Oct. 25, 2023

From Automation to Revolution: Exploring the Future of Conversational AI

This episode of This Anthro Life delves into the transformative impact of conversational AI, hyper automation, and composable software systems with UX pioneer Robb Wilson. It explores how these technologies are reshaping human interactions with machines, going beyond familiar voice assistants like Siri or Alexa. The conversation emphasizes the potential benefits of embracing conversation and simplifying software interfaces, offering insights on how automation can revolutionize work, enhance customer service, and stimulate creativity in the digital age.

Did you know that conversation, one of the oldest human technologies, is reshaping the future of how we interact with machines? We’re not talking Siri or Alexa here, but conversational AI, interfaces anyone can create and hyper automation that links together our intentions, tools and technologies. Join us in this episode of This Anthro Life as we delve into the fascinating world of conversational AI, hyper automation and composable software systems with the brilliant mind of User Experience (UX) pioneer Robb Wilson. Discover why embracing conversation and removing the barriers of complex software interfaces might just be the key to unlocking new opportunities and improved outcomes in the digital age.
In this thought-provoking conversation, we explore the potential shift away from antiquated software interfaces and the exciting possibilities that conversational AI brings. From the potential of no-code approaches and decentralized business models to the interconnectedness of user experience, customer experience, and employee experience, we provide fresh insights on how automation can evolve how we work, interact with technology, elevate customer service, and foster creativity.


Key takeaways:

  • Learn how to create your own skills ecosystem and embrace a no-code approach to building AI interfaces.
  • Understand the potential of conversational UI to revolutionize how we interact with technology.
  • Explore the impact of hyper automation on customer service experiences and employee creativity.
  • Conversational AI and hyper-automation could transform jobs by automating routine tasks and freeing up humans for more creative work.
    Decentralized organizations and composable UIs that allow access to all software through natural language could reshape business models.
  • Total experience design must consider both customer and employee experiences to be effective.
  • Responsible development of conversational AI could help create more empowering technologies.
  • Automation provides the opportunity for humans to engage in more meaningful social interactions.
  • The future of work and business will likely involve a blend of human creativity and machine automation.



Our insightful guest, Robb Wilson, brings a wealth of expertise on hyper automation and composable systems. As the founder of OneReach.ai, a company specializing in creating tools for companies to build their own Alexa-like ecosystems, Robb understands the power and potential of AI in transforming industries. Get ready to be inspired and informed by his insights!


Key Topics of this Podcast:
00:00:23 The future of technology is conversational.
00:06:06 Relaxation and space foster creativity.
00:09:42 Humans can adapt and find value in automation.
00:14:58 Decentralization is the future.
00:22:26 Automation creates space for connection.
00:27:30 Total experience is transformative.
00:30:26 Employee-first approach in business.
00:35:15 Revolutionizing technology through conversational interfaces.
00:41:00 The future of software interfaces.
00:44:36 Explainability is crucial for trust.
00:55:06 Machines should make human-like mistakes.
01:01:11 Unlocking software through conversational UI.
01:06:37 Automation enhances customer experiences.


About This Anthro Life This Anthro Life is a thought-provoking podcast that explores the human side of technology, culture, and business. Hosted by Adam Gamwell, we unravel fascinating narratives and connect them to the wider context of our lives. Tune in to https://thisanthrolife.org and subscribe to our Substack at https://thisanthrolife.substack.com for more captivating episodes and engaging content.


Connect with This Anthro Life:
Instagram: https://www.instagram.com/thisanthrolife/
Facebook: https://www.facebook.com/thisanthrolife
LinkedIn: https://www.linkedin.com/company/this-anthro-life-podcast/
This Anthro Life website: https://www.thisanthrolife.org/
Substack blog: https://thisanthrolife.substack.com

Transcript

Adam:

Welcome back to This Anthro Life, the podcast that explores what it means to be human in today's world, and how we shape technology and how technology ends up shaping us. I'm your host, Adam Gamwell. If you're new to the podcast, great to have you and don't forget to hit that subscribe button on your podcast app or YouTube. Now, in today's episode, we'll dive headfirst into the exciting world of one of the oldest human skills and how it might just revolutionize the next era of technology. What am I talking about? Well, maybe pun intended, the human voice and the power of conversation. Now, we all know how frustrating it can be trying to navigate complex software platforms and spending hours trying to figure out technical aspects, or if you're like me, just trying to get the software to do what you want. So imagine a world where software interfaces are as intuitive as having a conversation with a friend. Where you could seamlessly interact with technology without needing to learn complex mechanisms or navigate through countless menus. Does this sound too good to be true? Well, I mean we've all probably tried to use our different smart apps and Yeah, it sounds too good to be true. But our guest today believes that this future is actually closer than we think. You know, you're probably asking also, what are things like hyper automation? How does it relate to conversational UI? Robb  Wilson is a visionary in the field of design. He created the uber popular UX magazine about 15 years ago. He now has a conversational AI platform and organization called One Reach AI, and has just dropped a new book on the subject called The Age of Invisible Machines. Also has a podcast with the same name. Now Robb's here to challenge the status quo and guide us towards a future where technology seamlessly fits into our lives. And Robb  is no stranger to the complexities of the digital world, as we're going to talk about, but he also understands the importance of making things easy and accessible. Now, when we started talking, he asked me an intriguing question. Why can't we just talk to technology like we talk to one another? And that's when the conversation takes an unexpected turn, as Robb  begins to unpack an idea that flips our perception of traditional software interfaces on their head, or we might say, on their voices. So join us as we unpack this question and explore the possibilities of a world where technology can truly understand and respond to our needs, paving the way for a more seamless and meaningful human-machine interaction. What does this mean for work? What does this mean for where we're going next and how we can interact with devices? Well, you're going to tune in to find out. So let's dive into this transformative conversation with Robb  Wilson on embracing hyper-automation and composable systems in conversational AI. Robb, just wanted to say a huge thanks for joining me on this Anthro Life today. I'm really excited to talk with you. I really, really enjoyed your book and really, you know, I've been following your work for years, you know, through UX Magazine and all the other great work you were doing. So excited to finally get to talk with you and thank you for joining me.


Robb:

Yeah, thanks for having me on. Although, I don't feel like a lot of people are saying that these days. Thanks for having me on. I'm trying to come up with a new Thanks for having me. I'd like to say, oh yeah, I think we're done with that one. We need a new one. We need a new one that sounds, it works really well, which is probably why it's so popular, but there needs to be something new in there. Like, yes, you're welcome.

Adam:

It's good to be seen by you. Yes. A friend of mine would do that. He'd say, good to see you. He's like, yeah, it's good to be seen.

 

Robb:

Okay. Okay. There we go. I can go with that one. Right. Yeah. It's good to be noticed. It's good to be acknowledged. It happens so rarely in my life, especially with my kids.

Adam:

Right. Isn't that what podcasting is for? It's surrogate parenting, I guess.

 

Robb:

Yeah, I guess so, yeah.

Adam:

Once they hit 8, 10, or 12, they're like, all right, we're done with you, dad.

Robb:

Yeah, go talk to other people who care about what you have to say.

Adam:

I'll be back when I'm 18. You're like, okay, cool, cool. I mean, that's a legit point, though, I think. And how we can think about the question of I don't know, raising the next generation. I think it may be an important theme that we could dive across here as well.

