“AI is expensive. Let’s be honest,” says Anand.
Enhance versus safety
In October 2024, the mother of a teenager who died by suicide unlawful lawsuit against the technologies of the heroes, its founders, Google and Alphabet, claiming that the company attacked the son of “anthropomorphic, hypersexual and terrifyingly realistic experiences, during programming, during programming [the chatbot] To mislead as a real person, a licensed psychotherapist and an adult lover. “At that time a character spokesman. said CNBC that the company was “broken by a tragic loss” and took very seriously “the security of our users”.
The tragic incident set character. At the beginning of this year, US senators Alex Padilla and Peter Welch wrote A letter to several AI social platforms, including the form.
“The band has been taking it very responsibly for almost a year,” Anand tells me. “AI is stochastic, it is always difficult to understand what is coming. So this is not a one -time investment.”
This is extremely crucial because the character is growing. The startup has 20 million busy users per month, who spend an average of 75 minutes a day on conversations with a bot (“form” of character. The company’s user base is 55 percent of women. Over 50 percent of its users are a gene with or gene alpha. With this growth there is a real risk – what does Anand do to ensure the safety of its users?
“[In] Over the past six months, we have invested a disproportionate amount of resources in the possibility of under 18 years of age than over 18, which was not last year, “says Anand.” I can’t say: “Oh, I can hit the 18+ label in my application and say it to NSFW.” After all, you create a completely different application and a different small platform. “
Anand tells me that over 10 out of 70 employees of the company work full -time in the field of trust and security. They are responsible for building security, such as age verification, separate models for users under 18 years of age and fresh functions such as Parental observationswhich allow parents to see how their teenagers apply the application.
A model under the age of 18 launched in December last year. According to the spokesman Kathryn Kelly, it includes “a narrower set of characters that you can search” on the platform. “” Filters were used for this set to remove signs related to sensitive or mature topics. “
But Anand claims that AI will take more than just technical corrections. “Maintaining this platform is a sheltered partnership between regulatory bodies, USA and parents,” says Anand. This makes watching his daughter talk to the character. “It must be sheltered for her.”
Outside the company
The AI Society market is developing. According to estimates, consumers around the world spent $ 68 million on the AI company in the first half of this year, which is an escalate of 200 percent compared to last year cited by CNBC. AI startups shoot at a piece of the market: XAI released a terrifying, portioned companion and even Microsoft in July bills His Copilot Chatbot as a companion of AI.
So like a character. It is completely removed from this.