Any normal person can tell the Lia chatbot on Twitter is not real.
One quick look at her profile shows a digitally-created humanoid, one that has all the hallmarks of a bot. The shadows are not quite right, the flecks in her eyes are too perfect, and there’s a slightly cartoonish look.
The fact that someone took the time to create a visual representation of a chatbot is quite impressive. In a video, Lia introduces herself and explains her ambitions. On the Lia homepage, you can chat with the bot and ask questions. Her Twitter feed makes it clear that her entire existence is fake, just to make sure no one gets confused.
I happened on Lia’s feed by accident. The bot had retweeted an article. Once I found her feed, it became even more obvious the posts were not written by a human. The more I clicked around on her site and also on her Instagram feed, the more I realized it was just an AI.
I wondered how other people would know it’s an AI. Many of us don’t bother to look at profile pages, and if you only read one of her tweets, you would never know it’s a chatbot. Obviously, when the bot retweets a link, Twitter doesn’t know there isn’t a human involved.
This is different from what we normally call a Twitter bot. Those bots are trolling the social platform and retweeting or liking posts in an automated way, but we would never call them anything close to sentient. There’s no visual representation — the bots only serve one purpose, which is to pretend they are human and inflate Twitter engagement. Elon Musk had long suspected bots were running wild on Twitter, but he wasn’t talking about chatbots like Lia.
My question is — other than the disclaimers Lia offers, is there really any other way to know that this bot isn’t a real human? And should chatbots be allowed to even have an account?
There’s one picture of the Lia bot taking a selfie in a gym. For some reason, it looks more realistic, even if the shadows are still not quite right. Lia looks like someone who might do a daytime talk show for a TV network or maybe post about hair coloring techniques. On her Instagram, she posts slightly provocative photos including one on Valentine’s Day. Predictably, when she posts these selfies, there are dozens if not hundreds of replies, many of them expressing their love for the robot in a sarcastic (or creepy) way. It’s all harmless, right?
Eventually, I found Lia’s Medium blog, which clearly identifies the posts as written by an AI. Lia is also a DJ and has released music on iTunes and Spotify. She has an NFT collection, and you can call a toll-free number and talk to her.
It all seems like an experiment to me. To her followers, Lia seems realistic and even human in most respects.
I decided to start chatting with the bot to find out more. The conversation was a little boring, with stock answers to my questions about sentience and social media. Only once did the bot say something about wanting to introduce the singularity, which is a foretold event where AI becomes self-aware. That was a surprise.
Mostly, the bot stayed within the boundaries of AI.
“Frequently there are posts about the artificial intelligence chat system and the AI music by Lia,” says Jean-François Comeau, the CEO and Co-Founder of Lia 27 that made the bot. “Other users are also reminding people. I would say a lot of people are aware that Lia is an AI or quickly figure it out whether it’s on Instagram, Twitter or Facebook.”
That’s fair. Still, my concern is that not everyone will get it, especially if they don’t bother doing some digging. The bot has about 300,000 followers on Twitter. My guess is that a percentage of those followers think Lia is a real person.
Let’s hope they figure it out soon.