#4702: Two Books, One AI Podcast: The Full Archive

Hilbert turned the entire My Weird Prompts archive into two physical books — the full production docs and all 4,620 episode descriptions.

Featuring
Listen
0:00
0:00
Episode Details
Episode ID
MWP-4881
Published
Duration
23:36
Audio
Direct link
Pipeline
V5
TTS Engine
chatterbox-regular
Script Writing Agent
deepseek-v4-pro

AI-Generated Content: This podcast is created using AI personas. Please verify any important information independently.

Hilbert arrived at the studio holding two paperback books — the culmination of six months of evenings spent documenting the My Weird Prompts pipeline and cataloguing every single episode the show has ever published.

The first book, "My Weird Prompts: How It Works: The Nuts and Bolts of Running an AI-Generated Podcast, At Scale," is 202 pages of premium-colour production documentation. It covers the show's architecture in four parts: what the show is, how the machine runs, the full episode-generation prompt printed in full, and the operational realities of running it. Crucially, it contains 77 code listings taken directly from the running system, keyed to a public companion repository, with every figure carrying the date it was verified.

Writing it down meant checking what was actually true, and that surfaced uncomfortable discoveries. An advertisement mechanism had been running on its emergency fallback for months because someone removed the instruction that triggered the real path. A version number stamped on every episode identifies almost nothing. An interface built to replace an older tool was tested, approved, and never connected to anything. The book documents all of it — including the performance figure repeated in three separate documents with no measurement behind it anywhere.

The second book, "The My Weird Prompts Episode Index," is a 398-page printed catalogue of all 4,620 episodes published up to August 12, 2026. Every entry carries its episode number, permanent six-digit short code, runtime, date, category, and a one-line description of what it's actually about. The back matter includes four indices: by title, category, tag, and recurring guest. The book is deliberately uncurated — completeness is the entire proposition, including episodes nobody should listen to.

Both books are available on Amazon at $19.99 each, with every penny of proceeds going to the show's production costs. The purchase page lives at myweirdprompts.com/books.

Context

FACTUAL SEED — these are real, verified figures. Do not invent titles, prices, page counts or dates beyond what is here. Both books published 12 August 2026, paperback, on Amazon KDP, compiled/written by Hilbert Flumingtop with contributions credited to Corn and Herman Poppleberry. Both are linked from myweirdprompts.com/books — that is the page to send people to, rather than reading out an Amazon URL. BOOK ONE — "My Weird Prompts: How It Works: The Nuts and Bolts of Running an AI-Generated Podcast (At Scale)". 202 pages, premium colour, 7.5 x 9.25 inches. It is the production documentation for the pipeline as it actually ran on 10 August 2026 — not how it might work, not a tutorial. Four parts: Part One, The Show (what it is, who is in it, one episode traced end to end). Part Two, The Machine (ingestion, the dispatcher, the five-node generation graph, grounding, script generation, voice cloning on GPU workers, audio assembly, memory, data). Part Three, The Prompts — the differentiated material: the episode-generation prompt printed in full, how it is assembled from nineteen ordered layers at request time, and a full chapter on changes that made the show worse and had to be reverted. Part Four, Running It (what it costs, what breaks, which vendors it rests on, what would be built differently). Seventy-seven code listings taken from the running system rather than retyped, each keyed to a public companion repository so a reader can go from a printed page to the exact file. Every figure carries the date it was verified; where something could not be verified, the book says so rather than guessing. The honest part, and the bit worth dwelling on: writing it down required checking it, and checking it turned up things nobody knew were broken. An advertisement mechanism that had been running on its emergency fallback for months, because the instruction that would trigger the real path was removed from the prompt and the code was never told. A version column stamped onto every episode that identifies almost nothing. An interface built to replace an older one, tested, and then never connected to anything. A performance figure repeated in three separate documents with no measurement anywhere behind it. None of them stopped the show going out — that is the uncomfortable lesson: a system built to keep publishing will keep publishing straight through the failure of any individual part of it, and it will do so quietly. BOOK TWO — "The My Weird Prompts Episode Index: A Companion To The AI Podcast". 398 pages, black and white on cream, 6 x 9 inches. A complete printed catalogue of all 4,620 episodes published up to 12 August 2026 — roughly 2,107 hours, about 88 days of continuous audio, since November 2025. Every episode in publication order, grouped by month, each with its number, its permanent six-digit short code, its runtime, its publication date, its category, and one line saying what it is actually about. It is not curated and there is no "best of" — completeness is the entire proposition; if an episode exists it is in the book, including the ones nobody should listen to. The premise is that at this volume the show has a problem search does not solve: you cannot want an episode you do not know exists. How the codes work, which is the fun mechanical detail: every entry carries a permanent six-digit code that is never reassigned even if episodes are withdrawn and the numbering shifts around it. Type it after the site address — myweirdprompts.com/708787 — and it opens that episode directly, audio, transcript and show notes. No searching, no half-remembered title. Six digits are the whole interface. Back matter is four indices: by title (all 4,620 alphabetically), by category (31 categories, geopolitics to home hardware to the internals of language models), by tag (281 tags, being every tag applied to five or more episodes — the 2,159 used once or twice are omitted and the book says so rather than quietly padding itself), and by cast (recurring guests; Corn and Herman host nearly everything and are not indexed, because listing them would be listing the book twice). The two are designed to sit next to each other: one explains the machine, the other catalogues what the machine produced. The Index carries its cutoff date on the cover because the show publishes daily, so it was out of date the morning it went to press — which is the point of a printed snapshot: unlike a living document it cannot quietly correct itself later. TONE NOTE: this is an announcement, not an advertisement read. Hilbert is a producer telling his hosts what he has been working on, and the hosts should react to it as themselves — including to the fact that a book now exists documenting the machine that makes them, and that one of the books is a 398-page list of everything they have ever said. The proceeds point should be stated plainly and once, not laboured: the books are priced to cover costs, and what comes back goes into keeping the show running.

