TubeTalk: Your YouTube How-To Guide

How Hafu Go Doubles Views Yearly With Visual Science And Smart AI Workflows

• vidIQ

Share episode

Use Left/Right to seek, Home/End to jump to start or end. Hold shift to jump forward or backward.

0:00 | 42:19

Send us Fan Mail

Get vidIQ Boost for an exclusive price! https://vidiq.com/podcast

Want a 1 on 1 coach? https://vidiq.ink/theboost1on1

Join our Discord! https://www.vidiq.com/discord

Watch the video: https://youtu.be/_kUOrWhNynw

We sit down with Hafu Go to unpack how he keeps a channel growing year after year without getting crushed by day-to-day stats. We dig into evergreen strategy, visual-first storytelling, and the AI workflows he uses to speed up research, scripting, and editing without turning videos into soulless slop. 
• measuring progress year over year to avoid burnout 
• why evergreen ideas drive long tail views and back catalog traffic 
• building edutainment by starting with a visual hook 
• the exploding bubble demo as a perfect science package 
• using AI for scripts, shot lists, and repeatable templates 
• keeping a human in the loop so AI outputs stay high quality 
• automating the biggest time sinks instead of tiny tasks 
• outsourcing editing by systematizing creative decisions 
• using AI to analyse thousands of edit notes into a checklist 
• AI research that pulls videos, splits frames, and organises ideas 
• why AI shrinks the gap between small teams and big studios 
• what earns a bigger creator’s attention: proof of consistent improvement 
• quantity first for beginners to get faster feedback 
If you like that, you can hit that subscribe button. 
If that was something that helped you, of course, hit that like button. 
There's a link in the description below and in the show notes that gives you a special price on that. 


Cold Open Magic Trick Moment

SPEAKER_00

It almost looks like a magic trick. That's crazy.

SPEAKER_01

Ever since I started the channel nine years ago now, every year the channel has basically doubled in views and revenue. This is completely AI, but it's not AI slot. Because what differentiates good AI from bad AI is the input that you give it. Mr. Beast is a giant production company. I think in the future, like the gap between him, a huge production company, and a small team creator, is just gonna close with the help

Meet Hafu Go And The Channel

SPEAKER_01

of AI.

SPEAKER_00

Hey, welcome back to the only podcast that isn't afraid to test something. I'm Travis, and I'm here with a very special guest, Hafu Go, almost 19 million subscribers, six, I want to say trillion, but billion views, and has done something that a lot of content creators can't do. And we're gonna figure out how he does this. He actually has successful shorts and long form. Typically, when I talk to a content creator, they can only be successful at one or the other, but Hafu Go has figured it out. Welcome back to the podcast, Hafu Go.

SPEAKER_01

Very happy to be here.

SPEAKER_00

Yeah, we got a lot of secrets to share. I love it. And if you're new here, uh we're gonna listen, this is what we do every single week. We talk to content creators and help you figure out how to get your channel to the next level. If you like that, you can hit that subscribe button. But Hafu Go, real quick, tell us about your channel. Tell us what you do for those who don't know.

SPEAKER_01

Of course. Uh, our channel, we focus on science and tech. Uh, so we do science experiments, we explore tech gadgets. Recently we did a trip to China and explored the most innovative tech there. So we uh did flying cars, all the human uh robots, the dancing robots, and everything else that was there. And yeah, we overall, I think the my goal is just to make learning fun. Uh every video we try to learn something, but it's not boring.

SPEAKER_00

You know, one of the things I remember from our last uh interview is stuck

Year Over Year Growth Mindset

SPEAKER_00

with me. And it comes, uh, it talks about when we were talking about performance. And um, I want to bring this up because no one else that I've talked to in the many people I've talked to have ever said anything like this before. Most people think about performance as the one of ten, where on your studio page it tells you where your last video ranked, whether you know if it was the most viewed one out of your last 10 or less viewed. And I remember asking you about general performance of the channel. And you said, and I I found this to be very, a very good way of thinking of it, that you didn't really think of it in that way. You thought of it more year over year rather than even month over month. You were more concerned with the overall trend, because you know that there's dips and valleys and there's high points and low points. And you know, when we're recording this, for a lot of content creators, we're coming off what would normally be a low point, the January, February kind of slump for most channels. But you look at it holistically. Tell me more about that mindset and how have things looked over the last year.

