Now, Microsoft’s Cortana might have a lot more human touch. All courtesy to the acquisition of Semantic Machines, a California-based company that specializes in conversational human-machine interactions.

Now, in more clearer terms, Semantic sets itself apart from Alexa, Google Assistant, and other voice assistants, thanks to a multi-faceted approach to the intelligence at work behind the vocal interaction. Very recently, voice assistants have also worked primarily from responding to the task at hand, and only the task at hand, thus lacking the context of prior conversations, which makes them feel far from intelligent.


Now, even if Assistant and Alexa, both have the abilities to respond to immediate prior conversation points. For instance, when asked about a song, following the command of ‘play it’, obviously, the song would be played. However, these contextual responses are limited.

In a recent interview, co-founder and chief scientist of Semantic, Dan Klein, added that “Today’s dialog technology is mostly orthogonal. You want a conversational system to be contextual so when you interpret a sentence things don’t stand in isolation.”

Getting to know you:

This further refers to the fact that if users are looking for booking a hotel, the voice assistant would benefit from knowing whether they have booked a car or are planning on using public transport, knowing where the meetings are, and what his/her career is. So, with these, it would be able to tailor a recommendation based on TripAdvisor reviews by people in a similar field.

Of course, this level of contextual intelligence obviously leads to a far better user experience and makes the conversation feel like much more with a human assistant. However, it makes it overt how much information that company truly knows about you.


Now, there is already the existence of services like TripIt that collate the user’s travel plans from harvesting their emails, so the technology is nothing new. In fact, to speak frankly it doesn’t change the amount of personal data that companies actually have. It just makes it clear and repurposes it so that it’s actually useful for the users.

These improvements in Cortana would tie into Microsoft’s aims that Cortana would become a “situationally appropriate” assistant that moves with the user from work to home, and everywhere in-between, acting proactively to assist him/her.

Perhaps, this phase is the best and of course, an interesting time for digital-human interactions. With Google’s Duplex making headlines as a robot voice that imitates a human assistant, able to make calls on the user’s behalf, brings for the question whether we actually want our voice assistants to be more human?