I Never Shot 90% of My Videos
45sThe shocking reveal that most of the channel is run by an AI twin challenges viewers' assumptions about authenticity on YouTube.
▶ Play Clip"Delivers a solid tutorial on AI clones, but the title's promise of '6K subs' is not substantiated with evidence."
The video demonstrates how to create an ultra-realistic AI twin using the Heygen platform, and explains how to leverage it for YouTube automation and personal branding. The creator claims that most of the channel's videos are actually produced by an AI clone, and provides a step-by-step guide to cloning your own avatar and voice, generating videos, and using advanced features like the AI Agent and translation.
The creator reveals that 90% of the videos on the channel are not shot by him but by his AI twin. He mentions that other creators like Varun Maya are also using AI twins to gain millions of views monthly.
The tutorial begins by directing viewers to the official Heygen website. The process involves clicking 'Create Your AI Avatar' and choosing between 'Create a Virtual Character' or 'Clone a Real Person'.
To clone a real person, you record yourself via webcam or phone. The recording must be stable, with high energy and expressions, as this determines the realism of the final output.
After recording, you can clone your voice from scratch or upload an existing voiceover. Heygen integrates with 11 Labs for professional voice cloning, using models like V3.
Heygen takes video consent before creating a clone, ensuring no one can misuse your likeness. You have full ownership, and your videos are not used to train public AI models.
In AI Studio, you select your avatar, voice, and the Motion Engine (e.g., the new 'crazy' mode) to control smoothness and realism. You can also adjust background, layout, and crop.
To avoid repetitive clothing, you can add a new look by uploading a new recording or using 'Design with AI' to describe a new outfit and background, generating a fresh look for your avatar.
The AI Agent tool allows you to generate entire videos from a prompt. You select an avatar, voice, and style, then provide a detailed prompt using the CSAD formula (Context, Script, Attachments, Details).
The AI Agent opens the project in AI Studio, where scripts, frames, and elements are auto-fed. You can edit, add music, captions, and templates, then generate the final video.
The creator shares two tricks: provide at least 1 hour of raw recording data (like Varun Maya) and exaggerate your voice and expressions during recording, as AI tends to dull the output.
AI clones save time on recording, enable higher distribution (more videos per month), and allow translation into multiple languages, expanding global reach and potentially increasing earnings 10x.
The video concludes that using an AI clone in 2026 is a powerful leverage for content creators, enabling them to produce more content, reach global audiences, and build a personal brand without the time constraints of traditional recording.
What is the first step to create an AI twin using Heygen?
Go to Heygen's official website and click 'Create Your AI Avatar'.
00:27
What are the two options when creating an AI avatar?
Create a Virtual Character or Clone a Real Person.
01:21
Why is it important to record with high energy and expressions?
Because the energy level decides the final output's realism; AI tends to dull the output, so exaggerated energy compensates.
02:01
What does Heygen do to protect your likeness?
It takes video consent before creating a clone, gives you full ownership, and ensures your video is not used to train public AI models.
04:01
What is the CSAD formula for AI Agent prompts?
Context, Script, Attachments, Details.
08:39
What are the two tricks to make your AI clone more realistic?
Provide at least 1 hour of raw recording data and exaggerate your voice and expressions during recording.
11:04
What is the benefit of using the translation feature in Heygen?
It allows your AI clone to speak in any language, targeting a global audience and increasing reach.
13:36
AI Twin Revelation
The creator admits that 90% of the channel's videos are made by an AI clone, demonstrating the technology's viability.
00:01Consent and Ownership
Heygen's consent process addresses ethical concerns about AI likeness misuse.
04:01Secrets to Realistic Clones
Providing raw data and exaggerating expressions are practical tips that significantly improve AI clone realism.
11:04Translation for Global Reach
The translation feature enables creators to expand their audience internationally, a key advantage for growth.
13:36[00:01] But the shocking thing is that I have not shot 90% of the videos of this channel. In fact, this entire channel is being run by my AI twin. In the same way, Varun Maya is also gaining more than a million views every month on his channel.
[00:13] And on YouTube you will find thousands of big creators who do not shoot videos themselves. And outside of that, you are also able to build your personal brand Because they are all using AI twins.
[00:27] how to create an ultra-realistic AI twin and how to use it to grow your multiple social media handles. Make sure to watch the video till the end because at the end of the video which will increase your earnings 10x. And even this line you just heard was also
[00:43] spoken by my AI clone. The real tube sensor is here. To create A2 in, first of all you have to go to the official website of Heijn from the link given in my description. Because Hazen is the only AI tool that generates professional level clones.
[00:55] And all the creators I have mentioned use Hazen only. When you visit the official site of Heijn, you will see an interface like this. Inside Hezen you can generate a video scene by scene. You
[01:09] or a complete avatar from a photo. And there are many such features in Hezen which we will discuss further. But for now you have to click on Create Your AI Avatar. Now here you have two options. The first
[01:21] create a virtual character. By selecting Create a Virtual Character, if you have created any AI avatar or you have an image of someone to whom you want to give motion, then you can do that work. But if you want to create your own clone then you simply have to
[01:34] click on Clone a Real Person. And in just two steps your clone will be ready. First of all you have to record yourself through webcam or phone. After scanning this QR code in the phone, you can record from your phone there.
