Getting the transcript
Reading the captions from YouTube. A video nobody has opened here before takes 10 to 30 seconds; this page fills in on its own.
Getting the transcript
Reading the captions from YouTube. A video nobody has opened here before takes 10 to 30 seconds; this page fills in on its own.

Jonathan Acuña - Doctor AI · @jonathanacuna
Words
1,337
Runtime
7:11
Speaking pace
186wpm
Reading time
6min
186 words per minute, between the 181 median and the 201 75th percentile of 349 measured videos. That distribution comes from the 349-video hook study.
Opening (first 30 seconds)
Let's talk about some APIs that you could give to give your Hermes AI agent dangerous superpowers. The first one is going to be Firecall. Firecall, you create an account, and this allows your AI agent, or Hermes, to scrape the web and anything that's out there >> [snorts] >> that pretty much is like, "We block websites for bots." He'll be able to go and actually get the information, scrape web pages. Yeah, you can search for things like basically like Perplexity, scrape pages and information. You can map a website, so
93 words, the words spoken in the first 30 seconds at 186 words per minute.
Free, no signup. See how the first 30 seconds hold attention, with rewrites.
Sentence shape
| Measure | This transcript |
|---|---|
| Sentences | 91 |
| Average words per sentence | 14.7 |
| Longest sentence | 57 words |
| Questions asked | 4 |
| Sentences containing a number | 16 |
Most used terms
Filler phrases
44 in total: like 20 · basically 16 · uh 4 · actually 2 · um 2.
A literal whole-word count of the same phrase list the Prepublish browser extension uses, so a phrase inside another word is not counted and a phrase used in its ordinary sense still is. It is a count and not a judgement.
What this transcript is
Every word below is the caption track YouTube publishes for this video, pulled from the video itself and reproduced unchanged. It is not Prepublish's writing, not a summary, and not a re-transcription: it is the video's own published captions. English captions, generated automatically by YouTube, in the video’s original language. Source: the video on YouTube. A channel that would rather this page did not exist can ask for its removal through the contact page, and it is removed.
No Script X-ray for this video: YouTube shows a Most replayed graph only once a video has enough views.
Let's talk about some APIs that you could give to give your Hermes AI agent dangerous superpowers. The first one is going to be Firecall. Firecall, you create an account, and this allows your AI agent, or Hermes, to scrape the web and anything that's out there >> [snorts] >> that pretty much is like, "We block websites for bots." He'll be able to go and actually get the information, scrape web pages. Yeah, you can search for things like basically like Perplexity, scrape pages and information.
You can map a website, so then it will find out all the pages on it, subpages, hidden pages. Great for funnel hacking for competitors. Crawl will also crawl their website from here. This is all 1,000 They give you 1,000 searches a month for free. This is the first superpower that you're going to give to Hermes. The next superpower you're going to give to Hermes is Himalaya. Himalaya is a GitHub file that allows your AI agent to connect to your Outlook, Gmail, Proton, whatever your seven different emails that you have.
This will connect all of them to him. But, it's better than MCP because honestly MCP is dead. Long live APIs. MCP This basically allows him to basically send, delete, archive, reply, so that he can basically do all of the email automations from there. Next, this one is a little specific to members in my community. This is a 30-page prompt that goes through all of the tools necessary for Claude to complete Claude, Codex, Gemini, Grok, really doesn't matter, or more importantly Hermes.
This allows him to basically automate any browser, Safari, Chrome, really doesn't matter. But, this he can even defeat CAPTCHA, this type of automation. So, it's not just Playwright, there's actually more levels of browser automation because now that you have an AI agent, the first thing you're going to do is give them all the tasks that you do, that you have to go to a website and fill out some form, or download some information or something like that.
Yeah, this goes through all the different principles, what to install, what's the master prompt, what are best practices, what are fallbacks, contingencies, uh and then how to save the workflow so that he can repeat that automation and not burn basically uh Hermes credits every single time. So, basically he could run it and burn less tokens. So, that's browser automation. That's another superpower that you want to give to your Hermes AI agent.
