Meta Adds More Voices for User-Created AI Chatbots

While Meta’s not making a big noise about it, it is quietly adding more and more generative AI bot creation and interaction tools within its apps, with users now able to converse with AI characters on Facebook, Instagram, WhatsApp, and Messenger.

And it’s not just text-based chat, now, you can also create an AI chatbot with a voice, with an expanding range of voice options to consider.  

As you can see in this example, shared by app researcher Jonah Manzano, now, within Meta’s “AI Studio”, you can create a chatbot with not only custom personality traits, but also a custom voice, with Meta recently adding a lot more voice options to choose from.

As you can see in the second image, Meta’s even separated its different voice options into their own categories, making it easier to find the exact right tone, and pitch, for your AI persona.

Meta’s been developing its custom AI creation tools over the past year, for which you can also add interests, defined conversation parameters for your bot, and an avatar image.  

Last June, Meta CEO Mark Zuckerberg provided an overview of the company’s broader plan to enable the creation of “AI agents” in this way, which he says will eventually become key connectors for niche interests.

The main focus of these bots, according to Zuckerberg, will be to answer fact-based queries, with the more challenging element coming in answers that are more creative, and replicate the style of the creator. Zuckerberg says that creators will soon have the freedom to train their bots on different aspects of their social media presence, and through this, that should enable them to generate more life-like replicas of themselves.

Which also includes voice, and with so many options, you should be able to find a voice that matches your chosen style, and then enables people to have more present chats, even potentially through Meta’s Ray Ban smart glasses.

The update is another step in Meta’s broader plan to integrate many, many more AI characters into its apps, in order to facilitate more engagement.

Late last year, a Meta exec outlined the company’s vision of unleashing millions of AI-powered profiles into Facebook and IG, which will post, comment, and Like, just like real people.

Chat, then, is just the beginning, and with Meta investing $65 billion in AI infrastructure this year alone, you can see the framework being established for a transformative shift in how we interact in its apps.

So will that be beneficial?

I mean, many people are already chatting with tools like ChatGPT, using it’s a sounding board and a companion to check their reasoning, and come up with different perspectives on private matters. Logically, that could also extend to Meta’s AI bots as well, but then again, Meta’s last experiment with celebrity-influenced AI chatbots definitely didn’t work out like it hoped.

Back in 2023, Meta rolled out a range of chatbots on Messenger, which provided responses that had been guided by stars like Kendall Jenner, Dwyane Wade, MrBeast, and many more.

Very few people used them, and less than a year after their initial launch, which had been the major announcement of Meta’s Connect Conference, Meta shelved the concept. That’s also despite paying millions to these celebrities to use their likenesses.

But Meta’s now re-framing the concept, given the broader interest in AI, and it’s going to add in many, many more bots, by enabling anybody to create one.

It honestly doesn’t seem like a viable alternative, as it’s just doubling down on a concept that didn’t catch on in the first place. But then again, using Meta’s improved AI models, maybe now, people will be more enticed by these bots, and creators will also be more interested in experimenting with their own bot creations, which they’ll then promote to their audience.

Meta’s ultimate goal, as noted, is to increase in-app engagement, by getting more and more people to develop ongoing chat relationships with these AI characters. And some of them will undoubtedly catch on, and that seems to be the key revision in Meta’s approach here, in enabling everybody to throw their AI bot ideas at the wall, so it can then see what sticks.

And some will stick, with some users. I’m just not sure that they’ll see enough usage overall to justify their creation. But again, if Meta can eek out more engagement, and time spent in its apps, via these tools, it could well be worth it.

Of course, there may also be mental health impacts related to developing relationships with non-human entities by engaging with them like they’re real people. Though that’s a small consideration within Meta’s broader picture, and I’m guessing this won’t be an impediment to its AI progress.

Though I do think Meta AI characters should come with cigarette packet style warnings, alerting people to the potential dangers of such before they engage.

That’ll probably happen in retrospect, once more people have lost their grip on reality by establishing strong bonds with digital characters.

Till then, Meta’s seemingly pushing full-steam ahead.

UPDATE: When asked about the expanded availability of these tools, Meta replied that its AI Studio is now available Bolivia, Brazil, UK, Australia, Nigeria, Cameroon, Canada, Ghana, Guatemala, Jamaica, Malawi, New Zealand, Paraguay, Philippines, Singapore, Thailand, Uganda, Vietnam, Saudi Arabia, UAE, Iraq, Jordan, Libya, South Africa, Algeria, Morocco, Tunisia, Sudan, Zambia and Zimbabwe. It’s also several languages, including Hindi (Romanized), Tagalog, Thai, Vietnamese, Portuguese, Hindi, Indonesian, Arabic, Arabic (Romanized) and French.