Creating AI that Unites, Not Divides
Did you miss Personal AI’s latest Discord Discussion with world renowned AI Ethicist Nell Watson? Not to worry, we have a breakdown of the talk for you, with details on the topics that were discussed, the questions that were asked about creating AI for unity, and roadblocks on the path to the ethical future of AI technology.
Panelists:
Panelist: Nell Watson, Futurist and AI Ethicist
Panelist: Sharon Zhang, CTO at Personal AI
Moderator: Atima Lui, Strategy at Personal AI
Atima Lui: Welcome everyone. I know a lot of you know me in the audience. I’m Atima Lui. My pronouns are she/they. I am a strategy consultant here at Personal AI. I focus on marketing, diversity, equity, and inclusion initiatives. And it’s truly my honor to welcome you all to our first Discord event. Our mission, at Personal AI, is to empower every individual to own their intelligence and be their own thought leader. So one of the ways that we’re looking to achieve this mission is by holding events like this one, where our community has the opportunity to hear from incredible thought leaders, both inside and outside our company on our latest topics that are concerning AI and blockchain that we want to talk about.
So as you can see, on our virtual stage today’s event features our fantastic CTO, Sharon Zhang in conversation with world-renowned AI ethicist and futurist Nell Watson. Before I go any further and before I introduce them, I just want to take a moment to encourage all of us who are users of Personal AI to open up our Personal AI. On your browser or your desktop app and make sure that you’re ready to stack.
Here’s how our time together is going to go to set the stage a bit further. So that you’re prepared, Sharon and Nell are going to chat for about 30 minutes and they’re going to be focusing on the topic of Creating AI that Unites, not Divides and then there will be about 20 minutes or so at the end for the audience to ask them both questions.
I’m going to go ahead and introduce Nell and Sharon. So our wonderful panelist Eleanor “Nell” Watson is an engineer who lectures in machine intelligence research. She has contributed to ethics initiatives and holds fellowships with the British Computing Society, Royal Statistical Society, the Atlantic Council, as well as holding senior membership within IEEE.
Now introducing Sharon Zhang, the co-founder and CTO at Personal AI. Deeply passionate about humans and the technology that evolved alongside us, Sharon has spent the past decade building AI and natural language processing products that impact the lives of patients, employees, and everyday people. She has experience leading teams at large tech companies such as Nuance Communications, Kaiser Permanente, and LinkedIn, as well as at startups. She’s a graduate of MIT where she double majored in Biology and Electrical and Computer Science Engineering, with double majors in Economics and Biomedical Engineering.
Our first Discord event: on Creating AI that Unites, not Divides.
Our mission at Personal AI is to empower every individual to own their intelligence and be their own thought leader. One of the ways we’re hoping to achieve this mission is by providing a platform where our community can hear from individual thought leaders both inside and outside our company. AI brought us together, just like it did on this platform. Discord is great because I look at right now, how there are oceans between us with Nell being in Ireland, Sharon being in San Diego while I’m in New York.
Get Ariadne Bazigos’s stories in your inbox
Join Medium for free to get updates from this writer.
Question: “How can AI help us to bring people together, just like this event, instead of bringing us apart? In each of your own personal experiences, how and why is AI polarizing our society these days?”
Nell Watson: I can take a crack at that. I think it does seem clear that our society across the globe has become polarized in many ways, and people have become increasingly entrenched within certain identities or certain opinions. People seem less willing to give other people less of a benefit of the doubt if they don’t like the scent of their content, or the values they think they possess. I do consider the algorithm systems which have connected us with content and with people around the world have played a significant role in that and there’s quite a bit of research to substantiate that, as algorithms, which are optimized for engagement, tend to preferentially select engaging content such as that which makes people angry or afraid…that’s just one example of how technology can bring us further apart, but at the same time I do think there are other opportunities for the very same technology to help us bring our communities together to help us understand each other in new ways and to translate between people who might find it tricky to understand each other.
Sharon Zhang: I like the way that Nell puts it. I think technology has the power of doing both good as in bringing people together and also may have the side effect of separating people, which is a polarization problem…. Another thing that comes to mind is social networks. For example, we usually get recommended content that we like, which is the great thing about personalization. For example, I love how Spotify knows exactly what I want to listen to, but that also creates these silos if I’m only listening to my own opinions or people who are like me.
I think, perhaps the algorithm can be adapted to encourage diversity. Encourage you to connect with people who are passionate about the same thing but maybe unlike you. So one example that’s happening right now, with the new generation of social media is something like Tik-Tok. Tik-Tok actually has something that allows for the discovery of all of the different types of contents and creators and not just a few individual people. So I think that’s also what we’re trying to accomplish with Personal AI for each individual.