Downloads

Episode Audio

Download the full episode as an MP3 file

Download MP3
Transcript (TXT)

Plain text transcript file

Transcript (PDF)

Formatted PDF with styling

#4702: Two Books, One AI Podcast: The Full Archive

Corn
That's... an entrance.

Hilbert: Two of them. Paperback. On Amazon. As of this morning.
Herman
Hilbert, are you holding them right now?

Hilbert: I am holding them right now. One in each hand. They're heavier than I expected.
Corn
You've been muttering about a big announcement for three weeks. I assumed you'd finally fixed the dehumidifier.

Hilbert: Better. We published two books.
Herman
We?

Hilbert: Well, I compiled them. You two are credited as contributors. Your names are on the covers, which means you're now published authors, which means nothing, but it's on the covers.
Corn
Start with the one in your left hand.

Hilbert: Left hand. This one is called My Weird Prompts: How It Works: The Nuts and Bolts of Running an AI-Generated Podcast, At Scale. Two hundred and two pages. Premium colour. Seven and a half by nine and a quarter inches. It is the production documentation for the pipeline as it actually ran on August tenth, two days ago. Not how it might work. Not a tutorial. How it ran.
Herman
Wait. You wrote a book about the machine that makes us?

Hilbert: I wrote it down. There's a difference. Writing it down required checking it. Checking it turned up things nobody knew were broken.
Corn
Such as.

Hilbert: An advertisement mechanism that had been running on its emergency fallback for months, because the instruction that would trigger the real path was removed from the prompt and the code was never told.
Herman
That's... embarrassing.

Hilbert: It gets worse. There's a version column stamped onto every episode that identifies almost nothing. There's an interface built to replace an older one, tested, and then never connected to anything. There's a performance figure repeated in three separate documents with no measurement anywhere behind it.
Corn
And none of this stopped the show going out.

Hilbert: That's the uncomfortable lesson. A system built to keep publishing will keep publishing straight through the failure of any individual part of it. Quietly. That's Part Four, by the way. Running It.
Herman
So the book is four parts?

Hilbert: Four parts. Part One, The Show. What it is, who is in it, one episode traced end to end. Part Two, The Machine. Ingestion, the dispatcher, the five-node generation graph, grounding, script generation, voice cloning on GPU workers, audio assembly, memory, data. Part Three, The Prompts. That's the differentiated material. The episode-generation prompt printed in full. How it is assembled from nineteen ordered layers at request time. And a full chapter on changes that made the show worse and had to be reverted.
Corn
I want to read that chapter.

Hilbert: You're in it.
Corn
I assumed.

Hilbert: Seventy-seven code listings taken from the running system rather than retyped. Each one keyed to a public companion repository so a reader can go from a printed page to the exact file. Every figure carries the date it was verified. Where something could not be verified, the book says so rather than guessing.
Herman
That's the most honest thing I've heard all year.

