, and extravagantly feminine, and guys see that as the best option for a relationship. In the military, it's a trope: if you go overseas, you'll come back with a Korean or Japanese spouse, or, as Brian says, a Filipino one.
So I can see romance doing well in a Japanese setting. Cami Tang is at the forefront of that right now, because she's been writing those kinds of romances for a few years.
Thomas: Here's a more fundamental question, and since we're out of time, we want you to answer it in the comments: What makes a setting romantic?
We know what makes a setting not romantic: riots, fires, and chaos. That's what's hurting France. But why did France have a romantic image before that? As you find the answer, you can find the new location.
If you find a location no one is writing about, and you can find the romance in it, you may be the next hot romance author. There's a huge opportunity to discover the new Paris. Paris has fallen, but there will be a new Paris.
Jonathan: People want a new Paris. They want Paris to be Paris.
Thomas: Rome is romantic. Rome is always romantic, Jonathan.

