從我還是一個小女孩 第一次觀看星球大戰開始， 我就被個人機器人 這個概念所吸引。 當我是小女孩時， 我就愛這主意-機器人可以與我們互動 就像我們身邊有益值得信賴的助手， 能逗樂我們，豐富我們的生活 還能助我們拯救一兩個星系。 所以當那時我知道這樣的機器人還不存在時， 我就知道-我想要建造它們。
20年過去了， 我進入麻省理工學院的研究院 學習人工智能， 那一年是1997年， NASA剛剛登陸了第一個在火星的機器人。 但是，諷刺的是，機器人仍然不在我們的家中。 我還記得考慮過所有 爲什麽如此的原因。 其中一個原因特別使我震驚。 機器人技術一度就是關於如何和實物互動， 而不是與人類- 尤其不可能以一種對我們而言自然的社交方式 幫助我們真正接受機器人 進入我們的日常生活。 對我而言，那就是空缺-那就是機器人尚且不能做的。 所以那一年，我開始建造機器人Kismet, 世界上第一個社交型機器人。 三年后- 許多的編程， 和與其他博士生在實驗室的一同工作后- Kismet 開始能夠和他人互動。
所以在過去的這些年中， 我繼續探索著機器人的人際交流方向， 現在在麻省理工學院媒體實驗室 我建立一個團隊，由許多無比天才的學生組成。 我最喜歡的機器人之一是里納多(Leonardo). 我們與Stan Winston工作室合作研發了里納多。 現在我想向你們展示一個對我和里納多特別的時刻。 這是Matt Berlin 和里納多的互動， 給里納多介紹一種新事物。 因為是新事物，里納多並不知道對此該怎麼做。 但有點像我們，他可以通過觀察 Matt的反應來學習。
So just by a show of hands, how many of you all have a robot at home? Not very many of you. Okay. And actually of those hands, if you don't include Roomba how many of you have a robot at home? So a couple. That's okay. That's the problem that we're trying to solve at Romotive -- that I and the other 20 nerds at Romotive are obsessed with solving.
So we really want to build a robot that anyone can use, whether you're eight or 80. And as it turns out, that's a really hard problem, because you have to build a small, portable robot that's not only really affordable, but it has to be something that people actually want to take home and have around their kids. This robot can't be creepy or uncanny. He should be friendly and cute.
So meet Romo. Romo's a robot that uses a device you already know and love -- your iPhone -- as his brain. And by leveraging the power of the iPhone's processor, we can create a robot that is wi-fi enabled and computer vision-capable for 150 bucks, which is about one percent of what these kinds of robots have cost in the past.
When Romo wakes up, he's in creature mode. So he's actually using the video camera on the device to follow my face. If I duck down, he'll follow me. He's wary, so he'll keep his eyes on me. If I come over here, he'll turn to follow me. If I come over here -- (Laughs) He's smart. And if I get too close to him, he gets scared just like any other creature. So in a lot of ways, Romo is like a pet that has a mind of his own. Thanks, little guy. (Sneezing sound) Bless you.
And if I want to explore the world -- uh-oh, Romo's tired -- if I want to explore the world with Romo, I can actually connect him from any other iOS device. So here's the iPad. And Romo will actually stream video to this device. So I can see everything that Romo sees, and I get a robot's-eye-view of the world. Now this is a free app on the App Store, so if any of you guys had this app on your phones, we could literally right now share control of the robot and play games together.
So I'll show you really quickly, Romo actually -- he's streaming video, so you can see me and the entire TED audience. If I get in front of Romo here. And if I want to control him, I can just drive. So I can drive him around, and I can take pictures of you. I've always wanted a picture of a 1,500-person TED audience. So I'll snap a picture. And in the same way that you scroll through content on an iPad, I can actually adjust the angle of the camera on the device. So there are all of you through Romo's eyes. And finally, because Romo is an extension of me, I can express myself through his emotions. So I can go in and I can say let's make Romo excited.
But the most important thing about Romo is that we wanted to create something that was literally completely intuitive. You do not have to teach someone how to drive Romo. In fact, who would like to drive a robot? Okay. Awesome. Here you go. Thank you, Scott.
And even cooler, you actually don't have to be in the same geographic location as the robot to control him. So he actually streams two-way audio and video between any two smart devices. So you can log in through the browser, and it's kind of like Skype on wheels. So we were talking before about telepresence, and this is a really cool example. You can imagine an eight-year-old girl, for example, who has an iPhone, and her mom buys her a robot. That girl can take her iPhone, put it on the robot, send an email to Grandma, who lives on the other side of the country. Grandma can log into that robot and play hide-and-go-seek with her granddaughter for fifteen minutes every single night, when otherwise she might only be able to get to see her granddaughter once or twice a year.
Thanks, Scott.
The New Romo from Romotive on Vimeo.
So those are a couple of the really cool things that Romo can do today. But I just want to finish by talking about something that we're working on in the future. This is actually something that one of our engineers, Dom, built in a weekend. It's built on top of a Google open framework called Blockly. This allows you to drag and drop these blocks of semantic code and create any behavior for this robot you want. You do not have to know how to code to create a behavior for Romo. And you can actually simulate that behavior in the browser, which is what you see Romo doing on the left. And then if you have something you like, you can download it onto your robot and execute it in real life, run the program in real life. And then if you have something you're proud of, you can share it with every other person who owns a robot in the world. So all of these wi-fi–enabled robots actually learn from each other.
The reason we're so focused on building robots that everyone can train is that we think the most compelling use cases in personal robotics are personal. They change from person to person. So we think that if you're going to have a robot in your home, that robot ought to be a manifestation of your own imagination.
So I wish that I could tell you what the future of personal robotics looks like. To be honest, I have no idea. But what we do know is that it isn't 10 years or 10 billion dollars or a large humanoid robot away. The future of personal robotics is happening today, and it's going to depend on small, agile robots like Romo and the creativity of people like yourselves. So we can't wait to get you all robots, and we can't wait to see what you build.
