Speakers in tonight’s Society, Robots and Us at 6pm PST Tuesday Feb 23 include Henry Evans, mute quadriplegic and founder of Robots4Humanity and Aaron Edsinger, founder of Hello Robot. We’ll also being talking about robots for people with disabilities with Disability Advocate Adriana Mallozi, founder of Puffin Innovations and Daniel Seita, who is a deaf roboticist. The event is free and open to the public.
As a result of a sudden stroke, Henry Evans turned from being a Silicon Valley tech builder into searching for technologies and robots that would improve his life, and the life of his family and caregivers, as the founder of Robots4Humanity. Since then Henry has shaved himself with the help of the PR2 robot, and spoken on the TED stage with Chad Jenkins in a Suitable Tech Beam. Now he’s working with Aaron Edsinger and the Stretch Robot which is a very affordable household robot and teleoperation platform.
We’ll also be hearing from Adriana Mallozi, Disability Advocate and founder of Puffin Innovations which is a woman-owned assistive technology startup with a diverse team focused on developing solutions for people with disabilities to lead more inclusive and independent lives. The team at Puffin Innovations is dedicated to leveling the playing field for people with disabilities using Smart Assistive Technology (SAT). SAT incorporates internet of things connectivity, machine learning, and artificial intelligence to provide maximum access with the greatest of ease. By tailoring everything they do, from user interfaces to our portable, durable, and affordable products, Puffin Innovations will use its Smart Assistive Technology to provide much needed solutions the disabled community has been longing for.
This continues our monthly exploration of Inclusive Robotics from CITRIS People and Robots Lab at the Universities of California, in partnership with Silicon Valley Robotics. On January 19, we discussed diversity with guest speakers Dr Michelle Johnson from the GRASP Lab at UPenn, Dr Ariel Anders from Women in Robotics and first technical hire at Robust.ai, Alka Roy from The Responsible Innovation Project, and Kenechukwu C. Mbanesi and Kenya Andrews from Black in Robotics, with discussion moderated by Dr Ken Goldberg, artist, roboticist and Director of the CITRIS People and Robots Lab, and Andra Keay from Silicon Valley Robotics.
TRANSCRIPT OF THE FIRST INCLUSIVE ROBOTICS DISCUSSION (from video directly above)
Andra Keay 0:05
So welcome, everybody. Welcome to our first society robots and us for 2021. And I’m looking forward to a discussion that is going to help us set the agenda for robotics in 2021 and beyond. And I think it’s very important that as our technology emerges, we address the issues around how it is affecting society, and how it can have an impact both positive and negative on society. And so we have wonderful conversations. And we started doing this event in the early days of the COVID era, and we were focusing on so what actually does it mean? How can robotics and roboticists help in this time of pandemics, and it was a fantastic conversation, and we decided that it was time to expand the topic, and to start to talk about things like racism in robotics, global challenges and how we address those. So it’s one of my favorite events. And I’m delighted to see so many people. Joining us now, my role is to warm up for the speakers and the rest of the discussion, I’ll just give everybody a little bit of housekeeping as to how this is going to roll. I will introduce each of the speakers, and they will each share their thoughts with us. We will move from speaker to Speaker if you have questions specifically for one of the speakers, put them into the chat, and I can forward the question on and then we’ll start the general discussion once every speaker has had their time to speak. And in the general discussion? Well, I’m looking forward to finding out what is inclusive robotics? Why do we need it? How do we go about getting it? And even beyond that? What is the robotics agenda? For 2021? And beyond? What are the questions that we haven’t thought to ask perhaps, and perhaps it’s time that we started those discussions. And in the spirit of that, I would like to acknowledge that I and many of us are here on the lands of the aloni people who are an unrecognized First Peoples tribe of California. And I’m very pleased to actually see more events starting to acknowledge the first peoples as part of how things happen. And so I’ve given us the introduction and the housekeeping. I’d like to say a little bit about our speakers. And we’ll kick off the rest of the discussion. And I see this all because pointed out an event that’s coming up in the chat, a spring founders circle for the responsible innovation labs. I would like to also say, Silicon Valley robotics, and women in robotics, have regular events. So in women in robotics, we have a weekly book club, for example. And we have a slack community where we can meet together online, as well as having local chapters. If you’re interested in joining that, and I’ll pop the link in the chat. Please go to women in robotics.org. And it’s not a ht, TP s site, it’s still an HTTP site. So if you can’t find it, that’s the reason. But if you’re interested in signing up, please go there. Silicon Valley robotics, which is the organization that I call my day job not, although it is also a passion project is able to assist you if you are a robotics company at any stage. And we have mentor networks, we have events that are topic related, or that are related to helping you with your startup. And I’m just wondering, Ken, would you just like to say a few words now about CITRIS, where Ken is the director of the people in robots lab, and as well as the research there. There’s also the CITRIS Foundry, I believe.
Ken Goldberg 4:37
Thank you under so I want to say we’re really lucky to be partnering with you on this and so it’s a pleasure to work with you and citrus is a University of California, actually state level organization that connects for the campuses. They are said, Davis, Santa Cruz and Berkeley and the mission of CITRIS stands for the center is a center for Information Technology Research in the interest of society. So the mission of this series that Andhra is organizing is very much consistent with the with the mission of the center. And my my initiative within it is the is people and robots. So these are really come together very strongly. And this idea of inclusive robotics is something that we’re very excited about developing and expanding in the in the in the year to come and the years to come. So I really appreciate the the discussion we’re going to have tonight, I’m really looking forward to hearing your perspectives.
