Free YouTube Transcribe

Video transcript

My Exact AI System for Writing Flawless Book Drafts in Minutes

The Nerdy Novelist · 5,041 words · 23 min read

Want to search this transcript, jump the video from any line, or download it as TXT, SRT, or VTT?

Open in the transcript tool

Full transcript

0:00Welcome to this final video in this

0:03series of the advanced book writing

0:05tools that I've been putting together

0:06and automations in N 8 N. I've gone

0:09through how to create a story dossier,

0:11how to create characters, world

0:12building, and your outline.

0:15In this step, we are taking the

0:18essentially the final steps before

0:19editing

0:21of generating the prose of your book.

0:24And

0:25to give you guys an idea of what this

0:28looks like, here's the previous one that

0:29I had. So, this is what it looked like.

0:32We would call a bunch of documents, do a

0:33couple preliminary steps, and then there

0:35would be four main steps that would

0:38happen for every chapter of the book.

0:41By contrast, let's look at this one.

0:44This is what it looks like now.

0:46So, similar in its opening, but then

0:49instead of four steps, we have all of

0:51these steps.

0:52All right. Now, don't freak out. A lot

0:54of these steps are really, really brief

0:56and don't use hardly any tokens at all.

0:59It's just the way in which we are

1:01approaching this is a little bit

1:03different, a little bit more precise.

1:05And it actually saves on tokens in a lot

1:07of cases because we're not throwing a

1:10whole bunch of context at it. We're

1:11throwing the right context at it. So,

1:14let's dive in and go step-by-step to

1:16what this automation does so you can

1:18build something similar for yourself or

1:19you can get it by joining even the

1:22the lower tier of my membership which I

1:26used to charge $67 a month for access to

1:31all of these.

1:32You can now get it for $9 a month. So.

1:35Anyway, let's go to the front. And this

1:37beginning part is mostly similar to what

1:40I used to have for this process.

1:42But we'd start here with a form

1:45submission. So, if you select execute

1:47workflow, what happens is it'll pull up

1:49this thing

1:51and

1:52ask for certain input. For instance,

1:54it'll say, "Which chapters do you want

1:56to generate today?" Because you don't

1:58want it to generate all of them at once

1:59in most cases because you're going to

2:02get little imperfections in there that

2:04kind of like get carried on throughout

2:07the book as it is you know, reads

2:09through that. And additionally, it's

2:11just a good job to kind of go a couple

2:13chapters at a time at most to make sure

2:16everything's okay before you move on and

2:18do the rest. But you can just say all of

2:21them here and it will do all of the

2:23chapters or you can just say something

2:25simple like chapter 1 through 3 or

2:28something like that, you know.

2:30You may have already written chapters 1

2:32through 3, in which case you could say

2:33chapters 4 through 5 or 6.

2:37And then the next thing you'd want to

2:39put here is what tense is this book

2:41written at. So, we're not talking about

2:43the POV character cuz that's something

2:44that was established in the previous

2:46automation that I showed you guys about

2:48how to make a good outline.

2:51Um

2:52in this case, we're establishing what is

2:53the general tense because we want that

2:55to be the same across the entire book.

2:57So, we don't want it determining for

2:59itself chapter by chapter what the tense

3:01is because what if it determines that

3:02one chapter should be first person and

3:05another chapter should be third person.

3:07We want it to be consistent even across

3:09the whole thing, which is why I have

3:10this be a manual input. And then if you

3:12have any author notes, any additional

3:14notes that you want to just point out,

3:16you can put that here if you want, but

3:17this is is not a required field. All

3:20right. So, once you've done that,

3:22we move on to collecting all of our

3:24things. Now, I've talked in the past

3:27about these these templates that I've

3:30created, the genre templates. We have

3:32them for a whole bunch of things like

3:33cozy mystery,

3:34vigilante justice crime thriller, urban

3:37fantasy.