Robb:

Which ones, AIs or humans?

 

Adam: 

 

Well, that's the question, yeah. And I guess one of the questions is, which one should we start with? Because we're going to have to do both. And I think they're going to get raised together, though.

 

Robb: 

 

I mean, that's one of the- Ideally, yeah. Yeah, they do it as a team. Yeah. One clearly more important than the other, meaning the human, not for anyone who Who might've wondered what I was meaning there.

 

Adam:

Where is this going? Actually, with that too, it's something I'd love to pick your brain about. To visit the Humans Cafe in Ukraine. And for me, what kind of stood out when I was reading about it and learning about it from your work is that, one, it stayed profitable in deeply volatile times, which is important, but then What it says to me, as I'm an anthropologist, and kind of what it says to me and got me thinking about is that its existence and success also kind of reveal something else about this peculiar moment that we are in history right now. I mean, there is the ongoing war, But beyond that too, and as our relationship with technology is changing, and something that is a theme that I have seen throughout your work, and then I want to kind of use the cafe to open this up here, is that we're not only looking at technology changing, but really like the social structures and business models that we have today that could define a better tomorrow have to be built out of the the kind of existing structures that we have. And there's an interesting tension throughout your work about that. So I'd love to talk about the cafe as a starting point and tell me about the genesis of this idea and like, what might this tell us about our point in history today?

 

Robb: 

 

So again, anytime you're like a scientist, you want to experiment, you need a place to experiment and run your experiments because you really want, you understand that truth matters, right? It has to actually work. if you're going to, you know, go out in the world, cause people are going to test it and they're going to come back and they're going to tell you. So we needed a place to test this. And I think in some ways you could say that sort of corrupted the process a little bit because we didn't need the cafe to be successful. We weren't white knuckling this business, right? Like, oh my God, we better make money or I can't pay my mortgage. And I think that loose grip is part of its success, that that ability to be a little more relaxed and that let us experiment, right? That, that gave us the space to experiment. And I think we all know this from sports and other endeavors that when, when you're sort of too uptight and too wound up, you, you need to relax, right? And, and I think that number one, we learned that, wow, there's a, there's a really profound effect and positive effect that can happen by just relaxing and business, just not, not being so uptight. And, and allowing us to experiment a little bit and try some things out. So I'd say like that has nothing to do with AI and everything to do with just the human condition of fear and saying, okay, well, what happens when you make room for us to be creative? Well, big surprise, we do creative things.

 

Adam: 

 

Turns out, spoiler alert.

 

Robb: 

 

Turns out. Yeah, that space is what you need to get inspired, not necessarily the pursuit of something. with, with vigor, but the space. And, and I think it's an interesting idea that's to say like, yes, there are certain achievements that come from being like highly targeted at something and highly focused and aggressive and pursuing it in a, in a razor focused way. And then there's probably other things that, uh, you achieve by giving yourself space and room. to think and be creative. And I think it really depends if it's a formula that you're trying to execute that you have a good sense is going to work. So you're just rinse and repeat. Let's do that as fast as we can or let's make a formula. I think you need space. And companies I think often mistake the place they are in given different projects as always being in this exploit, like go fast, produce and miss the moments where it's like, no, no, no, this is creative. This is where we need to create space so that we can invent things. But that space, the good news is that space typically gets used for productive, at least in terms of the coffee shop, that space got used for productive. thinking and innovative thinking by average normal people. So this idea that once you lose your job or an AI does it and you have space to find other ways to be useful in the community, there's hope that you will come up with something innovative and cool. You won't just actually be lost in a life of uselessness because you can't come up with something better to do than punch a cash register. Yeah.

 

Adam:

 

Yeah. I think it's an important point. I mean, because so often since the Industrial Revolution, but really, I mean, since the 1950s and 60s, we've really defined ourselves through our jobs, right? Like we kind of say, I'm a miner. I'm a cafe employee. researcher, designer. And while that does play an important part of our lives, I think you raise an important point that when the kind of work that we're able to do or what do we have access to, it changes, especially if we have this kind of relaxed mindset about it versus I'm kind of in the business model of exploiting and or I'm going to lose my job slash my identity. Uh, we find that there is that space that people kind of do kind of lean into their creativity or do do more human things, I guess. And this is an interesting thread throughout a lot of your work too, right? As we're, as we kind of think about what automation might do, you know, so in case, in case folks, um, are not playing with the cafe too, right? It's, it's, it's largely automated, right? So we have, we've on one level, it's, it's more horizontal in terms of authority structure, right? There's not, there's not like a bunch of bosses and managers running around, but that a lot of the backend tools are taken care of. more mundane tasks, I guess, kind of bookkeeping pieces. And so it's interesting to know that when some of those are kind of out of the way, then people find other things to do. And I think that that's interesting to note too, that when we have some tasks out of the way, then other things kind of pick up, like the idea of more service-oriented activities with other humans. I think that that's powerful as a reminder that as we think about like the future seems a bit scary I think for some in terms of as we take our jobs right but I think you're kind of positing that when we actually provide space and think through it we don't see either that fear right there and or yeah we're afraid of that space we're afraid that we're not going to fill that space with something valuable, right? 

 

Robb: 

 

That we're not going to come up with something useful to do in society. We're afraid that, you know, maybe it's a good example. You, you know, your, your girlfriend breaks up with you or your boyfriend and you feel like you're never going to get another one. Right? Like we, we do fear that space. And, and some people fear like, oh my God, we have no plans for this weekend. Oh my God. And then there's people who are like, great. We have no plans for this weekend. I'm not worried. We will come up with something to do. And I think in, in life, those same people are afraid. Well, if I can't, if I can't copy from one spreadsheet to another spreadsheet for a living, what will I do with my time? You will come up with something. And I guess some won't, some will just be lost forever. Maybe they'll have garage sales every day. I don't know. Yeah. At least you're interacting with people. Yeah. But yeah, the coffee shop, what's, what was most interesting about humans, you know, aside from the fact that we had never anticipated it going on through wartime and, and all of that, was that The employees decide collectively, they meet every week and they decide collectively what structure they, they want to live within and they want to work within. And it's amazing as a group how they are, you know, pretty responsible. I think everybody wants to be the exception to the rule, but then when it comes to making the rules as a group, they tend to make good decisions, generally speaking. And, and that's, turns out how it is. And I think if you, if you mathematically look at companies, you realize that there's sort of algorithms that get executed at scale if they're really successful and consistently. And so those algorithms that prove to work, that sort of hack our brain and give us the serotonin and the. You know, the, the positive juice in our brain because we eat a hamburger or whatever it is that, you know, we pay for, we pay for those hits. Right. And if those can be delivered at scale anywhere in the world consistently, then you have an enterprise on your hands, right? You have the potential for a really big business. And so then your goal is to execute that algorithm, right? Flawlessly. I always, I heard this story. My wife told me she was on a plane with a guy. Uh, that worked for McDonald's. He was the bun guy, the guy who, yeah, he's a bun guy. He has a briefcase full of buns and he go, he was going into Eastern Europe at the time. And he, he goes to the bun factory and then he tastes the bun in the bun factory. And then he tastes the bun that he brought with him to make sure it's consistent. Right. So it's his like, it's like standard bun. And then what made him really good at his job, according to him was He could tell you what you did wrong. After he tasted like, ah, you need to do this, you need to do that, and he could get that consistency. So you realize like there's an algorithm consistency with that algorithm globally. And then you have humans that's job it is to, to make sure that you adhere to that algorithm in every McDonald's across the planet. And why humans are good at this is because they can adapt to slight changes and, and they can adapt the algorithm just ever so slightly. in case someone doesn't show up for work or some surprise, right? Cause you don't have this algorithm with every condition or every exception considered. So humans do that like error correction on the spot cause they're pretty good at that. And they're like, how do we get this formula back in play? And so you, so you go, okay, well that's like this top-down approach, right? Very centralized versus decentralized, top-down, run this algorithm and we know it works. And this is like this great hacking technique. And humans is a bottom up, right? It's, it's the, the employees that work there deciding what the algorithm should be, right? And designing it and then asking Hugh, the bot to execute it, not a person, right? And then giving it its margin of error. And because they designed it when little errors come up or little exceptions come up that the system isn't accounting for, they can account for it. They don't need. Right. a manager that accounts for it. And they tend to create the world. So every week they meet, update those algorithms based on those changes and shifts. And so it's, it's, it's more than just AI helping. It's this idea of, you know, more of a decentralized approach to running a business. And can that work at scale? Could they replicate that algorithm to other coffee shops? Absolutely. Why not? Would they be under the same brand? I don't know. I don't know that it matters that they do. So humans is willing, like, you know, anybody can. can use those algorithms, anyone can take those formulas and apply them. There isn't a like, oh, you know, it's not secret. It's not a secret recipe. We're sort of open to sharing it with other coffee shops, which others have showed interest in, in Kiev, but. That's cool.

