文字起こし 🔗
210 Intro: Welcome to Alter Everything, a podcast about AI, analytics, and the future of work. I'm Michael Cusick, one of the producers of the show. Right now, almost every company is racing to build and deploy AI agents. And when we talk about those agents, we usually talk about them like software. How fast are they? How cheap are they? Which tasks can we automate? But what if thinking about AI as just another software tool is the exact reason so many enterprise AI projects stall out? Today's guest, Asha Saxina, Has spent more than three decades building out data and technology businesses. She's an adjunct professor at Columbia University, the founder and CEO of World Leaders in Data and AI, and the author of the new book, The Digital Human Advantage. She makes a pretty fascinating case. Once AI starts reasoning, remembering context, and taking actions across your workflows, you can't just deploy it like software anymore. You have to onboard it like a new hire, complete with a job description, clear decision rights, and a path to earn your trust over time. In this conversation with Joshua Birkow, Asha breaks down the shift from AI agents to what she calls digital humans, how to move AI from an intern level assistant to a trusted advisor, and why so many organizations get stuck in the trap of merely optimizing old work instead of reinventing what they do. Here's Joshua and Asha.
Joshua: Hey Asha, thank you so much for joining me on the podcast today. I I did an extensive background search just to find anything about you I could because I got obsessed with this book that we'll talk about later, that I think is is gonna be end up being the foundation of this podcast. but it opened up all kinds of of questions and ideas. I see here that you're an AI strategist. You're Tech entrepreneur, board advisor, adjunct professor at Columbia University. I'm not even done yet. founder, CEO of the world leaders in data and AI, and the AI Factor Institute. And then, of course, you're author of both the AI Factor and this book, Digital Human Advantage, which I want to talk about. Did I did I miss anything? Are are are we screaming over cheever or that's amazing?
Asha Saxena: No, you got you got it all. You did get it all.
Joshua: And and you have time for family and and dogs and all this other stuff.
Asha Saxena: I do.
Joshua: amazing. Amazing. I'm impressed. can you talk about how you got into the moment before you decided to write this book, The Digital Human Adventure. And I'm really curious. I love backgrounds. I like understanding pathways, how people got into this this line of work and specifically what's what's intriguing about AI to you today that Maybe other people don't talk a about as much.
Asha Saxena: First of all, I think everything you said is was in works thirty-five years back when I went to school, did my undergrad grad now PhD in computer science engineering and and everything I studied I used during my work life, my professional life. So I started my career as an engineer, then built a data management consulting firm. I ran that for twenty-two years as an as a CEO, then I built an e-commerce company, then a healthcare software company. But as you're building software, You're learning about how technology is being used. and I think when you build a business, you also understand that technology just for the sake of technology is just a tool. It becomes really powerful when you use technology to really create a shift in your business. Shift in the outcome and produces results. I think I learned that through my last three decades, building companies and being a business leader, even though my core was engineering, I learned as a business leader that you need to use technology as one of the most powerful tools. In your toolbox to build your business, the systems, processes, and then train the people and educate the people. So I think my career of building technology companies truly helped me understand the value of technology. And then, of course, been teaching at Columbia University for 18 years. I started teaching entrepreneurship for data-driven businesses, and now I teach AI. And what what What when I'm teaching these graduate students, it's really so interesting because they're learning these concepts of building a business, but really thinking about how AI can give them the power from the very beginning, creating the foundation there, I think is critical. And I think that's what led me to writing the books because during the pandemic, what I realized that I built the businesses, I was working for a management consulting firm where I was coaching CEOs on building high grade. Companies and working with private equity companies. And I realized even though I was building these management teams to exit to a successful exit, technology was the core of it. how they they use technology to build the organization and scale is real critical element for any any business. And so during the pandemic, my whole effort was with the first book, The AI Factor, is to really bring business and technology leaders. Together so they understand each other's language. And I think the biggest gap is when technologists get so excited about their technology, they forget the business outcome. And the business leaders are so excited about the business outcome, they undermine the technologist, and they feel technology
Joshua: That's right.
Asha Saxena: is just a cost center. And so the AI factor was the first step into taking my 30 years of experience and putting it on the paper and saying, you can use these frameworks, And I think I learned that. being an educator to be able to share with the world that you can
Joshua: Yeah.