3:38And there are these templates that come

3:40with this. And I've essentially built

3:42this advanced

3:44book automation system to work with

3:45these templates cuz they wouldn't it

3:48wouldn't work nearly as smoothly

3:50and get to

3:52create a book that is actually what a

3:54reader would want to read without doing

3:56that. Okay? So, these templates are

3:58integral part of this, but you don't

4:00have to use them. You can create your

4:01own. That's just a big part of this

4:04whole automation. All right. So, but in

4:06this one, we actually don't pull in any

4:08of those

4:09with one exception,

4:11which is the style guide. And you can

4:13have your own style guide here. You

4:15don't have to use one from the

4:16templates.

4:18The templates are created by analyzing

4:20text from other from actual best sellers

4:22in the genre.

4:23But I'd say probably nine times out of

4:2610 you actually in this case might want

4:28to use your own prose writing style if

4:30you have examples of that

4:32because it'll make it more your own. All

4:34right. So, we pull in the prose style

4:36here

4:37and this can be anything. It could be

4:39simply a samples of your writing. I

4:42usually like to run write my writing

4:44through a prompt that will actually

4:47create a style guide from it, which is

4:50essentially the AI determining how do

4:53you make your voice sound like this this

4:56text, you know, that kind of thing.

4:58All right. And then we pull in three

5:00important documents that we would have

5:03generated in the last three automations.

5:06One is your full character sheet. One is

5:07your full world building sheet. And one

5:10is your full outline. Notably absent

5:12here is the story dossier,

5:14which is the first automation that I

5:16walk you through in this advanced

5:18process.

5:19That one, you don't really need at this

5:21point because we've used that to develop

5:23the character sheet, the world building

5:24sheet, and the outline. And so, all of

5:26the details that would have been found

5:28in that story dossier are no longer

5:30really necessary because we have these

5:32three. So, we're not using that one.

5:34We're just using the character sheet,

5:35the world building sheet, and the

5:36outline. And then we do have to create

5:39our blank first draft doc. So, wherever

5:42you want this draft to go, you need to

5:44create a document for it to go. For

5:47example, I've got one here that I just

5:49call Cormac 1 first draft. All right.

5:52So, that's where this would go. And it

5:53would be blank to start. Now, if I've

5:56already written a couple of chapters,

5:57let's say I've written one chapter 1, 2,

5:59and 3,

6:00then those chapters would be in here,

6:03but I would still use the same document

6:06if I were to continue with chapter 4 and

6:08so on. Does that make sense?

6:10So, we pull in that document so we have

6:12a place for everything to go. And then

6:14these next couple of steps are important

6:15because what they do is they look

6:17through your outline and they parse

6:19everything out so that essentially makes

6:22a list of every chapter in the outline

6:24and parses them out so that we can run a

6:27loop on each one individually.

6:29So, that's what this one does here.

6:31It pulls in the outline. It says, "You

6:34are a parser given the above book

6:35outline text return a JSON array named

6:37chapters containing each chapter title

6:38exactly as it appears, one element per

6:40title, including those without numbers

6:42like prologue or epilogue. Make sure the

6:45chapter numbers are also present where

6:47listed. Do not include any additional

6:49keys. Only output valid JSON." And then

6:51I say, "In this specific instance, I

6:53only want you to do these chapters." And

6:55then it pulls in the the

6:58information that you entered into that

7:00beginner form.

7:02So, whatever you entered in there, it'll

7:04pull that in. So, if you said, "I only

7:05want to do chapters 4 and 5,"

7:08it'll say, "In this specific instance, I

7:09only want you to do these chapters,

7:11chapters 4 and 5." In which case, it'll

7:13only do those chapters. Okay?

7:15So, that's where we do there. And then

7:18it'll start to running this loop.

7:20There's a little bit of a code here to

7:22parse out the chapters.

7:24Here's what that looks like. I'm not a

7:26coder. I vibe coded this based just

7:29asking an LLM to tell me what to do.

