Transcript 0:00 [upbeat music] Welcome to The Peggy Smedley Show, the podcasting voice of IoT and digital transformation with your host, Peggy Smedley. 0:13 [upbeat music] Hello, listeners, and welcome to today's edition of The Peggy Smedley Show. As always, I'm your host, Peggy Smedley, and what a great day to be on the radio. 0:24 As we march toward a world enhanced by artificial intelligence and automation, we need to make sure we're not marching forward too blindly as I wanna say today. 0:36 We need to think about what problems these technologies may create as they revolutionize all of our industries. 0:44 And while we won't be able to anticipate all the problems or even come up with all the answers to all of the problems we can anticipate, we can start talking about AI from every angle to help prevent us from being blindsided. 1:01 So one discussion we need to keep on having is AI and ethics. So today on segment one, I'm actually going to be kinda digging in a little bit deeper into the bias in AI and what that really means. 1:14 But first up, I've got some guests on the show. I'm pretty excited. 1:18 I've got Susan Dalton who's coming back, who's the vice president of Smart Technologies of IoT at GCP Applied Technologies, and she's gonna be joined with a colleague, Mark Roberts. So he's, uh, gonna be with us. 1:31 So the two of them are gonna talk about what's happening, uh, in the construction space and so much with big data. And then I'm gonna wrap up the show with Eyal Ben-Ishtai, uh, founder and CEO of IronScale. 1:44 So I've got a lot of really great things to talk about when we talk about what's happening in the world today, and we talk about security and all of the ethics on everything we're talking about. So let's get started. 1:54 We've got a great show and a lot of things to get underway here, but artificial intelligence is fundamentally changing our society. 2:04 And I think there's no question that you already know how great it is and how great it can be, but let's check our enthusiasm and talk about the ethical concerns that AI raises. 2:17 I mean, I'm sure most of you yesterday heard all about Google's new phone, and many of you, uh, have your own thoughts about the new recording device using AI. 2:28 So if you haven't heard about it, go check it out, but this also opens the door to many questions about the use of AI and ethics just in general. 2:36 But with that in mind, I wanted to really kind of think about where are we going in general with AI and all of these new devices and apps and things that are surrounding us every day and new things happen. 2:49 And some of these things we need to be thinking about in designing AI systems is transparency, fairness, accountability. How about safety? How about bias? 3:00 We talked to some PhD, PhD students who are doing some intensive research in this area, and one of them put it in a really great way, and I wanna articulate the messaging that the student said. 3:13 It said, "During the curs- course," I should say, "of making AI systems, countless decisions or trade-offs," think of that word, trade-offs, "are made, and this affects the way the algorithms behind these automated systems operate and make decisions." 3:32 I really appreciate that word trade-offs because it perfectly explains how bias ends up in AI, and humans are the ones building the algorithm- algorithms and training the AI. 3:45 Humans much-- must teach the AI how to make decisions, and they're going to do so based on their own personal worldviews. Now, we all know this. 3:56 We-- Maybe we wanna go blindly into thinking about this, but it's just inevitable and inescapable. But let's dig in even further here. 4:05 MIT makes the case for bias entering the AI equation before the algorithms even exist as far back as before the data that will inform those algorithms are even created. 4:18 Now, this is very interesting because consider how folks go about creating deep learning models in the first place. So their goal is to make a good business decision. Now we're getting into the business of this. 4:34 So the AI is going to end up prioritizing whatever makes the most business sense over other factors, and I think like fairness or discrimination. 4:45 So another sticking point to consider is this: Do we know exactly what fairness means in the context of machine learning outcomes? 4:55 I read a research paper earlier this year that effectively argues that no, we can't really do that. 5:02 And yet, even as I say that, I have to also say this, and you know, I know you're all gonna n- know exactly what's coming out of my mouth next, is we have to try. You know, we have to keep working at this. 5:15 The stakes are actually too high in my personal opinion. AI is already being used or will soon be used to make decisions regarding loans, insurance rates, college admissions, let's not go to that route, right? 5:29 Job placement and candid- candidacy and so on. We've all bet on this. We keep hearing about it in politics just all the time. But let's go back. Just think college admissions. 5:39 You know, we're talking about people paying, but where's the bias in that? I mean, that's gonna be a big one. But let's go in politics. We all know there's gonna be a lot of discussions in an election next year. 5:53 Bias is not just a problem even when AI is making big decisions either. How about when voice assistants can't understand a person's dialect? 6:02 Is that fair?Think about that for a second when there's a racial imbalance in the teams creating AI solutions like voice assistants it can mean those AI assistants have a hard time understanding people who speak differently than the people who predominantly are training the algorithms. 6:21 Think about when you can't pronounce just a simple word let alone a dialect. AI gets a lot right though, right? 6:28 We know that and we need to remember that even a biased AI system has the potential to be less biased than a human who doesn't necessarily try to be unfair but makes decisions based on factors that even that person can't always understand or explain. 6:48 I guess the scary thing if you think about it bias in AI versus bias in humans is the potential for AI to scale and the bias along with it. So that's what we have to think about. I think I wanna think about another way. 7:02 As McKinsey puts it, "AI can help reduce bias, but it can also bake in and scale bias." Now, that's an important point because that kind of makes you shudder just for a second, right? Think about that. 7:19 Think about the possibilities there. This is something that a lot of smart people are working on right now, and I would like to think a lot of smart people who are a whole lot smarter than I am. 7:31 But I have to raise these questions because I interview a heck of a lot of smart people every single day, and I ask them the very tough questions. 