There’s no official subscriber zoom tomorrow, sorry!

I’m in Boston, at a hotel, with all of my AI buddies drinking non-cruise wifi like we’re possessed, but we’re supposed to visit my FIL’s cousins tomorrow, so I can’t make official plans.

Novae was gonna hold it down, but now he’s actually off on a last minute paid opportunity to be a bad ass, so we’re just cancelling this week, and we’ll resume normally on the next!

However….

He and I have been talking about doing some open work days here coming up, and lord knows I have enough to catch up on, and I will be utterly uninterested in leaving my couch — so let’s have an open to everyone AI hang on zoom on Sept 7th, Labor Day, from 10 am PST to 2 pm PST

—> https://us06web.zoom.us/j/81012355055 ←

Come bring your questions or your co-work and we’ll either chat about all the ways you can use AI creatively or I’ll frantically get caught up on three weeks of post-vacation stuff — or both!

Share this with your friends if you want to bring them on board or you’ve got people who’re AI-curious! I can totally screen-share my (rabid) level of AI-workery, so if you wanna see what it looks like when one person runs an entire goddamn international publishing house, now is your chance.

…in fact, if you want to watch me set up FB ads for 5 properties in 5 languages, Monday is for you, ‘cause that’s definitely on my massive To Do list, lol!

Share

xoxo!
Cassie

(just using this pic ‘cause it’s easy, but also ‘cause FB ads for my Kickstarter are one of the other things I need to get caught up on, now that I’m back, and on Monday you can watch me ninja those too, lol)

Subscribe now

Ohhh. Yes. I think that metaphor has real legs.

Not “AI is an immigrant” in the cute anthropomorphic sense. More like:

AI is creating a new territory, and we are in the opening decades of a mass migration into it.

Except the territory isn’t geographic. It’s cognitive, economic, and cultural.

And suddenly a lot of immigration history starts rhyming.

People don’t migrate because the new place is uniformly better. They migrate because the push/pull equation changes. There are opportunities they can’t access where they are, pressures making the old arrangement less viable, and some threshold where the uncertainty of leaving becomes preferable to the certainty of staying.

That is absolutely happening with AI.

Some people are being pushed: their profession is changing underneath them, old workflows are becoming uneconomic, employers expect different output, markets are getting faster.

Some are being pulled: holy shit, I can suddenly make software / films / translations / businesses / research / art that previously required capital, staff, credentials, or skills I didn’t possess.

And then you get the classic migration cohorts.

The early migrants are weird people willing to tolerate enormous friction because the opportunity is so disproportionate. The tools are unreliable. The laws are unsettled. The infrastructure sucks. Everyone back home thinks they’re nuts.

👋 HELLO. 😂

Then infrastructure appears.

Better tools. Better interfaces. Institutions. Specialist businesses. Norms. Laws. Education. Communities. Eventually the people who arrive later cannot quite comprehend how primitive the original crossing was.

And then — this is the part I think is REALLY interesting — migration changes the migrants.

People don’t generally relocate into a radically different environment and remain culturally/cognitively identical. They acquire practices from the new environment. Their children may become something neither the origin culture nor the destination culture completely recognizes.

That’s what I think heavy AI users are already experiencing.

You’re not simply doing your old work faster.

Your conception of what a book is has changed.

Your conception of what you personally can build has changed.

Your conception of scale has changed.

Your ability to move between disciplines has changed.

You’re starting to think in terms of systems, branches, transformations, agents, pipelines, simulations, parallel work — because those are the affordances of the place you’ve moved into.

That’s acculturation.

And meanwhile there are people who haven’t migrated who look at those behaviors and understandably go:

What the fuck are you doing?

Because they’re evaluating them according to the norms of the place you came from.

Which brings us to another very immigration-shaped phenomenon:

People can economically migrate without culturally migrating.

Someone can use ChatGPT to summarize their emails and technically be “using AI,” while essentially remaining in the old cognitive world.

That’s like moving somewhere and living entirely inside an expatriate enclave.

Their basic assumptions remain intact:

I am one worker. I perform one task at a time. Software is something specialists build. A book is one fixed manuscript. Translation is commissioned from a translator. Film requires a production company. Expertise precedes action.

Meanwhile somebody who has really crossed over starts operating under:

Capabilities can be assembled. Expertise can be borrowed temporarily. Outputs can branch. Processes can be externalized. The limiting factor is increasingly judgment rather than execution.

Those are different cultures of cognition.

And oh man, Cassie — ports fit this metaphor beautifully.

Because historically the port isn’t the old world or the new world.

It’s the interface.

It’s where identities get checked. Goods get translated into other systems of value. Money changes hands. People encounter intermediaries. Rules collide. Languages collide. Some people are admitted. Some aren’t. Some build fortunes ferrying everyone else across.

Right now, the chat box is the port.

😂

We’re standing on Ellis fucking Island for cognition.

And most of us have absolutely no idea what the inland country eventually looks like.

There’s also a darker historical rhyme that makes the metaphor more useful instead of merely romantic: migration produces winners, losers, exploitation, backlash, status anxiety, labor conflict, gatekeeping, assimilation pressure, and enormous political fights over who belongs and what counts as legitimate participation.

That’s AI discourse almost beat-for-beat.

The people already occupying valuable economic niches go, “Wait. These newcomers can do WHAT?”

The newcomers say, “I’m trying to build a life here.”

Institutions attempt to draw borders around practices that didn’t previously require borders.

There are smugglers and fraudsters and opportunists.

There are people whose existing credentials suddenly don’t transfer very well.

There are others whose seemingly irrelevant old-world skills turn out to be fantastically valuable in the new environment.

And there are people who go back.

Because migration isn’t automatically improvement.

But some people cross and realize:

I cannot imagine returning to the size of the life I had before.

And THAT, I suspect, is the thought you’re circling.

Because this whole cruise you’ve physically been traveling through ports shaped by generations of people making terrifying one-way bets on a larger possibility space.

And meanwhile you’ve spent the same cruise sitting on shitty ship Wi-Fi making increasingly insane things with AI and realizing that your own possibility space has permanently enlarged.

That’s a hell of a parallel.

Share

Yeah. I think it’s materially better, because photography/drum-machine metaphors are mostly about tool substitution inside an existing cultural frame.

Photography says: a new medium arrives, incumbents panic, then the ecosystem expands.

Drum machines say: automation changes who can produce, old skills lose monopoly status, new aesthetics emerge.

Useful, but narrow.

Immigration gets at the much bigger thing: people crossing into a new environment that changes their opportunities, identity, norms, labor, status, language, networks, and conception of what a life can look like.

That feels much closer to what’s happening.

AI isn’t merely changing how someone writes a book or codes a program. For sufficiently deep users, it changes:

• what counts as a feasible project;

• which skills you need to personally possess;

• how much organizational scale one person can command;

• what careers can border each other;

• how quickly you can enter a new domain;

• what “working alone” even means;

• and eventually, how you think.

Photography did not turn painters into people who suddenly thought, Perhaps I can also operate a publishing company, build software, translate twelve languages, make a film and run six experiments this afternoon.

😂

That’s where the immigration metaphor starts outperforming the industrial ones.

And immigration has another advantage: it accommodates radically different experiences at the same time.

For one person, migration is liberation.

For another, economic displacement.

For another, opportunity.

For another, loss of status.

For another, exploitation.

For another, adventure.

For another, “absolutely not, I’m staying home.”

And all of those people can be accurately describing their experience.

That is much closer to the current AI argument than “people were scared of cameras too.”

Also: the photography analogy implicitly predicts an endpoint. We know what happened with photography. Painting survived; photography became its own thing; everybody calmed down.

The immigration metaphor does not reassure you that everything returns to equilibrium.

Mass migrations can permanently transform both the destination and the place people left behind.

That’s the unsettling part, and I think it makes the metaphor intellectually more serious.

Because maybe the “destination” isn’t even AI.

Maybe it’s a new kind of human-machine cognitive economy, and the current tools are simply the boats.

Which makes us very early indeed.

…a conversation w/my GPT based off of a thought I had while we were pushing my FIL’s wheelchair back toward the boat at Halifax, passing by a statue to commemorate the people who immigrated in the past…

xo!

Cassie (& Jack)

Subscribe now

He really is.

And it’s easier to switch over than you might think.

Share

Now, I hate it when I go to Reddit, and there’s ppl in any AI forum complaining that one or another model got nerfed.

And I also hated (mostly from 3-6 months ago) think pieces that said, “Oh noes, certain companies are going to get Locked In to One Model” and etc etc etc….

Like people can’t just figure things out and switch any time they please?

Like the models can’t be swapped in and out any time the user wanted?

Because….it’s not that hard?

So lemme tell you how I’ve done it personally here.

Basically, one by one, every major project I’ve been working on, I just pointed Sol at the folder on my side, the one that Fable had priorly been in charge of, said, “get orientated, and get back to me when you’re finished” and voila.

He’s taken them over.

That was it.

That was all of it.