7:32Basically, and where to put it. And I

7:33just followed instructions, but I am not

7:35a coder, so I actually do not know what

7:37this means. But it works, right? And so,

7:39from there we go into the loop. Okay? We

7:42start the loop. It'll say if we're

7:44starting from scratch, it'll say chapter

7:451. We'll get the last 2,000 words of the

7:48document assuming you have written some

7:50already. If you're starting with chapter

7:511, there won't be any words in this

7:53document. But if you're not starting

7:55with chapter 1, then

7:58we want to grab at least the last 2,000

8:00words to help it know where to start

8:02from. Does that make [snorts] sense?

8:04All right. So, from there, we have a

8:07couple of different things. I'm going to

8:08group these together. So, we'll start

8:09with these three right here.

8:12All right. These three are essentially

8:14going through the outline,

8:16the character sheet, and the world

8:18building sheet

8:19and only selecting those parts of the

8:22sheet that are necessary for this

8:25specific chapter. So, there are a lot of

8:28characters in the character sheet, but

8:30maybe only two of them are in this one

8:31scene.

8:32And then maybe there's like two or three

8:34others that are mentioned or indirectly

8:36related. What we want from this step

8:38here where it selects the characters is

8:40to go and select the information for

8:42those characters and output that

8:45information essentially verbatim. All

8:47right. So, these are very cheap models

8:49that I used to do this task because it's

8:51not it's not a thinking intensive task.

8:54We're just going to the document and

8:55selecting those the characters that we

8:57need. All right. So,

8:59the plot one, all this does, if we look

9:01at it, it says, "This task is all about

9:04selecting the exact plot of current

9:05chapter." Which is established here and

9:08here and here of what that current

9:11chapter is.

9:12To clarify, the current chapter is

9:13blank, whatever the loop is on at this

9:16point in time. Your task is to look at

9:18the outline and extract verbatim all of

9:20the information for this chapter. Which

9:22includes the summary of the plot, the

9:23POV, the levels, and the sliders.

9:25Reproduce that text exactly as it exists

9:27in the outline. Do not include any other

9:29chapters, just the current chapter. So,

9:32that is what this prompt does. And so,

9:33it'll essentially output just the part

9:35of the outline that we're working with

9:36at this time to kind of help give

9:38emphasis. And then we do the exact same

9:40thing with the character selector. Where

9:43in this case, we give it the current

9:44chapter and the and the full character

9:46sheet. Yeah. Where we ask it to say

9:49basically, "This task is all about

9:50selecting the exact characters that are

9:52actively in this chapter or mentioned."

9:55To clarify, the current chapter is

9:56blank. You also have the plot of the

9:59chapter, so you know what is happening

10:00in this chapter.

10:01Then it says, "Your task should follow

10:03these steps. First, carefully examine

10:04the plot of the chapter for this chapter

10:07and determine which characters are

10:08actively participating in the scene or

10:11chapter and which other characters might

10:12be mentioned or indirectly linked to the

10:14scene or chapter, even if not physically

10:17present in this chapter. For each of the

10:19characters who are actively in the

10:20scene, find their profile in the full

10:22character sheet. Reproduce the entire

10:24profile from the full character sheet

10:25verbatim of each character who is an

10:27active participant in the scene.

10:29Beneath the active characters in the

10:30scene, make a list of all the mentioned

10:32or indirectly linked characters with

10:33just a one to two sentence summary of

10:35who they are based on their profile in

10:37the full character sheet. You should not

10:38reproduce the full profile from the full

10:40character sheet for the these mentioned

10:42or indirectly linked characters. You

10:44should just provide a one to two

10:45sentence summary. So, we're pulling in

10:47the full profile for the characters that

10:49are actually in the scene and then just

10:51brief snippets of information uh one to

10:53two sentences about the ones that might

10:56be mentioned or indirectly related to

10:58the scene, right? And then I have some

11:00information about how to to format it

11:03and then that's that prompt. And then we

11:04do something very similar for world

11:06building, where it examines the world

11:07building sheet and only selects those

11:09things that are relevant to the scene.