SPEAKER_01

Of course. I mean, if you focus on like the minutia of the day-to-day, every upload, um, you're gonna be very depressed and you'll burn out very quickly as a creator. Um but it's not like I don't look at it at all. Uh, I do use it to uh adjust uh for the next video and also maybe to change thumb notes and stuff like that. But if we zoom out a little bit, this is for like a creator mindset type of vibe. Um, if we zoom out on a yearly basis, ever since I started the channel nine years ago now, every year the channel has basically doubled in views and revenue. So, like me worrying about the day-to-day is just giving me unnecessary stress.

SPEAKER_00

Yeah, I think that's a great way to look at it because for a lot of content creators, if you do look at it year over year, you probably are growing in some way. And there's just these moments you can't control because for the most part, 80% of your views come from like 20% of your content. Um, so worrying about all this other content may not really be a thing when your overall year was excellent. So that's great to hear from the other.

SPEAKER_01

And oh, here's another thing. Here's another thing from Travis. Uh this uh this is a stat that I

Evergreen Videos And The Long Tail

SPEAKER_01

uh discovered recently, which is a lot of our videos have a huge long tail. So they may not get like the most amount of views um from the first week or 30 days, but over time they get views. So like the average views on our video for long form videos uh is five million views over the course of a year. So that's the average views for long form, and then for short form, it's around like 15 to 20 million views. Uh so like they have a huge long tail. So I I don't know, like a short might pop off like half of a year into the cycle, you know?

SPEAKER_00

Yeah, which is unusual because uh before YouTube had shorts, TikTok was the only real place you went for short form content, a little bit of Instagram, maybe Snapchat, but I mean, let's be honest, it was really TikTok or nothing. And but TikTok is um in a lot of ways kind of like um news on YouTube. It has a very short shelf life. Like once your thing pops off, unless something weird happens and it goes viral weirdly six months later, 90% of those you get your views in the first 30 or so days, and then that's it. YouTube, like you mentioned, um, it it doesn't matter. Like it can happen a year from now. Like one of your shorts from a year ago could literally pop off today if you really wanted it to. Like you could make a new video about that subject and it could potentially bring that one back up.

SPEAKER_01

Yeah, um I think YouTube gives these like monthly uh roundups, they call them. They like email them to you and then you can see it in your studio. And one of the stats they give you is like the percentage of views that come from like back catalog. Uh and mine is very substantial. It's around like 60 to 70% of my views come from the back catalog. So that's videos that were not uploaded this month. Um, and like I think it's just the type of content that I'm making. It's very evergreen for a news channel, obviously it'd be different, but like I think for most creators, their content is pretty evergreen.

SPEAKER_00

Evergreen being the superpower here, especially on YouTube. And and you concentrate on that. And the cool thing about it is is you know, when most people think evergreen, they almost think boring, like uh SEO and stuff like that, but that's not what it means. It just means that it's relevant anytime. When I look at, for example, this video you have, I bought the world's uh biggest snacks, you're holding a huge MM, which is really a crazy image. I love it. That's gonna be just as cool six years from now as it is today.

SPEAKER_01

Yeah, exactly.

SPEAKER_00

I mean, it doesn't have to be educational videos that are evergreen.

SPEAKER_01

There could be entertainment videos. I would categorize ourselves as edutainment, but mostly entertaining, you know. Um that can be evergreen too, if your ideas are uh substantial enough to be interesting a year, five years down the line.

SPEAKER_00

I think it's interesting that you are so into science. You're I don't know that you'd call yourself science geek, but you certainly are in science enthusiast. But you make science interesting. And that's there's something cool about that. There's been a couple of people have been able to do Mythbusters has done it. There's been a couple of times it's happened, but for the most part, uh it's one of those things that have been relegated to kind of the the corner of the internet of like, oh, let me see this really cool thing. This thing is okay, I'm done with it. But you have taken it to a level that is like television. How how how do you come up with that process? How do you come up with something that's interesting scientifically, but also interesting to people in general? Because I think that's the big thing. How do you make something that seems like it might be for nerds for everybody?

Visual Hooks Before Explanations

SPEAKER_01

We try to always start with the visual. So when people think science, they usually think like explanation, we gotta understand why. But for us, we never start there. That actually comes after. We have to think about how do we package the facts in a way that people actually want to know about it. So, for example, if you want to explain a scientific theory, you should package it in an experiment that looks interesting. And that's basically like the key, you know. You gotta uh, for example, uh, do you want to see one? Yeah, I would love to. I was actually I was actually just uh I was just uh doing a video. Okay. Um I don't know for your podcast viewers, they might have to go on YouTube. Yeah, you definitely gotta go on YouTube for this. They might have to go on YouTube to uh this is cool to see this, but this is an exploding bubble. Oh, I need to see this. All we need is soapy water and lighter gas. All right, and then if we whoa if we now these bubbles are the exploding bubbles, and they're actually safe to light. Okay.

