TOM FUDGE: Maybe you know about Ted Talks, and maybe you have seen, or heard one. It is the creation of a group called the sapling foundation, now based in New York and Vancouver, whose slogan is ideas worth spreading. Later this month on Saturday, October 11, our local branch of Ted Talks will host a conference at San Diego State. At it, several intellectuals will present big ideas. We have two of them on the show today. Joining me in studio are Xavier Leonard and Marian Bartlett. Xavier, what do you do when you're not doing Ted Talks? XAVIER LEONARD: Lately, I have been doing a lot of talking in other places, about the same topic I am speaking about at Ted. Either working with groups, communities of interest who are really trying to figure out how to engage communities more through innovative civic programs, speaking a lot on that as well as this idea of having people in those engagements do their own DIY interventions in city process and space. TOM FUDGE: Tell us what is on your mind. It sounds that you have got into the subject of civic engagement. Why is that so important? XAVIER LEONARD: It is important because it is a big need that is felt by our population, by people in our country, and there has been a growing gap over the past sixty years between what people want to do in participating in government processes, and what they are actually able to do. Format we have now, which our city government follows with daytime meetings and the city council gatherings, and having people track the position of legislation and bills, and things like that as they go through, it is not working so much for people's current lifestyle. But you still have people with the same desires that folks had before, but now they do not feel that they can really participate. TOM FUDGE: In terms of what you are suggesting, is a look you get inspiration for your big idea from certain internet websites. XAVIER LEONARD: Certainly some of the ways that the internet is used particularly has a model for thinking about how to engage people in different ways. TOM FUDGE: You're talking about Wikipedia, and stuff like that? XAVIER LEONARD: More social media. Right now, through social media, we have for the first time an entire population people who were producers, just like the producers who work in this station, only they produce a very narrow band of content that is significant, meaning that there are folks who make things, and curate, and understand. They think of themselves as producers of some sense. When they are called upon to be engaged, we do not just want to offer a thumbs-up or thumbs down on a suggestion someone else has. They feel they can get in, the role their sleeves up, and do the dirty work. TOM FUDGE: You're talking about crowdsourcing, right? XAVIER LEONARD: Crowdsourcing is one tactic to get people's input. You state problems and get their ideas, and it's one way to come up with an aggregate solution, but there are also smaller interventions that are growing, and quite popular, and quite effective for making people feel like they are simply engaged. That might be just a small group of people coming up with their own ideas, and executing it, creating a prototype and seeing what people learn from it. TOM FUDGE: Let's say we're going from a website that people share, to a neighborhood that people share. How can we see this idea of yours applied to an actual, physical space? XAVIER LEONARD: One way is to do what we have in San Diego, have community members come together, and really design the way a new park is going to look, hands-on. That is different from bringing in professionals, and having them work with mockups or testing, and get feedback them a and bring back designs based upon that. But when you have people actually getting in, and rolling their sleeves up and getting their hands dirty in actual design, and sticking that forward to actually producing, and being there physically, and implementing the project. TOM FUDGE: I once heard someone say there are two ways to create walkways in parks, first you can go in and put down cement and hard scape and do it, and expect people to use it. Or, you can do grass, and see where the footpaths have been worn down, and then put cement there. Is that what you're talking about? XAVIER LEONARD: I've heard that story as well, so you end up with a product that is more like what people want because that is what they need, they show you what they need. In that sense, it is like that story, but instead of doing experience like that, we actually trust people to give us their input, and produce what they need, or at least what they think they need. I think an important part of all of this activity is that you do with not like you do on a big city planning scale. One of the things that makes it effective, you lower the amount of risk to taking chances by lowering the amount of money that is needed to activate spaces like this. TOM FUDGE: Marian, tell us about Emotient. What is the company about? MARIAN BARTLETT: Emotient is developing a system for automatic facial expression recognition. It works real time, and can measure facial expression from anyone across cultures and ages. By now, we have a system where we can upload videos that you took on any device and get feedback on what emotions are in that video. TOM FUDGE: So the computer reads the face of a kid, and the computer can tell better than I can what the child is feeling? MARIAN BARTLETT: We have had examples of the computer understanding better than the human observer about what is going on in the background. TOM FUDGE: This sounds like something that is already going on, it is not science fiction. MARIAN BARTLETT: It is not science fiction. We are able to measure certain things about emotion through the face. There are cases and circumstances where facial movement can indicate information about decisions, or whether an expression is real or posed. This will change and enable a lot of things were not possible before. In particular, and emotion that I have been looking at, in areas of education and medicine, this is enabling things that were not possible before. One of the things is having a way to estimate a kids level of pain. Through the university, I am collaborating with the Rady Children's Hospital to develop a system that automatically tracks and measures kids pain postoperatively. The idea is not to replace a nurse, the idea is to have a constant measure. The nurse can come in and take your pulse, but instead, we have a device that constantly monitors pulses to have a better way of tracking and responding. We can do the same with pain. TOM FUDGE: There is a scene from a Tom Cruise movie in which a video screen is talking to him as he walks into a mall. Is that what you're talking about, this relationship between machines and human beings, and the way it is changing? MARIAN BARTLETT: This technology could ultimately be used to change the relationship between humans and machines. I do not think the public is ready to have a machine talking to them as they walk into a mall in a personal way. I think that would not be something that would happen right away. TOM FUDGE: Is there any reason that we should be concerned about the ability for machines to read our minds and know what we are thinking, and what we are feeling? MARIAN BARTLETT: The machine does not read our minds, it detects facial movement. There are some things that can be inferred from that facial movement, but it does not read what you're thinking or read a motion. It reads expression. TOM FUDGE: What you think of Marian's idea? XAVIER LEONARD: I love it. I want a computer to tell me what I'm wanting right now. TOM FUDGE: What you think of Xavier's idea? Do you want to live in a neighborhood that has been created in a do-it-yourself way by people who live there? MARIAN BARTLETT: I think it's a fabulous idea, and it has great potential to give people great satisfaction. TOM FUDGE: Tell us more about the that you have coming up a week from Saturday. You guys are going to be giving Ted Talks, and will people be able to talk to you, and ask questions? XAVIER LEONARD: Tedx is an independently produced event, that includes here, something that goes far beyond the event. There is a community around the practice of joining people in Ted Talks ended enjoys content. It happens on the day of, with people beyond our talk, with other conversations going on. It also happens on the days whether or not presentations, so there been of activities leading up to it, and everybody joining up into activities under the idea of people who are excited about new ideas that can make some changes in our environment. TOM FUDGE: Have you been fans of Ted Talks up until now? MARIAN BARTLETT: Yes, I enjoy Ted Talks, and this is TedX, which is separately organized. But it is very similar. Several of my colleagues have been involved in Ted Talks, and I think that they are fabulous ideas. At the event on Saturday the eleventh, I will have a demo setup, so people can come by and try facial expression recognition on their own face. TOM FUDGE: Tickets to the October 11 Ted Talks event at San Diego State, tickets are still available. Go to KPBS.org for more information. My thanks to these two folks who are going to be at that event. [ END SEGMENT ]
Imagine a world where machines can recognize facial expressions including pain, or a computer game that can teach autistic children how to recognize emotion. That's the world where Marian Bartlett, co-founder and lead scientist at the San Diego biotech Emotient, lives.
Bartlett said her research has shown that the machines do a better job of picking up on human emotion than humans can.
Using machine learning to learn patterns of facial movements that are consistent with real expressions is just one of the big ideas that will be shared at the TEDx America's Finest City conference at SDSU on Oct. 11.
Speakers will share their ideas on transformation on topics ranging from civic engagement to education to inner empowerment.