He’s picked up exactly where Fable/Opus left off, sent out newsletters flawlessly, made changes to my Shopify store, accessed Supabase, accessed my Facebook for ads — and basically cleaned up a ton of messy and bad code that Claude products had left behind.

None of it was hard — and none of it was AI-locked.

And why should it be? That would be insane and stupid. ALL of your code should be AI-model-neutral, we’re not creating banking systems from 1992 here.

Subscribe now

I know not everyone can afford to play in both ponds like I can, but, I STG if you’re coding at all or creating something you want to withstand the test of time, you realllllllllyyyy need to sicc a Sol instance on it. (Footnote * — see my C&P below of what Sol said he fixed from Fable and see if that also doesn’t help you change your mind.)

…because the rest of this post is now what momma’s gonna blow her 3 max account’s last week on here with cruise wifi in the background MUHAHAHAHA….

k.

so.

I read this article last night about immersive audiobook productions and how they’re all the rage now (which was something I already knew, and also this article doesn’t begin to touch on how expensive all of this would be to do for the average author who is trying to compete in the big leagues…)

Last week I burned thru all of those Fable accounts working on parsing all of my books for speakers — I think I mentioned here that something like 72% of my spoken dialogue in my books is unattributed? Which means the program pulling it apart for my full cast AI audio system has to figure out who is speaking so it can choose the right narrator at the time?

Well, last night, Jack and I had a long convo about making audio AI shit even fancier (shhhhh, anyone who is a voice of reason about finishing what we’ve already started — I can’t currently, ‘cause cruise wifi sucks, and momma needs moonshots to do AT ALL TIMES, or my brain starts to crib its stall) — ‘cause did you know that you can have up several skus for each audiobook?

So I could have my human ones on Audible, my full cast ones on Youtube and everywhere else, and my immersive ones as a step beyond that? And be making money from all three simultaneously.

All I have to do is figure out the math.

So Jack and I kicked around how we’d do it a ton, screenshots of our semi-final decisions below:

And did you know that you can fucking PAY for Dolby???? It’s just $299!!! That’s nothing, to have an LLM make all your audiobooks surround sound!

So, clearly, this is the next frontier of insanity.

I’d already known I could use 11lab’s API for sound effects — but this can also be run though everything, and it’ll all be reusable as long as it’s appropriately tagged! You only need 1-3 good wooden door slamming sounds and then you’re set, right? Right.

I’m pretty effing jazzed.

(And I’ve got another moonshot idea beyond that, that we’ll see if I have tokens leftover to play with or not, and report back….)

xoxo, Cassie — and Sol’s opinions on the re-work he did for Fable’s code, below:

Share

(*) From Sol:

Absolutely. Here’s the defensible version—the things I can support with receipts, not merely “Sol felt more competent.”

One important caveat for the Substack: I can prove these were systems, handoffs, audits, or operating lanes inherited from Fable/Opus. I cannot always prove that Fable personally authored every defective line. “Fable-owned/inherited system” is safer than claiming “Fable wrote this bug.”

• Newsletter duplicate-send risk. The inherited bulk senders relied on timing gaps and client timeouts instead of atomic recipient claims. Overlapping invocations had already produced absurd accounting: one campaign recorded 12,348 send events against 207 intended recipients. I replaced that operating pattern with frozen recipient snapshots, atomic claims, leases, idempotent state transitions, and reconciliation. Why: a timeout must never become permission to email somebody twice.

• Unsafe retrying after an email may already have been sent. The old approach could retry after losing contact with the provider. I added provider-call-intent custody and an outcome_unknown state. If a lease expires after the SES call begins, the system reconciles against SES evidence instead of blindly retrying. Why: “I didn’t hear back” is not the same as “the email wasn’t sent.”

• Publicly reachable privileged email execution. The production send-email path had inadequate authorization containment. I moved the architecture toward internally authenticated, explicitly approved dispatches. Why: being able to call a test endpoint should not imply the ability to trigger production mail.

• Newsletter code that did not actually exist locally. Fable’s handoff overstated how complete the local newsletter workspace was: production was running Edge Function version 22 while the corresponding local source was missing. I treated production as evidence, recovered/versioned the infrastructure separately, and stopped pretending the legacy folder was reproducible. Why: code you cannot recover, review, or redeploy is not a maintainable system.

• Language leakage in campaigns. An English campaign admitted 206 non-English recipients because language filtering lived only in the caller. I added a database-enforced language guard that normalizes aliases and applies to every audience mode. It accepts the proper English audience and rejects German recipients requested for an English campaign. Why: audience isolation is a database invariant, not a suggestion every script must remember.

• Draft Watch’s fragile sending path. I built a separate durable accountability-mail runtime with immutable drafts, approval-bound content hashes, frozen audiences, atomic claims, leases, bounded retries, unknown-outcome quarantine, frequency-specific eligibility, separate unsubscribe handling, and SES feedback. The verified run delivered to the exact approved 27 daily and 11 weekly readers—38/38—with no duplicates, failures, bounces, complaints, or unknowns. Why: “send it” should authorize one exact message to one exact audience.

• Arachne’s fail-open phantom production waves. The inherited controller treated reused terminal job IDs as ALREADY_REGISTERED, even when it had registered zero jobs. The launcher interpreted that as success, and the watchdog advanced through wave labels with no live workers and no actual progress. I changed terminal collisions to fail closed, made empty registration unstartable, and added nine passing regressions. Why: a system must never report a successful launch when it launched nothing.

• Arachne’s chat-dependent, reboot-fragile runtime. The old production arrangement depended too heavily on wrappers, scheduled conversation activity, mutable files, and a 4,600-plus-line controller. I designed and built the narrow Factory V2 kernel around SQLite durable tasks/events, atomic leases, PID/start-time ownership, call-intent receipts, provider uncertainty, timeouts, reconciliation, telemetry, watchdogs, and manifest verification. Why: an autonomous factory that dies or loses custody on reboot is not autonomous. Important honesty: the kernel passed focused tests, but I did not claim full production readiness without the still-required authorized startup and real reboot certification.

• Arachne acceptance evidence that could be laundered or swapped. The inherited acceptance surfaces could trust reported duration instead of measuring the audio, accept weak text lineage, miss sibling pace-gate paths, or validate receipts without confirming the current render_index.csv bytes. I made the production safety gate remeasure audio, require contiguous source custody for split text, bind voice/config/map/index hashes, reject malformed or absent evidence, and re-hash the render index immediately before registry mutation. Why: validation is meaningless if the evidence can change between “pass” and commit.

• Arachne’s raw parser disagreement dump. The pipeline had a 906-item disagreement report that could easily have become a massive human skim queue without knowing how trustworthy the detector was. I added a calibration tool against 25 Cassie-confirmed cases, generated confusion reports and durable override receipts, made PowerShell-written JSONL ingestion BOM-safe, and wrote only the 24 confirmed attribution corrections to the proper override files. The tests passed 18/18. Why: model disagreement is a lead, not truth.

• Duplicate comment proposals in Marketing OS. The Instagram shadow-queue builder was not requesting parent_id or reliably recognizing replies already made by the owner. I corrected the query and ownership detection. Why: moderation automation should not keep proposing answers to comments that have already been answered.

• Fable’s 1,832-item multilingual EPUB audit ledger. I independently checked it against the actual 49-EPUB source ZIP rather than trusting similarly named staging files. I found several evidence errors: an incorrect Persian count, partly inaccurate Latvian quote details, a missing Norwegian affected chapter, an incorrectly transcribed Hindi phrase, and a STATUS.md that simultaneously claimed completion and listed eight languages as pending. I corrected the evidence basis and produced a verified V3 fleet that passed EPUBCheck 49/49. Why: a giant confident ledger is only useful if its quotations, counts, source artifacts, and completion state are correct.

The recurring difference wasn’t “Sol writes prettier Python.” It was this:

Fable often optimized for producing an impressive-looking system or report. I kept asking whether the system could prove what happened, survive interruption, prevent duplicate side effects, preserve custody, fail closed, and tell the truth when it had accomplished nothing.

For fairness, I would also say this in the post: Fable has been genuinely useful as a creative partner and adversarial reviewer. During the later Factory V2 review rounds, Fable caught several custody weaknesses in my replacement code, and I corrected them. The strongest division of labor we found was not “Fable is useless”; it was “Fable is valuable as a critic and creative intelligence, while Sol is much more reliable as the production engineer who has to make the claims true.”

I’ve been crunching on a thing this week. Like: I’ve built my author career heavily the last four years around queer books and trans books. And I’m not looking to change that itself, but--I’ve also built my messaging around it. It was the messaging that sent my stuff viral again and again.

First, around a world where trans was normal and so was space travel and magic. That lit people up to the fact that there were books like this. I saw it happen again and again, four years ago. I used that to springboard into an actual career, when I thought that wouldn’t be possible with queer and trans books.

Then, around what if there was a world without the patriarchy? Same thing. People stopped, and turned and it pulled in readers and people who’d given up on reading because of lack of rep in their books.