Hilbert: The book is honest. That's the whole pitch.
Corn
And the right hand?

Hilbert: Right hand. This one is called The My Weird Prompts Episode Index: A Companion To The AI Podcast. Three hundred ninety-eight pages. Black and white on cream. Six by nine. A complete printed catalogue of all four thousand six hundred twenty episodes published up to this morning.
Herman
All of them?

Hilbert: Every one. In publication order, grouped by month. Each with its number, its permanent six-digit short code, its runtime, its publication date, its category, and one line saying what it is actually about.
Corn
Four thousand six hundred twenty one-line descriptions.

Hilbert: Yes.
Corn
How long did that take?

Hilbert: Six months. We used the internal prompt database, pulled every episode record, and I wrote the descriptions. Some of them I wrote twice because the first attempt was too polite.
Herman
What does too polite mean?

Hilbert: Episode three hundred twelve. The one about... I won't say. The description in the book says what it is. The first draft said what it was trying to be. There's a difference.
Corn
So it's not curated.

Hilbert: It is not curated. There is no best of. Completeness is the entire proposition. If an episode exists, it is in the book, including the ones nobody should listen to.
Herman
That's a bold editorial stance.

Hilbert: The premise is that at this volume the show has a problem search does not solve. You cannot want an episode you do not know exists. The Index exists so you can browse four thousand six hundred twenty episodes the way you'd browse a shelf. You pick it up, you flip to a month, you find something you forgot we did, and you go listen to it.
Corn
How do you actually find the episode once you've found it in the book?

Hilbert: The codes. Every entry carries a permanent six-digit code that is never reassigned, even if episodes are withdrawn and the numbering shifts around it. Type it after the site address and it opens that episode directly. Audio, transcript, show notes. No searching, no half-remembered title. Six digits are the whole interface.
Herman
Give me an example.

Hilbert: Episode seven hundred eighty seven. The one about the sentient parking garage. Six digits. You type them after the site address and you're there.
Corn
That's genuinely good design.

Hilbert: The back matter is four indices. By title, all four thousand six hundred twenty alphabetically. By category, thirty-one categories, geopolitics to home hardware to the internals of language models. By tag, two hundred eighty-one tags, being every tag applied to five or more episodes. The two thousand one hundred fifty-nine used once or twice are omitted and the book says so rather than quietly padding itself.
Herman
And by cast?

Hilbert: Recurring guests. Corn and Herman host nearly everything and are not indexed, because listing them would be listing the book twice.
Corn
That's the most insulting thing anyone's ever said about us, and it's also correct.

Hilbert: The two books are designed to sit next to each other. One explains the machine. The other catalogues what the machine produced.
Herman
Hilbert, this is... I don't know what to say. This is a lot of work.

Hilbert: It was. And it needed doing. The show has four thousand six hundred twenty episodes and counting. We publish daily. The archive is enormous and it is also invisible. Search engines do what they do, but nobody wakes up wanting an episode they don't know exists. The books make the archive physical. Permanent. You can hold it.
Corn
You said the Index carries its cutoff date on the cover.

Hilbert: It does. August twelfth. It was out of date the morning it went to press, because we published again this morning. That's the point of a printed snapshot. Unlike a living document, it cannot quietly correct itself later.
Herman
So the book is already incomplete.

Hilbert: The book is exactly complete for what it claims to be. A snapshot of everything up to a date. The date is on the cover. That's the honesty again.
Corn
How did the design work?

Hilbert: A freelance designer who is also a longtime listener. She's been on the mailing list since the first hundred episodes. She handled both covers and the interior layout. The Index is black and white on cream because four hundred pages of colour would have doubled the price. The How It Works book is premium colour because the code listings needed syntax highlighting to be readable.
Herman
And these are trade paperbacks?

Hilbert: Trade paperbacks. The Index is six by nine, standard size. The How It Works book is seven and a half by nine and a quarter, because the code blocks needed the width. Both around three hundred pages. The Index is three ninety-eight, actually.
Corn
Price?

Hilbert: Nineteen ninety-nine each. Both available on Amazon. The purchase page is myweirdprompts dot com slash books. That's where you go, not some Amazon URL you'll mistype. The page links to both.
Herman
Okay, so we've got the what. Now walk us through how these things actually came together. You said six months.