SPEAKER_00

You sure about this? This could go wrong. I can't wait for this.

SPEAKER_01

It's great. Anything for content. So, like, this is literally what we do, right? We have to find a visually interesting thing to capture people's attention, and once they're there, they'll stay and hear you like yap about it. So here we just grab a little piece right there, and then it almost looks like a magic trick. That's crazy. Yeah, totally safe. And the reason why this works is because the propane, there's a couple of scientific theories at play here. The propane, it burns very fast, so the fire won't burn more than a second. And then also the heat carries a fire upwards so it doesn't burn onto your skin. And lastly, I coated my hands in water, which acts as a heat sink for the fire. So my skin stays completely safe. Look. And that's basically a video right there.

SPEAKER_00

Yeah, it is. It totally is. Yeah. I might have to make that into YouTube short. That was crazy. So, what you're basically saying is you come up with a concept of science. Um, you find the most interesting and hopefully visual way to show that off, and then you do the thing, which is, I mean, how fire is a thing that always works. I think humans are always gonna be fascinated by fire. So that was very visual and very, and then now you're like, okay, how did he not burn his hand? And then you do the explanation, and now you've kind of like satisfied the audience. Yeah, you gotta hook people first. That was fantastic. What I I want to see that video now. I can't wait to see it. So, what was one of the ones you've done that was that you thought you knew how it was gonna work when you went into it, and then something either wasn't the way you thought it was, maybe went wrong, or anything like that

AI Scripting And Production Templates

SPEAKER_00

ever happened? Oh, that happens all the time, man. Okay.

SPEAKER_01

That happens all the time. You mean this week we did a lot of research for our experiments, and then you would see stuff happen on YouTube, but then if you actually try in real life, it's fake.

SPEAKER_00

Yeah.

SPEAKER_01

So we we had to figure that out a lot of times the hard way.

SPEAKER_00

What's your research look like? Are you using AI for that? Are you for like uh ideation and stuff like that? How are you coming up with these ideas?

SPEAKER_01

Bro, uh, I've been cloud code maxing uh for the past couple months. I've uh yeah, I've been a huge proponent of AI. Um, it's honestly improved every single step of my production process. So, like when open cloud first came out, I tried using it. Didn't really like, didn't really get the hang of open claw. It was a little too complicated to set up. But then I started using claw code, and clawed code has honestly changed my life because I realized everything before filming and everything after filming could be improved by AI to some extent. If not fully automated, it could be done faster or done quicker or done better by AI. So that's what I've been focused on doing these past couple months, where I've been trying to implement AI workflows in every single production process.

SPEAKER_00

Can you give me an example of one that kind of was an unlock for you? There was like, holy crap, this is literally like maybe doubled your production or something like that. Like, what's an example of that?

SPEAKER_01

Yeah, for uh for sure. Uh you talked about research, right? Um, research, uh, one of the videos that we're doing is called Lego Vehicles versus Impossible Obstacles. First, we can take a look at the script. So AI has actually helped me with a lot of the scripting. Um, this is a script that this is a script for the long form video that we just did for the uh uh handheld fireball thing. And AI basically did this entire thing. Um and the thing I love about AI is that it's very detailed and it follows instructions very well. So you can change a template once, and then the next time it does it, it does it right every single time, right? That's what I love about AI. Whereas like if you train any employees or something like that, it's a lot harder because humans on a day-to-day basis they might not be feeling it or they might be tired. Um, and then also sometimes they forget. But AI, um, if you just teach it once, it'll do it properly. So basically, we have like the order of events, like a little overview of uh of what we're doing, and then we have the script itself, and then this is a template that I made for myself that I find is the most useful. So we have like what I'm saying, and then we have what we're gonna show. So this is uh the highlighted parts are uh my dialogue, and then a lot of it, there's like edit nodes and stuff like that. This completely AI, but um it's not AI slop because what differentiates good AI from bad AI is the input that you give it, the training that you give it, and then the templates that you give it, and then lastly, is just your revisions on the outputs, right? So this template has been iterated by me for like probably like 10, 15 times, right? This is like the 15th version of this template. I went through a lot of excruciating detail on like how it should write. Um, and I think now it's gotten to a point where if I tell it to write a segment, it's like 80 to 90 percent there. And then I could just come in and um like do my own little nose and adjustments. And uh if I do something during the uh filming that I uh where like if I if I'm filming and then I deviated from the script, I can leave comments here and then the AI can come and change the script. So it's basically like my assistant.