First, we want to allow everyone to be able to create their own content which creates this diversity in opinion and that diversity in creation.
NW: Yeah, I think there’s tremendous value in that serendipity of stumbling upon things that we didn’t expect, maybe something new that we hadn’t encountered before or it could be something that we have encountered, but maybe haven’t seen or haven’t thought about in a while. You know, I think about those photos that show up on Facebook, you know, that show you eight or ten years ago, what you were doing. That’s always rewarding, I think, to take stock of how far we’ve come in that time and remember happy memories.
AL: Absolutely and thinking about the good parts of social media always makes me happy but I also know Sharon that you and your co-founders, Kristie and Suman, talk a lot about how personal life AI is intentionally not a social network. So I’d love to hear how you envision humans and Personal AI’s interacting with one another in the future and being social without creating that polarization problem.
SZ: Yeah, that’s a great question. I think, at Personal AI, I would think more along the lines of creating a creator network instead of a social network. So, what I mean by that, is that we believe that every individual is a creator and your memory is what you create throughout your life, right? So Personal AI gives you a way of, not only interacting with yourself, like Nell mentioned, you can look back 10 years past, I really have these surprising or magical moments communicating with your own memories. We want to bring that to the individual level, but on the other hand, we can also make Personal AI accessible to everyone else who may be different or similar to us. So I think, in that way, we want to first democratize creation itself, so we’re building the creator network, where every individual is a creator. We also want to democratize that access. So for example, Nell is creating her personal AI with us and I will be able to subscribe to her opinions and I’ll be able to communicate with her even though she is not necessarily available at all times. So I think that is the interaction and the magic that we want to bring with Personal AI.
AL: Nell, I’d love to hear what your experience was with Personal AI? How was your onboarding and anything that you want to share there?
NW: Sure. I was very impressed with how it picked up elements that I was speaking about and was able to translate my voice into text and make sense of that. But then it started connecting things that I might not have connected myself and my mind. Concepts, which are similar but with different words and kind of put together a greater picture of the things that I was thinking about. Helping me to see things in slightly different ways, which actually brought me to a kind of a new understanding about what was going on in my life or the elements that I had put into it, and that was very interesting. It’s a little bit like in the Victorian Era before we had blogs, people used to share their diaries with very close friends, and they kind of sit around together drinking tea and they swapped the diaries around and it enabled people to have a much deeper understanding because it wasn’t just what somebody said, it was really about those things that are all about your values, your hopes, your dreams. What’s going on in your head? And I think that if we have an opportunity to collect those in a new way and to share those with others, it can enable much greater intimacy with those people that we care greatly about, or that we wish to understand better in new ways.
AL: I’d love to hear your opinion on: How do we build systems that are demonstrably pro-social and also more importantly, perhaps, how should we proactively think about building systems that may become antisocial and sort of erode that intimacy?
NW: I’m very glad that you mentioned the word demonstrable because to me that’s the really tricky and interesting part. We can create rules, standards and regulations. And I spent the last several years working on those kinds of problems, different standards organizations and that’s an important foundation because that’s how we can get machine intelligence, which is less likely to cause us trouble. It’s more likely to be safer to respect our privacy, etc. We need to go further because being truly trustworthy and taking care of the well-being and needs of others- it’s not something that you can put into rules. That’s not something that you can certify with a rulebook and a check mark, it needs to be demonstrable. As you said, it needs to be about taking action to demonstrate that one is pro-social, right? If you have a friend coming over and they have a pet hate about sound, like a noisy bar or something, you know, it drives them crazy. You remember that, you can take account of that and I’m trying to meet their preferences, right? And that’s part of being a pro social person, it’s part of being nice, it’s part of being friendly. Is understanding the norms and preferences of others. I’m trying to model those and trying to meet them. That’s what we need to be teaching AI about, the things that we prefer or the things that are generally welcomed, in a given situation, or given culture. And I actually find Personal AI very interesting because you are collecting a whole bunch of different ideas and experiences and people’s impressions of certain experiences so they can emulate certain characteristics or certain virtues or better understand what would be a preferable way to behave in a certain situation, according to us, right? Not somebody else’s opinion or another culture…I think that’s so important and we’re going to be needing to collect this kind of very subjective information. If we are to teach AI about the human experience and to put the very best of humanity into machines.
In order to stay tuned on more upcoming events and take part in discussion with other members of the Personal AI Community, join our Discord!
Visit Personal AI to create your own AI today!
