Presented by

What’s Even Real Anymore?

I think that 2024 will officially be the year where we no longer are able to discern between what is real, and what’s AI-assisted.

We’re already 98% of the way there.

On Friday I posted a video that is picking up steam on X and has been shared thousands of times across all platforms.

What I found most amazing about the video, is that none of it was possible for me even 9 short months ago. They were ultra-special capabilities reserved for expert ML engineers and seasoned Hollywood editors.

And on Friday – it took me only 3 hours from ideation to export.

Again. At the beginning of 2023, it wasn’t even possible. Two days ago – it took me 3 hours.

Progress this year has been dizzying.

Next year - we’ll see it elevate to a whole new level – and I’ll be here to keep you updated through it all.

The Process of Making the Video.

After posting the video, I was inundated with requests on how I made it. There was no easy way to explain it in a comment section or through DMs, so I made a video walking through it below.

If you’d rather learn through text – I got you. Let’s run through the steps below.

Video to Animation

Software used: Kaiber (Starting at $5/month)

Kaiber is a simple platform to use. It is also my favorite tool for transforming your videos into entirely new styles and aesthetics.

Simply choose flipbook to transform a video, upload your video by selecting “transform an existing video,” and then write a prompt.

This was the prompt I used to turn myself into an anime character:

Celebrity Voice Changing

Software used: RVC V2 (Free)

Changing your voice into a celebrity voice can be a little tricky. I would refer to the AI HUB Discord server for all things AI voice & music. It’s a great resource.

The software you want to use is called RVC V2. Here is a link to a free Huggingface version. For best results, you can duplicate this space.

You first want to go to the “Download Model” tab to download your preferred model. You can find a searchable list of models in the Discord I linked under the “voice-models” channel. This is how I found the Drake model:

Then, you want to copy the link to the model and paste it into the Huggingface space linked above and hit download:

Once you’ve downloaded it, go back to the Inference tab, refresh your models and select the appropriate one. Once you’ve done that, you drop in the audio file you want to convert and click on “Convert.”

And that’s it! For my video, I used a Drake, Tate and Trump model to get my desired outcome.

Translation & Voice Dubbing

Software used: HeyGen (Starting at $24/month)

HeyGen is remarkably simple and incredibly effective software. You can enter any video - and have it dubbed into whatever language you want. It even nails the mouth movements for the desired language you’re after.

To do this, start by clicking on their “Labs” tab and selecting Video Translate.

The next part is simple. Upload your video, and choose your target language. That’s it!

Scene Relighting

Software used: DaVinci Resolve (Free)

DaVinci Resolve introduced the AI-powered “Relight” tool with their 18.5 update. I learned how to use it with this simple 5 minute tutorial a few months ago.

Scene Changing

Software used: Photoshop Generative Fill ($21/month)

To do the scene changing, I used generative fill in Photoshop.

Basically, I took a screenshot of my scene, and used generative fill to expand the scene into the blank areas highlighted by the arrows below.

In Premiere, I then zoomed out from myself and into the entire scene that was generated in Photoshop:

Face Swapping

Software used: FaceFusion (Free)

For the face swapping in the video, I used OSS software called FaceFusion. It’s free to use – but requires a pretty strong GPU and an understanding of how to install Python and set up a dev environment.

The magic of FaceFusion is that you only need one single image as your source image to create a pretty convincing Faceswap.

There are a lot of web versions that let you do the same thing, but none have great output. FaceFusion is the best tool that I have found.

However, this is by far the hardest step out of them all for the average person.

Hence why I left it for last.

Below are the steps you need to take to install it to your machine. This is not an easy process. The FaceFusion Discord community can guide you in detail if you have any questions whatsoever.

1. Open command prompt as administrator: hit start and type cmd, than right click to command prompt and select "run as administrator"

2. Install python : command : 'winget install -e --id Python.Python.3.10'

3. Install pip : command : 'python -m ensurepip'

4. Install ffmpeg : command : 'winget install -e --id Gyan.FFmpeg'

5. Reboot computer and open command prompt as administrator again

6. Install Microsoft Visual C++ 2015 Redistributable: command : 'winget install -e --id Microsoft.VCRedist.2015+.x64'

7. Install Microsoft Visual Studio 2022 build tools. During installation, ensure to select the Desktop Development with C++ package. Command : 'winget install -e --id Microsoft.VisualStudio.2022.BuildTools --override "--wait --add Microsoft.VisualStudio.Workload.NativeDesktop --includeRecommended'

8. Open the folder you want to install facefusion

9. Type "cmd" in address bar and hit enter. it will open the command prompt with the folder path

10. Clone repository into folder : command : 'git clone https://github.com/facefusion/facefusion'

11. Create venv: command 'python -m venv venv'

12. Activate venv: command 'venv\Scripts\activate'

13. Install the dependencies: command 'python install.py’

14. install cuda toolkit 11.8: https://developer.nvidia.com/cuda-11-8-0-download-archive 

15. python run.py

16. facefusion should work on your browser at http://localhost:7860 

Once it’s installed, the process if fairly simple. You need a source image and a target video to swap the source image into, as shown below.

Make sure you have “face enhancer” selected for best results!

Note: This entire video took me 3 hours. The idea came to me at 6PM, and I was exporting the video by 9PM. That is a little misleading however. Because I already knew how to do everything above. So it doesn’t include the countless hours I’ve spent this year learning about AI and how to integrate it into my workflow as a creator. It also doesn’t include the years of experience I have video editing.

Once you’ve mastered the tools above once, you can re-use them at lightning speed once you get the muscle memory down.

By the way, the sound design and music for the video came from Epidemic Sound – which is my favorite sound resource as a video creator.

Why are all your favorite newsletters switching to beehiiv?

It’s because the founding beehiiv team were all early Morning Brew employees who helped scale that newsletter to over 4 million daily subscribers.

Years of trial and error went into building the precise tools, dashboards, and analytics needed to accomplish that. And now every newsletter on beehiiv has access to the same winning formula.

So what exactly does beehiiv offer?

• World-class growth tools like the referral program and recommendation network

• Monetization via the beehiiv Ad Network and premium subscriptions (i.e. beehiiv helps you get paid)

• Seamless content creation with a sleek collaborative editor

• Best-in-class inbox deliverability of 98.7%

• Oh and it’s the most affordable by a mile…

Take your newsletter to the next level — get started for free.