SPEAKER_00

Yeah, and I would imagine that this has saved a tremendous amount of time. If you had to like estimate per video, how much time do you think this saves?

SPEAKER_01

Oh, a lot. Um, I mean, this is like because like this is not just the script, this is um the setup, right? Like yeah, tells me like what I need, how to do things. Um it lays it out in this like format, right? If I like the most annoying thing is like if you were trying to come here and then you like color this and then like format this, bold that, like that's actually annoying. Uh and that takes a lot of time, but AI can just do it right uh right away. Um and then we have like multiple segments, so I don't know, per video, just a scripting process alone. Usually I would take probably like a day or two to script it uh with AI that comes down to like maybe two hours.

SPEAKER_00

It's a huge savings, it's a huge savings.

SPEAKER_01

What do you use? The thing is, the the output is better than if I scripted myself because I can focus my energy not on the minute details, I can focus my energy on the creativity element of it. Like I can think about like, okay, how can we make this uh more interesting to watch? What else can we add into it instead of like worrying about all the formatting and stuff like that?

SPEAKER_00

I think this is this is important to show because for the most part, when people think content creation AI, they're thinking like thumbnails and everything else, and a lot of content creators push back on that. But for something like this, and even scripting some people won't like, but but for something like this, you're saving so much time that even if it's only 70 or 80 percent there, you've saved enough time to get to that point that the human element, the human touch on the end is the part that's probably the most important anyway. Do you agree?

SPEAKER_01

Of course. I mean, I like you cannot fully rely on AI to give you the final output. I think that's where a lot of the error comes from. Where like you trust AI to give you a hundred percent of the output, at that point it becomes slop because you're not adding your own creativity into it. And then also AI kind of like degrades over time, meaning like if you just uh regenerate the same thing, uh same thing, like it kind of just like the output just kind of degrades. Uh but if you are adding a human in the loop, uh if you're giving feedback and then you're improving the template every time you work, then AI actually improves over time.

SPEAKER_00

And what so if you are a new content creator right now, what would be the things you would, and again, not not with a lot of um, you know, not with a lot of experience A, and but more importantly, not a lot of like money. What would you be using AI for to start a new channel if you were a new content creator, you're just trying to get started?

Human Taste Versus AI Slop

SPEAKER_01

I would think about this, right? This is how I thought about it. I would take a look at your production process, um, take a look at everything that you do on a daily basis to make a video, and then figure out where do you spend the most time? And I would start automating those with AI. Because, like, why would you automate something that only takes you a couple minutes a day, you know?

SPEAKER_00

Right, exactly.

SPEAKER_01

There's no point. So, like, you try to get the highest return on your investment.

SPEAKER_00

Yeah. Are you excited for um what will likely be the next kind of big step for AI, which would be editing? Is that something that you're looking into, or are you starting to use any AI editing at all or anything like that?

SPEAKER_01

Um currently we haven't fully used it in editing. The only thing that we've um used it in editing is for like the the scripting notes. Um but I do see some like uh videos on like remotion where like you can add graphics and you can edit the um uh video itself, but I just don't think the output is at the quality that I want yet. So I'll just wait.

SPEAKER_00

Is that now this is a good point because I really want to understand your thoughts specifically on this because you're you're kind of on cutting edge. I've talked to a lot of content creators your size, bigger, and lower, and they aren't using AI the way you are. They aren't they aren't thinking about it the way you are, which I so I definitely want to ask your your opinion on this. Uh editing, is that the one of the last portions of the art that might get taken over by AI? Because right now, thumbnails can be done by AI pretty well, scripting can be done, the acting can't be, maybe that's the last thing. But the last kind of thing that makes the story important is editing. And while it's maybe not there yet, it we might be only a year away from that. Is that gonna be one of the few last things you would actually give to AI, or are you looking would you be looking forward to it? I I just want to understand your thought process.

SPEAKER_01

I mean, yes, that that's a step of the process, but uh to me, okay, here's what I think, right? I think as AI gets better, the main differentiator between a good creator and a bad creator is their taste and their experience. Because if a normal person, uh like someone who's never made a video in their life, just used AI and told it to make them a video, the output is gonna be trash. Because they can't differentiate between good and bad. They don't know what a good video is, and therefore they can't give the feedback necessary to make it good, right? Um, but then a creator like me, who has had nine years of experience, we've hit like multiple hundreds of millions of views on videos, right? Like I know the good and bad. So if AI gives me like a really bad output, I can tell it to change certain things. That process is the main differentiator. Uh so I don't think AI will replace us creators, it's just gonna empower us to create more.

