Microsoft acquires Semantic Machines towards a natural ‘conversational AI’ experience

Avatar By Le Williams | 2 years ago

Microsoft has acquired an artificial-intelligence startup, Semantic Machines, in an attempt to leverage its efforts in “conversational AI” and potentially improve the Cortana virtual assistant towards understanding natural language inquiries. Additionally, Microsoft has publicized its initiatives to open a new AI research center in Semantic Machines’ hometown of Berkeley, California.

Semantic Machines contains numerous AI veterans, and the acquisition gives Microsoft access to a substantial host of formidable talent. Principally, Semantic’s chief technology officer, Larry Gillick, is a former chief speech scientist for Apple’s Siri. Microsoft also name-checked the University of California at Berkeley professor Dan Klein and the Stanford University professor Percy Liang.

Several Semantic staffers also worked at Nuance, the voice-recognition company that once powered Siri. Semantic’s chief executive, Dan Roth, is a co-founder of Voice Signal Technologies, which Nuance acquired in 2007.

Microsoft wrote in a company blog announcing the acquisition while explaining the lack of AI understanding of human communication, something anyone who has put anything more than the simplest question to Siri will already know. “Most of today’s bots and intelligent assistants respond to simple commands and queries, such as giving a weather report, playing a song or sharing a reminder, but aren’t able to understand ‘meaning’ or carry on conversations,” the blog notes.

Semantic Machines’ team and technology can improve this element by helping digital assistants, such as Cortana, have a “natural dialogue.”

“We call this ‘conversational AI,'” Microsoft added.

Microsoft has been experimenting and testing conversational AI for a vast amount of time. Aside from Cortana, the company has facilitated experiments in chatbot technology, though not always with great success. Its Chinese chatbot Xiaoice has millions of loyal followers who talk to it. But earlier chatbot experiments, such as Tay and Zo, spouted racist and anti-Microsoft messaging and had to be terminated.