Google announced that the artificial intelligence chat program Duplex, developed by the users, will warn users that it is not a human being but an artificial intelligence program.
Duplex’s speech talents have been featured this week at the Google I / O developer conference, allowing a user to make an appointment with the barber.
During the conversation, Duplex stops as people do, uses daily conversation, and does arguing.
Controlled and trained by Google’s DeepMind WaveNet software, Duplex is able to efficiently simulate human language use.
Some tech experts who followed Duplex’s performance said it could have “terrible” results.
Following the performance of Duplex, Associate Professor Zeynep Tüfekçi from the University of North Caroline described the humanistic imitation as “horrible and wrong” to be imitated by artificial intelligence so realistically.
The rifleman said it would fool people.
Making a statement about the subject, Google said that transparency is important to them, and that people who communicate with it will be alerted to an artificial intelligence.
With this practice, the company said that they aimed at getting people to deal with day-to-day tasks such as making appointments through a natural dialogue.