SPEAKER_00

I think that's an important thing you just said there because a lot of creators are scared of AI and for for good reasons for certain things. But what you just said there is really critical is that we need to use it as a tool uh because that's it, that is what it is. Um, and it can actually help you create better content, faster. And some of the things that aren't necessarily important to the creation process that need to be done, uh that don't necessarily make a video uh, you know, that much more better than it would be if you didn't use it, but saves you a lot of time. Definitely a good use for AI. If you were to go back in time and the Hafu Go that was doing the college videos was still doing college videos, what would that Hafu Go use AI for, do you think? My homework. You get an A's on everything there, right?

SPEAKER_01

I mean, that's very much the time-saving aspect. I mean, I was juggling like midterms and editing, so I would have dude no mid term.

SPEAKER_00

That's great. Since the last time we talked, um, you you know, you have a team behind you, which of course not all content creators have. So uh, you know, I only want to just touch base on this. Uh, who do you have doing what for you now? Because every content creator I talk to has so, for example, I just talked to a content creator uh earlier today with um I think 17 million subscribers, and only in the last two Years that they hire anybody. It's the first time they hired anybody. But you have kind of operationalized yourself to a way that you can get more out of you. What did you hire? Who did you hire first? Like, what is your thing? And who's like what are the things that you always need to make sure you have someone hired

Outsourcing Editing With Real Systems

SPEAKER_00

for?

SPEAKER_01

Uh editors. That's who I hired first, and that's who I always need. Um, I think editing takes up the most amount of time for most creators. So that's the first thing you should outsource. And often when I talk to creators, their main pushback for hiring editors is no one can edit like me.

SPEAKER_00

Exactly. We hear that so all the time on this on this show, yes.

SPEAKER_01

And uh the reason why no one can edit like you is because you haven't uh you haven't systematized the way you edit and your thought processes. Um you make creative decisions on the spot, but you're not analyzing your creative decisions and why you made those decisions. Therefore, you can't teach that. So if you actually want to outsource your editing, you need to like really sit down and think through exactly why you do certain things and then templatize your editing. And then the second pushback is if I templatize it, it's gonna be boring or it's not gonna be creative anymore. Um, well, that the like the template is kind of just like the AI thing. It's like it gets you 80% of the weight there. It gets rid of all the mundane decisions you have to make so that you can focus your energy on the higher level creative decisions.

SPEAKER_00

I think what you just said actually is just important to anyone. If you can explain to someone else your editing decisions in the way that you are making a story, even if you're not looking to hire an editor, you can then look at that and see how to improve those things that you're doing. Because you might be making decisions on things that make no sense. Like I've just always done it this way and I've just never changed it. But if you write it out, you go, oh wait, I'm always switching to uh, you know, zoom in of my face every five seconds because that's just the way I've always done it, but it doesn't have anything to do with the story I'm telling. Maybe you can learn that by just writing it out. It's really interesting.

SPEAKER_01

An example of this, actually, another example of how we use AI. AI is really good at analyzing a lot of data, so volume of data. So uh we review all of our uh videos on frame.io. It's like a website where you can make comments, timestamp comments based on the video, right? And then uh basically I told AI to pull all the comments I've made on all my previous videos and analyze patterns. So it pulled over 6,000 comments. I can't do that by myself, or at least I don't want to. It pulled over 6,000 comments, but then out of the 6,000 unique comments, there were 10 points that covered 50% of my time reviewing. So these were like very mundane things. It was like brightening up the colors, uh, adding b-roll when we're doing explanation, uh, equalizing the sound and the my my voice sound and the music, like uh changing the color temperature so every single shot matches. These are like very mundane things, they add nothing to the creativity of the video, but they sucked up like 50% of my time reviewing, right? So then we systematized that and made it into an editing checklist so editors can go through that themselves before they even submit the video to me. And that frees up my time to actually focus on the creative storytelling and stuff like that.

SPEAKER_00

Well, if you're listening to this podcast and you're thinking, man, I wish I had an AI to do that. You do. Uh VidIQ has that exact feature. You can have it, uh AI coach will look through your comments, give you uh feedback. Uh, there's a link in the description below and in the show notes that gives you a special price on that. So you should go check that out. Um, let's move on to Wait, can I show you one more thing?

