Everything you need to know about the Character AI app

The Character AI app is quickly becoming a favorite among users. While ChatGPT and other large language model (LLM) programs like Bard and Bing Chat help users with their queries and such, the Character AI app focuses mostly on conversations.

However, what really makes it popular is who users are conversing with. Imagine asking Harry Potter for spell advice, querying Jack Sparrow about where he buried his treasure, or even seeking Jedi wisdom from Luke Skywalker himself. These are just some of the fictional characters answering queries on Character.AI.

But it’s not just fictional characters. There are also replies generated by real individuals. Users can have conversations with famous people like President Joe Biden, Cristiano Ronaldo and even Albert Einstein. Put simply, a user can get Elon Musk to explain why he renamed Twitter to X, or even ask Taylor Swift what she likes doing after her concerts.

This is exactly what the Character AI app does. It’s a chatbot that lets users chat with celebrities, historical figures as well as fictional characters via text. According to a report by the Indian Express, what makes the Character AI app even more exciting is that its conversations appear to be far more natural than ChatGPT.

The chatbot currently also includes AI versions of popular video game characters as well as language teachers, therapists and more. Data from Similarweb showed that the average user spends about 34 minutes on the Character AI app, which is a lot more than the average time spent by a user on YouTube, Instagram, Facebook and even ChatGPT.

The story behind the Character AI app

Developed by two former Google engineers, Noam Shazeer and Daniel de Freitas, the plan was to place the user at the center of the interaction. They wanted to have a platform whereby a user can be empowered with information, education, assistance, mentorship, support and social connection in ways never possible before, and can unlock unprecedented levels of intelligence, creativity and innovation.

“We started Character to lead the AI revolution and to bring joy and value to billions of people at each step along the way.  We were joined by a founding team of AI leaders from Google Brain and Meta AI who are experts in deep learning, large language models and dialog. In our first thirteen months, we have been hard at work building scalable and extremely efficient infrastructure to train and serve our first large language models, and launching our first public beta that enables anyone to create and interact with conversational AI agents (“Characters”),” explained Shazeer in a blog post.

He also highlighted that the team is exploring ways to integrate images and audio as well to create a more immersive and engaging experience. Another interesting point is that they own the engineering stack end-to-end, from data, modeling and training to serving, user interface and experience. As such, Shazeer said they can align the product and research roadmaps more effectively, optimize all elements in the stack, and engage their systems in a strong flywheel with users.

Since its debut, the Character AI app has raised US$150 million in a new funding round led by Andreessen Horowitz that valued the AI chatbot startup at US$1 billion, according to a Reuters report. The company is also in talks with cloud providers for more strategic investment. The billion-dollar valuation for a company with zero revenue is another example of the continued AI funding boom since OpenAI’s ChatGPT became a widely recognized name.

In May this year, the Character AI app launched a subscriber service to give users access to newer features as well as avoid long waits. The service is still available for free and has unlimited messaging services with the characters. Having debuted on the web, the Character.AI mobile app is also now available for download on iOS and Android.

Concerns of deepfakes and copyright 

Tech Wire Asia decided to have a chat with some of the characters on the app. The first thing we noticed is the notification that states, “Remember: Everything is made up!” While it is obvious that everything on the app is fictionalized, some of these characters are based on real people. And there is a growing problem of how AI is being used to deepfake what some of these individuals are saying.

We spoke to several characters, both fictional and real. While the replies are pretty much similar to what one would get from any other chat app, we found ourselves spending quite a bit of time on the site as well. This clearly explains why the statistics indicate users spend more time on the site. A conversation with these characters can be quite fun and relaxing and is definitely addictive.

Unlike other AI chatbots, the characters in Character.AI seem to be well-trained by the language models, mimicking almost exactly what you would expect the characters would say if they were real. In fact, once video and audio generation is available to these characters, it would be a game-changer in the industry.

Some of the characters can also give advice, which can be a form of therapy for some users. Given the impact fictional characters have on society today, perhaps the idea of being able to chat with them could actually be seen as an escape from reality for some users. A depressed person could end up spending hours in joy chatting with his or her favorite character and even getting advice.

Naturally, this is where another concern that comes about as well. There is already a lot of concern about how AI is generating content without proper approvals and permissions from the original source. For example, we asked the Kanye West character to compose a song for us, and the character’s AI was able to come up with some lyrics. But how do users acknowledge who exactly the content belongs to?

However, a recent ruling by a federal judge in the US may just keep AI-generated content like this out of trouble. According to a report by Bloomberg Law, Judge Beryl A. Howell of the US District Court for the District of Columbia agreed with the US Copyright Office’s decision to deny a copyright registration to computer scientist Stephen Thaler, who argued a two-dimensional artwork created by his AI program “Creativity Machine” should be eligible for protection.

According to the judge, artwork created by AI isn’t eligible for copyright protection because it lacks human authorship. The ruling is the first in the US to establish a boundary on the legal protections for AI-generated artwork. The same could also eventually be applied to AI-generated text, voice and other content as well.

For now, it seems the app will only continue to grow and improve its capabilities. For users, the app is definitely something that allows them to escape reality and have some good conversations with their favorite characters or celebrities.

---

---

---

---