[01:47] you can also use Webcam for recording. How to record now ? Let me explain this to you first so that you can generate a good result. But you have to keep your position stable. If you
[02:01] record while moving or back and forth like this, the correct result will not be generated. Therefore, you have to record at a stable place with high energy and high expressions. The higher the energy, the more realistic your video will appear. After this
[02:13] After which, you have to read the script that will appear on your screen you want your final output to be. Because your
[02:26] energy level will decide your final output. After this you can review your video. If everything looks fine then you can click on Create Avatar. Otherwise you n't like this video so I will record one again.
[02:39] And after doing this, you will again have two options in front of you. The agent is basically giving you the option to clone your voice professionally. You can either clone your voice from scratch or
[02:55] click on clone voice here, again you will have two options. record your voice, or if you already have a voice over, you can upload it here. And Hegen will use it to create your voice clone.
[03:08] Heijn has tied up with 11 labs. So if you get your voice cloned, only 11 labs will be used on the back end. Which you must know generates. If you are recording audio,
[03:20] noise. And your voice should be realistic. Not edited, without cuts etc. If you have a microphone, use it. your desktop. Currently, I had uploaded my voice, on the
[03:35] basis of which Heigen has cloned my voice using two models of 11 Labs. But just so that your time is not wasted, I will not wait here. And here I will continue with the voice generated by 11 Labs using the V3 model.
[03:48] Once your voice is selected, your avatar is completely ready. And the person speaking to you right now is again my AI twin and not the real one. And this AI twin I
[04:01] that someone else will use your facial identity because you have given your because you have a twin so it will be used. But this is not the case at all. Agen takes a video consent from you before creating your clone.
[04:15] To make sure that whoever is making the clone is making the same. Which simply means no one can take your image and make a clone of you. Secondarily Hazen completely protects your likeness. Which means that no one else can use your facial
[04:27] identity. You have full ownership of your clone as well as the Hezen also makes sure that your video is not used to train any public AI model. video is not used to train any public AI model.
[04:40] but let's also see how to generate a video. select AI Studio under the Avatar section. After which you will come to an interface like this. And then you will have an interface like this. This is a complete video creation
[04:57] studio. Where first of all you have to select the avatar on the right hand side and After this, you are given the option to select voice below. If you have linked any voice with your avatar then it will be shown to you here. Like in my case,
[05:10] After this comes the main part which will bring life to your video. Motion Engine. It basically decides how many motions there will be in your final output, how much smoothness there will be and how realistic the overall video will look. And Hegen recently launched
[05:24] works absolutely crazy. So for now I will select this one. After which, if you want to remove or replace the background, you can do it from here. And then you can crop the corners inside the layout and zoom in and
[05:37] requirement. Once you have fixed these settings. After that you simply have to either copy paste your script on the left hand side n't have anything, you can also record a voice over instantly. As
[05:52] I'm going to do here. After doing all this, you simply have to click on Generate Video. After which you will reach a new interface. Where your video will have started getting made. And in a few seconds you will have a result like this. Whatever he is
[06:05] show you. This is my AI avatar. And this is not realistic. And I am not visible to you people. And I'm recording this audio only from my MacBook. There is no microphone installed either. But this avatar of mine will lip-sync
[06:19] well. Can you see how perfect it was? From my hand movements to my facial expressions, everything was copied perfectly. Now I know you people might be thinking keep using it in every video then people will come to know because the clothes will be the
[06:32] same every time. Well, there is a session for this too. You simply need to go to the Avatar section under the Avatar tab. Here you will see all your avatars inside My Avatars which you have created till date. Here you have to first select your avatar.
[06:44] After which you are given the option of add a look here. Here you can either click on Upload a Look and upload one of your recordings again with a new variation of yours, or you can simply select Design with a.
[06:56] see an interface like this. Here you can describe what new look you want. Like, what changes do you want in it? For example, here I write I want to wear olive color polo t-shirt and the background should have
[07:09] shelf with books. And then I click on Generate a Look and in just a few seconds I have a new look for my digital twin. Which I can use in the same way as we used the old avatar. All you have to do is select your avatar inside AI Studio and
[07:21] this time select its second look and you will be done. Now I know you must be thinking that this is a great thing. Now you can generate whatever you want. But somewhere, if you still have to generate multiple shots within a single video, it will take a lot of
[07:33] you still have to generate multiple shots within a single video, it will take a lot of provided another tool that will not let you do even this much hard work. This is the tool of AI Agent. If you click on AI Agent, you will come to an interface like this.
[07:45] Now this is your prep area. You with Facecam. And obviously you will not have to do any recording for this. How to do this? Let me give you a quick tutorial. You simply have to
[08:00] select an avatar here first. After which you have to select the voice. And then you are given another option here which is of style. shared with you in this. You can select here in which style you want your final video.