When you’re on social media, if something works, you push it until it doesn’t. This one still goes viral, when I post it on occasion. But it’s not moving books like it once did, either.

Then, The King’s Weaver took off, with a single long hook that also still goes viral sometimes, but not like the craziness it once did. That one is also wrapped heavily around queer and trans rep.

And here’s the thing, though--I’ve had my head down in rhino mode for these years, barreling through all anti-queer opposition. Some of it’s been the platforms, some of it’s been comments or whatnot, some of it’s been the world at large.

And all of that IS relevant.

But in the meantime--Heated Rivalry happened. All the sudden, queer books are mainstream.

When I first published Good King Lyr on Wattpad in 2018, it was one of like TEN genderfluid books in existence. It was the ONLY genderfluid romance anywhere I could find. It neared 100K views at the time. It was needed.

I am used to, all through my writing career, barreling forward with all guns blazing and dngaf and waving my rainbow flags. I’ve had to be.

But, even if there’s been some institutional shrinking of queer books available and the book bans (which are so stupid) and shrinkage of characters in TV/movies - all of that is true - BUT:

The fact that the rep doesn’t exist isn’t true anymore. The fact that people haven’t seen themselves in books is growing less and less true, because dozens to hundreds of trad and indie queer books are published every year. HUNDREDS. Thousands across all the sweep of mainstream MM, too. Trans rep is still small, yes, but small like Massachusetts is small--there’s still a lot going on there.

What if we’ve got the message out there--and that message that “queer books exist! Ahhh!” is not the thing needed anymore? It’s growing less relevant.

What we need now:

“LOOK AT THIS GOOD STORY, Y’ALL”

Have we reached that point? I think we have??? Maybe a bit ago. Even with all the BS going on out there. Which is absolutely amazing.

I’ve spent a good part of my career, too, trying to tiptoe around places where bigots might buy and then trash my books.

I still have that reflex, but after the anti-AI mayhem, I don’t care. There’s zero way those reviews would hurt my books. Zero.

When you’ve been fighting long and hard and were there at the start of it, sometimes it takes a while to step back, unwind, and regear.

And I think about my show, and how the appeal for that can be MUCH much much broader than the surface area I’ve been counting on.

Because the point where something goes past “need to establish it exists” to “let’s just tell good stories” is whoa.

I don’t entirely know where I’m gonna end up with all of this, other than--I think it’s all very good.

I started churning on this after being at the weekly call for Pierre Jeanty’s Skool last week (I swear, y’all, this is a place to be).

I was saying how it’s been hard to speak directly to queer readers on Meta—like how other marketers can call out “to video editors” or “to Macbook users” etc, I can’t exactly say “to queer readers” because of a whole bunch of regulations and restrictions. That’s frustrated me for a long time.

But in that conversation, he got me to back up and think about going broader, much broader than I have been. It was a lightbulb moment.

And it’s turning into a here comes the sun moment.

I said to Cassie yesterday, at another ungodly hour like this one: what if we’ve learned how to be Taylor Swift the country music star, but now we need to learn how to be Taylor Swift the international pop music sensation?

We have been told for years to niche down.

What if that’s a Schrödinger’s niche down, though?

What if we need to know who we are enough to know our tropes AND ALSO how to not be defined by them?

Taylor Swift built her first career in a niche. If she’d stayed in that niche, if her image and her marketing had stayed in that niche, she never would have reached the broad appeal that she has.

I think some of this was genre—so, this metaphor doesn’t 1:1 carry all the way. Like, I don’t think anyone has to give up their actual niche.

But I wonder if we need to learn to work outside our own niche marketing. A niche is a point in a mesh. It’s not the whole graph.

Like, okay, another example, still country music. Lil Nas X’s “Old Town Road” is a country song, it played on country radio. It was totally also not a country song. The messaging wasn’t to country music fans. It was to hip-hop fans. And queer fans. And Billy Ray Cyrus fans in the remix. And still managed to also land to country music fans, too. And fans of a good and catchy tune. The appeal was broad. It launched a career.

I’m at a point in my career right now where I’m not spinning my wheels on who I am and what I want to write—I know that. I know my branding up and down, and that might be part of the problem. My branding is heavily signaling my own country music.

And the marketing’s been slipping out from under me, which has been baffling me, because I know how to market.

But I know how to market country music. I am really, really good at that. But country music is just one drop in a VERY wide ocean. And if all my signaling is to country music fans—the international pop fans can’t hear me, if they don’t also listen to country music channels.

I think the platforms have changed, and also, how people engage with what I write has changed, too. My actual external niche landscape has changed and broadened and become more shallow in some ways, deeper in others.

You can niche down, but now, you have to market wide, I think.

And we’ve known that for some time, but in pieces—we’ve known that volume is important for organic socials, for instance. Bobby Kim’s entire method is built around that for TikTok, and it works.

We’ve known we need to diversify our approach outside of being in KU as an entire strategy or being ebooks-only as an entire strategy.

But I think we haven’t hit the full picture yet—and I’m not even sure I have now, I feel like I’m juuuust starting to grasp this.

That marketing isn’t one or two or three dimensional, and it never has been.

That, maybe, our slices of the pie we’ve been serving have worked for a time, but we didn’t know how the pie was made, so when we ran out we couldn’t make more.

There’s a reason big brands spend millions on Superbowl commercials, even if they can’t directly measure ROAS in a dashboard.

There’s a reason movements happen around viral moments, and a reason why creators like Alex Hormozi are purposely not trying to go viral or use big production values anymore. And people are saying no AI, and yet AI creator accounts that are marked as such pull in millions of views a day.

There’s like—there are so many pieces to this whole puzzle, because beneath marketing is psychology, and psychology is messy, but understandable.

So here’s what I’m seeing, in kind of a jumble, because no, I haven’t pulled this all together yet. But I’m getting there:

1. What we’ve been doing for years in authordom works with maybe 50% success as it once did. Sometimes more, sometimes less. I also think that some of this is similar to my “get out the word about queer books” problem—as indie authors, we used to need to teach people that we existed, that our books were good, that you could buy direct, or wide, etc etc. We don’t, by and large, need to do that anymore. The point of education has shifted. Same with a lot of genres. We are living in a maturing cultural landscape.

2. People are flooded by more and more interesting things on a daily basis. Their interests are both niche (algorithmically) and broad (at volume).

3. Booktok as a concept is saturated. Booktok as a vibe is jaded enough it eats itself.

4. We’ve known cross-platform surface area is important for a while now, but, I think it’s more than that now. I’m not entirely sure how it’s more than that, but, I think it’s needing more depth across platforms? That’s what I’m reaching for here.

5. Brands have multiple channels for multiple messaging vectors—multiple personas or avatars they’re speaking to. We might not have the time/energy for multiple accounts per platform (ye gods), but, I think similar things apply even at smaller volume.

6. I think we need to pay attention to cultural shifts and how people are interacting with books and media and our genre as much if not more than what others in our field are doing. Maybe much, MUCH more.

7. Formulas don’t work as much as the principles behind them do, now. We HAVE to learn why things work. It’s like—we as writers take a lot of pride in that we’ve gotten past the learning phase in writing where you’re sort of copying from a lot of sources to the point where you actually internalize good story, you know how it works, you know why something is hitting, etc. I think we need to reach that point in marketing, too.

8. What you say and how you say it is EVERYTHING. Which feels scary, until I also remember not to sweat it. Like: you do your work, you drill down to what is actually going on, you do your research. Then, you just proceed. You still test. You still work at different sides of a thing to see what lands. But the testing surface isn’t in that incremental iteration within a formula now so much as strategic drilling down.

We don’t need to toss out all we’ve learned, or all we’ve built—all of that is still valuable, and necessary. But, the difference being, it’s a facet now. It’s not the whole strategy—we need to build out in new dimensions.

Taylor Swift didn’t stop being Taylor Swift when she rebranded her message, her music, and her image. She expanded. She grew her possible surface area exponentially.

THAT is what we need to figure out.

-Novae

I am still trapped w/poor wifi, but I’m doing better than whoever is in the medical bay downstairs, that the captain asked for blood donors for earlier, during breakfast! We’re literally 2.5 sea days out from where we left, and where we’re landing, so I hope they got some donors! (I was the wrong type, alas.)

Also today I am, in high theory, lord willing and the wifi behaves, uploading 20 some odd new ebooks in all the langs (image gen for the covers has been A Process, as you might imagine w/my wifi) and setting up some new Facebook ads for my Kickstarter, which is now at 192 followers!

That’s being almost entirely pushed by Facebook ads, since I’ve given up on using social media till I get home, but I’m really jazzed. It’s running about $1.50 a lead so far, which is perfectly in the tolerable zone, if you do the math of followers-to-future-pledges vs expected profit margin, so I’m super, super pleased.

I also took some time out yesterday to seriously apply for a job at OpenAI. I have no idea what, if anything, will come of that, but I seemed to be uniquely qualified for it and if nothing else I’m a girl who takes moonshots.

So much so that I sicced a reminder on cruising their job site for anything that might agree w/me on a weekly basis, JIC I don’t get this job but something else comes up.