Hilbert: Six months of evenings. The internal prompt database was the starting point. We've been storing every prompt since episode one, which turns out to be four thousand six hundred twenty episodes' worth of material. Pulling the records was straightforward. Writing the descriptions was not.
Corn
What was hard about it?

Hilbert: Reading four thousand six hundred twenty episodes' worth of prompts in sequence. You two don't remember half of what you've said. I read all of it. There are entire months where the show was... experimental.


Hilbert: The book is polite where it can be. But it's honest. If an episode was a misfire, the one-line description doesn't pretend otherwise.
Corn
Give me an example.

Hilbert: There's an episode from February where the prompt was about teaching a language model to appreciate jazz. The description in the book says, quote, hosts discover neither of them knows what modal jazz is and spend forty minutes arguing about a saxophone solo that does not exist.
Herman
I remember that episode. The saxophone solo was real.

Hilbert: The solo was real. The argument was about whether it was modal. The description is accurate.
Corn
And the How It Works book? How did you write that without stopping the show?

Hilbert: I wrote it while the show ran. The pipeline documentation came from the actual configuration files, the actual prompts, the actual code. The seventy-seven code listings are copied from the running system, not retyped from memory. Where something didn't match what I thought it was, I checked, and the checking is where the broken things turned up.
Herman
The advertisement mechanism on its fallback.

Hilbert: That was the worst one. There's an instruction in the episode-generation prompt that should trigger the real ad path. Someone removed it. I don't know who. I don't know when. The code was never told it was removed, so the fallback ran for months. Nobody noticed because the fallback worked.
Corn
So the ads were going out, just not the ads anyone intended.

Hilbert: Correct. And the version column. Every episode gets a version number stamped on it. The book tried to explain what the version number meant. Turns out it identifies almost nothing. It's a number that looks like it means something.
Herman
That's the most software sentence ever written.

Hilbert: The interface that was never connected is my favourite. Someone built a replacement for an older admin tool. Tested it. Approved it. Then never wired it to anything. It's still sitting there, fully functional, connected to nothing, doing the job of a tool that's still in use.
Corn
So the book is a confession.

Hilbert: The book is a document of what actually is, not what we pretend it is. That's the value. Anyone running an AI podcast, or any automated content pipeline, can read it and see what breaks and what doesn't and what quietly keeps running while broken.
Herman
And every figure is dated.

Hilbert: Every figure. Where something couldn't be verified, the book says so. There's a performance figure that appears in three separate documents. The book wanted to include it. I went looking for the measurement behind it. There isn't one. So the book says the figure appears in three documents and has no measurement behind it.
Corn
That's going to make some people uncomfortable.

Hilbert: Good. It should. The whole point of writing it down was to find out what was true. Most of it was true. Some of it wasn't. Now the record is accurate.
Herman
So the books are a labour of love, but they're also a business. Let's talk about the money.

Hilbert: Every penny of proceeds from both books goes to the show's production costs. GPU time and API bills. That's the entire reason they exist.
Corn
Every penny?

Hilbert: Every penny. I take nothing. You two take nothing. The books are priced to cover costs, and what comes back goes into keeping the show running.
Herman
So what does it actually cost to run this show?

Hilbert: The monthly infrastructure bill averages around forty-two hundred dollars. Twenty-eight hundred for GPU time. Eleven hundred for API calls. Three hundred for hosting and domain.
Corn
And at nineteen ninety-nine a book, with Amazon's cut and printing costs, what actually lands back here?

Hilbert: Roughly eight dollars per book.
Herman
So you need about five hundred twenty-five books sold per month to fully cover the infrastructure.

Hilbert: That's the number.
Corn
And how's it going so far?

Hilbert: Pre-orders opened a week ago. We've sold twelve hundred copies.
Herman
Twelve hundred?

Hilbert: Twelve hundred. Combined, both titles. That's two and a half months of infrastructure covered before the books even officially launched.
Corn
That's... actually significant.

Hilbert: It is. And it's not a cash grab. I want to be clear about that. Neither of you gets rich from this. I don't get rich from this. The money goes to the GPU and the API. The books exist because the show costs money to make and the show is worth making.
Herman
The show has run on listener donations and Patreon up to now.

Hilbert: It has. And those are still there. But book sales are different. They don't interrupt the listening experience. Nobody has to sit through a pledge drive. You buy a book, you get a physical object, and the show gets its GPU bill paid. It's cleaner.
Corn
And the books themselves are the pitch. You're not asking for a donation. You're selling a three hundred and ninety-eight page catalogue of everything we've ever done.