[08:13] And even if you are an organization or a brand, upload it. So that your brand consistency remains in the video. After selecting the Now Style, you have to give a prompt in the Prep area as to
[08:26] what you want to generate. Usually people have a basic habit of having perfect expectations, which is wrong. You have to write a detailed prop for which you can use CS AD formula. According to this formula,
[08:39] first of all you have to give the context of what you want to generate, why you The better the contact, the more detailed it is, the better the result you will get. S means script. If you have written a script, some dialogues
[08:52] you can directly copy and paste it here. A stands for attachments. You can literally click here to
[09:04] need to include in the final video. And D stands for details. That means you have to specifically mention and what should be followed. After this you just have to turn on the seeding icon here. Due to which your video will be created using sequences.
[09:17] If you run the project in this way, Heigen's AI agent will understand your project. Will He will think accordingly. And then it will give you a final result which will be something like this.
[09:31] Content. They Fail Because Nobody Remembers Them. If You Want to Grow Online in 2026, Stop Copying Viral Creators and Focus on Three Things. Clarity, Consistency and Storytelling. Your audience should instantly understand who you help, what you talk about,
[09:46] and why they should follow you. One Viral Video Gets Attention, But Consistency Builds Trust, and Income. People don't connect with information. They connect with emotions and transformation. The Creators Winning Today Are The Ones Who
[10:00] Communicate Clearly And Consistently. So before posting your next video ask yourself. Will People Remember This Tomorrow? Because attention gets views. But identity builds a brand. Start Creating Content People Remember Not Content People
[10:13] Can you imagine how easy it was. And also very professional. I can run an entire channel without ever working on videos. That too only with the help of chat. If you want to change anything in the final output, you
[10:26] your project will open in AI Studio. And here your script, frames, AI spaces everything will be auto fed. Like you will find Bees in the media. In Elements you'll
[10:39] Music and captions and templates, everything is given here to generate the final video. After which, if you click on Generate, your final output will be ready to download. All your generations will be
[10:51] saved here in the project. Whatever you make inside the Hegen. Now if you think share with you guys the secret due to which my and Varun Maya's clone looks so realistic. Whereas
[11:04] all the AI models in the world generated by other creators work solely on the raw data given to them. Everything I've shown you in Heigen so far was recording in which I am doing natural movements, speaking, doing
[11:20] So Hezen will be able to understand my body language and my facial expressions very well and the final output will also be equally good. Basically, the more raw data you give to AI, the better results you will get. Varun Maya has provided
[11:34] recording as raw data to train his AI clone. I have also given 1 hour of recording to train my AI clone. I have also given 1 hour of recording to train my AI clone. you can see on Tube Sense channel and Praveen Finance channel. You have to use the same trick
[11:47] which will give you a much more realistic voice. The second trick is that you have to exaggerate both your voice and expressions. That means, if normally you move your hands only twice in 5 minutes, then this time move them four times. Whenever AI
[12:00] gives the final output, it always makes it a little dull. So if your raw data is a little more energetic than usual, then even after dulling it, the final output will be this will not happen with other creators and their videos will feel AI generated. If you
[12:13] do these two things then your clone will also look as real as me. Now let's move on to the part of the video for which you were waiting the most. I will explain to you logically how an AI twin or clone can Normally to record a video you need
[12:26] eye camera, studio, equipments and everything. After which, recording. And if you live in India then I am sure you will have can record inside your home. Otherwise you record the next day
[12:40] Due to which your final video also gets delayed. So first of all your AI clone because now you will not have to invest hours of time for recording and your script is ready then your recording part is also completed there.
[12:56] can directly give the video for editing. Due to which you get a lot of extra days in which you can do something else. The second thing is distribution. In 2026, the distribution. And the person who is seen more will be the only content
[13:10] creator the audience will remember. And if you process the recording manually, then no matter how hard you try, you cannot shoot more than 30 videos in a month. But Heisen Again solves this problem. Now if you want, you can generate
[13:22] And your face will be inside it. As a result, you will have the opportunity to appear in front of more people again and again, which will make them remember you. And third is value. I showed you all the features of Hezen but I did
[13:36] not show you one feature which is translation. If you use this feature, you can make your AI clone say anything in any language. That too very That too very realistically. As you can see right now. You can
[13:48] realistically. As you can see right now. You can
[14:09] generate CC videos in different languages and able to target the global level audience and your reach will also increase significantly. By doing just these You will increase your brand value. You will also increase your distribution.
[14:22] you will not have to record again and again for new channels and global opportunities will also open up for you. By confirming this, your earning will increase 10x very easily. I earning will increase 10x very easily. I
[14:35] In which I am using the faces of foreigners on more than half the channels. And the Foreigners don't even know they're looking at AI clones. If you're not using your AI clone in 2026, you're leaving a huge amount of leverage
[14:48] use it and defeat you easily. So start creating And if you want to know how people are earning lakhs of rupees per month by becoming AI influencers then do not forget to watch this video. See you in the
[15:02] then do not forget to watch this video. See you in the next video
⚡ Saved you 0h 15m reading this? Transcribe any YouTube video for free — no signup needed.