Andra Keay 5:35
Thanks so much again. And you’ve probably seen that we have a fantastic lineup of speakers tonight, we have Dr. Michelle Johnson, who is at the grasp lab in new pen, and actually is the director of the rehabilitation robotics lab at the grasp lab, and the Associate Professor of physical medicine and rehabilitation. We have Dr. Ariel Anders, and she is the first technical hire at robust AI, and is also a board member for women in robotics. We have Alka Roy, who is the founder of the responsible innovation project, and is working on building delight, trust and inclusion into technology and AI. Looking forward to hearing more about that. Then we have getting my notes out of order here. Then we have Kenny Chiku Nova nisi from who is a roboticist. And my notes are totally out of order now. and is a member of blacking robotics, and he will be able to speak to us a little bit about, like in robotics, as well, Kenny Andrews, who’s a computer engineering master’s student at the University of Illinois, and on the undergraduate committee of black in robotics. And, of course, Ken Goldberg, who is the director of the people in robots lab at citrus, and the distinguished William S. Floyd Jr. Chair in engineering at UC Berkeley, and is not only a roboticist, but an artist and I love the cross disciplinary perspective that that brings to the discussion. And without more from me now I think we’ve given all of the strikers time to join the conversation, I would like to introduce Dr. Michelle Johnson.
Dr Michelle Johnson 7:46
Thank you Andra. Thank you, everyone for attending. I was really intrigued by the title inclusive robotics. What is it? Why do we want it? And what do we need to do to have it and as I was pondering, the title to clear thoughts came to mind first, I thought, inclusive robotics means designing robots that reflect the diversity of society in terms of culture and race, as the first thought. The second thought was that inclusive robotics means to me, designing robots that are usable in all types of settings in low resource in high resource settings, in high income countries and low low and middle income countries that benefit all types of people at different socio economic status. And not just in kind of in the, in the US, or the UK or in Europe, but all over and globally. So those are two, two thoughts that were kind of really present. With me, as I thought about this, I wanted to just expand on those two thoughts a little, and explain a bit about what I mean by that. So go back, going back to the first idea that we should be thinking about designing robots that reflect the diversity of society in terms of culture and race. I, you know, when I’m in the healthcare area, and my PhD is in me, and I’ve been thinking about designing robots for people with disabilities for a long time, and as we suggest that robots will be taking care of us and being seen as some type of assistance to our clinicians or to our elders, or to people in general. I think that this is where we really need to start thinking about how we design them, how do we train them using the AI and their functional goals? A couple of examples of why this is important. I was talking to a colleague recently about his face recognition software. And something he said to me struck me he said, Oh, yeah, You know, our face recognition software’s really have a hard time detecting people with darker skin? And I thought, Okay, so we’re we’re the AI is how are the AI is trained. So you know, this idea that as we train some of our systems, they’re, they’re, they’re being trained, maybe not enough on a diversity of people. But they are then not able to function well, when, when meeting the diversity of people that we encounter. So that’s an example of what I what I mean by kind of designing them to really think about the diverse diversity of society in terms of culture and race. Another thought is, recently in my lab, we’ve been talking about social robot design, and thinking about robots for children with disabilities and doing remote telehealth. And one of the discussions that we had was, how do we make sure that this robot, when someone looks at it can be they can see themselves partnering with it, and they don’t automatically assume that this robot is white, or this robot is any particular race or color, but that they can actually form a collaboration with the system. And so we talked about developing a robot that kind of can be seen as multicultural. And in fact, we did an exercise with my class that I taught this past fall about asking them to interact with the robot and asking them about issues of diversity and gender, and race, and that whether interacting the robot with the robot engendered any of those thoughts. So those are that’s kind of I think we need to do a better job of including cross section of people in our development process in our discussion about what is ideal. Just another quick anecdote. There’s a paper that came out that said, robots social robots should have eyes with a pupil. And when you looked at the message section, the majority of the people that they had surveyed were white, and they had blue eyes, or eyes that you saw distinct pupil. And and so I was like, Okay, well, of course, then that makes sense that now you’re gonna say that robot should have you know, pupil, because the variety of people that you’re serving, you know, that’s what they’re used to seeing, while if you serve dade, you know, people with darker eyes, or brown or but you’re not going to see a distinct pupil. So that’s not going to be a big deal with those. So that’s just another anecdote of, we need to be careful as scientists and as developers to really consider who we’re talking to, so that we’re not, and that was a kind of well bred paper. And I’m thinking, Oh, Why didn’t anyone ask that question? Maybe because, you know, I think the responsibility is on the designer, and the person developing it to make sure that their, their population that they’re querying reflects this cross culture and gender. And I think if if we do that, we’re going to see robots that are more inclusive, and we won’t have these. At least we’ll have less of these anecdotal stories that I’m pointing out. The second point about low robots should be usable by all just quickly is that came out of kind of some work that I I’m really passionate about affordable robots in global health. Because as I look back at robot assisted therapy, Sara therapy and the systems that we develop, I find that wholeheartedly. They’re they’re quite expensive, and they really haven’t penetrated low and middle income countries yet, in terms of stroke, 80% of the strokes and functional impairments that are resulting are outside of high income countries. So there’s this disparity in terms of here’s this technology that we’re proving to be able to possibly support in areas that have low resource, not enough clinicians where you might be able to leverage technology to support recovery after stroke or upper extremity impairment. But yet still, we have not been able to penetrate these areas, because the systems are way too expensive. So I mean, my lab has been passionate about how do we develop systems that are not only effective but affordable and able to be used in these low resource settings. So that’s kind of my two things. And I think more of us need to kind of be thinking about those things. And I see sometimes we are thinking about like better and better and better and cooler and cooler tech, but the translation of that tech into communities Globally, I think is missing. So that’s my second point about inclusion in robotics. I’ll stop there, Andra, I think I made my main points.
