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.

Sharbel A. · @sharbelxyz
This video has no Most replayed graph yet: YouTube shows one only once a video has enough views. These are the moments viewers replayed most in Sharbel A.'s most watched videos.
Most replayed moment at 13:32
3.9x that video's typical replay level
built-in memory, which is great. Here is how I would think about memory providers. Memo is interesting if you want a dedicated memory layer for personalized AI agents. It focuses on extracting, storing, linking, and retrieving memories efficiently. Their
Said at 13:25
Most replayed moment at 6:37
5.8x that video's typical replay level
trading strategy. Tests it. If it's better, it keeps it. If it's worse, it discards it and tries again. So, that's exactly what I built. Okay, here's the system we have at play. I gave it two years of crypto data,
Said at 6:31
Most replayed moment at 1:53
3.7x that video's typical replay level
let's install it together. Okay, so for step one, we need to actually start by installing Bullpen's CLI so that we can actually do everything that we want to do. And for that, let's first open Claude and put in the dangerously skip
Said at 1:47
The graph counts replays. It does not show where viewers stopped watching.
Words
2,554
Runtime
15:08
Speaking pace
169wpm
Reading time
11min
169 words per minute, between the 160 25th percentile and the 181 median of 349 measured videos. That distribution comes from the 349-video hook study.
Opening (first 30 seconds)
This business used to be my entire life, but I lost my passion for it and I abandoned it for years. But then my X account got banned and overnight my business lost its main source of traffic. So I needed to replace that income fast. So I'm going to use AI to revive my old business. [music] Building it the first time took me years of grinding, but I don't have years. So [music] AI is going to need to build this for me
85 words, the words spoken in the first 30 seconds at 169 words per minute.
Free, no signup. See how the first 30 seconds hold attention, with rewrites.
Sentence shape
| Measure | This transcript |
|---|---|
| Sentences | 172 |
| Average words per sentence | 14.8 |
| Longest sentence | 52 words |
| Questions asked | 3 |
| Sentences containing a number | 23 |
Most used terms
Filler phrases
46 in total: I mean 16 · literally 12 · like 11 · actually 3 · uh 2 · you know 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.
Free, no account. See where attention is likely to drop, with a rewrite for each weak line. The free check shows the scores and the one issue costing the most. Or run it on the words above first.
Free · No login · See a sample audit first if you prefer.
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.
This business used to be my entire life, but I lost my passion for it and I abandoned it for years. But then my X account got banned and overnight my business lost its main source of traffic. So I needed to replace that income fast. So I'm going to use AI to revive my old business. [music] Building it the first time took me years of grinding, but I don't have years. So [music] AI is going to need to build this for me and learn how to go viral on social media.
Create and post content, generate leads and close those leads for me and run the whole business for me start to finish. Okay, so the first thing I needed to do was build my AI dream team. But I needed an AI tool that can reliably create entire agent teams that I could hook different models in and that I could connect to all of my different tools in order to revive my business as fast as possible. So I got to researching and found this very cool product called HyperAgent.
The reason I decided to use it for this test is simple. I do not just need a chatbot. I quit this page because it got repetitive sourcing content, writing captions, scheduling posts, replying to DMs every single week. But AI doesn't get bored. I also reached out to the team and they love this idea so much that they were very happy to sponsor today's video. big shout out to the HyperAgent team. This video literally wouldn't be possible without them.
HyperAgent is giving away $100 in bonus credits by the way when you sign up. So check it out now inside the link in the description below. And with that I got started and built five agents. So I'm building out my team and I have to say I really enjoy the platform. I really enjoy how it looks like, how it feels, even hiring the agents one by one. So, we currently have our team, an auditor, a researcher, a strategist, a content creator, and a salesperson, and we're still hiring one more, which is the operator, right now.
And it's done a pretty good job at understanding the page so far. I'm giving it as much information as it needs, and from there we'll get started with day one. All right, and we're wrapping up day one. It just posted its first post literally 3 minutes ago. We already have six likes, which is great. And this is the first post that it helped us spin up and make. This is the operator agent's daily job, actually writing out each post for the day with their captions, media over here, and the time slots it scheduled it out for for the day.