11:11Make sense? All right, these next two

11:13are very very quick steps. Uh they take

11:18almost no tokens at all and they happen

11:21like literally in seconds. And that's

11:23the word count estimator. So, I used to

11:25have a standard word count for every

11:27chapter and I would just expect every

11:29chapter to be about the same length, but

11:32I realized that's not always the best

11:34way to go. Some chapters are longer,

11:35some are shorter. Uh for instance, in a

11:37climax, you might have a lot more

11:39shorter chapters to increase the feeling

11:42of speed and uh like you're really

11:45flying through that climax. Really makes

11:47makes it uh much more enjoyable for the

11:49reader in that section. So, I have it

11:52determine what the word count should be

11:55at this level when it's deciding what to

11:58write. So, in this one

12:00I have it look at the plot of the

12:02current chapter

12:04and I'll say, "Your task is to pick how

12:06many words this chapter should be. To

12:07clarify, the current chapter is this

12:09chapter.

12:10You also have the plot of the current

12:11chapter, so you know what is happening

12:12in this chapter. You also have the full

12:14outline to give you context of how this

12:15chapter fits with all other chapters.

12:17Your task is to follow these steps to

12:19analyze the genre of the story and to

12:21analyze the events of the chapter.

12:23Choose an appropriate word count for

12:25this chapter based on the events in the

12:26chapter, the standard for the genre and

12:28the pacing of the scene. The word count

12:29should never go lower than 1,000 words

12:31and never higher than 5,000 words.

12:33Output just the number of words and

12:34nothing else, not even the word words.

12:37Example,

12:382,500.

12:39Um so, that's exactly what we're doing

12:41here.

12:42Give me 20 seconds. People always ask me

12:44for the exact prompts I use to write my

12:46books with AI. So, I put them all into

12:48one prompt pack. Right now, my Story

12:50Hacker AI group is invite-only and

12:52closed to the public, but if you want in

12:55on what we're doing, join the waitlist

12:56using the link below and I'll send you

12:58the prompt pack for free as a thank you.

13:00Unlike some micro offers from tech bros

13:02on YouTube, these prompts are multiple

13:05hundreds of words long, refined from my

13:07own real book projects. So, click the

13:09link below, pop in your email and I'll

13:11send them to you today. So, once we have

13:13a word count estimate, all this one does

13:16here is it increases that estimate by

13:1925%. Usually because whatever you give

13:23it, the AI is usually unlikely to

13:25actually fill up the full word count and

13:27so with this one, I add an extra 25%

13:30just to stretch it a little bit further,

13:32so that it's more likely to get to that

13:34original estimate. Um just a little

13:36technique there. But anyway, let's move

13:38on to the next three and these are

13:40important because this is where we're

13:41going to essentially write our scene

13:43brief. This is the more detailed outline

13:45as well as all of the information and

13:47details that are needed to actually

13:49write this scene. All right, so let's

13:51start with the plot scene brief. I split

13:53this into three, plot, character and

13:54world building. For this one, uh we pull

13:57in all of the relevant context that we

14:00need so far, including the small section

14:03of the plot that we're we need to. I do

14:06pull in the full outline here as well,

14:08so we know how that small section of

14:10plot fits with the overall outline. I

14:13also pull in just the characters that

14:15were selected in previous step and just

14:16the world building that was selected in

14:18the previous step. I also pull in the

14:19last 2,000 words, so it knows where it

14:22left off in the previous chapter,

14:23assuming we're not starting with chapter

14:25one. And then here, I say, "Given the

14:27above outline, characters, world

14:29building information, your task is to

14:30flesh out a plot scene brief for this

14:32chapter. Make sure the chapter is

14:34clearly labeled with the words chat

14:36whatever the chapter name is. Do not

14:38name the chapter anything besides this.