Andra Keay 15:12
No, that was that was excellent. And I think they were very clear points. I’ve been penciling down some questions myself. If anybody else has questions specific, specifically for Michelle, or questions about the subjects that she’s raised, you can table them in the chat, and we will definitely get to them. I’m looking forward to hearing what other angles on the discussion of what is inclusive robotics? And how do we get it that we’re going to uncover tonight. So without further ado, I would like to introduce Dr. Arielle Anders, who is the first technical hire at robust AI and on the board of women in robotics.
Dr Ariel Anders 15:58
So that was a really great talk, I’m excited to try to follow up, I think I have some similar ideas. And I really appreciate Michelle’s comments on your how do we you know, we keep pushing the envelope of what robots can do, but they’re not necessarily going to everyone. And so to start, I just want to say thank you for inviting me to talk and share some of my ideas about inclusive robotics. If you’ve been able to see some of my recent presentations, I’ve been trying to get into the habit of introducing myself and providing a little bit of background and context. And I think that for this talk, the most important thing, the important thing here is that I am a human being. And I think that when we think about inclusive robotics, and the questions that ondra asked the speakers to share our thoughts on, it should come from this part of our humanity. And they’re the three questions, you know, what, what is inclusive robotics? Why do we need it? And what do we need to do to have it? I think that my answers today, you know, I’m kind of excited that we’re on zoom, and it looks like this is recorded, because I am curious to see, you know, what do I think in a decade or so on some of my thoughts here. And I think we’re going to continue growing and learning. And just reiterating on this process. And, you know, we’ve really should start having these discussions more and more, especially from people like myself, who really generally did not work so much in the idea of human robot interaction, a lot of my previous work was in programming robots to do new capabilities. So I’m really excited to share my ideas and what my kind of first impression thoughts are here. And to start off with, when I think about, you know, what is inclusive robotics? I think, what is what does it mean to be inclusive? And my definition of inclusive is belonging. So, going with that inclusive robotics is creating robots that belong in our world. I think that belonging that word to me has a lot of implications on what type of robots will have. Sorry, I think I think my dogs also very excited about that idea. Hopefully, you guys can still understand we’ll try to speak over her in terms of what does it mean for a robot to belong? It should be safe, it should be a robot that’s comfortable around us, you know, we should be comfortable to have it or around. We want to make sure the robots are, you know, they belong, they’re probably not hurting people because out exclude others. I think there’s a question about the appropriate use of robotics kind of going along what Michelle said, you know, it should not be something that alienates more people, it should be something that doesn’t exploit all our resources, it should be something that’s accessible. And to me, it should also be something that society wants. We want robots that are trustworthy, and they work and we want to have them around. And most importantly, in this very kind of circular definition. When I say people, I really do mean all human beings. And so when I think about this, you know, robots that are trustworthy and safe, and they work we want them around. To me. It makes sense that we would want robots like that. I think the question really is, why do we want this definition? To include everyone. And I honestly can’t answer that for you. But I can refer to an expert, Maya Angelou has a wonderful quote on diversity, and that it makes for a rich tapestry. And we must understand that all threads of the tapestry are equal in value no matter what their color. I think that when we think about robotics, it is incredibly important to think about creating robots for marginalized groups. It’s not just for people who can pay for them. That being said, How do we get there, and I honestly don’t know how, but I think we should follow the principle of nothing about us without us. And that means we should try to make sure we have a diverse demographic creating robots, we need people from all backgrounds, otherwise, there’s no way we’re going to get there. So the other idea there is just to keep in mind as we continue going forward with our robotics development, is to remember that we are all humans. And I do want to share a little bit of a personal story here. Back when I was an undergrad, I had the opportunity to present my research at africamps. And our keynote speaker was Maya Angelou, which is incredible that I got to see her. And I can’t paraphrase I can’t even summarize exactly what she said. But she really solidified the point home that we are all human beings. And I think that’s that’s the message I want to share with you when we go to make our robotics more inclusive. That’s a kind of my short, short presentation on what I think inclusive robotics is.