AI-Powered Research From YouTube Frames

SPEAKER_01

Um Yeah, absolutely.

SPEAKER_00

You can show me everything.

SPEAKER_01

This is a research that I was doing uh for Lego vehicles versus obstacles. And in order to do research for this video, we kind of have to know like what kind of obstacles do we need in the video? What kind of vehicles do we need? What has been done? What can we iterate on? And in order to do that, I have to go through hours of footage uh on YouTube to see like what's been done already and stuff like that, right? Well, um I got really lazy one night and I'm like, I don't really want to do this, I want to go to bed. So I told AI to do it, and this was the output of that. So uh what it did is um it pulled like 25 videos from YouTube uh that are related to this topic. Um pulled 25 videos from YouTube that are related to this topic, and then it downloaded this video, split it into like I think 2800 frames, analyzed each of those frames, deleted like duplicates, and then categorize all the frames. So they turning to these pictures. So now what I'm left with is a um output that has the vehicles and then also um the obstacles, and then if I click on one of these, it goes to the timestamped link of that video.

SPEAKER_00

What software did you do to use to do that?

SPEAKER_01

So this is just Claude. Um and like I wanted a visual output, so I just coded the output into like an HTML website. But like you can just like this this is not it's like a temporary thing.

SPEAKER_00

Um no, but it's crazy because this actually makes a lot of sense. Because for what you're talking about for this particular video, if there's an audio podcast, he's making this video about uh Lego uh vehicles and and impossible obstacles. What I'm looking at here is a screen share of a bunch of basically thumbnails of a Lego vehicle and something that's obviously like an impossible object. And when you go there's screenshots from the video itself. That's the most thumbnails. Yeah, right. There's screenshots from essentially the timestamp. You click on it, it goes to the timestamp, yeah, and you can see the rest of the video. How do you what's even the prompt for something like that? Do you have to code that? Like, how does that work?

SPEAKER_01

Well, uh that that's actually you don't have to code it. Um, I can kind of show you. Uh it doesn't like basically like in clawed code. Uh this is what I mean. I like got like pretty advanced with clawed code, but I still don't know how to code. So that's fair. So I still talk in English to clot code. Okay. Okay.

SPEAKER_00

And a lot of times you just tell it like what you want to see.

SPEAKER_01

Yeah, a lot of times you just have to define the output in a very concise and clear way. Um, and then clawed code will figure out how to do that. And then over time, as you've like had more experience with cloud code, you can kind of uh push it in a certain direction. So, like here, I already have experience like downloading videos and analyzing frames like that. So I told it to download the videos, um splitting into frames and analyze those frames. But if you don't know, if you didn't know to do that, Cloud Code would eventually figure it out.

SPEAKER_00

At some point we'll figure it out, yeah. Because it'll understand what you're trying to accomplish and then try to bridge the gap between what you're asking and where it needs to be. That's fascinating. It really sounds like in a lot of ways, this is indispensable for you now. Like it's it's a part of what you do at Hafu Go and the things that you are moving forward with uh would almost not be possible without it? Is that is that fair, or at the very least, it would just take a lot more time? It is possible.

SPEAKER_01

I mean, we've done it for nine years, so I wouldn't say it's not possible. It's just making my life a lot easier, and it's uh it's allowed me to do more with less.

SPEAKER_00

And does that, do you think that that in the end will elongate your creator life?

SPEAKER_01

Here's a good example,

AI Shrinks The Big Studio Gap

SPEAKER_01

right? Um Mr. Beast is a giant production company. They have like hundreds of people. I think last time they said that like 300 or 500, I don't remember what he said publicly, but like hundreds of people, right, across his companies. Um, I think in the future, like the gap between him, a huge production company, and a single person creator or a small team creator is just gonna close with the help of AI.

SPEAKER_00

I think you're probably right about that. It's the the ability, well, in a lot of ways, it's easier to make YouTube videos now at a higher level than it ever has been before. And that's probably always gonna be the case from this point forward. Once upon a time, um, you couldn't use your phone and have high quality video, and now with AI, incredible graphics and thumbnails, it wasn't possible like three, four, five years ago. Uh, now, literally, you take your phone out, you can have super high quality, and as long as you're a good editor with everything else and you have good audio, you can have an incredible video um for very little money. Um looking at the case.