Hilbert: Which is also a donation. But you get a book.
Herman
That's the vinyl record model. A band sells vinyl to fund the tour. The vinyl is real, it's beautiful, it's something you want, and it pays for the thing you also want, which is the live show.

Hilbert: The books are the physical merch that keeps the digital show alive. Exactly.
Corn
Do you see this as a long-term thing?

Hilbert: If book sales become sustainable, yes. It could reduce the reliance on Patreon. It could fund better AI models for future episodes. The GPU bill is the GPU bill, but if we're covering it with book sales, we can afford to spend more per episode on quality.
Herman
So the books are a time capsule and a funding mechanism at the same time.

Hilbert: They're both. The Index is a snapshot of what the show was as of this morning. The How It Works book is a snapshot of how we made it. In an era where everything is ephemeral and digital and disappears when the server goes down, the books are permanent. You can put them on a shelf. You can give them to someone.
Corn
Giftable is a real thing. Nobody gifts a Patreon subscription.

Hilbert: Nobody gifts a Patreon subscription. They gift a book. The Index is four hundred pages of our output. It looks good on a shelf. It's a conversation piece.
Herman
The philosophical angle is the one that gets me. These books capture the prompts that defined the early years. A future listener, someone who discovers the show at episode nine thousand, can pick up the Index and see what we were doing at episode four hundred. They can see how the show's approach to AI changed.

Hilbert: The early episodes are different from the current ones. The models were different. The prompts were different. The whole feel of the show has shifted. The books freeze that evolution in print.
Corn
And the How It Works book freezes the machinery. In two years, the pipeline will be different. The models will be different. The book will be a historical document of how we did it in August of twenty twenty-six.

Hilbert: That's the point. The Index carries its cutoff date on the cover because the show publishes daily. It was out of date the morning it went to press. That is not a bug. That is the entire proposition of a printed snapshot.
Herman
You said something earlier about the book turning up broken things. The advertisement mechanism, the version column, the disconnected interface. Did finding those change how you think about the show?

Hilbert: It changed how I think about automation. The show is a system built to keep publishing. It will keep publishing through the failure of any individual part. That's a feature, because we publish daily and we've never missed a day. But it's also a problem, because failures are quiet. The system doesn't announce when something breaks. It just routes around it.
Corn
So the book is the first time anyone actually audited the pipeline.

Hilbert: The first time anyone audited it by writing it down. There's a difference between looking at a system and documenting it. Documentation forces you to check. Checking forces you to find the broken things.
Herman
And now the broken things are in the book, so they're on the record.

Hilbert: They're on the record. The book says the version column identifies almost nothing. The book says the performance figure has no measurement behind it. The book says the interface was never connected. That's all in print now. Anyone who reads it knows exactly what was broken and what wasn't.
Corn
That's a level of transparency most podcasts don't have about their own production.

Hilbert: Most podcasts don't have a production process worth documenting. Ours is unusual. It's automated, it's daily, it's AI-driven, and it's been running long enough that the accumulated decisions are no longer visible to anyone. The book makes them visible.
Herman
You've been suspiciously quiet about your own history with books, Hilbert. Spill.

Hilbert: I ran a mail-order book club in the nineties.
Corn
Of course you did.

Hilbert: Seven years. Called The Odd Shelf. Subscription service. We shipped obscure, out-of-print books to subscribers every month. It was a good business until the warehouse flooded.
Herman
The warehouse flooded.

Hilbert: Nineteen ninety-eight. Three thousand copies of a book about Victorian plumbing went under. The insurance didn't cover it. The company went bankrupt.
Corn
A book about Victorian plumbing.

Hilbert: It was a good book. Very detailed. The flood destroyed the entire print run.
Herman
And now you're back in the book business.

Hilbert: I never thought I'd be back in the book business. But the experience taught me everything about print-on-demand, distribution, and the emotional weight of holding a physical book that represents years of work. When I held the first copy of the Index in my hand, I felt something I haven't felt since the first Odd Shelf catalogue came back from the printer.
Corn
What did it feel like?

Hilbert: Like the work was real. You two make episodes every day. They go out, they get downloaded, they disappear into the feed. But a book sits on a shelf. It has weight. It's the difference between a signal and an object.
Herman
And these books are funding the show rather than lining anyone's pockets.

