Co-written by Howard Fields and Nita Drakatos, Voiceover Artists
image by Inspiring Alley
Unless you’ve been hidden under a rock for the last 10 years you’ve undoubtedly heard speculation that AI is bad for humanity and by extension, creative endeavors like voiceover. This blog piece offers a counterpoint suggesting that things are never as good or as bad as they seem. Frankly, we think the fear expressed in a recently-published survey may be overblown.
What does this have to do with voiceover? Everything. According to BloombergOpinion data gathered by a recent “Stop Stealing the AI Show” survey conducted by Equity, a UK union of 47,000 performers and audio artists, 93% feel that AI poses a threat to their careers.
This blog article speaks to the difference between human and AI voiceovers and why we believe there will be demand for human voiceovers for many years to come. It does not address the potential unfair practices or misappropriation issues that concern many voiceover artists. For more, see the Equity article here.
There are two important elements missing from this discussion. The first is the time it takes for technologies like AI to be perfected. And the second is any consideration of the human-only factors that create emotion and feeling.
So, let’s connect some dots and view this from another perspective… What do Taylor Swift, Japanese virtual pop star Hatsune Miku, ABBA, and the Gartner Emerging Tech Hype Cycle have in common?
Answer: They offer lessons and demonstrate that…
- You can’t stop progress.
- Perfection takes much longer than everyone expects.
- AI won’t be capable of replacing human creativity and emotion for quite a while.
Consider the following:
Lara Williams of “BloombergOpinion Today” recently published an article regarding Taylor Swift's newly-released album “Midnights.” Williams examined whether AI could be viewed as a threat to songwriters. Bloomberg’s conclusion, “there are other factors at play, including her celebrity status and large social media presence.” One of the factors Bloomberg missed is the depth of the ‘shared experience’ between Taylor and her fans.
Japanese pop star Hatsune Miku (translated as: First Sound From the Future) is a computer-generated hologram of a 16 year-old girl with blue pigtails who has more than 2.4 million Facebook followers. In this case, it’s not just the technology but the ‘shared event’ between Hatsune and her audience that creates a feeling of connection and a unique immersive experience. However, she does not have the same emotional pull as Taylor Swift.
Abba Voyage while also an event, is enhanced by a ‘shared nostalgia’ of the Swedish rock band during their prime the 70’s. The production, which requires a special venue, is a clever reverie that employs technology and the emotional pull of past personal memories to create a new concert experience.
The Gartner Emerging Technology Hype Cycle, is an objective roadmap that charts the path of emerging technologies as they mature in the marketplace. Innovators, investors, and consumers, use it to identify, predict, and strategize the timing and lifespan of emerging innovations. The cycle illustrates multiple inflection points. For example, the shift from overstated expectations to disillusionment when they are not achieved, to ‘shared adoption’ when the real value is determined.
Shared reliance on AI is a shared thread between the four comparisons made above. The visual diagram demonstrates how emerging technology evolves over long periods of time, in waves. It’s predictable. It’s repeatable, and it’s exactly what will happen with voiceover AI.
Shared experiences, nostalgia and adoption create emotions and feelings, but eventually human beings recognize the difference between the artificial and the real. Yes, innovation improves over time, but perfection usually takes decades.
Which brings us to the voiceover bogeyman of the day, AI, and specifically, the branch of artificial intelligence known as machine learning. Like almost every other emerging technology, voiceover AI, while functional today, will take a long time to be perfected. Text-to-Voice technologies like IBM’s Voicetype/ViaVoice products or Dragon Speaking Naturally, have been commercially available for over 25 years and yet, Siri and Alexa still need to be programmed to answer us with all those pithy responses. We all know and accept it, while wishing it was more of an authentic comeback. Perhaps knowing that it’s fake leads us to reject it more.
When you constrain the requirements, today’s technology may be adequate. But the real world is not constrained and that’s where the paradigm breaks down.
Think about it, we can easily envision AI prompts for constrained applications such as IVR telephone prompts, eLearning quizzes and feedback, digital surveys, lists, menus, and more. It’s a lot harder to imagine that AI - at least in the next ten years - will be able to convincingly persuade us to buy Nike running shoes, make us cry over an audiobook story like Never Let Me Go as read by Rosalyn Landor, or shock us by pointing out the absurdity of news as covered by John Oliver on "Last Week Tonight". AI may someday make us FEEL; but not anytime soon.
So, to finish connecting the dots, remember that:
- You can’t stop progress.
- Perfection takes longer than everyone expects.
- AI won’t be capable of replacing human creativity and emotion for quite a while.
Voiceover AI has a place today and, good voice actors are, and will, continue to be in demand for the foreseeable future. Which gives us the opportunity to hone our craft and continue to deliver inspiring performances which engage audiences.
THIS BLOG WAS CO-WRITTEN BY HOWARD FIELDS AND NITA DRAKATOS. WE ARE VERY INTERESTED IN THIS TOPIC AND KNOW MANY OF YOU HAVE OPINIONS REGARDING THE FUTURE OF AI AND VO. WE INVITE YOU TO COMMENT AND SHARE YOUR THOUGHTS, IDEAS, PREDICTIONS WITH US AND OUR AUDIENCE. THERE ARE MANY VOICES THAT NEED TO BE HEARD ON THIS TOPIC. PLEASE CONSIDER SHARING WITH YOUR NETWORK AS WELL.