Andra Keay 21:58
Thank you. All right, that was great and beautiful quotes from my Maya Angelou as well. And I’m getting such a lot of rich material from the discussion that’s going on in the chat about what is a multicultural robot that Michelle raised? And you know, you’re very clear points, nothing about us without us. And, you know, I’m thinking we have a lot of issues about where we cite the responsibility along the production of robots. And I think everybody kind of wishes, that it’s somebody else’s responsibility to do this. And quite often we’re reaching the production of robots with great big problems somewhere or other along this production process. There is nothing is working together collaboratively to develop appropriate robotics. So I’m starting to get some thoughts myself around this process. And you know, already, so thank you both for inspiring us there. What I should do is introduce our next speaker, Alka Roy is the founder of the responsible innovation project. And you’re currently a visiting faculty at Berkeley, I believe, as well. Okay.
Alka Roy 23:29
Yes, hi. I was just trying to figure out how to share my screen on my new computer. Sometimes, simpler is better. So I am, let me know if it works. No, it doesn’t work. So. All right. So I am. First of all, thank you for inviting me to this. I was actually more excited to hear what everyone was going to say then what I was going to say. I’m definitely not making robots, which I think I congratulate you for inviting an outsider to comment on this. But I do have a lot of opinions about it. Why not? And I think more and more people should have opinions about this. So I really am trying to see if I can still share my screen because I wanted to share a framework with you that I’m hoping we would think about my provocation tonight is very different. Because when I first got invited, and I heard inclusive robots, I said I don’t want to attend this event because or at least speak to it, because I’m not sure if robot to be inclusive and what does it really mean. And so I’ve been just thinking and thinking and thinking about it. And I because I actually believe that certain things should be excluded. And so let me let me let me explain what I mean by that. I am I’m working on or I have a responsible innovation framework. And that again, I says 1% of time, can I just share my screen or you should be able to do it on the green share screen at the bottom of your screen? Right. So my security settings are not allowing it. All right, so what I see. Give me one more second, if I can, I’ll just talk to it. Alright, so I will put my link in the website. And we’ll have another time to share this. But so I’m the founder of responsible innovation framework. And prior to that, I was at an innovation center in Palo Alto, where we were making a lot of the things that I’m going to talk about. And what happened for me was, I was the end of kind of 5g and AI and everything. So we were enabling anything that you know, mobility, can enable and immersive experiences, robotics. And in the middle of all of these discussions, I found a lot of voices missing, and which were community voices, or people or even when cities were involved, it was usually their CTO that was coming in, not people were thinking about the impact of society. And I kept raising the point, raising the point until I got so loud for myself, that I had to kind of feel like I had to take a break or two sides. And so I stepped out of that arena, and took a break. And this whole world happened with COVID, which has been a really interesting place for reflection for all of us. One of the things that I worked on last year is this responsible innovation framework, which is like a nested three by three by three, which is why I was hoping I could show you the visual, but just go with me. And and I will try to create it for you. So the first thing in there is the provocation for for wear robotics, which is really, I’m really excited to see what you’re going to think about it and how you’re going to disagree with me, and how are you going to tell me I’m wrong, is I think that we need to think about stakeholders differently. So we have people with everybody talked about and and really people, not customers, not clients, not you know, people, then we have the planet in the environment, physics, our laws, you know, all those things. The third stakeholder is technology, things. So I call it people planet things. And the reason I put that as a separate stakeholder is because I think, or my, my provocation is that the reason we design technology, so problematically including robots and chatbots. And digital twins, is because we don’t separate people and things. Because we try to build things in our own image, or the image of something living. So we’re taking an inanimate thing and creating it in that image, animation world is full of it. And that causes a lot of confusion for us. Because we transfer our feelings, we transfer our biases, we transfer our confusion to these inanimate objects. And now there’s so much research and where people are, you know, feeling things about the robot, which they’ve created a robot to sort of clear the minefield. But when they, they see this robot getting blown up, they stop and they’ve removed the robot because they’re feeling that this robot is doing more than it was designed to do. There’s a reason to create social and empathetic robots. I get it. But I think that in that process, we create teenage girls, we create women as our servers, and we just perpetuate the spice just sweep deep we can decouple. So my provocation is, why not design robots is things, useful things, but not in our image, not in our voice. And so that we don’t transfer, you know, just appears as a thing. And we treat it as a very useful thing that communicates and helps us and it’s technology. So those are, you know, that’s like basic provocation for me, the stakeholders. The other aspects of this responsible innovation framework includes the values of open and safe, which are conflicting values, but, but both needed, delightful and trustworthy, and inclusive and dependable. And their inclusion is about people and ideas and types of people. And the reason I say delightful and trustworthy is because whenever we talk about responsibility, you’re doing the right thing. We can vote like we get, we feel like we have to be very serious and boring. Whereas I think that’s the part of robotics design, which is very useful. lightful for us, and I love what Michelle said. And I think Ariel said about, we always looking ahead, and we feel like we have to be chasing the latest and the greatest. But there’s so much for us to learn from our weathering technology, the old stuff, and make things accessible, you know, around around the world. So I hopefully at some point, I’ll get to share this framework with you. But those are my provocations for this conversation. I’d love to hear what other folks have to say about that. So thank you.
