Semantic Machines, a conversational AI startup, is being acquired by Microsoft in an attempt to make intelligent assistants such as Cortana and bots sound and react more similar to humans. Machine learning is used by Semantic Machines, established in 2014, to make bots reply in a more natural manner to questions. Semantic Machines is directed by Prof Dan Klein from UC Berkeley and Larry Gillick, the previous chief speech scientist of Apple. Both are deemed innovators in conversational AI.
The acquisition by Microsoft will advance the Cortana digital assistant of the company and also it’s Azure Bot Service that is utilized by 300,000 developers. A human voice is already been used for its Cortana assistant by Microsoft. The software maker and Halo developer 343 Industries worked closely on the eyelike visual elements and Jen Taylor, the voice actress, for Cortana’s voice back in 2014.
Chief technology officer at the Microsoft AI and Research, David Ku, said, “With the acquirement of Semantic Machines, we’ll set up a conversational AI unit of excellence in Berkeley to shove forward the limits of what is likely in language interfaces. Merging the technology of Semantic Machines with AI advances of Microsoft, we aspire to provide natural, powerful, and more creative user experiences that would take conversational computing to an all-new level.”
On a similar line, Google is also functioning on such human-sounding AI. The company displayed its Google Duplex effort recently that enables the Google Assistant to make calls to businesses and schedule hair appointments or reserve restaurants tables on Google users’ behalf.
The demonstration of Google has surfaced several worries over casual human speech and strong worries over ethics as well. Microsoft has had its own controversial issues already with its own AI attributable to its racist Tay Twitter bot; thus, you’d picture it will be extremely watchful with whatever it builds from its acquirement of Semantic Machines.