What is the Future of Intelligent Assistants?

TRANSCRIPT:
This is an exciting time to be involved in artificial intelligence. We’ve reached the point where the data sets that computer scientists have access to and the amount of computation that we have access to has gotten large enough so that for the first time in human history we can actually build systems that are surprisingly intelligent, and in some cases, even able to demonstrate human-like intelligence, to a limited extent.
And we’re fortunate there are all these new intelligent assistants that have become available in the past couple of years. For example, there’s Apple’s Siri intelligent assistant, there’s Google Voice search, Microsoft is rumored to be building a product called Cortana, which is their intelligent assistant. Nuance has a technology called Nina, there was a company called EV which was acquired by Amazon. So there’s a lot of intelligence systems. All the big players have intelligent assistants. And they’re all investing in this area, primarily because we’re entering this era where the data sets are large enough to do some very powerful things..
The problem that we see is we think the model the intelligent assistants is currently using is flawed, and and personally, I mean, I think that in order for these intelligent assistants to become a part of our everyday life, they’re going to need to improve along four different dimensions. I’m going to create a video blog post about each of these dimensions: they are speed, these systems have to get faster, one. Two, they have to get more accurate. Three, they’re going to need to get much smarter, and four, we’re going to want them to anticipate what we want. The good news is I think in the next three to five years all these improvements will happen and I suspect five years or maybe even sooner, these intelligent assistant technologies will become a part of our everyday life.