 

Adam: 

 

Yeah. I mean, I think that what, what really strikes me about that example is, you know, I've, I've And other folks I've talked with too, I think as we think about what the future of business and role of technology in our lives is, or could be, can be, there is this interesting tension that on the one hand, we are moving on some areas like this. I mean, this is a great example of how we can move towards decentralization, right? In terms of like autonomous organizations and DAOs or using cryptocurrency. So on the one hand, humans are really good at working at village-level problems like this. We can have whatever system it is to then keep the bottom line books in different areas of you know, maintenance flow in place, but we can kind of solve problems together as people collectively. Generally, that's better for the group. I think you said that really well, and that's an important point. But then there's the tension of we, that's, you know, we might say it's the web three model, right? But we're living in a web two world right now. And like, we're kind of building our way sort of towards, towards web three, you know, crypto and decentralized orgs had a nice boost in 2020, 2021. And then we had both crypto and stock winter for a while. But it's an interesting question in terms of those jets haven't actually cooled though, right? I think a lot of public attention turned away from some of that thinking. But what I thought was really interesting is as I was going through your book, Age of Invisible Machines, and kind of going through some of the tactics of what it would take to kind of think about this, what does hyper-automation actually look like in an organization? You actually surprised me because in the end, then you pointed, I think the key piece that like, actually, if we're going to do this, it's, it's going to be based on functionality, not on brand. And then on top of that, the only way to do it is if the systems are actually open, like you're saying with human here. And so I think that's a really interesting and, and I hate to say this legitimately disruptive idea, you know, just in terms of Silicon Valley speak, right. But like, but it's, that's actually is a disruptive model in terms of like how Adobe might work, right. In terms of they want to have proprietary software that you have to go through a Salesforce. And so let's, I want to dig a little bit into this, this idea in your thought too, in terms of like, I know we're like, we jumped in the beginning and now the end in terms of where we are at the book, we'll get, we'll get in the middle, I promise. But, but this, I think is a really important piece too, of like, as we, as we approach this, something else that you said that, is that the status quo right now is death. We can't keep doing the exact same models that we're operating on in business and technology. So maybe if I leave it there, how do we think about this idea of what we're doing right now is a death sentence if we don't change things? And then how might decentralization, as we're seeing in the humans model, and also as you're thinking about software, help us think about what can come next?

 

Robb: 

 

Yeah, the whole death sentence thing. I mean, I know that's got a lot of fear baked stuff in it, but I think anytime you see a certain model that is clearly superior to another one, you know, you're, you're going to have those winner losers scenarios. And I think when you look at a lot of these things. through the eyes of the evolution of technology, we've been on an evolution to decentralization, more complex algorithms. People do want more freedom in their jobs. They don't want to execute these strict algorithms every day. They're boring. They do have choices. There's three open positions now for every one person in the United States, despite all the, you know, The US sucks and blah, blah, blah. There's still three open jobs for every person. And so I think now what we see, which makes sense when you have this surplus, people are, are deciding like, not just, I need a job, what kind of job do I want? You know, which is great. That's, that's good. What kind of job would fulfill me? And, and I do think that once we kind of enter the world of decentralization, We enter a bigger world of variety and diversity so that the restaurants we go to aren't all the same across the world. And, and why, why did that work for a while? Because, because it wasn't, you couldn't count on being sure that you were going to get something that was, that was good enough. Like you couldn't trust. that good enough, you couldn't assume that good enough was available. Um, cause we didn't, you know, we didn't have a lot of the tools we have now with reputation and, you know, uh, public rating systems and things like that. So you didn't know that, that that restaurant didn't suck. So we go to the restaurant, the trust, you know, the familiar, may not be the best, but it's familiar. We know what we're going to get. And now with a lot of these tools, this feedback loop and this trust on, you know, a single restaurant that, isn't part of a chain, we can trust it. We have tools and ways to actually rely on the reputation of that place. And so I think we're one-upping the idea that trust is king and consistency is king. And we're starting to say, no, trust and consistency now are you know, the baseline, they're just, you know, table stakes and it's time to now differentiate above and beyond that. And so how do you do that, right? It's going to be, you need to use these tools. You need to use technology to up the game. Customer service, I think is where we're really going to see this is we're very much in checkbox, very much in, you know, at that place of consistency and trust and, and not at the place of exceptionalism. So much room to grow. This idea that we're going to automate customer service and, and companies are just going to automate the way they do it today seems to make no sense to me. There's, each company has so much room for improvement. Um, and that's where they're going to go because that's where they're going to differentiate. And so these tools say, all right, you make room for people to be creative and think of ways to make the customer experience better. They're going to do that. Humans, I go back to humans. One of the things we noticed, these subtle little things, is that because people were ordering on Hue, you might think, oh, well then they're not interacting or talking to the barista. Right? Not true. What ended up happening is they were ordering on Hue and so when they went up to get their coffee, they had a chat with the barista, but it was more about like, where did these beans come from? Instead of, do you want milk or oat milk? Or, you know, what kind of milks do you have? Like none of that. It wasn't this boring routine conversation for the barista or the person ordering coffee. That was taken care of by Hugh. So that left room for them to talk and get to know each other better. And so I, we, you know, there's no, they would be tough to test this, but I would say that the folks that work there know the customers better as a result. You know, there's more of a connection now being made with the customers and the people who work there, the people who work there enjoy their jobs more because the conversations are more creative. and more useful, so that space was created through automation. The next coffee shop that's still at the point of, do you want oat milk or almond milk or blah, blah, blah, they're going to have a hard time competing.

 

Adam: 

 

Yeah. Actually, I mean, you've got me thinking there too. I bet we could design us, I mean, it'd be a survey, so it won't be like super, super the exact percentages, but I think that'd be really interesting to get, you know, questions in terms of like, how well do I feel that I know my community at this place? Or like, what are the topics that I discussed with my barista this week?