Andra Keay 30:33
Thanks, Alka. And I could see some people are very positive about what you’re saying there. I have to agree that the good design framework is to not create something that is perpetuating stereotypes, because we do anthropomorphize and it triggers our unconscious biases and stereotypes, and I see some more conversations coming. But I think it’s time to introduce Kenechukwu Mbanesi from the Black In Robotics undergraduate committee. And you’re currently, he’s currently doing a PhD in robotics engineering at Worcester Polytechnic Institute in Massachusetts. And I’m looking forward to learning more about that over to you.
Kenechukwu C. Mbanesi 31:29
Alright, hello, sorry, I was just trying to get my sharing going on here. Just give me one second. Can you see my screen?
Andra Keay 31:47
Yes. Looks good.
Kenechukwu C. Mbanesi 31:49
Awesome. Yes, thank you so much for the privilege to be to be here to join this conversation today. I, I’ve benefited a lot just by listening to what the previous speakers have talked about. I want to take this discussion, somewhat piggybacking on what Ariel ended with, about how one way we can achieve inclusive robotics is by getting people from all backgrounds involved in robotics engineering. For a long time, I’ve been very passionate about the power that robotics and technology has to, to provide improved quality of life and socio economic benefits for people in underserved and underrepresented communities, especially on the African continent. And that’s really where my passion lies. And you see it, I see. I see exclusivity, which is the inverse of inclusivity is caused by two things. And it could be more but in my understanding is two things. One is discrimination. In other words, on equal opportunity. And second is limitation. More or less on equal access, right. And my passion really lies in addressing the Equal Access problem. And how I envision that is using education. Right? So how can we get more people who otherwise would not have a seat at the table of discussing and developing robotics technologies to be able to do that, by providing them the means through education. And I’ve been very privileged to be part of interesting projects with this goal in mind, and I’d like to share a few of that with you. The first project is math and science for Sub Saharan Africa. This project was supported by the World Bank, and our audacious goal was, was to run a continent wide train the trainer program to upskill stem teachers to teach math and science and robotics, and really to understand the interconnectedness. Right, so how does science and math inspire people to pursue engineering and robotics? So I think it was that if we could inspire and educate teachers, well, they would in turn, educate and inspire students. Right. And so we started this in 2016. And we’ve been able to partner with government education agencies on the continent, to run training programs in person and virtually in several African countries. And just to put a plug in here, one of the major challenges that we’re faced and this project is I know we really still do is accessing affordable, educational robotics kits at scale. We were very fortunate to get Vex robotics to donate a couple of kids. But beyond that, we still see that as a significant bottleneck to getting students and teachers access to kids that can really help them learn and participate. The second project is called bad for kids or co bad is stands for public collaborative robots. And this project was supported by the advanced robotics for manufacturing Institute here in the US. And the project was designed really to combat the current gaping shortfall of skilled manufacturing talent in the US. And particularly trying to address that in an inclusive way, right. So what we’re doing really was to swim upstream, to inspire and train middle and high school students, especially from underrepresented backgrounds, to be passionate about robotics and passionate about advanced manufacturing by taking them through several week after school program, so like after school program, where we teach them about how to program collaborative robots, reasonably UI robots, you can see on the screen, and to actually teach them how to manufacture items using CNC machines with machines and other things. And really, our vision was that this hands on experiential learning opportunity would inspire them to, to be passionate about this and to help with getting them into this programs, or in college, and to really retain it to increase retention so that they can actually go and pursue robotics in the future. And the last project is when I got involved in very recently, and is very dear to my heart, it’s the pan African robotics competition. It started in 2015, really small, but right now it’s grown to be probably the largest robotics competition on the continent. And really was designed with inspiration from from first robotics, if you’re familiar with that here in the US, really with a goal to inspire the next generation of roboticists on the continent. And we have seen very strong participation and significant impact on how students are motivated to not just be consumers of this technology, but to have the confidence that they can be part of the producers of the technology, right, they can be part of people who actually develop this technology. And that’s naturally where my passion lies. And so this competition is going on every year, and we get students from all the way down from middle school, all the way up to college level students, and, you know, to program robots to compete with each other, and to just really inspire them. So really, to summarize, I, I am a very relentless believer in the future of inclusion. And one, in one way I’m striving for that as, as this short presentation shows is by working to provide young people, particularly people from underrepresented communities, with access to education, that and skill development that allows them to not only be at the table where they can develop technology in a way that’s diverse, but also benefit from the dividends of, of the technology. Thank you.
Andra Keay 38:11
Thank you very much, Kenechukwu that’s, um, I’m looking forward to learning more about all of those initiatives. Particularly, I think Ari raised an interesting point, there can be a lot of education initiatives, how can we maximize the benefit, as well as the access rather than splintering or fragmenting success. And I know that open robotics is very keen to democratize access to robotics by putting forward access to simulations, to get around some of the costs of having access to physical robots. But of course, that requires access to internet. And we’re seeing even in the United States that there is there is a complete gap between those that have access to internet for education. And those students who for the last year have really struggled because of lack of lack of lack of access to the internet. And I don’t know what steps we need to take. I think we’ll discuss that a little bit more as the discussion moves on. But I would like to introduce our next speaker, Kenya Andrews, master’s student in Computer Engineering at the University of Illinois, and on the black and robotics undergraduate committee.
