Keynote Speech: How will voice AI change our life or work - Sam Liang | SVIEF
5:41PM Sep 30, 2018
My name is Sam of the founder CEO of aisense and our flagship product's otter
I like to talk about how voice AI it's actually changing our life and work
You know their billions of people in the world. Everybody talks a lot.
You know, think about it, maybe you didn't pay attention. If you do some statistics, everybody talks several hours a day and listen to people several hours day. That's what we're doing here right now. There are some statistics that, you know, over your lifetime, you may speak like 800 million worlds words. Sorry.
That's how people communicate every day, you know, tons of sound and voice that people interpreted words.
However, you know,
even you listen, a lot you speak a lot. It's really hard to remember everything.
After today, within 24 hours, you're probably going to forget 70% or 90% of the information.
So this is why we started this company in Silicon Valley.
We're based in Los Altos. It's a lot of experts from Google speech, Google AI team. And we have team members from Uber, Airbnb, Facebook, Yahoo,
I'm the founder CEO and also Yun fu is our co founder.
We actually work together in another startup before this was which was successfully acquired by Alibaba in 2013.
So our mission just to borrow Googles mission statement, we want to organize world's voice information. And to make it universally accessible and useful.
You may ask, you know, what's the difference between Siri Alexa and Google Home?
So it's actually a huge difference.
If you think about it. I don't know how many people have an Alexa or Google Home at home. Just think about it. How many times do you talk to it? Three times five times.
It's a short question. You ask Alexa. You say, hey, Alexa, what's the weather tomorrow? But then the robot answer a question
the total amount of time you talk to your Alexa or Siri is probably no more than one minute day. But again, as I mentioned, how many minutes or how many hours you talk to your colleagues, talk to your family, talk to your friends.
So number one, the information needs to be understood and number two, it should be remembered as searchable.
This is why we think this is still a new market that nobody has been serious about. You know, Google is not working on it. Alexa is not working on it. Apple Siri doesn't do this
when you're talking to other people it's multi speaker You know, when you talk to Alexa it's a single speaker. It's a short question. But when you talk to other people, it can last for hours. You know, how does the system understand a one hour or two hour conversation with multiple people talking to each other.
in addition to the words you see there, there's a lot of meaning behind it. You know, even the same question or same sentence spoken by different people mean different things. So it's important for the system to understand who the speaker is.
So we think that, you know, we're taking a different approach, than Alexa, than Google Home, or Siri, we build this system to use AI and speech recognition and natural language understanding to stay in the background passively. You don't have to say a special word like, hey, Alexa.
Hey, Siri, you just keep talking
our product called Otter. You can download it right now from App Store and Google Play Store, you can use it
for your daily life for your daily work. As mentioned by Ryan earlier, Tim Draper is using it. I don't know whether it's every day well, at least every week in all his founder pitches, even some of his board meetings to take meeting notes, then everything is searchable.
Let me just quickly switch to the
demo here. This is actually a special group created for svief.
If you download otter or you just go to otter.ai website,
you can get access to it. All the speeches are available there in addition to the live transcript. Remember yesterday when Kai Fu Lee and
they they talk about creativity. So you may not remember the sentence or you're not here. But when I searched for that it find the speech quickly
network a little slow. Okay, so you see on the top it has the title has the a bunch of keywords identified automatically by the AI engine we built
by the way we built the speech recognition engine all by ourself. We're not using Google Voice API or nuance or IBM actually we beat most of them
that's why you see the accuracy there the key words actually you know that identified by the AI automatically it summarize the key topics you see the picture there
and Tim Draper Kai fu Lee and and, and
the professor Zhang is also here. So I can click anywhere
jobs are going to get more abstracted. So as you know, as we've used
iPhones, our jobs have changed, we've been able to do more, we've been more flexible. We
So you can play the audio, you see the transcript anywhere. This is where Tim mentioned creativity
Where's your creativity? There are what will I mean, I always think what would you do if AI took your job? Well, you would find something more creative and more innovative and more interesting to do. Because AI is now doing the rep the thing that you can replicate over and over. So I, I do feel.
So this is a system that not only useful in an event like this, this is actually useful for your daily work for your daily
life. You know, My son is actually a high school student, he uses this to take lecture notes,
it's going to be immersive, it's going to be everywhere, it's going to be, you know, pervasive, you put it on your cell phone,
you can turn it off, you can leave it on for a while, you know, when you walk around talking to people here it can listens and remembers, you don't have to take notes like crazy, you know, this is your personal note taker. And again, the
the intelligence behind it can understand a lot of things for you longer term, you know, even for doctors this can be useful because by listening to your voice by listening to the conversations, you talk to other people, it can detect your emotion change,
the system could tell
help dr. D to diagnose depression even before the
patient recognizes depressed. So this this was the one of the best apps by Mashable, one of the best seven apps and also if anybody use zoom This is licensed by Zoom. Any video conferences can be automatically transcribed. We're working with large conferences like Bridgewater associates it's hundred $60 billion
hedge fund. They want to use this to handle all their meetings, everything go meeting
yeah, this is a AI for voice conversations and
just to pitch a little bit. We are hiring you know, if anybody wants to join this is a billion dollar opportunity. It's really going to change other people's lives. Thank you.