AI is Revolutionizing Video Creation. Here's Why This is Good for You

Becoming a video creator has become easier than ever.

Photo by Sebastian Pandelache on Unsplash

I love reading articles and listening to podcasts, but let’s face it, video is the type of content most people consume on a daily basis.

Whether it’s a short TikTok clip or a long YouTube video, video content is the go-to for both education and entertainment. That said, making videos isn’t easy. You need to buy expensive gear, show your face on camera, and spend hours making the video and editing it.

But guess what? AI is increasingly making these obstacles a thing of the past.

Thanks to AI now we can use avatars in our videos, create our own avatars, clone our voices, turn images into talking photos, and even translate our videos into multiple languages.

I’ve tried most of these AI features, and while they’re not flawless, the results are still impressive. Let’s have a look.

Can’t Talk? Let AI clone your voice

Now it’s possible to clone your voice using AI. This means a world of new possibilities when it comes to recording audio and even videos.

To do this, you need to train the AI with some minutes of your voice. The more minutes of your voice you give to train the tool, the better the results. Then the tool will create a realistic clone of your voice, and you can “record new words” by simply typing them in!

• Writers can turn their articles into audio using their cloned voice

• Video creators can write scripts and let AI speak for them

• Whenever you can’t talk, you can create audio using your cloned voice

What AI tools have these features? ElevenLabs and Descript

I’ve cloned my voice using Descript. To train the model, I used one of my 10-minute YouTube videos. Descript recommends 30 minutes of English-language speech in your voice though.

I’ve used my cloned voice to read the first sentences of this section. Here’s how it sounds.

Probably you’ve heard words such as “jibber” and “jabber” in the audio. That happens because I’m using the free plan that has a 1,001-word vocabulary limit. Any word that isn’t in this vocabulary will be replaced unless you upgrade to the pro plan.

Overall, I think this clone sounds a bit like me, but I might be biased, so if you’ve heard my voice in one of my YouTube videos let me know if it actually sounds like me.

This voice cloning by Descript is good but HeyGen has taken this feature to the next level. They’ve powered their new video translation tool with ElevenLabs’ voice cloning and the results are good.

AI Video translation is mind-blowing and will impact the creator economy

Imagine you could translate your videos with one click just like the way you translate text to multiple languages with Google Translate. Now that’s possible with HeyGen Video Translate.

To use it, you have to upload a video that is less than 500 Mb and 5 minutes long and select the language you want to translate to. Here’s one of my YouTube videos I translated from Spanish to English.

If I hadn’t watched the original video in Spanish, I wouldn’t be able to tell that the video was AI-generated.

I have to say that the translation is good and although the speaking style and voice generated don’t sound quite like me, the video translation is still good. Probably the voice doesn’t sound like me because when selecting the target language I chose “English-American Accent” instead of “English-Your Accent” (my bad).

Anyway, this is huge for content creators because the language barrier will be less of an issue. 

Here’s how content creators might use this tool:

• Non-English-speaking creators will use this tool to translate their videos into English and publish them on platforms like YouTube.

• English-speaking creators will translate their videos into multiple languages and publish them on YouTube.

• Online instructors will translate their videos to reach more students on platforms such as Udemy and Skillshare.

That said, this tool might not fit everyone. Here are some drawbacks.

First, this isn’t free. HeyGen has a $29 monthly subscription that gives you 15 credits per month (1 credit equals 1 minute of video). This tool might be expensive for new creators, but it’s still cheap if you compare it to professional video dubbing. Creators who are already generating income online will invest in this tool to reach a more diverse audience and increase their earnings.

Second, the tool isn’t perfect. After doing tests with my own YouTube videos, I found that the translations have some issues that currently you have to deal with manually. This is a video translated from English to Spanish using HeyGen. After some edits, the video translation looks decent, but the raw translation provided by HeyGen had a few bad translations, had long pauses between some sentences, and some words/phrases were oddly isolated/separated.

Last but not least, this video translation feature might not fit all types of content. This tool is a great fit for tutorial content, but I’m skeptical about its utility for entertainment, comedy, or gaming channels. The translation just isn’t good enough to preserve the video’s original vibe and emotion.

Don’t want to show your face on camera? Use AI avatars or turn images into talking photos

Now there’s no excuse to make video content. If you just don’t want to show your face on camera, use AI avatars!

To generate a video, you only need to create your own text and choose from the dozens of avatars available. These AI avatars will fit content such as tutorials, product videos, outreach videos, and training videos.

Here’s an AI product marketing video that HeyGen made for Notion.

These AI avatars seem very realistic, but if you want a less realistic avatar you can turn an image into a talking photo. Here’s an animated avatar I created using Midjourney, ChatGPT, and ElevenLabs.

Are you as impressed with these AI videos as me? Let me know in the comment section.