Creative ways next-generation AI voices are being used

What is Replica?

In case you're not familiar with Replica, our product allows you to type some text and generate speech for use in your projects. You can choose from a variety of voices, all of which we replicated using AI.

Cool, right? But wait, there's more.

You can also use Replica to replicate your voice or your actor's voice to use in your project.

How can I use Replica in my existing workflow and what kind of projects can Replica be applied to?

Typically, when you're creating a project that requires a Voice Actor or Voice-over Artist, your process will look something like this:

1. Find your voice actor.
2. Negotiate a contract.
3. Send them your script.
4. They record your script.
5. You give them feedback.
6. They re-record parts based on your feedback.
7. You receive the files and use them in your project.
8. Payment made and contract complete.

Phew... This process could take weeks to months depending on the size of your project and, steps 5 and 6 may repeat a number of times.

You may also finish your project and realise days or weeks later that you need to make changes to your script, pushing your budget and deadline out further.

This is where Replica comes in.

With Replica, your process would look like this:

Using Replica, you can quickly experiment with different voices for your characters and script, and export audio to use in their projects.

Need to make a change to your characters script? Not a problem. Type or paste the new lines in, listen, tweak the pacing, pitch, emphasis and export it. All of which can be done within a matter of minutes, not days.

If you already are working with a voice actor but are fed up with how long everything takes, we can help you speed up your process by replicating their voice and using it in Replica.

That's cool, but what about the ethical use of such a product?

Since Replica's inception, we have been mindful of the ethical use of AI synthetic voices and how technology like this could be used for malicious intent.

First, we never enable public access to anyone's Replica voice without the owner's consent.

Second, we never let our users replicate anyone else's voice—we have security safeguards in place to prevent this from happening.

Finally, when we work with a voice actor and provide their Replica voice to any user of our product, there's a licensing contract in place to ensure they are adequately compensated for ongoing use of their Replica voice.

While this is a step in the right direction, we are conscious there is still more work to be done. The technology to synthesize human-like voice isn't going away any time soon and there will always be people with malicious intent. We're working with the voice acting community to find the best way forward, together.

It looks like you're going to put Voice Actors and Voice-over Artists out of business.

Quite the opposite. We think the market for voice is actually going to grow a lot. In addition to the regular work actors do in a studio, we'll also have a marketplace where talented actors and voice-over artists can license the use of their Replica Voices and make a passive income - these voices will be used for dynamic customisation and personalization of voice driven messaging.

Our goal is to allow anyone to create performances by telling compelling stories and the limits to the type of content one can create with a TTS product are boundless.

Now, let's go through some of the ways in which Replica is being used.

Animation as a medium for storytelling has been around for some time now. And while fiction podcasts are relatively new within the last few years, the art of storytelling through audio has been around since the 1920s.

Whether you're an indie animator, writer, YouTuber or a production studio one thing is for certain; Replica easily fits into your workflow reducing your production time and costs by allowing you to prototype scripts and create performances fast.

The following short scene was written and produced in under an hour with the use of three Replica Voices and a bunch of free sound effects from freesound.org

Audio advertising has come a long way since the first radio commercial hit the airwaves in 1922. Nowadays there's a plethora of platforms to advertise on and an overabundance of products competing for our attention.

To compete with each other, companies started personalizing their ads, so they're more relevant to their potential customers. For the most part, advertisements are based on the content you browse. So if you've been watching a lot of videos on YouTube about self-improvement, it's very likely you'll start seeing ads related to this subject. These ads will also drip into your favourite music streaming service and your email service, and so on. And with the advent of streaming video and audio, a single ad can be heard around the world in multiple locations and at various times of the day.

80% of consumers are more likely to make a purchase when brands offer personalized experiences

Cited: Epsilon study on personalization

Using the Replica API, you can take ad personalization to the next level.

For example, Grace has recently been looking for some new warm clothes for Winter. She's on her morning commute to work listening to her favourite music streaming service. Between songs, she is served an ad that not only recognised she was recently searching for warm clothing but also greeted her by name, time of day and recommends a clothing store that's having a sale nearby.

Here's an example:

Because the voice-over is produced dynamically using Replica API, an advertisement can be personalized on the fly to the audience in a way not possible before, thus making it feel much more relevant.

Using AI voices in games will empower developers to create a more personalized and unique experience for the player.

While AI in games is nothing new, the process of creating a voice for a character is much the same as animation. Therefore Replica can be used in the same way to help independent game developers tell their stories.

Some day soon, AI could be used in games to create unique, unscripted storylines and events from scratch as you play the game, ensuring that no two players ever experience the same thing. This will result in slightly different dialogue each time.

That's where Replica comes in. Using Replica's API developers could generate dynamic voice over to any number of dynamic events and scripts.

Here's an example of a short scene for a fictional game.

Voice-activated applications are used every day around the world. How many times do you ask Google or Siri to do something for you?

Once? Twice? Thrice a day? Maybe more?

Traditionally, voice apps have been used as personal assistants. However, this is only scratching the surface of the potential of AI applications.

Smart speakers like Alexa and Google Home allow users to do pretty much anything. You can find a restaurant, schedule a meeting, check your credit card balance, control the thermostat or even get Alexa to start your car. There is a voice app for pretty much everything.

As with smart speaker apps, there are no limits to what you can create with mobile and smartwatch apps in conjunction with AI voice.

For example, how about a guided meditation or yoga practice? There are already a bunch of great meditation and yoga apps out there with pre-recorded guided practices. But what if your Headspace or Calm app could be more personal?

How about a custom, personalized guided meditation?

Let's say Indigo has a smartwatch and it's monitoring her resting heart rate. The device notices Indigo's heart rate is elevated more than usual. Judging from the time of day, 10:30 am on a Monday which is usually when Indigo is at the office; the device determines using GPS that she is not at the gym and that because she isn't moving, she is not out for a run. Based on all of these variables, the device determines that Indigo may be under a lot of stress right now and gives her a gentle bump.

She taps her device and receives another message:

Indigo replies 'Yes'

She is then served with some options that she can select to help the AI determine what kind of meditation would best suit her.

For example; 'I'm at work', 'I am feeling inadequate', 'I'm under a lot of pressure'.

Based on these variables, the AI tailors a meditation that will focus on not only reducing Indigo's stress but also give her affirmations to help her cope with her feelings of inadequacy and boost her confidence and morale.

Now you might be thinking 'Okay buddy hold on, this sounds great but a bit too fantastical'. To which I say; yes it does. But it's entirely possible, which leads me to my point.

AI voice can be used for anything. Your creativity is your only limitation.

Apart from the right tools and skillset, to achieve any of the use cases covered in this article, you're going to need an AI voice that speaks naturally and sounds realistic.

Try Replica for free

If this has inspired you to create something using AI voice or you need an AI voice for your existing project, sign up for a free account at replicastudios.com and start creating today.