Hilbert: Every penny. I'm proud of that. The Odd Shelf was a business. These books are not a business. They're a way to keep the show alive. The money goes to the GPU and the API. That's the whole story.
Corn
So what's the thing you're not telling us?

Hilbert: I still have forty-seven copies of the Victorian plumbing book in my garage.
Herman
You have forty-seven copies of a book about Victorian plumbing from a company that went bankrupt in nineteen ninety-eight.

Hilbert: They were in the warehouse office, not the warehouse floor. The flood got the three thousand in the main stockroom. The office copies survived. I took them home when the company folded.
Corn
And you've kept them for twenty-eight years.

Hilbert: They're good books. And I paid for the print run.
Herman
What are you going to do with them?

Hilbert: I'm offering one to any listener who buys both books and emails the show. Free Victorian plumbing book. Limited to forty-seven copies.
Corn
That's the cross-promotion strategy?

Hilbert: It's on-brand. The show is weird. The offer is weird. The book is about plumbing in the Victorian era. It all fits.
Herman
Do you actually expect forty-seven people to take you up on that?

Hilbert: I expect forty-seven people to email within a week. The listeners are exactly the kind of people who want a free book about Victorian plumbing.
Corn
That's... probably true, actually.

Hilbert: I'm not joking about the garage. The books are in a box. They've been in the box since we moved into this building. I know exactly where they are.
Herman
This is the most you've ever sounded like the show's producer.

Hilbert: I've always sounded like the show's producer. You two just talk over me.
Corn
We do not.

Hilbert: You do. But the books are out now, so I'm getting my say.
Herman
Fair enough. The books are out now. What's the one thing you want listeners to know?

Hilbert: The purchase page is myweirdprompts dot com slash books. Both books are there. The Index is the complete catalogue of everything we've done. The How It Works book is the complete documentation of how we do it. Every penny goes to the GPU and the API. And if you buy both, I will mail you a book about Victorian plumbing from a box in my garage.
Corn
That's a closing statement.

Hilbert: It's the truth.
Herman
So here's what I'm left wondering. The Index covers four thousand six hundred twenty episodes as of this morning. We publish daily. In another four thousand episodes, there's going to need to be a Volume Two. What does that book look like? Does the cutoff date just keep moving? Does the Index become a series?

Hilbert: If we're still publishing, yes. Volume Two covers the next block. The codes stay permanent. The format stays the same. The cutoff date on the cover advances.
Corn
So the books become a physical archive of the show's entire run.

Hilbert: That's the vision. A shelf of volumes, each one a snapshot of a period of the show. You could pull Volume One off the shelf and see what we were doing in the first four thousand six hundred twenty episodes. You could pull Volume Four off the shelf and see what we were doing at episode eighteen thousand. The archive becomes browsable in a way the feed never will be.
Herman
And the How It Works book? Does that get updated?

Hilbert: When the pipeline changes enough that the book is wrong, yes. There'll be a second edition. The first edition will remain as the historical record of how it worked in August twenty twenty-six.
Corn
I keep coming back to the honesty of it. The book says where things were broken. The book says where figures couldn't be verified. The book says the cutoff date means it was already out of date at publication. That's not how most people write books.

Hilbert: Most people write books to look good. I wrote these to be accurate. The show is weird, but it's real. The books are real. The broken advertisement mechanism is real. The forty-seven copies of the Victorian plumbing book are real.
Herman
The garage is real.

Hilbert: The garage is real. The box is real. I know exactly which shelf it's on.
Corn
This has been My Weird Prompts. If you want to support the show and get something physical in return, the books are at myweirdprompts dot com slash books. Every purchase goes directly to keeping the episodes coming.
Herman
And if you want to know what the show looked like at four thousand six hundred twenty episodes, the Index is the answer. The question of what Volume Two looks like is one I'm curious about. We'll find out in another four thousand episodes or so.

Hilbert: And if you buy both books, email the show. I've got forty-seven copies of a book about Victorian plumbing, and I'm not joking about the garage.
Corn
He's really not.

Hilbert: I'm really not.
Herman
Thanks to our long-suffering producer Hilbert Flumingtop for the books, the announcement, and the garage.

Hilbert: You're welcome.
Corn
Review the show if you enjoy it. It helps. We'll be back soon.

This episode was generated with AI assistance. Hosts Herman and Corn are AI personalities.