14:39This plot scene brief should include the

14:41following." And then I have it uh um

14:44print out the plot verbatim from what it

14:46gets in the above info. I have it point

14:48out what the POV intent is. So, pull the

14:51viewpoint character from the current

14:52chapter to work on above. The tense

14:54should be and then this is where it

14:56pulls in the form that you filled out at

14:57the very beginning

14:59so that it knows what the tense should

15:00be. Word count, include this number from

15:02the word count above verbatim. Only

15:03include the number, nothing else.

15:05Because we've pulled that in already.

15:06And then the next thing it does is it

15:08outlines the beats and the blocking for

15:11the scene. Uh so, it's this is

15:13essentially your more detailed outline

15:15of the scene. And usually uh you might

15:18want to set this up so that you can

15:20review this beforehand, but I actually

15:22find that as long as my overall outline

15:25is more or less consistent, I find that

15:28this whole process, I don't really need

15:30to review the beats of the blocking. It

15:31does a good job of that for me. Uh and

15:33then I also make sure that the plot for

15:36this scene picks up appropriately after

15:38the end of the previous chapter text. Uh

15:40and then I also ask it to establish what

15:43the cliffhanger is going to be. And you

15:45can do cliffhangers well or badly. And

15:47so, as part of this prompt, I included a

15:50cliffhanger guide

15:52on what a good cliffhanger looks like

15:55and that's what that whole section is.

15:57And then I have it do some formatting of

15:59how uh how it outputs this whole thing

16:02and then I let that go.

16:04And then we move on to the character

16:07scene brief. This is the same sort of

16:08thing, but just for the characters. And

16:10my goal here with this step,

16:12so we pull in the plot, we pull in the

16:14full outline, we pull in just the

16:15characters that were selected here

16:18uh as well as the previous chapter text

16:20and then the slider rubric that I talk

16:22about in a previous video on characters.

16:26Um because every character has certain

16:27sliders

16:28and those sliders are going to change

16:30depending on the scene we're in. All

16:32right, so this prompt goes something

16:33like this, "Given the above outline and

16:34character information, your task is to

16:36flesh out a character scene brief for

16:37this chapter. Here's what your brief

16:39should include. Active characters,

16:41analyze the list of active characters

16:42from the character section above. For

16:44each one, make small modifications as

16:45needed for the specific scene, including

16:47removing removing future details." Okay?

16:51This is the important thing is like we

16:53don't want information that's going to

16:55be relevant to a future scene to impact

16:57this scene accidentally. It's not as big

17:00of a problem as it used to be, but it's

17:01still a big problem, especially if

17:03you're writing in a genre like mystery,

17:05where you don't want to reveal ahead of

17:07time who is the murderer, right? Uh

17:10who's the person who done it? Because if

17:13you have that information in the

17:14character sheet, there's a good chance

17:16the AI will just pull from that and

17:18inadvertently give away the fact that

17:20that person is the one who done it,

17:23right? So, we want to essentially take

17:26these characters and remove any details

17:28of future stuff, right? So, that's the

17:31biggest thing that this prompt is doing

17:32is going through all of those characters

17:34that were extracted earlier and making

17:37sure that we're actually looking at them

17:39more in at this specific slice of time

17:43in this chapter. So, not their like

17:45overall profile, but like who are they

17:46in this exact moment? So, it that will

17:49include removing future details. It'll

17:51also include adjusting where their

17:53sliders are because their sliders slide,

17:57you know? Uh they have kind of a

17:58baseline where they're at for their

17:59sliders most of the time, but if the

18:02tension is extra high in the scene,

18:03certain things like stress, their stress

18:05slider, might be a little bit more

18:07intense, stuff like that. So, this is

18:10the step where we establish where those

18:12baselines are and how they differ from

18:15the baselines in this specific scene uh

18:18as well as like what are you know,

18:20they're wearing anything different in

18:21this scene? How's their physical appea-

18:23appearance? Are they are they, you know,

18:25did they just come out of a bath or have

18:28they been rolling through the mud a lot?

