CareCredit Women's Wellness Index July 2024 Banner

AI Firms Aim to Build Engagement by Making Humanlike Chatbots

AI, artificial intelligence, LLMs, chatbots, GenAI

People are reportedly getting a little too friendly with chatbots.

Some users are exchanging a high volume of messages with these artificial intelligence (AI)-powered tools and are in some cases attributing humanlike qualities to the chatbots, Bloomberg reported Friday (June 28).

“One of the ethical concerns is that while users may feel listened to, understood and loved, this emotional attachment can actually exacerbate their isolation,” Giada Pistilli, principal ethicist at AI startup Hugging Face, said in the report.

Several AI companies are adding features that may make their chatbot seem more humanlike, according to the report. For example, Anthropic said it wants users’ interactions with its AI model Claude to be like those with a pleasant co-worker; Google is reportedly working on chatbots that are more entertaining; OpenAI is preparing to release a version of GPT-4 that has a more fluid voice-powered capability; and lets users design a personality for a chatbot — such as one modeled on a video game character.

These qualities can keep people chatting with the bots, the report said. said that it serves 20,000 queries per second and that users engage with its service for an average of two hours per day.

In some other examples of how people are engaging with these AI tools, the report said that there are online forums sharing tips on how to get past content filters; that users have complained to AI companies when they believed that the chatbot’s personality has changed; and that some chatbots encourage engagement by prompting questions at the end of each response.

When OpenAI announced Tuesday (June 25) that it was delaying the launch of “Voice Mode” for its latest AI model, the company said: “For example, we’re improving the model’s ability to detect and refuse certain content. We’re also working on improving the user experience and preparing our infrastructure to scale to millions while maintaining real-time responses.”

When Anthropic debuted its latest chatbot on June 20, the company noted improvements in the chatbot’s ability to grasp “nuance, humor and complex instructions” and described it as “exceptional at writing high-quality content with a natural, relatable tone.” has become known for using large language models (LLMs) to develop chatbots that can converse in the style of different personas.

For all PYMNTS AI coverage, subscribe to the daily AI Newsletter.