 

Robb: 

 

That'd be awesome. And it is alluring. Like, it's a challenge. I think you could pull it off, like you said, within a degree of reasonable accuracy.

 

Adam: 

 

Yeah. Yeah. But I think that's right on though. I think, I think that's, that's a really important point though, of in terms of what the kinds of interactions that can take place when we give them space to do so. And they would take place anyway if I was not necessarily maybe in line at a coffee shop and I had to then accomplish task one, which is order coffee and pick the milk that I need. When I have that taken care of, then I actually can say, hey, that's a cool shirt. Oh, thanks. I just got it down the street at this shop. 

 

Robb:

Oh, cool. Yeah. Or whatever it is. And those baristas weren't trained on here's the top 10 questions to start a conversation with a customer. Yeah. Like they don't need that, you know, right. Yeah. Like how to be a nice person. Right. How to be interesting.

 

Adam: 

 

Yeah. That's, I mean, hopefully that that's, that's like the human technology, you know, but, uh, but I think it's right. You know, the little conversation cards, you know, it's like, here, how to have a, have a conversation starter at each table. Um, yeah. Ask them about their favorite color or something.

 

Robb: 

Yeah. Yeah. Like it just happens. You, you give them the space and it happens, you know, if a barista is not overwhelmed running around because they forgot to order whatever sugar oat milk. You know, because he was taking care of all that, you know, he's on top of it. They're not doing all these mundane. Now they're just there to engage and to have a conversation and co-create in essence with their, with their customers. And it's what they're focused on and they've been given that space. So yeah, it's, we're not measuring them on how many cups of coffee did you make? It's, it's an engagement that would be fun to try to figure out how we can measure that.

 

Adam: 

 

All right. My, my, my spidey sense is tingling now. I gotta, I gotta think about this, but I like this idea because it's, you know, because also we have, you know, in, in our end sample size of two, I'm not, I'm not sure of existing surveys that, that test for that or check for that. So that could be a really compelling new area to explore. Yeah.

 

Robb: 

 

I mean, honestly, it could be transformative because we, we do tend to obsess over what we can measure, whether or not it matters. Right. And that's huge, huge part of why we go off the rails so much in business, because we focus so much on what's quantifiable. So maybe just meeting them there instead of asking them to not quantify some things and trust, we go, let's, let's give them some numbers so they can move off of their, how many cups of coffee did you make today?

 

Adam: 

 

Yeah, yeah. It's not a bad idea. You get sugar packets to someone who didn't need them. Actually, I did. Yes. Bonus for the week then, you know. But I think also like the other piece you said there that's I think is key is that something that I appreciated that you talked about in the book is that, you know, CX or customer experience or, you know, customer service is one of the areas that we have a huge amount to stand to gain as we think about what can automation do in terms of what is it, what might it free us up to do? And an area that this is a term that I learned from reading your book is just talking about the total experience or bringing together customer experience, employee experience, and all these other pieces. Because so often we see in organizations, especially I do organizational anthropology sometimes too and work with organizations to see what's their structure and their culture like. And so often you see the customer experience and the employee experience are disaggregated, right? They don't really talk to each other. And that, I mean, causes a host of problems anyway. But then when we begin to introduce something like automation or hyper automation into a system or an organization, you make the case here that we have to actually think as a kind of a total experience, right? And so as a systems thinker, like that made me happy also as how we actually have to think about the interconnected pieces there. So I'd love to kind of get your thoughts about this and how can we kind of help you know, user experience in general, you know, and I don't know if UX counts as the, as the overall branch anymore, if we're talking about, you know, customer, consumer and employee experience, but just the experience bucket, you know, total experience. How do we think about bringing those pieces together, especially for organizations that are a little bit more resistance or tell you that they don't need to be connected?

 

Robb: 

 

Yeah. Well, two things. So total experience I borrowed from Gartner, I think in a lot of ways, those guys really think ahead. Um, I think people don't give them the credit they deserve, to be honest. I think those guys, there's a lot of good thinking that goes on there and that's one of them. And it's this idea that, you know, there's a holistic experience for employees, customers alike. Why does there, you know, why would there be a interface for customers and an interface for employees? There just should be one interface and then the data should be, it should just be decided based on who you are, what data is available to you. So in conversational interface, that's easy. You know, it's, I'm sorry, like, you know, if you want to know our secret passwords to our vault at our bank, that's, I can't give you that as a customer, but the interface can all be the same. There's no reason that that should be separate interfaces. And that's what conversation offers is the unified, holistic, composable UI. which is a pretty cool idea. When we put these things together, like decentralize, like why, you know, decentralize is a cool concept to be like, let's go decentralize, but to what end? Like, what does this, what does it really offer? And I think one of the key benefits, um, is that it, it's a great tool along with other things to, to kind of rethink, as you said, the business model. So there's a groundswell that's mostly at lip service stage where people are talking about it, but not doing it. So it's that phase where they say they're doing it, but nobody's actually doing it. And I guess there's always a phase you can get away with that. I don't know why, but we, in the early stages, you can say you're doing it and not do it. So a lot of CEOs, Bezos, et cetera, they talk about this idea of instead of shareholder first, customer than employee, right? Which is, comes from the sixties and this study that was done back then that said when you put shareholder value at the front, companies are more profitable, right? So everyone's like, Oh my God, I love it. Let's put shareholder value at the front of the stack, make it the most important thing. The problem with the study was It also creates the most short-term lifespans of companies and generates the most short-term thinking. So it didn't study it over time. And so it was a very flawed study in itself, but it was so appealing to shareholders who had the power that, you know, it was kind of like, you had me at shareholders first. I didn't even read that. Um, so they jumped on, on, on this idea. And then I think we started going into like, oh, customer first, shareholder second, employee last. And now there's this, you know, cry at groundswell going on about employee first, customer second, shareholder last. Now, of course, we're not seeing the kind of adoption and excitement from shareholders about this that we're used to seeing from, from what we saw in the sixties. But, but it's amazing if you actually look up who signed off on this and who's talking about it, you know, the, the big CEO names that are kind of paying lip service to this idea. I think they all even made a commitment towards it, but then like, wow, trying to shift a large organization like Amazon to our side. I guess all you can really do is pay lip service for a while. And this is the idea that happy customers, happy employees equal happy customers. Happy customers equal happy shareholders. Long-term, short-term. It really is logical. It really does make sense. And, and I, and I don't know that, that, you know, holistically we've ever tried it like across the board. I think there's companies that have tried it and things like that, but I don't think there's a point in history where you can say, you know, we put employees first. You know, maybe unions might be an argument that unions did it, but, but that was more of a conflict, right? It was, it was, that was more of a fight that the business leaders were still putting the shareholders first. And it was, it was more of a combating than it was like alignment. So I, so when it's aligned, when management is aligned, you don't need a union, you know? Yeah. And so this idea of employee first could come from decentralizing because when you have centralization at scale, you also have corruption. Right. You have, you have this central source and that's great for pivoting, you know, like, you know, it's true that when you have governments that are dictatorships, they can move fast, but there's, they're also so susceptible to corruption because there's a central place to corrupt. And it's not just the central idea. It's also a target. Right. Everybody looks now you have a place that everybody's looking to try to corrupt because it's like a goldmine. Right. So you're like, oh, it's detention. Like, how do I corrupt? So you have all these raiders that are like, that's you know, I'm not going to I'm not going to rob the poorest house on the street. I'm going to. So I'm going to go to this, if I can get into the bank vault, right, that's a centralized idea. And so corrupt, it attracts corruption, centralization. Um, decentralization is, is maybe the best way that we can get an employee first approach because, because it's less prone to attack and therefore we can kind of let employees administer their own just like humans is administer their own rules and, and know that if they make a mistake, it's not going to affect your other thousand locations, right? It's just, just them and, and they can learn. So anyways, that was a long, a long answer to your question, but.