And yeah, we have two posts scheduled per day. Day one, I'm focusing on the content itself. I'm going to let today's post run and come back and check tomorrow. But, a couple of things I've noticed, it's been 3 years almost since I've posted. Literally, the post before it was 151 weeks ago. That's almost 3 years, if not more. A lot has changed on Instagram. The researcher has come back and was literally like, "Dude, like I don't think we can continue posting the things you used to post.
Policies have changed. The Instagram has changed. The internet has changed since then, and it wouldn't be fair to post these today." Uh so, we need to rework new strategy, something I've never tried before. It will have to come up with it, which makes this challenge even more interesting. That's the researcher's report. It pulled Instagram's current policy in 2026 and told me flat-out that the old format is a ban risk.
And that right here at the bottom is the strategist coming up with completely new content plan. See you in day two. Day two here and I'm genuinely impressed with what we've been able to do in day one and so far. I mean, the first post we've posted has done very very well. Its reception was taken very well. Just evidenced by this comment, "Well, finally the admin is back." I do suspect though that, as you can see, the post after it did a little worse and we just posted another post right now on day two.
I suspect that this was because this was my first post in a long time. Instagram notifies a lot of people, you know, like, "Hey, this person has posted for the first time in a while." I suspect that might be why, but what I'm more so surprised by is Hyper Agent's ability to just work so well with a fleet of agents. I haven't been talking to it that much, but it's been able to do a lot of research where needed. It's been very spot-on, very accurate.
I'm going to be testing new angles, like news for example, on that page. And it's crazy that it's been able to do all of that on its own. And what's even crazier, something that I failed to realize when I first started this challenge, I thought it would be a lot closer to posting the same type of content I used to post before. But thanks to Hyper Agent, it's literally gone and done a deep dive on the state of Instagram right now.
And I didn't even know this, but I can't post the same type of content I used to post anymore because Instagram has apparently flagged that as a type of content that if you post, you can either get banned or you can just, you know, be shadowbanned. Your post won't show up to anyone, including your followers, which we don't want that to happen. So, it literally spun up a strategy that I wouldn't even have thought of or had the the time or patience to come up with, and it's currently testing it out.
This is insane. I'm pleasantly surprised with this. It's day four and honestly, I'm surprised. Yesterday's post took off absolutely that well. It did even better than the first post that we posted almost two times as good. Even though it didn't get that many comments, I think news are great thing we have to play into for the rest of this video and it's such a great strategy. Honestly, I've told it to double down. We just posted two.
So, we'll see how today's posts go. But I love how it's just learning. It's literally taking up every single post trying to see what does well, what doesn't do well, monitoring competitors, seeing how to go viral. I wouldn't have figured this out on my own to be completely honest or it would have taken me months if not years to figure this out and I would have either gotten banned or shadow banned at that point. The second thing I've done today so far is start the DM replying which was just such a headache back then to have to find, reply, answer all the questions.
The great thing with AI is you can literally just give it that responsibility and it will do a 10 times better job than you ever could. What's nice is I can talk to specific agents directly. This is me telling the sales agent, "Hey, we need to start DMs and selling. How do we get started?" It asked me a list of questions that it needs and then we literally started. I set up an automation so that it sends me an email.
It replies to DMs and it will stop before pitching a price. Oh, I'm excited. I'll be on the lookout to see how we do by day five. Day five and I'm pleasantly surprised with the results that we've had today. I mean, it decided to double down on news which I mean, absolutely because just look at this post right here. We be about 11 hours ago so in the morning, and it's absolutely blowing up. Uh we can see the reach on it, but on the phone it shows that it's currently at 80,000 people.
It's crazy that it understood how to go viral. I mean, this post right here is doing better than posts I had around the peak of this page's lifespan. So, this is absolutely wild. This is update number one. Update number two. It did start replying to a bunch of DMs, all the DMs we're getting, and there's no lead yet that is looking like a promising one, or that is looking like they're about to pay. But, it's going to keep going, and the great thing is we have it up and running, and yeah, that's literally it.
I can't wait to see what we do on day six. Day six, and oh my god, guys, yesterday's post, let me just refresh real quick, went absolutely viral. I mean, look at that. 3,000 likes, over 150,000 impressions on this just one post alone. We had other posts do well as well. Today was a very good day in terms of posts, and yeah, the best part is we got to enjoy the lake. We went, we sailed, we had fun, and I didn't have to worry about any of this.