My prompts are 29 pages. Like Like I'm not doing like one pit Like there are levels to this game. I want full browser automation. I want to give one prompt to one model and have the guarantee that he can automate anything on a browser. Takes 30 pages. The next skill that you want to give is kai.ai, k i e.ai. Um this is basically all image and video so that your Hermes AI agent can create all of your content for your website, for your social media, for your books, for your promotions, your ads, all of it is going to be here.
We don't use Higgs field. If you're still using Higgs field, you're far, far behind. My God, who's still using Higgs field in 2026? Anyways, do you want mandatory everything has to be an API. If you're not Higgs field just released their API because they're like, "Oh shoot, humans are not the whole are not the next economy. Agents are." And it's like, "Yes." For me, I have 50 AI 50 AI agents. I got to give them an API.
I'm not hiring human beings. Like no, it's too much like W-2 and labor and time off and like a payroll I forget that. Like I'd rather just pay an API, replace a human being, and then give this to an agent to basically automate that tool. Because if you purchase tools that do not have an API, you have to hire a full-time employee, W-2, payroll, 9-5, 40 hours a week to manage that plus 14 other tabs across their uh in their role.
So, that's why this is very important. So, this will basically generate um Now, there are tons of image models and tons of video models. John, which are the best? It's going to be Chat GPT-2. Basically, to generate any image from there. How much is it going to cost? Roughly 3 pennies for every 1K image. And you could go baller. You could go 8 pennies for every 4K image, which is great. And then if we go down over here, you want the image-to-image model right here.
The image-to-image model, if we go down, this is how you replace you face swap clones. Cuz everyone wants to basically avatar clones. This is going to be the first thing that you'll use to basically upload a photo and it will clone that person in there, for example. So, now that that's there, then from here you're going to be using Let's see. But Higgs Field has an environment to make managing the pri- Like Master AI?
You don't you No, like Keep playing with Higgs Field. Keep playing with Higgs Field. Enjoy Higgs Field. This is for agents. This is for so we can give this to our team of agents and they can basically fully automate it. The video from here is going to be SeeDance. SeeDance is basically text-to-Hollywood film from here. This is going to be very high quality, which is crazy. So, if you need quick and fast and dirty, just like their 720p fast one is very popular.
Basically, you get the same quality, but it's super cheap. It's almost like you can go down to like 50 cents for like a 15-second video with your AI avatar in it. Super cheap. This is the cheapest way to do avatars. Then in here, so that's image and video from there. The last API, because I've given you too much dangerous APIs for your Hermes AI agent are ready to replace your graphic artist, replace your personal assistant, and now it's going to replace your social media manager.
The next API that I'm going to give you is Zerneu. This does all social media publishing. Also, hashtag not sponsored by any of these tools. These tools just make me and my clients more money, so that that's the only reason why I keep them around. Uh this, look at all the platforms you are wasting your time trying to post manually. I post 13 times a day every single day, and I generate millions of views every single month.
I don't do that by logging in one at a time and posting one piece of content, and then doing that every single day. What a waste of my time. So, in order for you to post and and keep up with a social media, that means that you have to be posting multiple times a day. Do not do that manually. You're going to be giving this to your Hermes AI agent, hooking up all of your social media profiles, and then from here, you'll be able to post content.
Then, you can even give your Hermes AI agent a phone number, so that he could respond through calls and SMS texts and voice calls. Okay, so that is the third superpower that we're going to give to your Hermes AI agent. That means in 30 minutes, we set up a Hermes AI agent, connected them to his Gmail, connected them to his calendar, gave him the ability to replace a graphic artist to generate all the images and videos for your social media.
Then, we gave him the ability to post to all social media channels at the same time. So, that means in 39 minutes, this guy is going to make you an extra six figures a year by saving you from hiring additional labor and employees.
The words are the caption track's own and nothing is reworded or re-transcribed. Paragraph breaks are placed between sentences so the text reads as prose.
Free tools for your own script. No signup, no login.
Paste your draft and see where viewers are likely to drop off, with a rewrite for each weak line.
Paste the first 30 seconds of your own draft for a hook score and rewrites.
Check your draft against YouTube's advertiser-friendly guidelines before you record it.
Read this channel's public videos and transcripts, and download a writing brief for it.