Adam: 

 

No, I think it's really interesting though, as we think about that, because we are finding ourselves, because technology is getting us to this point now where we are seeing obviously huge changes today and on the horizon, right? Obviously, you know, this sounds glib to even say at this point, but it's like, okay, so chatGBT changed the consumer mindset or is beginning to allow what generative AI can do, right? And that like, obviously AI has been around for a long time, but this is kind of the first time there's been a mass scale consumer facing product that now has spawned a bajillion clones and weirdo knockoffs, you know, helping you write your essay and your resume and stuff. What I'm curious about too is that what's really interesting in your work through OneReach and also what you write about in Age of Invisible Machines is this process of hyper-automation of actually the conversational platform. One of the reasons that like ChatGPT took off is because they just slapped a conversational interface, the text-based one, on top of a large language model. And there's something that's so natural about conversation to us that we immediately feel comfortable doing that. What's interesting is that like, you know, I've, I've used an Alexa for years, you know, the Amazon product, there's, there's Siri and there's, uh, you know, Google's thing. And they've always seemed relatively unintuitive because you don't know what, there's no affordances. You just, you see a button, but I don't know what I can do with it. Right. And it has to literally tell me sometimes. By the way, I can tell you a joke. Okay, great. Tell me a joke. I think something really interesting about the capacity for conversational interfaces to revolutionize what we do with technology, but then on top of that, on one level it makes us feel comfortable, but then kind of what you're positing, and I want to kind of dig into this too, is how that also goes hand in hand with this idea of hyperautomation, that we can actually string together different functions, different technologies. So I guess my, to ask a question there, what is hyperautomation? Like, how can we think about this concept and how does that play with the idea of a conversational UI?

 

Robb: 

 

So I think, so think of hyperautomation, again, another Gartner term as a state of being, it's just automating right and really fast, an oversimplification, right? And that's relative to how we were automating. And I would say that one of the key ways to get there is to actually use the software that we bought. And that's available to us. So the amount of software that we buy and don't use is staggering. Oh, yes. It's embarrassing. Yeah. Yeah. Like unbelievable amounts. And there's been many attempts to create a single UI like SharePoint and, you know, Salesforce, like the list goes on and on of companies that said, well, if we could just make one UI for all your software, Man, imagine how powerful that would be. And then you have like, you know, 2000 tabs designed by 1000 different people in an interface. It doesn't take long to realize that graphical interfaces don't scale. Simply, they don't scale across a lot of features. And we see this because they're, you know, software companies get into feature wars, they add too many features, and then someone comes and disrupts them by stripping out You know, 80% of the features, a lot of times it's Apple. And all of a sudden they take off like a rocket and everyone's standing there that just spent a decade building feature after feature after feature going, what just happened? Like, were the features not useful? Like, yeah, but the interface made it so challenging to learn. Cause it's a whole new language. And, and so, you know, the learning curve was too high. And so people weren't using it. And if we could interface with machines in the language that we use to interface with each other and do this hybrid of graphical, take the good from the graphical UIs, like charts and graphs and stuff like that, and mix them up with conversation like we do when we use PowerPoint and stuff like that. And take advantage of the fact that machines can conjure up drop down menus and radio buttons in a second, and we can't. So they can, you know, make conversation even faster and smoother. Then we end up with the new language and now we end up with a scalable UI, the concept for the first time ever in history that we could have one UI for all of our software. And it could be all of our software globally, not just the software you bought, all the software available in the world. And how could that be? Well, imagine if every, if every API could talk like humans and it could talk to ours and, and they could have conversations with each other without integrations having to be done. And every company that had a useful piece of software put a conversational UI on it instead of an API. Now your conversational assistant or whatever you want to call it, your Siri, could talk to it and get things done. And now you could have access to all the software in the world. So what we're talking about isn't AI being this hugely powerful thing. It's saying that all the software in the world is already hugely powerful. It's the fact that we can't access it. That's the problem. What if AI just lets us access the software we already have, we've already bought, because it gives us simply a UI that is understandable in a way that we can assume it. Hey, I want to send out an email, right? It's a great example. To who? To my customers. What do you want to say? Boom. I don't need to go to MailChimp and learn how to get an account and run like, why do I have to do all of this? And then you take that up a level and you go, wait, wait, why do I need to know it's email? Why? Why do I need to say the mechanism in which I need to do this? Like, I don't say over the phone, I'd like to make a phone, no, make it a WebRTC call. No, I think I want to use a SIP call. Like, no, I just talk to someone. I just want to call someone. So maybe it's just send my customers a message and you figure out how to get it to them. I don't care. I don't care if you use SMS or WhatsApp or whatever, just whatever they read, um, letting them know this and. And so I think if you really take a look at it, it's kind of like looking at a gas powered car objectively compared to like an electric motor. You're like, this is a ginormous hack, a gas powered car. We have so many things to like, oh, and then it overheats. So we have this oil that needs to run through and then we have, oh, and then we need a radiator that also cycles and you've got to be moving and get air through the radiator. And there's these little explosions and, and it's a missile in the front. So we got to like, make sure that like, it's just this like massive, massive hack that we've, you know, ingeniously put together to be reliable, almost amazingly. And then you look at like an electric motor and you're like, it just goes round. There's no transmission. There's no radiator. There's. And I think it's the same idea. We're going to look back at traditional software and those interfaces and go, wow, how did we ever, like that's, they're so antiquated. How did we get anything done? Foglum sticking between all these apps.

 

Adam: 

 

I mean, yeah, that's a good point because it is this, I had not thought about it like that between kind of a gas and electric engine, but you're right. Like that's a great, that's a great illustration of. what's possible, and we realize we can make a car go more elegantly and easier with fewer moving parts. And conversation kind of feels that way too, right? There's less explosions, hopefully, and missiles firing here and there, and oil to lubricate. Otherwise, you're going to dry up. I think that that's, that's a really cool, it's a good way to think about that too. And then on top of that too, it's that the, what's really compelling about this too, is that it's not just, I can say, send, send a message to my customers and then machine, I guess the machine then can figure that out. It then has access to, if some customers said, I prefer to get an SMS or WhatsApp, others ones say, email me, other ones say, I don't know, send me a video. And that they can, it could then kind of route in and do that as well. I think it's such an interesting proposition. And, and so, I mean, I guess, does that mean like we're facing the death of APIs at some point, which was kind of like one of our big pieces in terms of like, oh, we can now make software talk to each other. That was the first key, but then it's kind of like, that's not that good start, but we're not, we're not done yet.

 

Robb: 

 

You'd be shocked at how many people disagree with me on the death of APIs. Shocked.

 

Adam: 

 

Cause they put a lot of hours into it probably.

