You Won't Believe This Video is AI-Generated

Today, an AI-generated video fooled thousands of users, including myself.

Good gracious, this one actually got me as I was just quickly scrolling past the original post. My eyes and brain are trained to spot AI content after over four years of working with AI — but this one slipped right past me.

If it fooled me, an average person doesn’t stand a chance.

It’s a testament to how convincingly AI can mimic reality, creating moments that resonate on a deeply human level.

The video, titled “Mother’s Love is Universal,” was first posted on a popular subreddit, where it quickly took off and got over 12K upvotes. Its theme of a mother’s love struck a chord with people everywhere, bringing out strong emotions and sparking a lot of conversation.

If it were you, would you assume that the video were not real?

The post is now removed on various channels, likely due to concerns about misinformation.

It’s impressive but also worrying to think about the potential of AI-generated videos to evoke genuine emotional responses.

These AI tools can be revolutionary for storytelling and content creation, but they also open the door to ethical issues around misinformation and manipulation.

How to Generate AI Videos

There are currently hundreds of AI video generators out there, but only a few can generate highly realistic results.

The most popular tools use techniques like text-to-image-to-video pipelines, diffusion models, and generative adversarial networks (GANs) to produce videos that look and feel natural.

In this case, it’s unclear which tool was used to create the viral “Mother’s Love is Universal” video, but some of the top contenders in the market are Runway, Kling, and Pika Labs. Each of these platforms uses advanced machine learning algorithms to produce videos that can often pass as real footage to an unsuspecting viewer.

Let’s try an example.

First, generate an image on AI image generator tools like Flux Labs AI with this prompt:

Prompt: A national geographic image of a mother bird protecting her five chicks from the heavy rain. They are all inside a nest on top of a tree. Mother bird’s wings are closed to cover all her chicks

Once the image is created, we can upload it into a video generator like Runway. Runway allows users to animate images, turning still frames into 10-second video clips.

Here’s the final video:

Although it may not look as realistic as the viral Reddit video, this approach is a great way to experiment with AI video tools.

Both Flux Labs AI and Runway offer free credits, so you don’t necessarily have to pay for these services.

You can also experiment with various prompts and parameters to improve the quality and realism of the output. Adjustments to lighting, motion intensity, and scene transitions can make a big difference in achieving a convincing effect.

How to generate AI sounds

If you thought the audio in the original video was from stock media, you’re wrong. There are now AI tools that can generate sounds from either a text description or a video reference.

Example with Eleven Labs

You can use Eleven Lab’s sound effects generator to create sound effects, instrumental tracks, and more by simply describing a scene.

Prompt: A mother bird in the woods protecting her chicks from the heavy rain

The AI processes the prompt and produces a few sample audio files. Each file gives a unique take on the description, allowing you to select the best match. The results are often impressively realistic, capturing the ambient sounds, background noises, and subtle shifts that give life to the scene.

This audio can then be added as a layer to the video, creating a synchronized and cohesive final product.

Google’s Video-to-Audio (V2A) Tool

Another way to create an audio file is by simply uploading the video and letting the AI analyze the scene and generate the sound for you.

Google’s Video-to-Audio (V2A) tool starts by encoding video input into a compressed representation. Then, the diffusion model iteratively refines the audio from random noise.

This process is guided by the visual input and natural language prompts given to generate synchronized, realistic audio that closely aligns with the prompt.

The tool is not yet publicly available but you can explore sample results on Google’s blog.

People on the internet are easy to fool

It’s surprising really, how quickly we can be tricked online.

I’ve noticed tons of videos featuring enormous, exotic creatures popping up on my Facebook feed, shared by friends and family who think they’re real — even though they’re clearly not.

And if you hop onto YouTube, it’s even crazier. There are these nature shorts with full-on narrations, showcasing “animals” that don’t even exist. But the comments? People are like “Wow, I’ve never seen anything like this!”

Here’s the thing: a lot of this stuff isn’t meant to inform or help anyone. The people making these videos do it for clicks, likes, and maybe a quick cash grab from ad revenue. And sure, some creators slap a disclaimer on there, but that’s rare. Most just let viewers believe it’s real, and people eat it up without realizing they’re being fooled.

According to reports, synthetic content could make up 90% of what we see online in the next decade. This opens a door to all kinds of messed up stuff — deepfakes, fake documents, non-consensual porn, etc.

Why does this matter?

So what exactly is my point, and why should you care?

The fact that it’s getting harder and harder to tell what’s real and what’s not is such a huge deal.

It’s one thing to watch an entertaining video, but another when you can’t trust what you see on the internet anymore. This tech is evolving super fast, and the line between real life and AI-generated “reality” is almost gone.

Years of research have brought us here, and now we’re looking at a whole new world of possibilities — and risks.

The question now is, does it make the internet better? or does it make the internet worse?

Final Thoughts

The debate around AI-generated content isn’t going anywhere. If you’re moved by a video, does it matter if it’s real or AI-made? That’s something we’ll all have to think about more as AI gets better at creating content that feels genuine.

Video creation is about to get way easier. Soon, anyone could make high-quality, unique videos straight from their phone and share them on platforms like YouTube or TikTok.

In fact, according to a report from European law enforcement group Europol, experts estimate that as much as 90 percent of online content may be synthetically generated by 2026.

Whether that’s exciting or unsettling is up to each of us — but it’s clear that AI is changing how we experience “reality” online.

‍