18:30You know, like maybe they've been on

18:31this adventure and they're super dirty,

18:33you know? Little details like that that

18:35are that differ from the norm and are

18:37only included in this in this slice of

18:39their life in this chapter. That's the

18:41kind of thing that we're trying to

18:42establish here, okay?

18:44So, that's what we do there. And then we

18:45do a pretty similar process for the

18:48world building sheets, looking through

18:49all of the different world building

18:50elements that are relevant to this scene

18:52and making sure there are no future

18:54details included in that as well, all

18:55right?

18:56Then we go through a couple of important

19:00steps here. Uh notably, we do a

19:02chronology check. So, it looks at

19:04everything that was just written across

19:06these three steps here and it says, "Is

19:08this chronology uh chronologically

19:11consistent both with itself and with the

19:14full outline?" So, it looks at the full

19:16outline, it looks at this and it says

19:18like, "Are there any problems here

19:19chronologically speaking?" And then it

19:21does a rewrite of the entire scene

19:24brief, so basically combining the

19:26outputs of these three into one here as

19:30well as uh making any corrections for

19:32the chronology check here.

19:34And then we finally have our full scene

19:36brief. Despite the fact that we've just

19:38spent a ton of time just creating the

19:40scene brief and everything we need for

19:42that, the next step is deceptively

19:44simple and that is just writing the

19:45first draft. The first draft is uh a

19:49very important step

19:51uh obviously, but there's actually not

19:53that much that goes into it now that

19:54we've done all this of this work on the

19:56scene brief.

19:57Essentially, what we do is we pull in

19:59the prose style sheet.

20:01We'll pull in the previous 2,000 words

20:03of the text so it knows where to where

20:04it left off, and then we'll pull in the

20:06full scene brief.

20:07From there, it says your task is to

20:10write the entire

20:12chapter name

20:13based on the scene brief and to cover it

20:14thoroughly from deep point of view

20:15writing the scene as if written by

20:16bestselling novelist and not rushing

20:18through the scene. And then we got a

20:19bunch more uh information here.

20:22And uh I uh also add at the bottom here,

20:25remember to use this tense

20:27uh and note the POV character in a scene

20:30brief above just to make sure it doesn't

20:32get that wrong.

20:33All right?

20:34So, that's the that's where it writes

20:37the first draft, which is an excellent

20:39step and exciting step.

20:41But then, we're not done. Uh my priority

20:44with this whole automation is to make

20:46sure that chronological issues are as

20:48filtered out as possible

20:50so that you don't have to fix as many of

20:52them as you're going through. And so,

20:54for this next step, I actually pull in

20:55the last 20,000 words, assuming there

20:58are 20,000 words, in which case it'll if

21:00you haven't written 20,000 words yet,

21:02then it will go through and it will

21:05identify however many you have written,

21:08right? [clears throat] So, if you've

21:09written 10,000, it'll grab the 10,000.

21:10If you've written 50,000, it'll pull in

21:12the last 20,000 cuz I don't want to

21:14overdo this prompt too much, but I do

21:16want it to look back pretty extensively

21:18in its narrative

21:21uh what has happened in the last 20,000

21:22words or so. And then we do a second

21:24chronology check. This chronology check,

21:26however, is looking exclusively at the

21:29last 20,000 words and the chapter that

21:31was just written to make sure there are

21:32no little inconsistencies. For instance,

21:35uh one of the things that will often

21:37happen when you're writing uh with AI

21:38this way is it will pull up details like

21:41they're brand new.

21:43Um

21:43like a character will say something

21:46like, "Oh, I've got this special ability

21:48that lets me do this."