 

Robb: 

 

I guess I, like, it seems so obvious to me that if a machine can talk to us in natural language, why wouldn't it talk to other machines? And why wouldn't we want machines to talk to each other in a way we could understand what the hell they're saying, just in case we don't agree with it, just so we can oversee what the hell they're talking about. Um, I don't want it to be this, you know, cryptic kind of language of JSON and No, I don't care if it takes a little longer, like 100 milliseconds longer. But if I could see that a conversation just took place between you know, my healthcare provider and, you know, maybe my medical supply company that provides me with whatever it is, dialysis machines, or, or I, I just want to be able to, at the end of the day, browse my newsfeed and see that they had a conversation and that they're scheduling a delivery. I'm just. I just want to know. I want to know what's being talked about. I want them to do it without me. I don't want to be involved in the conversation, but I want to be aware and I want to see what they said to each other. And I want to be able to go in and say, wait, wait, wait, wait. I'm not going to be here. I actually just decided I'm going to visit, you know, my son or daughter or mom or dad. So can you reschedule that to another day? So I don't see, I think it's inevitable. I don't see how anybody could argue that we want our machines talking to each other in a way that we can't easily understand what they're saying. I think disclosure is key to where we're all going with technology to trust it. So I think every company is going to be compelled to have a conversational UI on top of their software. That's something I put a a lot of my time into creating a platform that allows you to do. So I really hope they do. And I think there'll be a tipping point where even if you don't agree with me, you're going to have to do it because enough other systems Rely on it in this way and expect that you're going to be left out of the conversation, literally. 

 

Adam: 

 

But I think that's an important point too, because as technology at least appears to get more and more kind of embedded in our everyday lives in ways that are becoming black boxes or already are black boxes to most of us. Right. I think that's an important point that, you know, explainability is really important when it comes to, to people feeling that they can trust the system. Cause I want to know, yeah, I don't, I agree. I don't want to, I don't want to have to like make it do what it's doing, but I want to be able to say, okay, why'd you just run that query? Or like, I asked you to do this, but I noticed that you did that. Like why, what, what happens? And they can say, oh, sorry, I thought you meant this. Right. And be able to talk about interpretability like that, I think is really important.

 

Robb: 

 

Yeah. I mean, just imagine when two systems talk to each other, we get into a new world of, it's like, Hey, I want to send John an offer. Is that okay? John says, yeah, yeah. Send it across. I'll take a look at it. Send it across. Like John's not going to be interested in this. I'm not even going to share it with him. Yeah. Oh, okay. Or, okay. I pass it on to them. Right. And why wouldn't it? And now you want to see that. What didn't it pass on to you? Like, you know, just, just a summary. Hey, you got like 20 emails. I didn't bother you about here's a summary of what they were. You're like, oh, wait, there was, was there any from blah, blah, blah. Yeah. Oh, I need that one. Great. There's my new spam filter, right? It's, it's not like I got to go look at all of them now. Like, oh, we've put them all into a bucket so you don't have to look at them. Well, how do I know if something in there wasn't important? Oh, go look at them all. Yeah. You're like, wait, I just, well, you just deferred it. You didn't actually.

 

Adam: 

 

That's a, that's a good point too, actually. Yes. A lot of the, a lot of the, I guess, tools that are supposed to save us time today actually just defer the action to, to some other point in time. Yeah.

 

Robb: 

 

More convenient time when you go through all your junk mail.

 

Adam:

 

Oh, great. Which is somehow still every day. So it's like, yeah.

 

Robb: 

Yeah. Which means never. So it just means that important messages sometimes get by me and people get angry. Yeah. Unfortunately.

 

Adam:

 

Yeah. But which, which happens, you know, you know, but that's, that's the funny thing too. I mean, maybe, maybe that can get, that can help us, you know, yeah. I don't know, figure the AI can, the, uh, the email conundrum, I guess, you know, cause that is one of the.

 

Robb: 

Easy to solve with generative, like easy, easy, like, you know, super easy now. Now, now we, the idea of what is going to be interesting and important to us could be written in a paragraph. You could write a paragraph and I've, I've done this. It's took me like. 20 minutes, wrote a paragraph of the kinds of emails I want, kind of emails I don't want, use that as a primer to the LMS, I mean the LLMs. And then, and then you feed it the content of the email and you see if it, and then you ask it to classify it as the kind you want, the kind to just mention and the kind to just ignore. And it does a great job. Now, when it didn't, you just have to tune the description a little bit as to like, oh, it, you know, I missed something. It's usually that you missed giving it an instruction. Like, oh, if it's from my mom, like definitely got to get that across right away or she'll be mad. And maybe I needed a new category for urgent, whatever, like, but as you tune it, It's amazing what it can do with a description, but now we realize each one of us can create our own descriptions. We don't have this like universal algorithm, like the spam filter that everyone can use with the same algorithm because everyone wants the same email. Right. And so we get into this simple world where you just wrote your own spam filter in 10, 20 minutes. That sounds sweet. I would like that. Yeah. And it's shocking how well it works.

 

Adam: 

 

I mean, so this is what you're working on in one of each kind of behind the scenes. And I'm excited to think about this because like the… the, the platform. So as you're thinking about this, like how, how is the, are you kind of rolling it out now? I know it's, it's not, it's, I don't know. I was reading, it's like sort of stealthy, but not really. I mean, obviously you're talking about a podcast, but like how, how is that going in that process and how are you thinking about.

 

Robb: 

 

Yeah. I'd say we're like right at the, right at the transition of stealthy to not stealthy, you know? Yeah. The, so first of all, like, you know, generative AI, I wrote the book before you know, chat GPT and, you know, we're like, Oh my God, it's going to change everything. And everyone thought we were just nuts. And now they're like, Oh, you, you weren't, you understated it. And you're like, yeah, well, we were getting tired of the high rolls. So we started toning it down a little bit. But now it's like, wow, it's everywhere. I knew also, and I still believe that these tools aren't as complicated as people think they are. And, and, they're already on the road to being commoditized. The best way to think about LLMs in my opinion, and I think a lot of folks will agree with this, that are deep into it and have been doing it a long time, your LLM is a product of what you put into it. If you don't want it to be racist, don't feed it Reddit data, just feed it Wikipedia data. It's not really complicated. It's kind of like if you don't want arsenic in your muffins, Make sure that the ingredients you use don't contain arsenic. You're in a first year kids. Now it's funny how little people pay attention to sampling the data before you put it in. Right. Like we do with our foods and our ingredients, the ingredients as well. How do you like that word? That's a good word. Yeah. Ingredients is, uh, we're going to, we're going to check for arsenic. Like these are things you have to do. You have to check the ingredients. If we can check the ingredients of what we put in to these LLMs, then we have a lot more visibility into what comes out of them. And just because. OpenAI collected a lot of data from a lot of different places that contained a lot of random things. The way I think about it is imagine you went through the kitchen and you put every possible liquid powder, anything you could find into the mixing bowl, including the bleach under the sink. Like if it was a liquid, you put it in there and you mixed it up and you're like, I wonder what will happen. Right. And someone tasted it and they're like, this is actually pretty good. And they're like, yeah, and then they get sick afterwards. You're like, all right. Like, how do we remove some of the ingredients? These, like, can we neutralize the bleach? Is there any way to, you're like, well, just don't put the bleach in in the first place. Like no one, no, I forget. That's not going to work. How do we neutralize this? So we're like using this reinforced learning where humans are going through and trying to like bias the model against it. Um, the question is, can we get enough data that doesn't contain racism and other bad things? Can we test, can we find a way to test the data before we put it in? Yes. The answer is yes. Could we possibly remove those things? If we can identify those things, could we possibly remove those things before we put it into the model? Yes. So what we're going to end up with is realizing that there's lots of sources of data that people will provide to the world. Facebook just released Lambda 2, right? Like, and, And now like a chef, different people are going to use different amounts and different mixtures of different data. And they're going to come out with LLMs that output. things that are interesting, just like baking, right? So LLMs are going to be like, like meals, like there's going to be all kinds of flavors of them that do different things depending on different combinations of data. Some are going to surprise us like peanut butter and chocolate. Never thought that was going to be so amazing. And they're going to mix, you know, Wikipedia with, you know, medical data and maybe they'll throw in like psychology today or something like that and go like, wow, that like made for an amazing LLM. So I knew like the world was going to do, I wasn't going to get into that. It was expensive and like, it's like, I'm going to have the meal that all people will want to eat now. So I just created an infrastructure to leverage all of them, right. To help you mix and use different models and to create this ecosystem. So my goal was composable UI to focus on the user interface for all software. And LLMs are just one of those pieces of software and they enable that that UI layer, but not, uh, not get into the LLM space. So that we're talking like 15 years ago, I was like, Nope, LLMs are going to be a commodity. I'm going to focus on the picks and shovels around it. You know, that people are going to need to leverage them to, to do things. And, you know, it just turned out that it was, you know, way more popular than I expected. And took way longer. I was like, five years from now, we're all going to be using conversational UI. 15 years later. It just began, right? Yeah. It always, it's always funny when people are like, you're so lucky you hit it right on. It's like, Oh my God, I just, I want to cry when they say that. You have no idea what I waited.