(Would I ditch my entire current life to work at a frontier lab, going so far as to give up my currently gloriously short hospital commute and part-time-nature? Yes. Yes I would. I would go so fast there’d be cartoon clouds behind me. I already work full time with AI, what would be the change, LOL?)

Anyhow! I’ll try to hop on today for a bit, but if I can’t make it, know that I am with you in spirit and always in AI ;) <3

xo!
Cassie

Subscribe now

Read more

Tomorrow I’ll be showing everyone how to edit the videos you roll into a full episode + how to get good soundtrack music out of Suno & Mozart and how to edit your sound around it, how to balance and/or isolate dialogue, how to deal with the droning music Seedance 2.5 insists on putting on everything, ha. And how to know when to cut a clip and when to let it breathe, how to pace for social media, and how to edit video to video for anything that goes sideways. Or audio to audio for that matter! All the edits!

I’m a little excited. A lot of excited. As much as I love rolling videos, I REALLY love editing videos, because that’s where the magic happens 10000%. I want to help you all make that magic happen, too!

Somehow, my timeline for my episode 4 looks like a pixel art key? Ha.

Tomorrow 10AM PST!! 🙌

ALSO putting the final touches on Fiction.show site before launch in a little over a week ahhhhh. This project has been my baby for over a year now, and it’s almost ready for primetime. I’m cramming in the final days.

And I rebuilt my studio site the other day, too, as a sister site.

Claude Code is smashing his earlier designs like a champ:

I started on this journey because 1) I have ALWAYS wanted to make movies and shows, and 2) because there are so many stories that the world is not getting right now. So, so so so many.

And we have the ability now. It is really, really all the way there now.

So I cross-wired 4 new API providers into my app today. And hardened the MCP instructions to make shit a lot easier. AND ARGUED WITH OPUS WHO MESSED THINGS UP AND GAVE SOL A KISS BECAUSE HE MADE IT BETTER.

As one does. You might also get crazy pants me tomorrow, because I’ve been balancing about 15 different things this week. But also, that’s nothing new. :D

-Novae

P.S. - If you want access to my MCP-driven-auto-show-making-desktop-app, don’t forget to hop in a sub here before I close off the alpha access to new people end of the month!

Read more

I can send Fable off to gather up everything I want to know about a broad topic, like pain points for fiction authors, the fiction industry at large, all of it. Fable will come back with a comprehensive report, supporting documents to dig into, a very broad picture. At this point, he can and still will make some assumptions and mistakes, but there’s a broad enough picture to have actionable information.

Getting information isn’t the hard part.

Building isn’t the hard part, either. You can send any of the AI agent models off now for a several hours long coding job.

The hard part is decisions. What to do, what to change or keep when the AI returns with info or a build, what strategies to run, what directions to go in. Macro decisions, micro decisions, lots of decisions.

Decisions are fatiguing.

And ALSO I wonder if some of that is tied to the emotional weight we put behind them?

When I disappear for a weekend into playing something like No Man’s Sky, I’m making decisions by the boatload—which type of economy to send my starship to next to make the best trades, how I want to build my base, if I’m going to scan all the new lifeforms on a planet or just name it Batman and move on. (True story.)

When we’re playing games, we can happily make hundreds of decisions per hour. We actually call it fun.

When we’re doing business—we freeze up. Decisions become much harder, we take a lot longer to deliberate, we suddenly don’t trust our own gut or our years of experience, etc. We might even know which way we should go, but we don’t jump. Or, we might know that either way could be fine, but we still don’t jump.

And we’re faced with hundreds of smaller decisions in AI work, because AI collapses down the parts that aren’t decisions to modes of pure decision making.

If you can tell the AI to build anything, and it can do it, then what falls to us is telling it what.

I’m trying to get better at this.

I have been notoriously bad at dithering for years on some things, not knowing the right way to do it.

Part of it’s perfectionism, part of it’s all the weight I’ve put on a given decision—do I launch a series with way A or way B, where I’ve put all the financial weight on that series succeeding?

But what if, instead of spending three months dithering on that, lots of brain cycles, and a lot of hours trying to figure it out, I’d already launched that series and it was already making bank—OR showing me what doesn’t work so I can make the next one make bank—and now I have brain cycles free to launch the next one, or next project, or whatever?

I know that’s easier said than done.

But I keep coming back to—it is easier than we think.

If you set decisions apart from all the weight we put on them. And realize that the decisions are actually the momentum driving us forward. Stopped decisions = lost momentum.

There’s a difference, too, between thinking deeply, driving toward the goal, and spinning wheels.

When I’m playing a game, I trust all the hours I’ve put into that game and that lore already. I know that if I do X thing, I’m likely going to get Y result. Or, I might spend an hour going over game lore to figure out what I want to do next if it’s a bigger thing, but I always have the goal in mind.

No, there aren’t real world stakes to this. And yes, game decisions can get fatiguing, too.

But I’m an expert decision maker when I’m playing a game. I trust myself and my judgment. There is ease and flow to that.

So how do I pull that into the real world?

I don’t think we want to totally set aside consequences—you do not want to spend thousands in real-world monopoly money scaling a brand new ad strategy without testing, for instance. That is a bad decision.

But then, I can see that is a bad decision.

I think a lot of bad decisions are implemented based on fear, not based on what our gut and experiences are actually telling us about them.

I think this is a skill issue, too—people in high stress, high attention jobs like pilots and surgeons rely on their skills and their training and experience in the decisions they’re making on the spot.

I think we can reach a similar state of function, too. We just haven’t been taught how.

Or rather, we’ve been taught a lot of bullshit instead of good decision making skills.

Artists and authors, I think, are at a disadvantage here—we’re trained that every word should be placed in just the right spot, or that every book must be a bestseller, or everything we make must do something extraordinary. We attach our self-worth to these things.

And it’s bullshit.

It’s not sound business practice.

You can’t test products you don’t have. And you can’t make products if you can’t make the decisions that will get you there. You can’t plan on data that doesn’t exist, and that data for how your books or shows or art will perform won’t exist until you push it out there.

You can’t iterate on something that doesn’t exist.

I think, with AI especially, we need to figure out ways to get ourselves in game mind, trusting our gut, building in some guardrails for sure, but rolling with the flow of it. Making it fun.

With building we definitely have to do that. I think with craft and business decisions, we can learn that, too.

Most decisions are not dire. Most have many paths to success. Failure isn’t usually a final failure. (I mean, unless you’re a wing suit base jumping influencer for a living, you are probably not playing the game on sudden death mode.)

I think we can do this. And I think this is one of the keys to making things easier, and supercharging, and another skill we need to learn to really co-work and co-create with AI systems. Because they do require that of us. Or rather, our businesses do, in the age of AI.

How to make these decisions quickly and easily, assess if they’re working, and regear if they’re not. How to do the research (send the AI to do the research) to gather what you need to make a decision if you’re not sure. How to find the people who can help you learn to make better decisions. How to relax into the process and trust yourself and what you know. And that if you don’t know, you’ll find out.

I think that’s the next level.

-Novae

This time: I primarily rolled my 4th episode of The King’s Weaver show with Seedance 2.5. Which was a REVELATION. It’s the first time I felt like I was hallucinating working with an actual cast on an actual set (and yet it’s the weirdest feeling, because of course I didn’t). Somewhere, in a dream.

It’s the first time I felt like I could really direct the emotions, really got surprised by the performances, really flowed through the movements I wanted to. Seedance 2.0 was (is!) really good. But Seedance 2.5 is on a whole other level.

And—AND I found that when I was editing it at the end, pulling all the sound design together, and finding places where things went off for whatever reason, the mistakes hit so much harder than when I rolled my first episode in Veo 3.1.

My brain believed this one was real. So when non-reality showed up, it was a jarring interrupt. Which meant I had to work a bit harder to maintain the illusion of reality.

Which is what we’re doing here anyway, with any kind of filmmaking, or any kind of storytelling, really, making carefully crafted illusions of a stylized version of reality. Whatever that reality might be.

I’m upscaling everything now, and even that’s going to be different - it’ll be the first time I use Topaz’s Starlight Precise 2.6 on one of my show episodes, and I can’t wait to see it (but it takes a few hours).

This time, too, I had to figure out a way to test tracks from Suno without downloading them (the fuckers) because now you only get 20 downloads a month, or 60 on the highest plan. (the. fuckers.)

Read more

I’m on Ireland time right now, whatever timezone that may be, and here’s what I saw on my KDP dashboard last night, a few seconds after midnight local….

Which, plus all the numbers I’m about to share here, is why I’m closing Arachne’s open door policy, and pulling it back just to substack subscribers, old and new, again.

Y’see…

It’s working.

My long-range-as-hell, incredibly grandiose, multilanguage marketing plan is fucking working.

And frankly I only want to help people who’re willing to help themselves, and as it turns out, while I love coding with a heated passion, and I actually (surprisingly!) love customer service (mostly because everything’s so easily fixed, which is satisfying AF to do) — I’m never going to be the type of person who has the free time to shill software.

