Despite all the excitement surrounding AI and its clear potential in numerous applications, it appears that social media platforms are still struggling to find genuinely valuable use cases.

Perhaps I’m missing something, but for me, AI bots that respond in the style or voice of celebrities—like those introduced in Meta’s latest update—just don't cut it.

Meta’s initial attempt at creating celebrity-themed AI chatbots in text form didn’t resonate with users, leading to the shutdown of that project.

After all, the concept isn't particularly compelling, right? Engaging with a chatbot that gives the impression of conversing with a celebrity loses its appeal once you realize that it’s not truly that celebrity—or any human, for that matter—on the other side. It’s just you interacting with a programmed response system.

This holds true for celebrity-voiced bots as well. Whether it's a voice mimicking Billie Eilish or just generic responses attached to a standard profile picture, the answers will remain fundamentally the same. Sure, having John Cena respond to random questions might be amusing with friends initially, but that novelty will fade quickly. Ultimately, it turns into a solitary experience, with you trying to convince yourself that you’re actually “friends” with Kristen Bell as her robotic voice echoes in your mind.

Moreover, these interactions lack the essence of social engagement as we understand it. The term "social" typically refers to human connection, which is the foundation of social media itself.

In fact, these offerings are counterproductive, as they foster interaction with a non-human entity.

While it's possible for people to develop a certain bond with an AI system (and many already do), this can be potentially unhealthy. It's something we should approach cautiously until we fully understand the implications and risks involved.

But this isn’t the limit of Meta’s ambitions with AI bots.

During his Connect keynote, Meta CEO Mark Zuckerberg also unveiled the next phase of the company’s AI character platform, which will eventually allow creators to develop AI representations of themselves in video format.

So, you could be video chatting with your favorite celebrity, and they would be able to respond to any question you ask them.

Sounds cool, right?

But once again, this isn’t truly social, as you’re not interacting with real people. It’s essentially a video game featuring recognizable characters that increasingly resemble real-life individuals.

This concept strays far from the traditional appeal of social media.

One of the primary advantages of social media is its ability to connect you with people from diverse backgrounds, including celebrities. In the past, your chances of interacting with your favorite stars were virtually nonexistent. Now, you can send them a post or DM, and there’s a slightly better chance they might actually respond.

This interaction has significantly enhanced fan engagement and helped build audiences. However, soon you might find yourself unsure whether you’re communicating with the real person or an AI bot. This uncertainty could ultimately diminish the entire experience.

As reported by MIT Technology Review:

“Since 2022, a wave of Chinese startups and major tech companies has been providing services to create deepfake avatars for e-commerce livestreaming. With just a few minutes of sample video and a cost of $1,000, brands can clone a human streamer to operate 24/7.”

In this setup, businesses simply provide the script and product details, and the virtual streamer handles the sales on their behalf. More sophisticated versions can even analyze comments for specific responses and interact with viewers, adapting their approach based on the number of viewers present at any given moment.

There is certainly a demand for these digital doppelgangers. However, live shopping hasn’t really taken off in Western markets, and it doesn’t seem like this development will be the catalyst to change that.

Essentially, Meta is still experimenting with AI concepts to see what resonates. However, I struggle to understand how moving away from the "social" aspects of its products could be a sustainable strategy for future use.

Encouraging more bots to impersonate humans could lead to a less engaging social media landscape. Meanwhile, Zuck himself recently stated that he believes:

“Every part of what we do is going to get changed in some way [by AI]. [For example], feeds have shifted from being primarily about friend content to largely creator-driven. In the future, a significant portion of it will be AI-generated.”

So, Meta envisions an increasing presence of AI-generated content in our social feeds, one way or another.

But is that beneficial? Will it enhance engagement?

AI serving as a practical assistant to help you find relevant answers is one thing, but AI as an engagement option or a virtual relationship feels like something entirely different.