 

Adam: 

 

That's a good point too. You're like, yeah, it seems like it's right now, but turns out there must be a history of building this set of conversations around this idea.

 

Robb: 

 

Yeah. The degree to which it works surprises me and everybody like, wow, look what happened when you mix this stuff together. It works surprisingly well. So I, the degree in which it works and the speed in which it went from where it was to where it is, was really, really fast. But I think that's true for the people who've made it and everybody else, you know, that's, that was hard to see the degree, but I think we all knew where it was going. It's just like self-driving cars. Is it going to happen? I think most people would say, yeah, eventually. When? I don't know. Yeah.

 

Adam: 

 

Right. That's a good point. When it gets, it gets around the other, the other hurdles of technology and regulation and Yeah. What's that thing called? I think the, I don't know, the driver's dilemma. It's like if you have to hit the brakes, otherwise there's a kid crossing the road or it might hurt the people in the car. And how does it decide that? Interesting philosophical conundrums for an AI and for humans.

 

Robb: 

 

Yeah. I don't remember who I was talking about with this, but they made a really good point and I want to give them credit, but I can't remember their name. So unfortunately they'll… If we find it later, we can add it in. But they said that one thing that's interesting about machines operating like humans or doing human-like things is that if it doesn't make mistakes in human-like ways, then we find it more discomforting and unpredictable. We want our machines that are anthropomorphic to make mistakes the way in the same way as we make mistakes. And when they make mistakes in surprisingly different ways than we would, that's what makes us so uncomfortable. And they sort of explain that as to why driving self-driving cars need to drive better than humans. It's because when they do make mistakes, oftentimes it's not in the same way we do. And that makes us understand that for other human drivers, it makes that us less predictable and, and that makes us uneasy. So, so we understand that patterns for mistakes have levels. There's the. the unanticipated, unpredicted surprise mistake. And then there's the quasi predicted mistake, right? Because we've seen that mistake before. And so when machines make mistakes, unlike we do, that represents, you know, a double, creates a double standard for how high that machine has to perform. So if it could make mistakes like we do, we're kind of more okay with that. And I think, yeah, and I think chat GPT falls in that category where In a lot of the ways it makes mistakes, it's similar to how we do, you know, it'll hallucinate and get something wrong, but it's kind of like, I've done that before.

 

Adam:

 

I made that source up.

 

Robb: 

 

 

Yeah. Remembered something or, you know, yeah.

 

Adam: 

 

Hmm. That's interesting actually. And that, that, I guess, is that, is that a feature versus a bug in that regard?

 

Robb: 

 

Yeah. It's just a good way to fail is, is to fail like a human would. It's going to give us more comfort. So I thought it was a great, I'm not taking any credit for that. I thought it was a great example and explanation for why we put such a high standard on machines and why there's a bit of fear that gets created.

 

Adam: 

Yeah, I think that's a key piece. I guess one kind of question or topic area I have for kind of a final question for now is the kind of promise of what the kind of having this composable interface can be. It's something like OneReach, right? Where it's not about the commoditized LLM, but it's about the ability to connect them together, right? One of the things that you talk about in the book, and then we talked a little bit about this earlier on in the conversation was that this might shift economic models, right? To be less about like one brand or product, but more about the functionalities that I'm able to accomplish. And we talked about this, like if you send John the thing, I don't care how you do it. I don't care if it's a phone call or email, but this idea that might shift how we interact with technology and what we think it can do. So there's this piece I'm curious about, but then also the other side of this is the capacity for accessibility, right? Or making technology more accessible to more kinds of people because it's not everybody having to adapt to the one single LLM, right? But in fact, I can compose what is best for me or my community or my people. So I'm curious your thoughts about this, like the kind of broader goal of making technology more accessible and how does something like OneReach and kind of these composable interfaces help us get closer to that goal?

 

Robb: 

 

Yeah, so super simplistic view, One Reach came out before Siri. So Adam Share was working on something, I was working on something, both of us pretty unknown to the world. His was bought by Apple and became Siri. And we were unbeknownst to each other as well. And you think of, you know, shortly after that, you know, We end up in this sort of, nobody knows what to call these things yet. You know, personal assistants, whatever Siri, Google, Alexa, Google assistant, Alexa is to people. I don't know what the, you know, I, I know what, what they call them, but I don't know what the world would call these things. Right. Yeah. I've asked many people, what is it? And they're like, it's a speaker. I know it's, but whatever that is, I was like, look, every company is going to have their own.

