Acquiring Semantic Manchines, a company specializing in artificial intelligence applied to the conversation, Microsoft seems determined to make Cortana more human.
Remember, it was only a few weeks ago: Google presented its voice assistant called "Google Duplex", the reactions so human that they became disturbing for some. Today, Microsoft seems to want to follow suit with the American giant.
Microsoft acquires "Semantic Machines"
Founded in 2014, Semantic Machines is an American company led by Dan Klein, a professor at the University of Berkeley, and Larry Gillick, a former "chief speech therapist" for the apple brand. Both are very well regarded in the field of artificial intelligence, their company having in particular developed the basic technology serving as a structure for Siri and Google Now. Microsoft has just bought the start-up.
This acquisition will allow Microsoft to improve Cortana so as to make it more "human", but also to boost its performance in terms of natural language, speech recognition, semantic understanding, language learning or even deep learning, skills developed by Semantic Machines.
"Creating a Conversational AI Center of Excellence"
The acquisition will also help Microsoft improve its Azure Bot service today, which is used by more than 300,000 developers around the world. As part of this acquisition, David Ku, Director of AI & Research at Microsoft said: "With the acquisition of Semantic Machines, we will establish a conversational AI center of excellence at Berkeley, to repel the limits of what is possible at the level of linguistic interfaces ".
Following the slippage that occurred in 2016 with his Tay Twitter bot, Microsoft will have a lot of work to not reproduce the same mistakes as in the past. As a reminder, Tay Twitter was a conversational robot able to chat with Twitter's users, who had at the time made various racist remarks that had pushed Microsoft to remove it from the social network.