SPEAKER_01

And the best thing, and the best thing here, um what I've really enjoyed about AI is is the fact that I don't need to hire people that I don't enjoy working with. Meaning a lot of the mundane tasks can be automated with AI, uh, which means the people that are in my company are the higher level creative decision makers. And I love having conversations and calls with them. Uh like all this research stuff can be done with AI. So we can just get together with the other creatives and we can analyze the research and talk about what we're gonna do with it rather than spend the time going on YouTube and doing these like mundane trivial tasks.

SPEAKER_00

And that this applies to any size creator, really, to be honest. What we're talking about now, which I think is really important to point out, it doesn't mean you have to have multi-million subscribers. Like, this is something you can do as a content creator who doesn't even have a channel yet. Like, you can automate all this stuff and find niches that work and find topics and and ideas that work right now. You can do that right now. There really is less excuses now than ever before.

SPEAKER_01

Yeah, it doesn't cost so much money. I mean, like the block subscription is what, like 20 bucks? Once you reach the limit, I mean you can upgrade to like a higher tier plan, but the most expensive one is like 200 bucks a month. Like that that's yeah, I mean, that's not that much money compared to hiring an employee.

SPEAKER_00

If you're investing in something, yeah. So what are the what are the employees that you have now? Like what kind of what are they doing? What what are the jobs? How many people do you have?

SPEAKER_01

Um, I have editors. Uh so I have long form editor, short form editor. Um, I still have some thumbnail designers who uh like touch up the thumbnails and also like do more photo manipulation. Um and then I have a ideation person that I work with um who like me and him go back and forth. And then um I have a manager, um and then uh producer. I'm actually looking for a producer. Um I I need someone who who's actually um like produced YouTube videos before who's really good. Uh ideally someone in LA. Uh LA or Toronto, either one.

SPEAKER_00

Okay. Well, if you're one of those people and you're listening, the uh maybe I'll have a uh link down in the description to help you uh get in touch. Or just email me. Speaking of email, email me. Or email email me. Yeah. Um we'll have an email link or something like that. Okay, yeah. Uh let me ask a question here that uh I've never really asked any of the other content creators, but it just came to mind because uh this is the second time we've talked, and um I I feel like you're the guy that if someone is on the come up and has the drive, you have a lot of the knowledge you could share with them. And I don't feel like you would gatekeep it. I feel like you would be a person who would share it. Am I right in that, first of all? Yeah, yeah, of course. I mean, I'm sharing all this internal stuff that I don't think one else is showing you. People in the company are like, what are you doing? So if that's the case, tell me if I was a up-and-coming content creator and I think my stuff is good, and I wanted to maybe not maybe not work for, but maybe work with or or get in contact or uh get your your interest.

Getting Noticed Through Consistent Improvement

SPEAKER_00

What would that be like? What would that look like? Would that be an email? Would that be a video? What would be something that would make you go, wow, I actually want to talk to this creator because I feel like I can either help them or something? What is something you would see in an upcoming creator that would make you want to take time out of your day?

SPEAKER_01

Get my interest to work for me or specifically.

SPEAKER_00

Well, specifically, you would you would either want to maybe not mentor because that seems like it's too high of a lift, but like you, if you let's say you're watching a YouTube video, you're like, this creator is really great. They just need a couple of things. I think I'm actually going to reach out to them and tell them something. What is it that would make you want to do that?

SPEAKER_01

What would you see? I think they need to have evidence that they're doing what they're saying. Because a lot of people talk a big talk, but then that's all they do. So what I need to do is I need to go on their channel and see that they've been putting in consistent effort and then see that they're actually improving every single video. Whether they're getting results or not is not that important. I just need to see that they're putting in the effort and also constantly learning and improving.

SPEAKER_00

That's so it's funny. You answered the question in a way that most other countries will answer how they'll collab with someone. And I think that that's kind of the same answer. It's like, oh, I want to collab with this larger creator, but I'm not that big. They probably won't re-answer out to me. I remember, listen, when I first started, I got like a 2 million subscriber channel, my first half a year to collab with me. But because of what you just said, not just that um I asked them and then I had a really great email that you know convinced them, but they were able to look at my channel and go, okay, yeah, this guy's actually really trying. He really wants to do this, and it's a quality that I'd be okay being associated with.

SPEAKER_01

Yeah, I think that's actually a great point. Um, quality that I'm okay with being associated with. Um if you're not there yet, keep trying, you know? Yeah, yeah, yeah, yeah. That's why I like learning is the most important skill because it's like a lot of people they stop learning once they graduate from school.

unknown

Yeah.