Kenya Andrews 39:41
Okay, thank you so much. Hello, everyone. I’m very excited to be here. And I didn’t bring slides so I’m just going to, you know, speak. So, first, I’m gonna first talk about a little bit about my background and Kind of where I fit in the space, and then I’m gonna go into the topic. So I’m a first year Masters student, and I live in this space where we make algorithms, right? And those are the things that that are the, the decision making properties of robots, right. So that’s, that’s the space that I’m in I do. I, my focus in my research is in machine learning fairness. And within that, my, my passion areas and that are algorithmic justice, algorithmic bias and decision making. Okay, so my current project is looking at fair distribution of COVID-19 vaccinations amongst vulnerable populations at the state of Ohio. And right now we’re looking at different measures, trying to build different models. So some of the some of the things that we’ve kind of tossed around are like, well, should should there be equal hospitalization rates between vulnerable and non vulnerable people? Would that make it fair? Or would it be more fair, if every every, like census tract or, you know, center would have the same amount of vaccines distributed to them? So, and we’ve been, you know, having to work around things like, like, distrust between historical you know, because of historical injustice and things like that, how that creates distrust and less vaccines and how that affects distribution. So hopefully, that gives you a little context for for where I’m coming from. Okay, so when I first saw inclusive robotics, I thought First, well, oh, that’s a, you know, loaded topic list. So let’s break down some definitions. So first, I wanted to look at what is the robot? And is the programmable machine that does the task, you know, with or without human assistance? And then I want it to look at, well, what is inclusion? That’s where I started having some some issues, because we, we don’t have a have a real agreeance, on what inclusivity looks like. And it’s not because, you know, it’s for different reasons, right? Like, maybe maybe it’s like, you don’t agree with someone’s lifestyle, so you don’t want to, to include them in things or, you know, I think I should have more because I do more, I work more, right. So it creates, you know, these biases around if people should be or should not be. So, I went to Oxford, and some of their definitions, were not excluding any of the parties or groups involved in something, and aiming to provide equal access to opportunities and resources for people who might otherwise be excluded. So putting those two things together, I thought about, well, we want to have robots that makes smart decisions to achieve, you know, different goals in this world, right? And it’s not like, it’s not just that we want them to do things, you want them to do them well, right. So unless they can do them, well, then maybe they shouldn’t do it at all right? That’s what that’s kind of what we were talking about a little bit earlier. So this analysis of if robots are doing something well, should encompass, you know, are they being just are they being fair in the decisions that they’re dishing out? And in order to look at that it kind of needs to start from the beginning, from the beginning of algorithm in that, you know, at the end, there are several stages of that people look at fairness, and one of them is at the end, they all the outcomes at the end come out good. What about when you were making the decision? So that’s another another scheme of fairness, like, Where were they in between stages, between the decisions good, where they fear that’s what I’m good. So I do want to hopefully you guys can hear me a little bit I want to talk about a few definitions of fairness. So hopefully, that can bring some context into where we are. So right. So some definitions of fairness are unawareness. So that that’s when we exclude different attributes of people. So that from data, even though it’s there, we know it’s there, we don’t look at it, we just ignore it. And that can be really dangerous. Because if you ignore those things, then you would be ignoring the historical injustice that comes with it. Right? So you’re if you ignore the fact that someone is a black or brown person, then you would be ignoring the injustice that they had to face to get where they are, and how that could have affected what we look at as their resume. Right. So maybe they don’t look qualified to you. But if we, if we, you know, considered the things that that affected their resume, then we can start to see, you know, actually, maybe they are qualified or maybe even more qualified than, than someone you know, who has a similar or equal resume. Another definition of fairness is demographic parity. And best when you look at if all the demographics and a data set, have the have equal outcomes, right, like, if white, black, Asian, Indian people are accepted at the same rate, or rejected at the same rate, and grouping demographics together, can completely cancel, completely cancel another group demographic that they could be a part of, right. So it’s not only race. What about you, though? Oh, they’re short, they’re tall. What about age, you know, things like that. So you could completely be ignoring someone else’s a part of their a part of their person, if you group them together like that, that’s the danger in that. And then there’s something called equal last odds. And that’s if you’re qualified, that you’re equally as likely to be accepted into something as someone else. And another one is equality of opportunity. So regardless of your demographic group, you’re equally as qualified to have something as someone else or reject it as someone else. And then there’s predictive parody. And that’s when the is very similar to demographic parody. And is when the positive rates are the same. Okay, lastly, I’ll go over firearm safety. And lastly, I’ll go over counterfactual fairness. And that’s if the outcome for you is the same. If you’re in this world, where you are, who you are, versus, you know, if you had a different set of demographics. So instead of being a short black woman, then I would be a Caucasian woman, and Mother, do I have the same outcome? Right? So those things are important, because when we look at robotics, and the decisions that they’re they’re pushing out, are, are you what measure of fairness? Are we going by, you know, and which one is most appropriate for our for our context, it changes, it changes all the time. So we have to be we have to be very careful. Um, when we’re designing robots and understanding what kind of space are they going to be in? And who are they going to be working with? Sorry, I’m reading my notes one second. Okay, yes, we want to make sure that we’re not promoting disparity but that we are minimizing your or mitigating it, right. So how can we do that some things that I really think we should do is just kind of take a step back and slow down, we need to look at where we are right now. And, and come to come to agreement on, on what inclusivity looks like. And once we all can sign off on that, we need to understand how we make decisions, because we don’t even have a great understanding of that. We need to know, you know, what is it that we’re that we’re saying is good or bad? And why are we saying that? If the things that reason are Why are bad and maybe that shouldn’t be in our design, right? Or the reasons that are good, you know, maybe we need to superimpose those and make them even larger, make them count more. And another way, after we after we have a good understanding of that and we’re developing good things. I think we have to start investing, right like you, you maybe you if you’re designing things for community or robots For a community that’s going to be specific to a community, maybe you should go to some of the community meetings and sit and understand their struggles, the things that are actually going on there, so that you have a real understanding. You need to support organizations like nesby, right? The National Society of Black Engineers or women in robotics, right? support them with your funds, support them with your knowledge, volunteer to speak. Even students tell students about robotics, and you know what it looks like. But while you’re telling them about robotics, take them to the African American museums, take them to the Holocaust museums, show them, you know, teach them moments in history, where we weren’t so fair, we weren’t so just we weren’t so inclusive, and how those things can can translate over into our lives, start a scholarship plays, you know, be a mentor. Yeah, those are, those are my thoughts on. So thank you so much. I appreciate you guys.
