Ramona chatbot avatar

May 11, 2011

KurzweilAI has launched Ramona 4.1, an audiovisual chatbot avatar. This version adds new knowledge sources, including one with Wikipedia-like knowledge and another that enhances Ramona’s whimsicality, along with an improved interface, forming Ramona’s updated unique personality.

Ramona is capable of synthesized speech and changeable accents. Her personality can also be adjusted to make her less or more nerdy or whimsical by typing “Nerdiness” or “Whimsicality” plus a level between 0 and 99 (e.g., “Nerdiness 87″).

Users can hear her speak in 14 international accents, ranging from Baltimore to Bangalore. Users can also converse with previous versions of Ramona.

Ramona is a long-term project. Ray Kurzweil projects that Ramona version 11 circa 2029 will pass the Turing test. This is one aspect of Kurzweil’s plan to win his bet with Mitch Kapor that a computer (a machine intelligence) will pass the Turing test by 2029 (Kapor is betting this will not happen).

Chat with Ramona now

A short bio

The Virtual Ramona Project was begun in 2000 by Kurzweil Technologies to envision the future of lifelike avatars in virtual spaces, and prototype their remote control by human operators.

Ramona conducts conversations with visitors, responding to typed questions or comments with a human face, lip-synched speech, and appropriate facial expressions. For the curious, she graciously answers questions about herself and her life story, including her burgeoning career as a virtual rock star.

A live-performance version of Ramona debuted at the TED 2001 conference, where she entertained the audience with live music, a dance performance, and music videos via state-of-the-art motion capture, real time photo-realistic animation, and pitch alteration software. Ramona’s virtual life story and downloads of her music and videos are also available.