SPEAKER_01

But then what, you're like plateauing high school or university? Like your life is like over at when you're 20? Yeah, yeah. Uh yeah, yeah. So I think learning is the most powerful skill you can have. That's what our channel is about. That it's like if you're constantly improving, then a year from now you'll get so much better.

SPEAKER_00

Do you have a mentor or some someone that talks to you about YouTube that you look forward to, or do you look at certain creators and try to learn from them? Or are you just kind of doing your own thing?

SPEAKER_01

Uh yeah, I would say like I never had mentors. I like I would hear people say, get a mentor, but I'm like, where do I find one? Or do I even start looking? How do I get the never get? I just like learn myself from uh like watching videos or podcasts like this, uh watching what people are actually doing, so analyzing their content. And then also uh the most important piece of that is the uh iterative loop of doing something, getting feedback, and then improving and then trying something different. It's the experimentation, right? You have a hypothesis, test it, see if it's right. If not, what can you change? Test it again.

SPEAKER_00

Yeah, and I think as one of the final questions, that's really important as a content creator, especially when you're first starting out, because you don't necessarily have a lot of views. So it's hard to know if you've done something that was good because sometimes those videos won't take off for months or or you know, sometimes years, depending on the content creator. What would be a way that a content creator could self-evaluate uh in a in a way that's not you know uh overly pumping themselves up to know rationally that, hey, this video actually was pretty good. It's just that I don't have a lot of um back catalog for YouTube to know who the audience is yet. Because that's a hard thing when you're first starting out. You think everything is good that you're putting out, but that's not necessarily the case.

SPEAKER_01

Yeah, I think the algorithm right now, in particular, has been pretty good at promoting small channels. I've been seeing a lot of small channels on YouTube browse recently. Also, the same thing with Instagram. Um, with Instagram trial reels, you can get very uh fast feedback. Like I started a new account uh this week, and like my first couple of videos got hundreds of views, not like one. So, so like the the new accounts are actually pretty good at um being pushed out there. So you get some initial feedback, and also I would just say um maybe it doesn't matter because volume in the beginning is the most important. People say quality versus quantity. It's not a versus, it's a spectrum. And which one you focus on depends on what stage you are in your creative career. And I think when you're just starting out, you really need to focus on quantity because the more you put out there, the more you learn about what you like making, what you don't like making, because a lot of times at the beginning, you can't decide on a niche, you don't know what you want to focus on. So, like you just try all these different niches, see what sticks with you, but also maybe through that process, you'll see some videos get more views than others. So that's uh gonna tailor your direction. So, like the faster you can get that feedback, the faster you can learn and improve. So, like I would just focus on quantity in the beginning and just don't worry too much about like, oh, is this video good? Just like move on to the next one.

SPEAKER_00

I feel like you've given information and a lot of tips for people of all uh ranges, whether you're first starting out or you've been established. Um, pretty much even the high-end stuff with AI can be used at a beginner level. So if that was something that helped you, of course, hit that like button. There'll be information in the show notes for the audio podcast and in the description of the YouTube channel for more Hafu Go goodness. Uh what's a what's another video that uh we let me ask you

Dream Video Science Experiments In Space

SPEAKER_00

this. What's your dream video you haven't made yet that you'd like to make?

SPEAKER_01

I want to do science experiments in space.

SPEAKER_00

Well, it almost seems very possible nowadays. That's very interesting.

SPEAKER_01

I want to like the the coolest videos I watched growing up was um uh those like uh what was that guy's name? The the the guy that was in space. Um like the astronaut Rogers? I mean uh here you know you know this one? No, I haven't seen this. No? Okay. Well, uh, I want to do science experiments in space just like that. I think that's so cool because like as a kid, I would like to watch he he made so many of these. He's like, how do we use toothpaste? How do we drink water? How do we uh like eat and not leave crumbs? And all things are just cooler in space. We're actually in talks with NASA um to maybe do some collab videos, but obviously going to space is a lot bigger of a project than just doing a collab video on Earth.

SPEAKER_00

Get a good collab videos first and then try to work your way. Hey, by the way, I wouldn't mind going to space.

SPEAKER_01

But yeah, that's a dream video. I think that would be so cool.

SPEAKER_00

That would be dope. That's definitely a hundred million view video on the very low end. Uh and if you want to see that video, of course, hit that like button. Maybe NASA will see it and they'll put Hafu Go in the video. Or if anyone works at NASA, listening, please hit me up as well. Hit up Hafu. He would love to go. If you're new here, I got another interview just like this one. Fire from Hafu Go. Thank you so much, and we'll see y'all in the next one.