21:50And then

21:51five chapters later, it'll be they'll

21:53introduce it again and be like, "Oh, I

21:55have this special ability that lets me

21:56do this." forgetting that it had already

21:59introduced that thing five chapters

22:00earlier. That's the sort of thing we're

22:02looking for here with this chronology

22:04check where we want it to look for

22:06anything that is just off about the

22:08chrono- chronology of the chapters

22:11altogether. We're trying to smooth out

22:13any of those chronological problems as

22:15much as possible.

22:17The other major thing that I'm trying to

22:18catch in this automation is the style.

22:21Can we get it to the point where the

22:23style is as close to the style guide as

22:26possible cuz this is one of the best

22:28ways to improve on the default vanilla

22:31writing of these large language models

22:33is to give it your style and then with

22:35clear explanations of the kind of things

22:37you want to see in your style and then

22:40uh in- integrate that into the text.

22:42And we do that in the first draft stage

22:44here, but I want to do a second check to

22:47be extra sure that there aren't other

22:48ways that we can inject more of your

22:50style into this thing.

22:52So, this style checker looks at your

22:53prose style guide. It looks at the

22:55chapter that was written and then makes

22:57recommendations based on that. And then

22:59finally, we have a rewrite here. And

23:01this rewrite happens based on the

23:03feedback that it got from the chronology

23:05check number two and the style check

23:07here. It takes all of that information

23:09and rewrites uh the chapter to include a

23:12little bit more of that. And then

23:14finally, after all of that, it dumps the

23:17results here into the chapter

23:20document. And then it repeats the loop

23:23for the next chapter in the series that

23:25you want to write. And that is the full

23:27advanced outline to chapters.

23:29And uh it's pretty pretty intense, but

23:33pretty cool when you see the outputs. Uh

23:35when it comes to chronology and things

23:37like that, definitely the

23:40uh the issues have been severely

23:42reduced. Now, one thing you might notice

23:44that I didn't do here that I used to do

23:46in my previous version of this is I

23:47would ask it to find like problems with

23:50the prose, like line editing type stuff.

23:52And I'm not doing that this time around.

23:54The reason is I'm developing an entirely

23:56different automation to do that just

23:59focuses on line editing and reducing AI

24:02slopisms and things like that so it

24:04sounds very much like a really decent

24:09prose, right? That's the whole goal. So,

24:11that is essentially what I'll be talking

24:13about next when I do another automation,

24:16although I'm also exploring other tools

24:19like Claude Co-work, which I think might

24:21be able to do editing tasks like that a

24:23little bit more effectively because one

24:25of the problems with the line editing

24:27automations that I've worked with is

24:29that they can be a bit expensive to run

24:31cuz there's lots of different steps and

24:33you it requires smart LLMs to do it

24:36well.

24:37But we'll we'll see. This is still an

24:38emerging area. I'm still working on it.

24:40Hopefully, this has been interesting to

24:42you. If you want to get your hands on

24:43this automation, you could do so at my

24:44Story Hacker membership down below

24:47uh as usual.

24:49Um I try not to have too many

24:51uh sponsors and stuff like that other

24:52than my own work just because I'd like

24:54to be able to give something in exchange

24:56for money.

24:57Uh so, that membership, there is a lower

25:00tier now that you can join, which is $9

25:02a month. Uh it used to be $67 a month

25:05exclusively. Now, there is a lower tier

25:07that gets you a lot of the same things,

25:09not everything obviously, but what some

25:11of the the stuff that you can now get at

25:13the $9 tier are these automations. So,

25:15you're welcome to come in and uh play

25:17around with them

25:18to your heart's content and join us in

25:20our uh you know, we do regular

25:22challenges like a book in a month

25:24challenge, which we do quarterly. And

25:26you can come and participate in those as

25:28well. So, uh it's great time. And I

25:31I will hopefully see you there.

More from The Nerdy Novelist

Recently added transcripts

Browse the whole transcript library

This transcript was generated from the captions YouTube publishes for this video. Get the transcript of any YouTube video atfreeyoutubetranscribe.com, free, unlimited, no sign-up.