AI is everywhere these days. It's all over our social media feeds, and I think we are all curious about how this will affect our current industries and creative professions.
Today I wanted to talk about what is happening in the world of Voice Over or Voice Acting and Artificial Intelligence. It's progressing fast and already is changing the ways we are thinking about and produce content. How will AI affect how we work with human voices, in advertising, film tv, or video games?
TEXT TO SPEECH
You're probably aware of the way you have been able to have your phone speak back to you. Like Siri or Google assistant or having a block of text read back to you by Alexa or adding voice to your tik tok videos.
This is called text-to-speech technology. it getting better all the time but it still sounds like a robot is talking. When will voice-to-text start to sound more like a human?
VOICE SYNTHESIS/CLONING
The most interesting and scary side of this technology is in the area of voice cloning or voice synthesis or voice augmentation. being able to create new content using someone’s actual voice. Are we going to be able to trust our eyes and our ears ever again?
So many questions and ideas come to mind. How will this be used against us? how can we trust anything we now see or hear?
I'm not smart enough to deal with those questions so let’s chat about what this might allow us to do….
We can now take recordings of someone’s voice and use them to create a voice model that then can say anything we want.
APPLICATIONS
Let’s say I want people to take my blog posts on the road and listen to them on their commute to work.. I can do that right now with text-to-speech….but again it sounds like Siri or Alexa.
But what if I wanted them to hear them back in my voice….the author's voice?
Traditionally I would have to go to a studio or set up a mic at home and record them, edit them, and upload them. Well…..with the help of AI, I don't need to do that.
Eleven Labs is a company where I can upload some previous recordings of my voice and train it to make a voice model of my voice. It does a pretty good job.
LANGUAGE TRANSLATION
A mind-blowing aspect of AI and voice is the area of language translation.
What if I wanted to take my podcast and create versions of it in different languages….but still sound like me? No problem. See video.
Traditionally I would have to transcribe my podcast, translate it into the language, hire a voice actor, and then record. Now AI can do this at a touch of a button.
FILM
Think about how this changes translating your film or tv series into different languages—no need to find actors that sound like the actors on film.
What about ADR? Automated Dialgoe Replacement in films? That is the process of re recording the actors in the film for scenes that were too noisy or did not pick up their lines properly during filming. We’ll be able to synthesize actors' voices as part of their deal and create models that can handle their own ADR without them being in the studio. They already are using this tech in Hollywood.
In the past when they needed to recreate lines or recreate an actor's voice they would splice together old dialogue to create new lines. See video
But now with AI, they can recreate even Young Luke Skywalker, despite the fact that Mark Hamil is still alive. See video.
VIDEO GAMES
An area where AI and voice will thrive is in the Video Game and Animation industry. Imagine having your voice actors ready to go even before you write a line of dialogue. Being able to test scripts out with the actual voice of your character. To be able to iterate on the fly!!!
speaking of on the fly there are tools now that can mask our voice and help us sound like our favorite podcaster, actor, or politician….IN REAL TIME!!! See video.
APPS AND ADS
Let's talk about Advertising! Draft Kings, the leading betting app just signed a deal with comedian Kevin Hart to synthesize and use his voice likeness for their app. We will be hearing Kevin’s voice speak to us in real time, based on our clicks and betting decisions. Up till now, he would have to record every response but now that they have his voice synthesized it can generate any type of response in real-time….and sound just like him.
VOICE LICENSING
This opens opportunities for a voice talent to sign deals to license their voice likeness to a company or brand. To get a lump sum for a brand to use their voice for a set period of time, and never have to go to the studio. Imagine AI news anchors giving news in real-time with a voice from a professional voice actor. Or imagine our ads and apps speaking to us IN REAL time based upon our decisions.
TEMPING
Another use case would be for agencies to use AI for temping. Many times in advertising the animator or editors like to work to record the script for the spot or video. Usually, it's someone in the office who will read a “temporary” version of it.
SUMMARY
So where does this leave us? Should all my voice artist friends retire and find new jobs? Hell no! For the foreseeable future, most established voice artists will be unaffected. In fact, for them…..AI has the potential to create new opportunities for them.
Most ads and brands will still require hiring real voice actors…. to have that back-and-forth and be able to dial in that perfect take. AI still struggles with nuance and the tools we have to finely tune words or annunciations aren't there…….YET. I stress yet.
Where I see AI being used most is in the areas of pre-visualizing ideas and concepts (like video game characters and temping), where people need mass amounts of content created like audio books, training videos, and real-time content that needs to be created cheap and fast.
Our culture continues to choose convenience over quality and our clients are usually no different …but the benefit of speed will allow businesses to save time and money. To be able to iterate ideas fast is what will set companies apart and help them survive this next wave. Progress is inevitable and we need to figure out how to change with the times and use AI for our good. But Can it help us create better art? Will AI help us get to the final version faster? or Will it cause everything to become vanilla and lifeless? How do we protect our industries and our craft? So many questions to ask and dark rabbit holes to go down.
Let me know what opportunities you see and what things we need to be careful of by commenting below.