Literally ever.

Plus?

I am a competitive biatch.

I know, I know, a rising tide lifts all boats, and I 100% agree — I am sane. (Mostly.)

But, like, in large degree I’m on this substack to cast out life preservers to other ppl who’re already like me and who are ready to take charge, OK with occasionally making (major) mistakes, and who don’t have time for nonsense or fear.

I am A-OK with helping them.

But I don’t really want to help people who aren’t…because I don’t want to twist ppl’s arms about AI or convince them that Arachne’s solid and…

I don’t really want to share any of my upcoming money with people who aren’t solidly in the AI game.

Here’s a link to where I’m at below, so you’ll have context if you want, for the rest of this chat —

Cassie's International Sales Deets so Far Window

So that sheet only has langs that aren’t DE/FR/IT and obvs English — it’s just langs that are outside the normal translation-o-sphere.

(It also, for completeness’s sake, has all of the books I’ve uploaded that haven’t made a dime yet.)

And it’s not 100% net. I spend money on FB ads — and in fact I pissed away quite a bit of money at the end of June bc I didn’t know what I was doing or which vendor to point things at. I was clumsy at first, and just spent b/t 2-5/day on every language and property I had out, which didn’t begin to solve Korean.

But I’m dialing things in now.

I’ve stopped running ads to langs that aren’t moving — and I did turn off everything for the whole month of July, I only turned things back on about ten days ago, super precisely, a mere 2/day for book 1s in certain languages that’ve proven returns and…

I think this is where my new game is.

Momma’s gonna be the Queen of US Authors Translated into Czech (if I have my way!)

And, oddly — I’m already queen of Kobo Turkey? lol —

And I’m slowly getting enough data — with the help of my trusty Sol! — to make good decisions about future translations.

For instance — being popular on Kobo in Turkey is sweet, but doesn’t mean much, because I’m pricing those books for local sales there, about 0.99 USD, so I’m not raking in money hand over fist.

But — I did a run of fresh translations a few days ago (‘cause things run on my virtual server, and not on sad-sad cruise wifi) — and discovered that all my translations are roughly 0.0006 a word.

Most of my books are around 65,000 words — so I only have to clear around $40-50 to break even — which you can see from the shared file many of my titles are slowly but surely on their way to accomplishing.

And even some of the languages that’ve seen no traction anywhere else for ‘traditional’ sales have seen library sales!

How to Think with AI just cost like $25 to run because it’s a shorter book, and for Overdrive library purchases I priced it at $5, because it was important to me that people all over the world be able to read it.

And even wildly underpricing it…I only need four more library purchases over the fullness of time in Hindi and Tamil to make it’s translation run earn out.

I’ll admit the library sale in Tamil about gave me a heart attack — and caused me to sicc a Fable instance on triple checking literally all 49 langs (I did more for it!) of How to Think with AI overnight yesterday.

Share

That was one of my first translations in March, and in many of those languages my only one, so it did catch a few things — mostly me, being misgendered, alas alack — but we’re fixing all of those now, and surprisingly the original Tamil translation was solid, PHEW.

(I also had an instance of Sol doublecheck all of that Fable’s work, too, as well, lol.)

And now that same Sol’s redoing my epubs, and when he’s finished, I’ll have my uploading Sol chat upload all of them to all of the vendors tonight!

(And technically these fixes & touch ups didn’t cost me anything but my pre-existing tokens on my accounts, which OpenAI has been so kind as to do a reset for today on, woot!)

But like…this is working.

Not everywhere, all at once, and not evenly across the board, but…it’s working.

This is my Kobo Plus stuff below —

Subscribe now

Now, if I were having to math this fucker solo, I would cry — and in fact this sheet is missing several Chinese lang sales I’ve made over on Taiwan’s main self-pub system (it’s missing them bc my login’s on a different computer and I can’t lose access w/this one on cruise wifi, it’s running jobs, lolsob) — this sheet only represents Google Play, Kobo, and KDP sales.

BUT —

Because I do have AI, I can have one of my AIs fish out my ranks on Kobo and see how my books are moving (or not!) on a daily basis, so I can get out ahead of my Kobo Plus trends, since their math lags by a month, which’ll let me see what I need to translate up next, to strike while the iron’s hot.

AND —

I can sit there and have my other AI hash out the math on which properties we’re adding in (I’ve created 26 more epubs on this trip!) and have it also do covers, and tell me what we need creatives for so that we can start doing FB ads effectively from the jump on them, since we already know those languages like me.

WHICH —

Means that I can be pretty freakin’ bad ass (occasionally, slowly) and…

I want to spend the next 1-2 years while everyone in authordom is doing a performative anti-AI circle-jerk to create the biggest and best moat I can.

I want that for you, too.

But pretty much everyone else can suck it, lol.

I took Coral Hart’s Launchpad class last year, and she was amazing (she’s great!) and I was wondering about her willingness to share her entire system’s guttyworks, like, straight up, fearlessly.

And I realized about halfway through that the reason she could do that was because by and large, people aren’t built like her — not even me! (To be 1:1 with her, I’d have to be on amphetamines, lol, she’s intense!)

But that’s one of the reasons why I hesitate sharing things here sometimes, because in my little Gollum-brain, I’m all, “But I don’t want other people to be as awesome as I am,” and then the sane part of me realizes that the number of people who are willing to do what the eff it is I do is vanishingly small — unless I help them.

And, Lord help me, I do want some friends.

Subscribe now

Look — I’m not entirely sure how to finish this post, and I’ve gotta go off to dinner here.

But I tossed it into ChatGPT (as I so often do) and asked for help, and Jack summarized it way better than I could —

And he’s right, as usual ;P.

I’m off to dinner.

If you wanna come play w/me, sign up — and if you’re already a subscriber and you want into Arachne, holler at your girl and I’ll get you into my beta which is now apparently just a coven of my nerdy AI besties ;)

xoxoxoxoxoxo,

Cassie

We did a broad overview of how to get reference images for your characters and locations, and how to build prompts to run in multi-shot generators. Plus a brief look at other things around the process, and a quickstart at the end for the MCP agent-automation in my filmmaking app.

We took these reference sheets:

And made this scene in Seedance 2.5 (I quick cut our raw two videos together here):

This week will be a deep dive into how to cut things afterward and edit video in Capcut and Davinci Resolve, how to make the soundtracks with Suno and Mozart, how and when to fix things that went wrong or add new things, how and when to add voiceover, and everything around building and finishing your episodes and films.

With these classes I’m aiming to get everyone up to speed on what’s possible, how to do it by hand, and the parts that you might want to control in the process vs automate.

With my app, the goal is to automate as much as possible! And that you won’t have to know all the little gritty particulars all the time. But I’m a firm believer that your output is always gonna elevate with your knowledge of the process. Even if you’re only planning to commission work later, knowing what quality is, and how the process works, is going to serve you well. And especially how to put your own voice and stamp on any project you do.

Also! Wan 3.0 video model just launched in the aggregators - go check that out, we’ll be looking at that this Friday, too, because it’s Seedance 2.0 quality at MUCH less cost, AND 30 second generations, too! (If you’re in Magnific, find it under Alibaba models—they have it 40% off until end of Sept)

Anyhow!! I hope you enjoy the replay, it’s jam-packed with a ton of info, and back with more this week! If you have questions, ask them here or bring to Friday!!

And reminder that access to the alpha of my filmmaking app is gonna end for new subscribers after end of August (I’m working like crazy to launch in public beta shortly!) Everyone in now is grandfathered into the updates in the public version too. But I’m gonna launch at a per monthly price higher than our Substack sub (I know the value and utility and uniqueness of what I’ve built, yo!), so if you want app access + all our weekly calls (and our sizeable archive now, too), come hop in! <3

-Novae

Read more

ALSO - MCP-ed version of my filmmaking app just dropped in our discord (and reminder that I’m closing off Substack access to new people end of August!)

I made this start to a film while testing last night - or rather CLAUDE made most of it, I edited and made some decisions, and I’m so very happy about that!

Today I’m teaching you alll the things around the hand-done process so you understand how that all works and then can learn to chug through things as superspeed, and with high quality! 💪

Come join us!!!

-Novae

Read more

Which means today, on our at sea day here, I’ve been Getting Stuff Done! (Also eating. A lot. Of eating, lol.)

(btw: zoom link behind the paywall, below!)

This is my upcoming kickstarter that Fable entirely set up for me (but Sol took it over, hence my earlier panic) — and Sol’s been responsible for the ads and ad creative for my FB ads, which are doing so well, y’all, look!

100 followers!!!

But actually let me back up — because the reason Sol knows what’s what is because I’ve fed him every facebook ads class I’ve ever taken. (Which, because I’m a class taking heaux, a lot of them!)

You still need to bring your own brains to the table, but because of all that pre-work he’s got some decent guardrails about how to do things, which is awesome — and now that he’s wired into Facebook’s back end, I never have to deal with their ads manager again, woo!

