Character.ai allowed sensational test of its Ki chatbots

Character.ai allowed sensational test of its Ki chatbots

8. October 2022 0 By Horst Buchwald

Character.ai allowed sensational test of its Ki chatbots

San Francisco, Oct. 8, 2022

Character.ai has allowed the public to test AI chatbots modeled after famous people such as Albert Einstein, Sherlock Holmes and Elon Musk.

The startup was founded by Noam Shazeer and Daniel De Freitas. They were also involved in the development of Google’s LaMDA (Language Model for Dialogue Applications) AI chatbot model.

Shazeer and De Freitas launched the beta version of Character.ai last month. The idea was to develop a product that could help “millions and billions of people” especially as more and more people became isolated during the pandemic. The chats include a disclaimer that everything the characters say is “made up.”

Like LaMDA, Character.ai’s technology was developed using large language models. In LaMDA’s case, it was built on Google’s Transformer neural network architecture and trained specifically on dialogues.

LaMDA hit the headlines when a software engineer, Blake Lemoine, was fired by Google after claiming the technology was sentient.

Hits: 4