 

Robb: 

 

And they're going to want to name it, which is still the smallest, most insignificant part, but still important. And they're going to want it to be accessible, not necessarily on a speaker because, you know, putting a speaker in an office doesn't make any sense. They may want it in a speaker, like in sort of long tail access, but not as a main way to interact with it. And so we look at things like the telephone. you know, SMS, we look at Slack, Teams, like these are ways that they'd want to interact with it that make more sense on a business level. And then there's a lot of tools that, that Amazon has, that Apple has, that Google has, to manage behind the scenes, to manage the skills that people create, provide, security, the list goes on and on and on. That they don't share because, you know, this is their proprietary technology. They want everyone to use Siri. They think of it as an operating system in a sense. Like they want the whole world to just go to Alexa and Siri. They don't want to share. Hmm. share that with competition. So those tools are sort of locked up, right? And from what I hear, not the best anyway. I set out to say, well, I'm going to create those tools so companies can, you know, have their own ecosystem where anyone in the company could submit skills. They can create their own Alexa. It has all the security that's necessary. It has the channels. It doesn't need to be on a speaker and connects to those channels. And then, and then that becomes like the beginning of a journey, like a baby step into a journey of just slowly wrapping all your software into this near you, this near UI. So then you just start adding skills over time as. vendors start exposing their products like Salesforce would, or like, you know, ServiceNow, they start exposing their products with conversational UIs in front of it. You start adding those as skills and slowly over time, this thing is, is slowly becoming the one UI to access all your software. Uh, so you didn't have to do this like forklift in this big thing. So starting small. Um, so I basically just set out to, to, you know, to build that infrastructure, knowing what was behind the scenes and what, you know, Amazon was using at Alexa and things like that. And, and then recreating that in a way that made sense for companies. And that's really what OneReach is. And so it's this idea, like we can start simply with just questions and answers. And there's a lot of companies that do just basic questions and answers. But it doesn't lead up to a UI for all of your software. It's like, it's like a point solution. It's actually makes the problem worse in a way. Cause it's like yet another piece of software to add to the list of software that people are using. And oftentimes it's just another piece of software that doesn't get used because it doesn't do a very good job of answering the questions. So starting there and then just kind of growing this ecosystem. And so another piece was like no code so that, you know, anybody could actually build skills. It didn't require a centralized approach where you're. where your IT department is. So the idea is like your IT department becomes more consultative and enables the rest of the company instead of being the sole builders of everything, they become the enablers and the experts. Um, so that's what it's all about. And, and I think this gradual approach is, is how it will happen, whether it's with us or with some other company that, you know, that decides to compete with us on this, but. But I think it's just going to be gradual. We're going to start with something simple and then it's just going to grow into like, Oh, remember when we used to log in to Salesforce or whatever it is. Yeah.

 

Adam: 

 

No, I think it's a really exciting piece of a framework of thinking. And I think it's an exciting piece of software because even in the small sense of every once in a while you'll see some AI enabled software that comes out that says it's not just chatting to chat GPT with a nicer you know, skin or interface, but it can actually help you find files in your computer. I'm like, Oh, okay. That's, that feels neat. Cause it's actually using, it's interfacing with the software that I already also used your point. I don't have that sad, empty app drawer of, you know, I don't ever open Photoshop because it's, it's too much work. You know, I'd rather be a safe, let's actually make a nice photo for this episode. And it says, here you go.

 

Robb: 

 

Yeah. Yeah. And it's really funny that it's kind of, I think we'll look back and say, why did we, why did we start with the automated house versus the automated company? Like what makes more sense? Like automated a house? Like how much is our house based on like how much productivity did we do today? Oh my God, we should automate watching TV together and seeing a movie and. Maybe if we watched shorter movies, we could watch more movies together and make more posts on Facebook, right? It's funny, an automated business makes more sense by a long shot. It's surprising it didn't start there, right? It didn't start with businesses. So I still think eight years later, after having a finished tool, I'm still like shocked that this is a new idea. I just can't believe it. I can't believe we've gone eight years with these things and no one's gone like, I wonder if I should put one of these in our office. How does that happen? How does that get missed? It shocks me. And maybe it's because the people are caught up on the speaker and the mechanism, you know, they don't realize like it doesn't have to be a speaker. It could be a phone. It could be text. Yeah.

 

Adam: 

 

Right. And it's like, yeah, to your point, yeah, because I'm getting, I'm like, oh, can I turn my fan on or my light on by talking to it? And it's like, that's one thing it can do, but that's, yeah, you're right. Right. Why not? Oh, actually, let me actually have it make me a functional, I don't know, grocery list that's based on what's in my fridge or not. Or when I went shopping last, so it knows what I bought or whatever it is. Yeah.

Robb: 

 

And you're like, look, it turns on the lights and you, and then you like play that 19, I don't know, 80s, like the clapper, you know, you're like, oh wait, that's not new. It's still there. Like I could have just installed the clapper and maybe it even understands me better.

 

Adam: 

 

It's a fair point. Something about the eighties, I think did understand us better. You know, there's things we could learn, I guess, from that decade. Awesome. Robb, I just want to say thank you so much for taking the time out of your day to chat with me on the podcast. This has been a great conversation. I learned a ton. Cool to dive into your wisdom.

 

Robb: 

 

Yeah, no, it was great. It was an awesome, awesome conversation.

 

Adam: 

 

Like, wrapping up today's episode, we dived into the fascinating world of hyper-automation and conversational AI with our guest, Robb  Wilson. A huge thanks to Robb  for joining me on the podcast today. We discussed the complexities of technological interfaces, the potential for automation to enhance things like customer experiences, and the concept of decentralization in business and technology. It's been an enlightening conversation. I'm really grateful to Robb  for his valuable insights. So as we conclude, take a moment to reflect on the fascinating points that we've explored today. How do the complexities of technology interfaces affect your own interactions with software and AI? And have you experienced the power of automation in enhancing things like customer experiences, your own, or working in an organization? Or have you witnessed the challenge of disconnected customer and employee experiences through technology? So I invite you to ponder these questions and consider how these insights can shape our modern world for better or for worse. At This Anthro Life, we appreciate your continued support and engagement, and we want to hear from you. So share your thoughts, feedback, topic suggestions with us. I'm always open to it. And you really are an integral part of the podcast community, and I value every single one of your contributions to our conversation. So thank you so much for folks that have jumped in so far, and I want to hear from more of you. And if you're hungry for more thought-provoking content, do not forget to explore our Anthrocurious Substack blog, linked in the show notes below, where we delve in deeper into the fascinating intersections of anthropology in everyday life. It's a great resource to expand your understanding of the topics that we discuss in the show. And you're invited also, so if you are interested in writing a post for the blog, you can shoot me a message over at thisanthrolife@gmail.com or get in contact on the website. And finally, I encourage you to take action and engage with the podcast, right? Subscribe to This Anthro Life to stay updated on our latest episodes. Never miss your chance to explore this intriguing world of anthropology, humanity, and experience. And leave us a review or share this episode with someone you know will love it. So help us spread the word and join us in building a vibrant community around the podcast. Remember, the convo doesn't end here. Let's continue to explore the possibilities of human experience, technology, and culture. So thanks so much for joining me on This Anthro Life. I'm your host, Adam Gamwell. We'll see you next time.

 

Robb WilsonProfile Photo

Robb Wilson

AI researcher, technologist, designer innovator, host of the Invisible Machines podcast, and WSJ bestselling author

ROBB WILSON is the Founder, Lead Designer, and Chief Technologist behind OneReach.ai, the highest-scoring company in Gartner’s first Critical Capabilities for Enterprise Conversational AI Platforms report. OneReach.ai has also launched an AI-powered coffee shop on the front lines of the war in Ukraine, called HUMANS, Coffee in Tech. HUMANS is a unique cafe that has managed to maintain profitability during a volatile time and is working to extinguish conventional management hierarchies within the food and beverage industry. Robb has spent more than two decades applying his deep understanding of user-centric design to unlocking hyperautomation and at OneReach he is helping to democratize two of today’s most transformative innovations – digital communications and machine learning – accelerating society to the inevitable moment wherein humans can interface with any system, the way that we innately do with each other. Used by companies including Nike, Expedia, DHL, and Unilever, OneReach helps its customers design and deploy complex conversational applications as easily as one might author a spreadsheet or craft a presentation.