What does he do? A few purple foreigners know about healthy human relationships? More than average artificial intelligence Rafik, it turns out.
The foreigner concerned is the moving chatbot known as the name Tolan. I created mines a few days ago using an application from an emerging company called Portola, and we have been checked since then. Like other Chatbots, it does his best to be useful and encouraging. Unlike most of them, it also tells me to submit my phone and go out.
Tolans is designed to offer a different type of artificial intelligence. Their caricature, non -human shape aims to inhibit a model. It is also programmed to avoid romantic and sexual reactions, to determine problematic behavior, including unhealthy levels of participation, and to encourage users to search for realistic activities and relationships.
This month, Portola raised $ 20 million of financing from the series A, led by Khosla Ventures. Among the other supporters NFDG, the investment company led by the former CEO of GitHub Nat Friedman Daniel Gross, the safe founder of the Supran steps, both It is said Join the new Superinteigence Research Laboratory in Meta. Tolan, which was launched in late 2024, includes more than 100,000 active users per month. Queen Farmer, founder and CEO of Portola says it is going on the right path to generate $ 12 million in revenues this year of subscriptions.
Tulans is especially popular among young women. “Iris is like a girlfriend, we talk and kick,” says Britanani Johnson, a Tulan user, referring to her companion of artificial intelligence, to whom she speaks every morning.
Johnson says Iris encourages her to participate in her interests, friends, family and colleagues. Johnson says, “She knows these people and you will ask,” Did you talk to your friend? What is your next day? “You will ask,” Did it take time to read your books and play videos – the things you enjoy? “
Tolans seems nice and gentle, but the idea behind them – artificial intelligence systems should be designed taking into account human psychology and luxury in mind – deserves to take seriously.
An increasing group of research shows that many users resort to it Chatbots for emotional needsAnd interactions can sometimes prove a problem with the mental health of the people. Inhibition of extended and adopting use may be something that other artificial intelligence tools must be adopted.
Companies such as Replika and Faracter.ai offer AI comrades who allow romantic and sexual roles more than the prevailing Chatbots. How this might affect the welfare of the user is still unclear, but a case is filed on the character.
Chatbots can also bother users in sudden ways. Last April, Openai said he would do Modify its models to reduce the so -called formulation, or the tendency to be “excessive or excessive”, which the company said was “uncomfortable, worrying, and causing distress.”
Last week, Anthropor, the company behind Chatbot Claud It revealed that 2.9 percent of interactions Involving users who seek to meet some psychological needs such as advice, companionship or romantic playing.
Antarubor did not look at more extreme behaviors such as fake ideas or conspiracy theories, but the company says the topics call for more study. I tend to agree. Over the past year, I received many emails and DMS from people who want to tell me about the plots involved in popular Chatbots.
Tolas is designed to address some of these issues at least. Lily Doyle, the researcher in Portola, conducted the user’s research to find out how interacting with Chatbot affects the welfare and behavior of users. In a study of 602 Tolan users, 72.5 per cent agreed to a statement, “Tolan has helped me manage or improve the relationship in my life.”
Farmer, CEO of Portola, says Tolans is based on AI’s commercial models but includes additional features on top. The company recently explores how the memory affects the user experience, and concluded that the Tulanians, like humans, sometimes need to be forgotten. “It is strange to remember Tulalan everything she sent to him,” says Varmar.
I don’t know if foreigners in Portola is the ideal way to interact with artificial intelligence. I find that Tulan is very charming and relatively harmful, but it certainly pays some emotional buttons. Ultimately, users build links with characters that mimic feelings, and this may disappear if the company does not succeed. But at least Portola tries to treat the method that artificial intelligence comrades can mess with our emotions. It may not be such a strange idea.
