Full transcript
0:00So, I've been talking a lot about NADN
0:02automations for authors on this channel,
0:04and a lot of people have trouble getting
0:07started with NAND because they view it
0:09as being really technically challenging.
0:12And to be fair, there is a lot more
0:14technical involvement with any kind of
0:17automation platform uh but particularly
0:20NAN when compared to a software built by
0:24someone else like Sudorite, Raparite,
0:28uh Novelcraft, all of those uh or even
0:31just a simple chatbot which is the
0:33easiest thing of of all. And Naden
0:35automation is obviously a lot more
0:36complicated than that. But I want to
0:38just assure you that you can absolutely
0:42use NADN. And I'm going to give you one
0:45of the best technical helps that I have
0:47found in my process that should
0:49hopefully help you solve almost any
0:51problem you come across when do in your
0:54NAND automations and trying to build
0:56things. So let's get into the meat of
0:59the video. The tool that will help you
1:01the most is a tool like Perplexity. Uh,
1:05so I've done videos about Perplexity in
1:07the past. It's, uh, no secret that this
1:09is my favorite chatbot at the moment, at
1:11least for now, until things, you know,
1:14unless we see something improve
1:15elsewhere. Uh, I won't get into why it's
1:18my favorite, but you don't have to use
1:20Perplexity, but you do need to have any
1:22chatbot that has web access, which is
1:24most of them these days. You can usually
1:26get out of Gemini or Claude or Chad GBT.
1:30They all have a web access option, but
1:33it's important that you have web access.
1:35And the reason I use Perplexity is
1:37because it was initially designed
1:40as a web access tool, as a way of
1:43searching the web for things. Uh, so
1:45that's why I use it for this specific
1:48use case. But let's go to NAN and let's
1:51take a look at an automation. So this is
1:53the uh brain dump to fully fleshed out
1:58story dossier automation that I showed
2:01off in the previous automation. And uh
2:04we're going to I'm going to demonstrate
2:05something really cool that you can do
2:07with Perplexity. So first of all, let's
2:09go ahead and download this automation
2:11just like as if I were to share it with
2:13somebody. We're going to download the
2:15JSON file for this. So you hit download.
2:18And once that's downloaded, we're going
2:19to actually upload it to Perplexity. So
2:22you just go ahead and select local
2:25files, select the file. So now it has
2:27this and all you have to do at this
2:29point is say this is an N8N automation
2:34I've built and then add something like
2:37please explain step by step what it is
2:41doing. And this is really cool. It's
2:43going to be able to read the JSON and
2:45give me an accurate description of
2:48everything that this automation does.
2:50Now, if you recall, this automation
2:52grabs a number of documents for it to
2:54pull from, different templates and
2:55things like that. Then, it brainstorms
2:57ideas. It picks the best of those ideas.
3:00It creates a complete story dossier, it
3:03critiques that dossier, and then it does
3:05a rewrite of that dossier before dumping
3:06it into a new file. So, let's see what
3:09it gets out of this. All right, so it
3:11has a trigger. Then, it does some
3:12research, gathering up all of the
3:14templates right here. Then it goes
3:17through the ideation phase. Node
3:19brainstorm AI model Google Gemini 3 Pro
3:22via open router. So it knows all of this
3:25action. It takes your brain dump on the
3:27genre tropes and plot template it just
3:28read. It then generates 12 distinct
3:30story premises. Each premise is a one
3:32paragraph pitch that tries to stick to
3:34your original idea while effectively
3:35using the requested tropes. Like bam.
3:39That is absolutely spoton on what that
3:42part of the automation does. and it goes
3:44through and it just gives me an accurate
3:46reading of everything that has happened.
3:49So if you want help
3:51with your automation, all you have to do
3:54is download the JSON file, load it into
3:57Perplexity or another chatbot with web
3:59access
4:01and then ask it what it's doing and then
4:04you present it with whatever problem
4:06you're having. Okay? So if you're
4:08running into any issue, which I'm not in
4:10this particular automation, but if you
4:13were, then you could just describe what
4:16the issue is that you're having and
4:17asking ask it for help. And since it can
4:20read essentially the entire automation,
4:22it it uh like nine times out of 10 gives
4:25me the exact answer that I need to solve
4:28that problem. So here is a past
4:31conversation I had where I said the
4:33attached is a JSON file from an N8
4:35automation that I've built. please
4:36explain what happens. It went and
4:38explained what happened and then I said,
4:40"Okay, that's correct." So, I'm running
4:41to a small problem. The original Google
4:43doc that it's pulling from in get
4:45original doc has formatting in it like
4:46italics and the like, but when it's
4:48pulled into this automation, that kind
4:50of formatting is not seen by the AI step
4:52that needs to proofread. So, it can't
4:54necessarily get the right italics. How
4:56do I get it to properly see the
4:57formatting? So, this is an issue I was
4:59running into where I needed it to see
5:01that certain words were out were
5:03italicized or it had headings or
5:06something like that. And then it said
5:08something like this. This is a common
5:10issue with the standard Google Docs node
5:12which pulls clean text and strips out
5:13all formatting before your workflow can
5:15even use it. To fix this, you need to
5:18change how you retrieve the document so
5:19that you get HTML version which contains
5:21the formatted tags and converts it to
5:23markdown so your splitter can handle it
5:25safely. And so I was just like, great.
5:27And then it gave me specific
5:28instructions, including a little bit of
5:30code on what to do. And I had to go back
5:34and forth with it a couple of times, but
5:36by the end, I had exactly what I needed,
5:38and it was doing exactly what I wanted
5:40for this particular automation. So once
5:42again, I know this is a little bit of a
5:45short video, but hopefully this has been
5:48really helpful to you because this is by
5:50far the way I've been able to achieve
5:52lots of things when I had zero idea what
5:55to do or how to code or anything like
5:57that. You just turn to an AI that has
6:00web access so it can access Naden's uh
6:03files and documentation and stuff like
6:05that and it just it really helps the
6:08process. Now, as a reminder, my Story
6:10Hacker Gold membership is open just for
6:13this week and will be closed at the end
6:15of the year. So, that at December 31st
6:18at midnight, US Eastern time, it will be
6:20closed. And in addition to that, the
6:22price will be going up the next time it
6:24opens. So, this is the last time that
6:26you have to get it at its current price
6:28of $47 a month. So, just thought I'd
6:31lead that with you and I will see you in
6:33the next video.