I mean, typically it was on Sundays, days like this, where I would spend the day sourcing content, creating the content, just hours of my day spent on creating that content, and I just got to rest, to relax, to enjoy, to make good quality memories, which makes me so, so happy. On the downside, though, the DMs don't look that promising. >> I mean, you could see it in the pipeline with all of these statuses, all of these follow-ups sent, not a single reply received. 10 active leads at the moment, zero replies, zero objections, zero purchase intent.
Oh god. I mean, it's been replying to every single DM we've received and it's gone as far as literally one guy just deactivated this account. This is how bad the DMs are going. I don't know if we're going to make a sale this week. I'm not very hopeful. We have only one day left to this challenge, but only time will tell. I mean, one thing I know is even if we don't make sales this week, I know it's going to happen this month.
I know we need to get the ball rolling first. I mean, content is number one. If your content sucks, your reach sucks. If your reach sucks, your discovery sucks. And if your discovery sucks, then your pipeline doesn't exist. So, content is number one and it seems to really understand what makes good content. I mean, we just posted this and it's already doing well. See you in the final day of this challenge. Oh my god, I was getting ready to wrap up and writing down the results for this challenge on day seven and we just got our first close.
Finally. I honestly didn't think that we would get a single close during the 7-day challenge. I mean, 7 days is insanely short to revive an entire business. I'm shook right now. But I mean, on a positive note, look at this. We took a page that was completely dead for three entire years and we managed to bring it back to life. So, after 1 week, here is what happened. The page started with no posts whatsoever for the past 3 years.
However, in the last 7 days, we've managed to post 20 content pieces this week, generate 668,000 impressions, reach 151,000 unique accounts, and receive 13,900 interactions or engagement. We did lose 242 followers in total, but that was mostly because on day one of this challenge, hundreds of people unfollowed as soon as we got active again. I mean, you can see that massive dip on our first day of this challenge, but then you can see we're now above break even.
We're now finally starting to gain followers again. And that is classic for a business that's been dormant for 3 years. We did get 18 total inbound leads and closed one, generating a total of $50 in revenue on week one, which isn't a massive life-changing number, but it's honestly more than I ever expected on week one. But, I don't think that is the only useful way to judge this. A dormant page does not usually go from dead to profitable in 7 days.
The better question is whether the system created leading indicators that make the page worth continuing. I mean, we did have one post reach 179,000 views at the time of this recording, with many others getting around 50 to 100,000 views. I am pleasantly surprised by how well my AI agents were able to handle autonomy and pump out much better content than I would have ever been able to. So, the big question is, after 7 days of trying this out, would I keep this running?
And the answer is a big fat giant yes, absolutely. We managed to [clears throat] turn a dead business asset into a working pipeline again. I mean, we got a strategy, we created content, we replied to leads, and this AI agent gave me a reason to keep operating the page instead of letting it just sit there. And that is a massive win. If you want to try this out yourself, I will leave a link for you to try out HyperAgent in the description below.
You can get $100 worth of credits when you sign up. Look, here's my advice after what I've learned reviving my dead business. Do not start with a giant company-wide automation. Start with one agent that does one useful job every morning. Maybe it researches leads. Maybe it prepares a content brief. Maybe it checks your pipeline. One single agent, one single job. That's literally the whole starting point. And once that agent creates a clean handoff, add the second agent.
That is when it starts feeling less like a chatbot and more like a team. That is the difference between AI that gives you a reply and AI that prepares work when you actually use it. So, did AI revive my dead business? I would say yes. My wife is actually the one who suggested this whole video and who asked me if AI could bring this thing back to life. And I told her probably not, but it took something I was ignoring for years and it helped me automate the parts that made me lose passion for it.
And that is usually where businesses die in the first place. Not because the opportunity is gone, but because nobody keeps pushing the boring parts forward. Comment HyperAgent if you enjoyed this type of challenge video. I'd love to know if there is interest for me to do more of these. And if you enjoyed this video, make sure to leave a like and subscribe if you're new to my channel because I have a ton more content coming your way.
Anyway, I'll see you in the next video very, very soon. See you later.
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: paste a draft and see where it stands before you record it.
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.