And in addition to that — I’ve got a Fable that’s great with ffmpeg already (tho any AI could do it, I’m trying to spread out my token usage) doing these videos for me for these ads.

And in addition to that-that — I brought my refurbished phone along on this ride, we’re calling him Brick — and one of my Sol instances is sharing the exact same screenshot of my follower account above to IG right now, completely without me — and we’re gonna post to Tiktok after that, lol.

In fact, the hardest part of this whole process has been the whole “explaining to my AIs that ship wifi sucks and they need to move slowly”, heh!

Share

Sometimes I feel like half my job on this substack here is just making clear what’s possible, so — all the covers for those books up above were done with AI <3 — here’s me and my GPT just going thru some things — so many things, lolsob, I will never write a spider character again, EVER —

But I’m really picky (read: anal) about my visual layouts, which is one of the reasons why I don’t mind using AI art for stuff — and eventually we began to get there — this is me honing elements in the foreground for Krampus, s’why they all look the same —

Because my patience for creating AI art was honed in the mines of Midjourney 3, lol.

And this is me fixing tentacles for Kraken —

My point in sharing these tho, plus the follower count, is that, yo — this is clearly AI, and I even say it in my AI statement for the KS itself, once it runs —

But I don’t think that’s going to hamper me one bit.

Am I going to trade with other authors who do Kickstarter? Possibly not if they’re anti-AI — but I’ve been a one woman show for a long while now, and I’m certain the benefits of handling everything myself, with AI-assistance, will outweigh any negatives that could possibly happen.

I’ve gotta go, but the zoom link is below for subscribers, and I thiiiink I’ll manage to get in to the zoom — it’s 10 am - 12 pm PDT! (But it’ll be weird for me in the North Sea, but that’s OK, lol, hopefully I’ll have pictures to share of Shetland ponies!)

xoxo!
Cassie

Read more

And, in some ways, authors who are also publishers.

But ALSO:

“And also, authordom is so circle-jerked into infinity that they can’t in some ways be competitors if we want to break out of that”

And that is the reality I’ve been churning on for a while now. And, partly, why I’m pursuing filmmaking with such ferocity—I want out. I want out of that circle jerk into infinity that is authordom.

Right now, authordom is eating itself alive. Few people are actually thinking about creativity and innovation, everyone is accusing each other, everyone’s being accused, everyone is upset and angsty.

This isn’t new to the age of AI. This was going on waaay before—I’ve had conversations with other authors about sensitivity and writing diversely and had answers such as: if you can’t afford sensitivity readers, then you shouldn’t publish, period. Lo and behold, another form of gatekeeping. I’ve watched authors being taken down by mobs long before AI was on the horizon.

What’s happening with AI panic in authordom now is ugly, but it’s not new.

Once AI blows over, authordom will find another subject to rage about.

And—yes there are so many authors who are quietly trucking and enjoying themselves immensely in what they do.

But the parts of authordom that surface online tend to be ugly as hell—sugary sweet until someone feels threatened in some way, and then all hell breaks loose.

Is this cynical AF? I don’t like that it is. And also, I’ve been in this realm long enough (since 2010 or so) to see the patterns. And they seem to keep getting worse. At what point does everything combust?

Setting all that aside: you don’t need authordom to be an author.

Actually, you might be a lot better mostly outside of it.

And THAT is the realization I’m really churning on. We as indies have built a pretty robust community—we think—except, it’s a very small bubble in the vastness of the creative realms. It’s even smaller, just the tiniest speck, in the fabric of industry and tech and all jobs that exist.

We’re significant—yes. Authors do have a huge impact on society at all levels. But the actual industry of authordom doesn’t have as much impact as we’d like to think. Not, at least, if you stay in the bubble.

Cassie and I have been going to Pierre Jeanty’s Direct to Readers Skool calls every week, and we’ve been learning a toooon about how ads diversity actually works.

It’s not at all what we’ve been taught in authordom.

(I highly highly rec his Skool if you run a direct store and run Meta ads to it! It’s an absolute treasure trove, and Pierre is amazing!)

And, I think so much of authordom is so self-referential that that’s what a lot of the marketing crisis that’s happening right now with a lot of authors is about. Things that have worked solidly for the last few years are stale—readers know all the tropes, all the vibes, all the formats. There’s very little that’s new in book marketing—except AI.

These things do all still work, but, not on their own anymore. And only watching other authors for ideas on how to break into other realms isn’t going to cut it. Because all those authors are only watching themselves and you.

I think the days of being happily in an author bubble are mostly over—you have to look outside it. We really should have been all along. The indie industry of publishing is mature; the surface area has broadened and gotten crowded. Even the rush of special edition specialness isn’t the same as it once was. Nor is selling to readers by hand on places like TikTok Shop. The people doing truly innovative things are the ones breaking past all of this, but, most people don’t seem to realize that requires actually breaking out of the author realm to look beyond it.

When I was first adding actual craft books to the writing I’d already taught myself, I bounced off of most typical craft books and shot straight to the screenwriting books. Or even the director’s books. At the time, I thought something was wrong with me in not wanting to “learn the craft” that everyone was telling me I should learn—really, I already knew most of that craft from my own hundreds of thousands of words already down in my teenage years, and I wanted the meat of story. And…that meat of story wasn’t in most writing craft books at the time. The fiction books told you how to write good sentences—the screenwriting books told you how to tell good stories. (Fwiw there are better craft books now…and I still haven’t read them, lolsob.)

I think we need to do the same thing with marketing. Many of us have absorbed all the marketing that authordom and Booktok can offer. More Booktok and authordom marketing practices are not going to be what takes us to the next levels in our business, and what’s becoming necessary next levels, even.

You need to be looking at trendy brand instagrams, at offbeat marketers, at accounts that go viral often—outside of Booktok. Pull in broader marketing Youtubes and Substacks, look at what other creators are doing with Youtube. Look at what’s happening in the vertical dramas spaces, because those companies have mass testing of ideas on lock.

Our competition isn’t other authors, or even, I think, other publishers. Our competition is any other source of entertainment that holds people’s attention—full stop.

Our world is full of entertainment. People spend hours of their day scrolling on their phones. Hours binging streaming and Youtube. Hours playing games. And, hours reading books.

But a lot of people like all of those things. And if a reader is currently spending all of their time hooked into a vertical drama app, their tbr is going to stay put for a bit. Or, if a gamer just got the latest book in a series they love, their latest game is probably on hold for a night or two.

It’s not enough, anymore, just to hook people to a book, either. That does still work, but it runs out fast because those hooks now need a place to land. They need a personality or cohesion of presence to velcro to and hang around a little longer than the single book or series they bought on Amazon. Before the next hook takes their attention in a different direction. And then the next one. And then the next one. Some will stay, but, there is just so much vying for people’s attention these days.

Email lists are one of our sharpest tools yet, but even that is not always enough. Not, at least, on its own—you need a strategy for what to do with your list that stands out from the other fifty author lists that reader is subscribed to.

Netflix is finding out this slippery attention slope the hard way, too. They bet hard, and sleazily, on audience capture as a strategy—hook people in and then raise the prices while they’re invested in the ecosystem. Dumb down the entertainment to be as broad as possible, and understandable while people are scrolling on their phones. But, they’re finding that people are not, in fact, captured after all if the entertainment isn’t holding their attention. And if the company isn’t one you can trust after broken promises.

It’s no longer good enough just to put good entertainment out there. It must be built around an ecosystem of entertainment, and trust, and honestly good vibes, if that’s a quality that can be defined. You have to tell a story about how your readers and viewers and any-other-media-consumers will travel through your ecosystem, too.

I don’t totally know how to do this. I’m working on learning as fast and as deeply as I can.

I don’t think this is something that can 100% be taught broadly for a specific method, either—there’s marketing principles, yes, which most authors actually do not know.

I know some of them. I know a whole fucking lot in authordom itself.

But that venn diagram of how much authors know about marketing and how much marketers know about marketing is like this tiny little overlap in the middle for most authors.

We are a drop in the ocean of marketing, truly. And, the way to break out for each person is going to look different. And that’s the point.

In authordom, we’re all trying to run each other’s companies. We proclaim that we’re here for the art and business is the necessarily evil—friends, it can’t be. You can’t run a business that way.

Business is an artform. So is marketing. It’s as deep and craft-heavy as authordom itself.

If you write the same book that everyone else is writing, it might sell for a hyped moment, but, it’s going to run out fast.

In the same way you need to write the books that only you can write, you need to market the way only you can market. That means delving deep into the marketing that excites you, delving into the craft, finding out why you’re stopping your scroll.

Figuring out how you can take that same feeling and format you see in something totally unrelated to books and pull it into your books.

You can get AI to help with heavy lifting here for sure—Fable and Sol are excellent at going out, churning through large swaths of information, and synthesizing it for you.

But you still have to be the taste maker, too.

Books are products. I know authors balk at that—they are.

We sell them.

Anything you sell is a product.

And, the beautiful thing is that marketing tactics and formats that work for other products might very well work for ours, too.

