Day 24 of 30 Day AI Challenge – Supercharge Your Social Media Profile with an AI-Generated Video Intro

For those just joining us, every day during this challenge, I’m going to try and do something different or better, using currently available Artificial Intelligence tools.

Today’s challenge will use a combination of tools to create an intro video for social media profiles, using your own voice and a representative image, but without actually having to record real video or audio. Let’s get started…

First, The Intro Copy:

For this we’re going to go with ChatGPT, feeding it a summary resume from LinkedIN and asking for a succinct but impactful intro draft that we can tweak.

With a few more tweaks and personalization, here’s where I ended up:

Hi, I’m JJ! Great to meet you!
I’m a hospitality leader with over 20 years of experience in hotel operations, marketing, and digital strategy.
I bring a unique blend of hands-on and strategic expertise to my field and have a passion for loyalty, marketing and new technology.
I have led efforts across various parts of the world, including Europe, Asia, Africa, and the Middle East, delivering data-driven marketing and bringing innovative loyalty programs and strategic partnerships to life.

Second, The Profile Picture

For this bit, I tapped into Midjourney to create a somewhat recognizable photo of myself, in different styles. This step proved to be quite difficult – despite providing the bot with an actual profile picture and varying prompts, the results almost always ended up looking like a different person. Here’s one I finally settled on, as a somewhat lovable Pixar style character…a tad younger that I’d like but probably the best of the lot (I’ll spare you the horrors of the other images generated).

Third, The Voice Recording

Now, I’m not a big fan of recording my own voice or videos, so I decided to use ElevenLabs to clone my voice. I provided the system with audio inputs to create the voice and then used the generated voice for my audio recording.

The first time I tried this with a short 30-40 second audio recording, the voice generated was decent but wasn’t quite as similar sounding. The second time around I provided three separate recordings with variations and the result was a lot better. You can also tweak the Stability and Clarity/Similarity settings as needed.

Fourth, The Video Recording

For this step I used D-ID, which does a tremendous job of creating videos with available or AI generated characters and voices, plus text or audio file inputs.

The system is pretty quick and the results quite good, especially when the quality of the character and audio inputs are high.

Finally, The End Product, Ready For Use

Here’s what the final video recording looks like:

And here’s what the video looks like on LinkedIN, after I uploaded it:

All in all, I felt pretty satisfied with the process and final result. I look forward to hearing from you…have you tried any of these yourself? Any tweaks or improvements to workflow or prompts that gave you better results? DM or leave a comment to let me know!

Till tomorrow…

Important Note: Given the power of generative AI, it’s important to use it responsibly and ethically. The use of AI-generated images, voice, and video, as seen in this blog post, should only be used with your own personal identity or with explicit permission. It’s crucial to respect privacy, comply with any applicable local laws, and use common sense when utilizing these AI tools.