Earlier this year, we launched our first podcast series. Our first episode, explored current and future developments in AI technology with four industry-veteran guests:
Momin Mirza is the Senior Manager of Strategy and Partnerships at Verizon, where he focuses on innovations in commerce, AI, and conversational user interfaces.
Esther Crawford is the CEO and co-founder of Olabot, a company working to make it simple for everyone to have their own bot.
Raj Ramaswamy is the CEO and co-founder of ShopInSync, a company that uses deep contextual bots to provide a comprehensive shopping experience across top retailers.
Mounir Shita is the founder and CEO of Kimera Systems. His company’s Artificial General Intelligence (AGI) system named Nigel™ will change the way apps are conceived and digital services are delivered.
Here’s what they had to say…
”The next era of social identity is having more intelligence behind the digital identity that's attached to you. With so much content across so many different platforms, there must be a convergence where all the multiple identities come together, so that who you are is more representative in one unified context.”
- Esther Crawford, Olabot
The discussion began by reflecting on the evolution of social media: anonymous usernames, using your social profile to login to third-party sites, and then escalating quickly to a world where a large portion of your social identity is attached to a mobile device. Some, like Esther Crawford, CEO and co-founder of personal bot company Olabot, believe that social identity is moving in the direction of convergence.
But there can be a utility to having multiple identities, such as maintaining professional, personal, and parental profiles and being able to keep them separate when preferred. Varying avatars and photos spanning multiple platforms often result in a display of very different behavior from the same entity. As younger generations spend more time on social media, this may cause confusion between behaviors of the digital and physical self. The question is should technology strive to converge those identities or should the focus be on making separate identities easier to manage? Senior Manager of Strategy and Partnerships at Verizon, Momin Mirza, compared the ability to maintain separate digital identities to living in a big city and retaining some form of anonymity whereas a convergence of all identities might feel like living in a small town community. For some, that may feel a bit repressive.
“You're going to have to have some sort of a digital surrogate that's on the network somewhere representing you. It thinks like you, has the same political views as you, has the same religion as you, and it can actually represent you online.”
- Mounir Shita, Kimera Systems
The idea of “digital surrogates” was another topic of discussion - a result of device and login overload. Mounir (Mo) Shita, CEO of AGI company Kimera Systems, suggested the idea that we are moving in a direction where we will no longer want or need to operate our devices, and that autonomous digital surrogates would take this responsibility off of our hands by becoming our digital representative. Mo went on to illustrate a world where digital surrogates communicate with other surrogates, services, devices, and anything that’s available on a network.
We’ve seen the idea of turning yourself into a bot in certain startup ideas and pop-culture shows like Black Mirror, but is it possible to have an online version of ourselves that exists posthumously, or even in tandem with our real life persona or offline persona? This might sound like “the dream of science fiction,” or the human unwillingness to accept that we are not immortal, but we have progressed to a place in technology that makes a version of immortality possible.
There are more practical business implications for this kind of technology, as Momin pointed out, such as replacements for employees who are no longer with a company. Instead of emails bouncing back, you could interact with a bot that catalogued that person’s work and could provide useful information after his/her departure. The obvious downside is the implication that employers could then find bots to be sufficient replacements vs. temporary transitional support.
“One of the key things that we constantly think about is how we can replicate more and more of the experiences that feel near human, where it doesn't feel like you're talking to a digital entity.”
- Raj Ramaswamy, ShopInSync
But could digital surrogates have interactions outside of devices, such as with loved ones or sales assistants? Raj Ramaswamy, CEO and co-founder at ShopInSync, a company using AI to improve the shopping experience across retailers, suggested that we’re already getting to a point where digital entities are starting to feel more human. We have enough information and preference data to inform interactions with digital entities, allowing them to feel quite personal and human.Esther provided the analogy of digital extensions, or assistants, being “...like APIs with personality. They are characters that you could have an interaction with, but essentially they're just pulling data from multiple databases. They have a centralized brain and can respond, hopefully in a playful way. I think the best bot experiences right now are those with character that matches with who you are."
Other examples that were discussed included interactions between humans and bots in making dinner plans. A bot can accommodate allergy restrictions and/or provide recommendations for cuisine preferences, based on recent restaurant visits and orders. These kinds of preferences can be easily stored with no need for advanced AI. Mo predicted that in the next few years we'll see AIs starting to understand a person's’ goals, and start to ask the questions that you may not think of, and give the answers that are relevant to change behavior.
“If you go into the future, I don't think having a huge marketing budget is going to give you much advantage in a world of AI.”
- Mounir Shita, Kimera Systems
The topic of bots influencing human behavior brought up a few questions: How can an assistant or a digital surrogate nudge people to make changes? Should they be nudging people, like the way that brands nudge people to buy their products? It all comes down to trust, and not just from the privacy perspective. Mo suggested that we want to trust that our devices can just observe us and learn from their observations. We want to trust that our device knows what we want to share, what we don't want to share. And we want to trust our devices and their suggestions, with the expectation that they will be completely aligned with who you are as a person. But he brought up some huge consequences.
Mo believes that fifty years from now, brand loyalty will be a thing of the past since we will trust our digital surrogates more than brands. Digital surrogates will be able to provide you more trustworthy advice on what to buy than any brand marketing could possibly affect.