7:40 So the Algorithmic Justice League is a collective that aims to highlight algorithmic bias and increase awareness about it, and I find this organization or this group very interesting because they provide a space for people to voice their experiences with and concerns about bias in AI, and they help people develop practices for accountability during the design, the development, and deployment of code systems. 8:09 Now, we all know organizations like this mean well, but how much do people really get involved? That's what I'd like to understand. 8:16 And on their website, users can request a bias check to help testing their technology with diverse users, and I like the idea that they're trying to do a checks and balances here, and they can report bias in someone else's technology. 8:32 Now check that out. You can report someone else who's got bias, okay? And they offer a lot of res-resources that help raise awareness about bias in AI through media, art, and other science outlets. 8:46 So if you're a company that's using facial recognition and facial analysis technology, I, I would encourage you to go to the Algorithmic Justice League's website and take the Safe Face Pledge. 9:00 Maybe you wanna be a part of it. Maybe you wanna say, "Look, we don't want it." I know people do these pledges all the time, but do they really mean it? But the idea is we have to start somewhere. The Safe 9:11 Face Pledge that they have, Safe Face Pledge, is an opportunity for organizations to make a public commitment kinda to mitigating the abuse of facial analysis technology. 9:25 I mean, SAFE stands for Show Value for Human Life, Dignity, and Rights, Address Harmful Bias, Facilitate Transparency, and Embed Commitments into Biased Practices. 9:40 I think it, it, it's kind of interesting what they're trying to do, and the question is whether everybody wants to stand up and do it together. But here's my tip for today. 9:49 IBM says AI bias will explode, but only the unbiased AI will survive. Very interesting commentary, right? So make sure yours is in the right category. Make sure yours is the unbiased AI that's gonna survive. 10:06 We've talked about how it's anything but straightforward to create a fair and unbiased AI, but the key will be to do everything you can to avoid bias and then deal with the issues as they come up. 10:19 I think that's really what you gotta think about. So... And I guess my opinion, too, today is this. 10:25 As we raise the bar for automated decision-making, let's raise the bar for human decision-making, too, because that's important. The best way to prevent and spread of bias in AI is to hold ourselves to a higher standard. 10:39 That's essential. What's the standard? Let's raise that bar. Think about it, and I wanna know what you think. The more we raise the standard, raise the bar, the better we're going to get, but I wanna hear what you think. 10:52 I started this discussion on LinkedIn, and we're just starting to dis-discuss it. Chime in. I mean, let's get your thoughts. The more you chime in, the more we can get others to think about this. 11:04 I know it's a tough question, and I know we don't have all the answers, but unless we start challenging ourselves, that's the only way we're gonna raise the standards. And we've gotta actually talk about these things. 11:15 And I know lately some of you have been talking with me on LinkedIn and on social, in other ways, in Twitter, so let's keep it going at connectedwmag and on LinkedIn. Let's keep the discussions. I love it. 11:26 I love when you challenge me, and I love when you challenge each other. [upbeat music] All right, we're out of time for this one, but stick around. We've got more to come. 11:32 Check out our website at connectedworld.com or thepeggy smedley show.com. Stick around. This is The Peggy Smedley Show, the podcasting voice of IoT and digital transformation. 11:41 Remember, with great technology comes great responsibility. Go get that cup of coffee. We'll be right back after this commercial break. [upbeat music] IoT is here. 11:59 AI and machine learning are catching on, and they are all changing our world. But what's next?How will you be impacted? Let The Peggy Smedley Show keep you up-to-date and looking ahead for your business. 12:11 Tune in every Tuesday at noon Central Time at peggysmedley show dot com. Too much to do? Not enough time to get it done? Call on the experts at Another Eight Hours for your business support needs. 12:24 By partnering with Another Eight Hours, we allow you to focus on the more important matters, like being in front of your clients, doing what you do best, rather than being stuck at a desk poring over paperwork, rummaging through emails, returning phone calls, and struggling to get everything done by yourself. 12:38 Meanwhile, your family and social life are going down the drain. Go to anothereighthours.com or call Eight More Hours. That's eight six six seven three four six eight seven seven. 12:48 Constructek TV, your fierce advocates for construction. Constructek TV digs into infrastructure, drills into global economic conditions. 12:59 It is changing the talent gap narrative, adopting a new vision and a voice to the new skilled worker in construction, all with the help of technology. 13:09 Constructek TV, everything you need for infrastructure, labor, and technology. Like, share, and subscribe to Constructek TV at youtube.com forward slash constructektv. [upbeat music] Hi, Scale listener. 13:26 This is David Finkel, co-host with Jeff Hoffman of Scale Your Business. I wanted to let you know that our newest book, Scale, was just released and to encourage you to get your copy. 13:35 The book will give you seven proven principles to grow your business and get your life back. Scale will help you work less by getting your business to produce more. Get your copy online or at your local bookseller. 13:48 For more information, visit us online at scaleyourbusinesstoolkit.com. [gentle music] Can you imagine a world without children? A world where children don't play or dream or imagine. At St. 14:04 Jude Children's Research Hospital, we're working every day to find cures for diseases that strike down children, because we can't imagine a world without children. Can you? Finding cures, saving children. St. 14:17 Jude Children's Research Hospital. To learn how you can help, log on to our website, stjude.org. I've heard this is, like, one of the best pizza spots in town. Yes, it is. 14:28 I'll do a slice of pepperoni, slice of vegetarian. You got it. And I will pay for all of that in three days. In three days? [laughs] What's that mean? Well, wait, you accept credit cards. 14:36 That money's not gonna hit your account for three days anyway. I need my money quicker. At Chase, we hear you. With Express Funding, card payments are in your Chase account the next business day. 14:44 Go to chase.com/expressfunding. Chase for Business. So you can. Compensated participation. All businesses are subject to credit approval. 14:51 Not all clients are eligible for next business day funding and additional terms, conditions, and restrictions apply.