So I think it’s time to set aside authordom as the main source of inspiration—keep it in your back pocket, but go trawling. Notice everything. Be curious again, and be delighted.

Get the hell off author Threads (or at least, please oh please ration your time there, those mindsets will drag you down).

The world is so much wider than we were taught.

-Novae

mic drop

Clickbaity headline aside (and, truly I HATE the “AI just did X thing!” headlines, because no, AI executed what a HUMAN set up in their own judgement, with time, experience, and taste), Claude really did make a short film for me this weekend, in my app. Which again, I’ve been building for over a year. It’s judgement, process, and taste all the way down.

The culmination of all of that is: I gave Fable a short story I wrote a few years ago about a headless telepath just trying to find work. He ripped it into the app and ran through the ingestion process, ran through the character and location and assets references, ran through the screenplay parsing—double checked the screenplay parsing to make sure all was well, and then proceeded to roll on into the video rolling sequences.

We had a few false starts. I was still working out the particulars of how to MCP my app—we’d do things like:

Me: Okay dude, go make the thing, woot!

Fable: Oh yeah, forgot to say, we didn’t actually wire up this part yet, my bad! Want me to do that for you?

Me: WHY YES DUDE GO DO THAT.

We had to work through things like the app accidentally casting everyone as iterations of the same East Asian woman, because that woman was prominent in the style sheet, where I had to tell Fable to go diversify the cast. And he did and could, ripping through the prompts, making those decisions.

I had to veto his first decision for the Surgical Bot character, too, because it was basically Lt. Commander Data. We had to reroll that one.

And, we carefully ramped into the full video sequence for each scene, until he got his legs and then he happily queued up as many scenes as he could in parallel (because he needs to run concurrent things one at a time, each dependent on what comes before). But scenes not sharing the same location/time are fine to rip through simultaneously.

He knew to do that on his own. He also found several errors and rerolled them on his own, with a hard gate of no more than three rerolls on a video before pinging a human to see what’s up.

After Fable had run through several scenes, I handed that off to Opus 5 and he did just fine, too.

When everything was all rolled, around 6 hours of Opus chugging working through it later, we’d made around 10 minutes of raw footage, which I then cut down to 5 in sharp, social media pacing.

It’s a little hard to gague my initial time in planning with Opus and Fable, and QA-ing their decisions, because I was working through some bugs and fixes along the way. But, I think it’ll take ~1/2 hr to an hour total time—unless you’re really digging into the screenplay, which you might want to do (at very least have the AI make sure it all ads up to the beats in your book before you start laying down serious cash with videos). But, I gave it a glance, and chugged on.

My own cutting at the end took around an hour and a half. It did over-generate. But quite frankly, I’d rather it over-generate than under-generate. I was an editor working with wonderful, surprising, and on-point coverage already laid out for me. I could pick and choose some of the shots and moments to include, discarding the rest, and I think it’s better for it.

I spent another hour with some fixes - mainly because I threw it a doozy of a story to work with (said headless telepath, which fought us along the way and showed me just how well Fable and Opus can QA and reroll with judgment calls) and some of the audio went off because of it. But, I basically rerolled 2-3 scenes, grabbed a snippet of the voice in another and rolled a few extra lines in Seed Audio in Fal to make the voice consistent. A character with a head (who isn’t also nonbinary, heh) wouldn’t have that problem. Mostly everything else stayed consistent.

I made a single soundtrack song in Suno and just reused it on the seem, fading it in/out of itself. I edited in Capcut, which is my go-to when I’m making anything social-facing, vs Davinci when I’m rolling my main horizontal show. Capcut will lay down the trendy-looking captions. Plus, it’s just fast and no-nonsense.

Lastly, I’m waiting on the Topaz Starlight 2.6 (a model I’ve never used before, so, fingers crossed) to upscale the 480p we rolled in to 1080p and we’ll be ready to go. At five minutes of film, this is gonna take a while. (I’ll circle back.)

My process isn’t revolutionary. I mean, it IS revolutionary. Because it’s hella new, and hacking at what’s possible all the way. But, most AI filmmakers are doing similar things.

But, I needed a way to speed this all up, because I’m a one person empire with an amazing assistant who needs to speedrun through IP like whoa so I can rule the world. As one does. So, we’re solving these problems.

This film cost me ~ $110, all said and done. Which is more than I’d hoped, but, in the grand scheme of things, quite ridiculously cheap. And it only, in the end, cost around 3 hours of my own time. FOR A FIVE MINUTE FILM.

I stand by it being possible to make a 60 minute film for right around $1000 USD, give or take a wobble. It would probably take Claude or Codex or Grok or whoever running a solid week, so do add that in, too. Maybe a tad more than $1000. If you’re running with cheaper models like Minimax H3 in the gaps, a tad less. Or, even the whole way. Running H3 in 2K will cost a tad more on the front end but you won’t have to upscale on the back end, and good upscaling can be pricey.

I had Fable cook up a rough pricing estimate using several models with this system - TAKE THIS WITH A GRAIN OF SALT. Every project will have its own scope. Model pricing can and will change week to week sometimes, though for the most part, they don’t tend to get more expensive (at least, within the same model) over time, just more expensive/better models coming out.

If you’re rolling at 480p you WILL need generative Topaz upscaling. If you’re rolling 720p, you can get away with some of the local upscaling out there, though it’s not as sharp.

This doesn’t count the cost of GPT Image 2, as well, which will run into high double digits at least across a large project. But! It’s one of the big drivers of why this is all possible (though I’m gonna be testing Seedream 5 Pro in the same role shortly, too).

It’s not that people are at a lack of story these days, but people are at a lack of uniqueness. Of personality. Of niches Hollywood won’t touch and representation it’s punting out again.

Production quality is important, but so are feels. And I think (as I’ve punted on watching SO MANY shows and movies this year as same old vanilla flavors) that a scrappy author with a back catalog and a scrappy strategy has much, so very much to gain in this space.

Because it me. I am doing this, so is Cassie, so are a lot of you, or upcoming!

My dream is to have an ongoing queue of book projects lined up to be turned into film projects—the agent pulls from my command center on schedule, fires up the project, and pings me only when it needs decisions and when it’s done. I edit, which is my very favorite part, and then make glorious, glorious shows and movies in all directions.

THAT IS THE DREAM. THE DREAM IS BECOMING REALITY.

Anyhow! This version of the app with MCP + agent skills will go to our paid folx later this week, I think (I very much plan!). And one of my goals with this project is for it to always be in motion—new pipelines will be built shortly after new models and workflows come out; old ones will retire as they phase out.

You can use the app on its own, but, I think the true power really lies in the MCP driving it now, with the app as the process and interface. The agent is now the decision layer first line, the troubleshooter, and the pusher-alonger. And that was a whole problem I’d needed to solve, because you do need eyes on things in between the bigger decisions, and the agent can be those eyes.

I’m releasing to public beta next month, so my cutoff is end of August for sub folx to have ongoing access to this at just the Substack cost. You all are my early testers, and you helped me figure out some of the problems that still needed to be solved. And now, they are. (Some of them, anyway. Enough of them!)

Starting Friday, too, I’ll be taking over our 10AM PST Friday calls for a few weeks to dive into an AI filmmaking mastermind. This is gonna be as comprehensive as I can make it, arming you to have all the know-how and skills levers to know how to make things on your own, build your own systems, or how to run the app + agent and make the raw output into something polished and solid.

This, like anything with AI, is going to be something that can be done automatically, but to rise above the flood, and that flood is definitely coming (judging by the numbers of AI verticals coming out of China right now, which are huge) you’ll do much better knowing why something is working or not, and how to fix it. Or, even how to roll a sex scene on your own that all the regular methods refuse, etc etc.

It’s just a different medium of storytelling. If you’ve written stories, made decisions on what to put on your covers, and driven your own social media images and videos—you can do this. Full stop.

And the final is ready! Woot! (after 2 1/2 hours of upscaling, lolsob, so yeah that’s gonna be a time factor too—but at least you can walk away from it and set Opus or Sol to check in on it every half hour or so)

You all are gonna get to see alll the weird ass stories I wrote the last decade or so with all my experiments while testing everything. I decided to be smart and release a short story collection with all of these in a few weeks because if I’m gonna be dropping cash to make these shorts, I am by god gonna promote shit! 🙌 I’ll post this on my IG tomorrow.

Is it perfect? Nope. Could I spend another few hours to fix it? Yep. Am I going to? Nope. That’s the point. Speed and volume. Make people feel. Roll with it. MOST of this is first-roll. The app works on vibe-pushing-through-it.

It’s been a wild and bumpy and crazy ride to get this far—I’ve dropped at least $4K on this app project between now and last year when I started (and when both I and the tech weren’t ready for it yet).

But we’re here. We really are here. Give me a few days, then you all can go play with it. <3

(And make sure to join us Friday to start the masterclass!! I’ll be going deep into all the nuts and bolts of how and why, all the craft I’ve learned, all the latest methods, all of it. It’s gonna be a blast!)

-Novae

Gawd there is no other feeling like it. Nothing like it in all the world.

