How to Spot Artificial Intelligence in Your Shorts
To Be…or Not to Be.
To ‘A’ or not to A…I, is certainly the question…everyone has been asking. It’s been at the center of one of the longest strikes in our nations talent union’s history ( Sag/Aftra). Because it presents an answer that could potentially end an era in the voice acting industry. (This is where, if we had the budget, we’d see the Fontaine Bleau over here come careening down…pretty much the same way the by-gone Las Vegas hotels did at the end of the movie, Casino.)
Ah, but before we start talking to the CGI folks over at Skywalker Studio’s, let’s take a closer look or better yet, listen to some differences between humans and machine. Here’s a link to a Utube Short of The Top 10 Largest Hotels in Las Vegas ….Click link and listen if you can spot the difference!
Then there are the not-so-obvious examples, like “Crazy Historical Photos You Won’t See Anywhere Else” with Tab Hunter and Gwen Verdon. The AI voice-over narrates Gwen’s sir name as if it’s separated in the middle — Ver’ Don ala Pierre Car’don. So, you hear two distinct syllables and in the end, something doesn’t quite sound right. The result is that it becomes less recognizable…unless it’s Gwen’s new way of pronouncing her name.
I also remember that as a generational ‘Boomer’, it was Jacqueline Bissett who set the stage to alternate celebrity ‘sir name’ changing. It seemed that every couple of years someone previously-know-as would step forth and make the announcement. Biss’set begot Biss’it, Francis Ford CaPolla became Co-pa-la. And we’re still not all that sure about Mr. Scorsese’ s preferred dinner call.
Or how ‘bout this gem from the same folks over at Crazy Historical Photos You Won’t See Anywhere Else, where the AI voice pronounces actor George Peppard’s name so mangled, that it becomes a WTF moment. George was pepper’ed rather than Pep’pard (ed). And no, he didn’t have a stutter. Just bad or no redundancies to detail on the production end. No brainer stuff — you know, “When the light is green the trap is clean”. We saw Ghostbuster’s four, last Sunday. Second time in less than a month, the theater was way under occupied.
The Most Challenging Training Ever explores what happens when there’s a give-a-way word or phrase that well, sorta let’s the AI outta the bag, so-to-speak. In this case it’s the word; string’gent. Click on the red title to listen for yourself. This brings to mind when I was a teenager in Holland. My travelin buddy, Hans Niehout touted a story or proverbial phrase about how a German could not say a certain word in Dutch without giving away his nationality. The word sounded something close to a contrivance like — schlaivningha.
Thing is — throughout the years I’ve run into several people who’ve confirmed & offered other nationality tongue twisters.
“As artificial intelligence continues to evolve, the law must progress as well…. Beyond state laws, there needs to be a push for greater protection of the common person against vocal deepfakes.”
This comes from the pages of IP Watchdog dot com. It’s entitled: Voices, Copywriting and Deepfakes. The article dives head first into the murky waters of protecting ones iconic voice. In short, is a voice copyrightable? Here’s the law.
And if you are in the music industry or the comedy industry…welp, get a load of this sh — ! George Carlin’s estate had to bring suit against some AI-Generated Comedy Special (George Carlin — I’m glad I’m Dead) over an AI-generated imitation of the late comedian. The creators agreed to remove it from their YouTube channel and podcast feed. SAG-AFTRA has pushed for federal legislation that would make it illegal to create a fake digital replica of someone without their consent.
In the music world, some 200 artists, organized by the non-profit Artist Rights Alliance, including Billie Eilish, Kacey Musgraves, J Balvin, Ja Rule, Jon Bon Jovi, Katy Perry, Miranda Lambert and more, are speaking out against artificial intelligence-related threats in the music industry.
The artists sent an Open Letter to AI developers, technology companies, platforms and digital music services to “cease the use of artificial intelligence to infringe upon and devalue the rights of human artists.”
The letter highlights AI threats including deepfakes and voice cloning, as well as “irresponsible uses of AI” such as the using AI sound to diminish royalty payments to artists and the use of musical works by AI developers without permission to train and produce AI copycats.
by:+Tom Clifford, VoiceOver & Creative Services
FaceBook Business Page
tomcliffordvo@gmail.com