Andra Keay 51:09
Thank you so much, Kenya, you raised great points. And it was wonderful to have the definitions there. And what I like most is that you took it back to saying starting with the algorithm. And this is something that is both critically important and also crucially problematic at the moment. Because right now, robotics has become a subset of AI. And that means that federal policies on AI, are incorporating robotics, it means that the agenda is being driven around the discussion around AI and AI ethics. And it is often being done in complete ignorance with robotics. And some of the problems there is that if the discussion is only about robotics, then it may only be about safety, rather than algorithmic transparency. But at the same time, if the discussion is primarily about the algorithm, then it’s going to be excluding the impacts of physical robot. And they just have a completely different and expanded way of inter intersecting with us in society. So, you know, I love that you started with the AI in that discussion. And let me just see if Ken would like to speak now. And
Ken Goldberg 52:31
Thank you. Good, thank you. I appreciate that. I am really inspired by by a lot of this discussion. And I also want to take a moment to acknowledge that the event tonight is sandwiched between two major events, at least in the United States. One is Martin Luther King Day, which we celebrated yesterday, and tomorrow, which is the inauguration. And it will be approximately 15 hours or so we will have a new president of the United States new ministration. Which I don’t know how others feel about that. But I for one, I’m very, very, very happy about it. The I think that it’s important, because is going to we are at an opportunity, a new chapter in in American history, which I think will affect a lot of events globally. I think as as you know, one thing that that struck me is that the that is Kenya, I’ve just mentioned that the COVID vaccines process is going to be very interesting reexamination of our sense of inclusivity. Because we are going to have to think very carefully and deeply about how we prioritize the the vaccine. It’s been very interesting to me that the that seniors and, and healthcare workers, prisoners, incarcerated individuals have been prioritized with good reason. But it’s very interesting, because, you know, they’re not often they’re often not considered in, in our priorities. And so it’s been a forcing us to reconsider. And I think as the more vaccines become available, we’re gonna have to do some really careful thinking about how this is rolled out is going to cause a reexamination. And I want to note that this pandemic has, has, has, has woken us up in so many ways. It was 100 years ago that the 1918 pandemic occurred. And I was reflecting on the idea that the word robot was coined in 1920, right after the end of the pandemic, and I’m still trying to wrap my head around the idea that it was such an interesting context where they had just gone through World War and this horrendous threat to humanity. And that’s when the playwright Karl che back in Czechoslovakia basically comes up with a story about robots rebelling against against the the totalitarian regime that was basically forcing them to work. And that that’s the word that’s where the word robot originated. So 100 years later, We think that thinking about robots in this context of our political, economic, social environment is so important. And so I think that the the points that were raised here from the beginning from Michel characterizing, you know, what are we? What is our definition? Because I think that’s so hard to actually wrap our heads around. I mean, we can talk about, you know, seniors and children age as a sort of inclusivity. Right, that’s one very big factor, then there’s a one we didn’t talk about tonight, but gender, and LGBTQ T, right, there’s all the gender issues that have come to the fore, actually this year. And race issues, I mean, Black Lives Matter. I think that the bipoc, the whole idea of thinking about in new ways that has created a lot of shifting of attention and priorities in a really positive way, I can tell you that black and robotics, and black and AI have had a big influence this year on our admissions process at Berkeley, we’re right now reviewing applicants, and we are getting a lot of attention. We’ve got more applicants than ever before, more black applicants than ever before. And two of them are here, actually. And I want to say it’s wonderful, because they’re there. What’s really important is that we’re, we’re learning and educating the faculty that what’s important is not just looking at the scores, how many papers they’re reading, but what is their trajectory been? So if a student has come from a, from adversity in a small village, and in Africa, and now is an undergraduate at, you know, doing Greenham? Well, in classes, that’s a huge trajectory. I mean, that means they’re there, you know, imagine what they had to overcome to get there. So really think about that, in regard to how you’re evaluating that student. They may not have a published paper, but they’re on a trajectory to do, nothing’s gonna stop. Right. So I think this is really fascinating. And it’s a really powerful and important time, that I also think in terms of other you know, races, we talked about, you know, you open tonight ondra with a with a story about Native Americans and Aboriginal people, I think it’s really important. We also consider Hispanic, Indian, though the full spectrum of races that are out there. And languages, by the way, a big disparity that excludes many people is their language they speak. There’s a, you know, there’s an emphasis on English in a lot of the publications. But that is very difficult that it’s not your native language. So you have a barrier to overcome and the way you read and write that is, we need to think about how to overcome that even our conversation tonight is in English. And then the translation, nowadays, some of these tools and