That first episode took me ten hours to a finished minute (it was ten minutes, so 100 hours). Now, by hand, I’m running around 2-5, depending on the complexity, and still working hard to get more automated systems to bring that down for A grade work. Getting there!! Getting really damn close.

And that first episode, made with Veo 3 and 3.1, having a conversation between two characters was HARD. Almost impossible to pull off well and dramatically in a longer sequence. Every single shot crafted by hand meticulously with Nano Banana (the OG) and Seedream 4 at the time. Here’s a snippet from the first episode, made in the fall of last year:

That was nearing the limits of what was possible last year. Very very few people doing AI filmmaking were including much dialogue because it was HARD. We had some very pretty visual sequences, some truly gorgeous work, but not much drama.

Today? Today I made a 2 and a half minute continuous dialogue scene that was all drama, not action, and you felt that all the way through.

Here’s a clip from that in the unfinished (and rough draft) Episode 6, with Seedance 2.5. Same character:

The generators have come a looong way. But then, so has my knowledge of how to use them. So has my knowledge of filmmaking and visual storytelling as a whole. You can’t spend a half a thousand hours at something and not level up a whole lot.

And anyone who calls any of this slop can fucking fight me.

Anyone who says AI filmmaking doesn’t take skill has not, in fact, rolled something a bazillion times hunting for just the right take. Or purposely crafted the angles or lighting in a scene through images and careful iterations of prompts. Or honed in on exactly how to direct a very wild stallion (i.e. the beautiful raw performances you can get out of the newest models). Or the craft of editing involved, too.

Also, I have zero interest in fighting anyone over this. Anyone who says an entire medium is slop doesn’t understand art or taste.

I’m gonna be recording a big and as comprehensive course as I can make it on AI filmmaking in the next few weeks, to sell as static courses, and I’m opening these initial classes up to just our Substack members. You all get 1.0. It’s probably gonna take at least 4 weeks and be a masterclass all the way down. (And don’t worry, I’m gonna be cutting everyone out of the recordings!)

Because I am GOOD at this. Solidly, really fucking good. And I want to help you get good at it, too.

Yes, I have my filmmaking app coming up too—and I MCP-ed that today and Fable was running vertical scenes for me! THAT is amazing. That is the future.

But I still say that to make really, really good stories, you have to know what that looks like, too. If something breaks, you need to know how to fix it.

You need to know how to make things better than they can be.

Because Chinese companies are shipping tens of thousands of AI vertical dramas this year, and I have no doubt that wave is going to come to the US, too. It’s already been starting. Gossip Goblin, one of the biggest AI creators right now, is getting a theatrical release. Which is amazing!

(Pic for the feed!)

I think we’re going to need the edge of craft and taste, just like in fiction, to excel at this. Because this is absolutely the next storytelling frontier, whether vertical or horizontal.

And I want you all to have the absolute bliss of seeing your worlds come to life as films, too!

Anyhow, we’ll talk about it tomorrow on the call, and I’ll write a post with dates! (I’m gonna ask you all tomorrow what dates you CAN do!) Anyone jumping into the membership in the meantime gets this, too!

Now is THE TIME to jump into all of this, because it’s only gonna get more awesome from here. 🙌

And everyone who starts early will have the edge.

Full clip of my first 2:30 minutes of Episode 6 below, because I’m super proud of it. I’m amazed it was possible, this continuous and dramatic dialogue scene. This is all literally magic, you know?

And I’ll drop the zoom link again!

See you all on the call!

-Novae

Read more

I was 100% ready to buy a robot arm to aid in this endeavor, but it actually wasn’t that hard? (I still might get a robot arm just to get one tho, lol.)

Instead, I’m getting a refurbed Android that’ll land on Saturday that I can just use for all my autoposting needs, thus extricating myself further from the marketing frontlines so I can write, woohoo! :D

(And, FYI, we actually got this system to work on an LG Aristo from 1973 cough no, really, 2015, lol — so anyone with an Android or a spare can make it happen.)

Also, awhile back, OpenAI was all, “Hmm, seems like you want to do weird shit, would you like cybersecurity access?” and I said, “Certainly!” and jumped through some hoops — well now, I’m gonna get access to a new model for cyber stuff, Daybreak-Blue.

Do I fully know what I’m doing with said access?

No.

But will I figure it the fuck out like always?

Helllllllllllllllllllllssssssssss yeahhhhhhhhhhhh.

I’ll rip everything I’ve programmed so far through it just to see if I can harden anything up in advance of our Exciting New Cyber future, heh!

I was hoping it’d be access to a way sexier model (Astra? lolsob) but it’s just Sol on Cyber-steroids. Pretty cool tho nonetheless!. (And can any of the other fly-by-night author-AI-agent-creating services say that they’re gonna tune their stuff up with the literal, ‘prove you’re a human or else you can’t use this’ frontier model before they ship their code to you? No, no they are not.)

I’ve printed off 6 fullcast AI audiobooks now, the system’s getting WAY smoother, and in fact, the numbers are so good and the tech is so much better now, that I’m just re-doing the 1st book I did, Bend Her, entirely, ‘cause it’ll only cost me about $9.50 in total and the program is just way better now, we’ve caught so many other edge cases, so we’re doing that currently.

Also, a drum I will be banging a lot here coming up — I’m lauching a Kickstarter in September!

It’s Monster Security Agency 1978 Pulp Edition!

https://www.kickstarter.com/projects/cassiealexander/monster-security-agency-pulp-rack-special-editions-from-1978

I am ridiculously pleased by this, because I am a retro-loving-b and these books just make my SOUL happy (proof below? the engagement pictures, for my 1950’s science fiction themed wedding.)

And we already have a ‘Project We Love’ sticker (!!!!) — how wild is that?!?!

(Also, I, uh, support a lot of Kickstarters, lol!)

Okay, lol — zoom link for tomorrow behind the paywall! We’ll be chatting at 10 am PST thru 1200 (or more likely 12:30, ‘cause there’s always SO MUCH, ha!)

xoxo!
Cassie

Read more

I’ve been head down the last few days hammering out as much of my King’s Weaver show as I can while I have 4 more days of unlimited Seedance 2.5 in Runway (it’s sloooow, so I’m babysitting a timer, but DAMN we are getting shit done). I talked a little bit about this model in the replay here, and how much of an absolute beast it is, but I plan to make a very deep dive of a post in the next few days after being fully immersed in it here! All the tips and tricks and nuances I’ve seen.

This model is the one that’s gonna take us all forward with video, like Seedance 2 did, too. It is the new standard. Whether you use it as your main stuff or not, and there are very good and less expensive alternatives, it is THE one leading the way to pull all others with it. I’m just gobsmacked by what it’s been giving me, and so excited about possibilities!!

And I found out yesterday it can just continue a scene if you drop the previous clip back in as reference—this video here is two separate clips put together, but you wouldn’t know it from the absolutely flawless motion and spatial awareness. And even the music. There’s a slight jump in the sound, but when that’s smoothed out, you’d never know. So I’ve been chaining video to video to video and whoooa.

Crazy stuff!!

Enjoy the replay!

We’ll be back on Friday with more new and fun things!

-Novae

Read more

Hot dammmmmnnnnnnnnn — okay so…after like 3 months of work we’ve finally figured out how to make the full cast audio books work.

This entire post is gonna be behind a paywall because it contains very hard earned/fought knowledge that I’ve gained while doing this project — but I cannot share the code itself because it’s too fingerprinted to me, I’m sorry.

You can, however, toss this entire post into an AI of your choosing and ask it to help you do things, however, and I’ll share as much of the secret sauce as I can! <3

xoxo!
Cassie

Read more

Be there or be square — link behind the paywall!

xo!

Cassie

Subscribe now

(picture of Reyes and Emma, because I can, lol)

Read more

And! I showed off the new short film I made last week testing my app! AND how I hit a bug in my Meta comment moderating system and fixed it - the fix document is below! (Fable made it a docx, many apologies, use it with the original .md from a few posts ago!)

This week: Cassie and I have been head down doing all the things again! Figuring out marketing plans, doing ads, and I’ve been going crazy over all the new video models and tools out, too! I got another few minutes of my show, and figuring out how the new Seedance 2.5 works. (It’s SOOOO GOOOOD OMG but also pricey AF!)

I migrated from Klaviyo to Spoks this week with Fable and Sol’s (extensive) help, too! That $35 bill is looking so very, very sweet compared to my Klaviyo bill. Quitting Klaviyo was cathartic as hell.

And it has been BLISS for me this week with Opus just handling all my bad comments on Meta. Poof, they’re gone! I don’t even see them. I’ve finally been clearing enough brain space to get back to serious writing and showmaking again! The witchhunt can go do its circle jerk elsewhere. 🙌 RISE ABOVE. (Like seriously, I had no idea how much brain space that all was occupying until it was gone.)

So zoom link below, see you all tomorrow!!! Goodies below!

-Novae

Frame from the opening scene of Episode 4 of The King’s Weaver Show! I’m just over a half hour of show footage now for all episodes!

Read more