AI, again, comes into play, are going to open up these doors, and I hope they will increase in quality so that we can have simultaneous translation. And for example, for for people with hearing disabilities, having ability to have automated translation, Closed captioning is wonderful thing we’ve been using in our classes, I have a student who’s on hearing disabled, and we use this for all of our meetings. So it’s been it’s opened a huge amount of doors for all kinds of disabilities with regard to cognitive, neuro cognitive diversity, or neuro diversity. And people have learning disabilities, we find out today with COVID-19, that these kinds of learning disabilities are much greater than we thought before, students have all kinds of challenges. And that is also important to acknowledge the and also, as Michelle noted this socio economic variations, right, we are oftentimes targeting this kind of particular people who can afford these kind of robots and tools and even have who have Wi Fi in their homes. Right, but many people do not. So how do we think about that? And also, I also think that intellectually, we also tend to target, you know, in the inclusivity, in terms of the people developing are oftentimes, you know, nerds like me, engineers, right? We’re all people who feel pretty good about doing science or math stem, but many people don’t, they just don’t have that, that they’re uncomfortable there. And they, but they feel excluded. So how do we engage with people who are the artists and the humanists, and the writers, the journalists, you know, who are so engaged across the board. So all these things, and workers are oftentimes the affected by the robots that we’re, we’re developing, so we need to be engaging with with workers, and really thinking carefully about how it’s going to affect those workers, especially minimum wage, who are, you know, and most vulnerable to these technologies. So, um, there’s so many things that this is sort of, you know, engaging for me. And I also have to say, I’m Kenny, I have not met you before, but I’m so excited to follow up with you because we have a common background in Nigeria. I was born there. And in the in the 60s, and we would those programs you mentioned I’m not aware of. But it was fascinating because we had we started something called the African robotics network with a professor in Ghana. And it’s, it’s, it’s, it’s been a little bit in with the first objective of it. This was in 2012, was to build an ultra affordable robot for education to design an ultra affordable. And the challenge was to design something under $10. So we thought nobody could ever do that it’s programmable robot for under $10. Anyway, it turned out that someone did. And it was a it was a it was a hobbyist living in Thailand, basically came up with what he calls a lolly bot. And if you look it up on the internet, it’s l o, l, l, y BOT, a lolly bot, and it costs $8.64. So you can build it from an old Sony game controller. Anyway, what I want to say is, I love what you’re saying, because I completely agree with this, there is a big opportunity, I am very excited about Africa and its potential, I think that is a major continent, and that we are that there’s going to accelerate into the future. And one of the things that African students, I found a really incredible ability is to know how to think outside the box in a way that they think differently because of their experience. So they’re very, very attuned to how to make something affordable and sustainable, how to make something that works, even when the electricity goes out, which nobody in the West usually thinks about. But those kind of things are really important. And so and there is engaged and interested in robots as any kid anywhere. So that’s why I want to be able to bring them to robots and the programs that you’re talking about, like the pan African robotics competition, I absolutely love it. So I want to connect with you because I would really like to follow up. But that is one of the things I think we can do. As the group of us tonight, which is to break this we’re forming a community I mean, what I feel is that there’s a there’s a real sense of, of some ground ground grassroots thing happening here. And I’m so excited, I want to thank you Andra for putting that putting this group together. Because there’s a spark here that I want to support. And I mean, I really want to see that grow over the next few years. And I think we are at a moment in time historical moment when this is an opportunity for us to step forward and really take take this opportunity and do something with it carried forward in a really meaningful and sustainable way. Thank you.
Andra Keay 1:02:03
Thank you so much, Ken, that was a very wonderful note to finish on. And sadly, for tonight, we are out of time. But the problem space is enormous. But it’s fitting to realize that the opportunity space is even larger. And I personally had thought a lot. What would a robot look like if it was designed by women for women, and I could imagine things being different. But imagine now what it would look like if your language models were for languages other than English. Or if you had to rise to the challenge of developing language models for multiple languages, as is the case in Africa. And I love the examples that can gave us there as well about really thinking outside of the box. If it’s been proven, fairly scientifically, that diversity drives innovation. And it might not be as comfortable. But it is certainly far more productive. If you’re looking to make change, as well as creating something that’s inclusive. So we shall have to continue this discussion and extend this discussion into our workplaces and to the rest of the people around us because we Yes, we need to get everybody in the room. And I’m looking forward to taking that journey with you all and thank you so much tonight. That’s wonderful speech. Okay, I’m going to stop the recording and say goodnight to everyone.
Dr Michelle Johnson 1:03:48
Transcribed by https://otter.ai