Asha Saxena: use these frameworks to learn about
Joshua: Yeah.
Asha Saxena: having these conversations and having a common ground.
Joshua: Yeah, I I that was the first thing it jumps out to you so anybody who who goes through the book is that it's filled with so much structure, so much
Load full transcript
frameworks and and concepts that aren't just diatribes of of information of all the stuff you need to know. It's this working models which we'll we'll get to later on. I I I think that's probably the the disadvantage of you being a professor and having to explain concepts to to folks that that don't have your thirty years plus experience, you have to get that down into the the most important sort of aspects that people can operate from.
Asha Saxena: Exactly. I think that's the advantage, right? Have being an entrepreneur, being an engineer, being a coach, the C C level coach, and then being a professor for eighteen years. I think has given me a really great mix of learning how to communicate the value of technology for the help of, building growth and scale in a business.
Joshua: Do you you find that's harder now with with AI? i you and I both know AI's not new, but this new world that we're in in say the last three to four years since Chat GPT came out, is it hard to tr try to think about this in frameworks because it hasn't had the benefit of ten, fifteen, twenty years?
Asha Saxena: And even though AI feels new, I think the biggest challenge is that everybody has their own definition. And everybody has their own expectation of this technology and perception of what this technology can do and the fears.
Joshua: That's right.
Asha Saxena: So there's a lot of different flavors of
Joshua: Unbelievable, yeah.
Asha Saxena: fear and flavors of expectations you get in the market. And that's where the confusion begins. It's everybody's thinking and and and that's why I think having a common ground becomes really important. And in my book both books, my intent was to create that common ground so people understand what the definitions are. And once you start with the the fundamental common terminology, then you can start building.
Joshua: Every author I've talked to has said the same thing that you have to beat up your own ideas, right? You have to, you put down your operating model and you gotta you gotta check it. You gotta validate it. You gotta talk to to
Asha Saxena: Washed.
Joshua: leaders that are implementing it. Can you talk about that a little bit? Were you way off from where you ended with the book? Did you start off with the this one premise and then ended up somewhere else?
Asha Saxena: Well of course. Of course, always. Always I
Joshua: Right.
Asha Saxena: would say that right that's the evolution of your thought process. It's when you're doing research, you start with a hypothesis and then you're like, oops, I was completely going in the wrong direction. And so I think
Joshua: Got that wrong. Yeah.
Asha Saxena: that's the fun part of writing the book is the research you do to validate what your hypothesis is. Is that accurate or not? And if it's not, and you hope that it's not accurate because of course you're trying to figure out am I going in the right direction or not? So I think the research piece becomes really really critical when you're writing the book and actually implementing your framework and seeing does my framework actually work becomes really critical and I think that's the fun part of writing a book for me because you become a student of life, student of what you think and then figuring out, my God, I was totally wrong as I was thinking about it. So I think it's
Joshua: Yeah. Yeah.
Asha Saxena: it's a journey and for both my books, actually I I wrote three books, the AI Factor, the AI Factor workbook and then the Digital Human Advantage. But During the AI factor journey and the digital human advantage journey, I think it was such a critical time when I as an engineer, as a practitioner, was hunting for answers. and so it took me on this journey to find these answers that hopefully the industry leaders will find helpful because it gives not only them, but using it with your business leader and technology leader can give ground for them. to get started.
Joshua: Sure. I I I think there's something I find really fascinating about this and it's the reason why I asked this question was we live in a world now where information is so easily accessible but these aren't problems or ideas that that are solved overnight or that are Google queried or Chat GPT is gonna tell you the answer. These are things where you you really gotta dig into them. and investigate it and and get into it versus, hey, I'm just gonna write a business plan because it's it's easy to do in cloud.
Asha Saxena: I I agree with you. I think it's all the years of experience, my thirty five years of industry experience that I bring to the table. So when I'm thinking about a problem, I'm being curious about a solution and I'm actually having a conversation with industry leader, the place you come from is because of that experience you have. So I agree with you completely. Even though the book took me two years to write, it didn't really start
Joshua: Sure.
Asha Saxena: at two years before. It started Fifteen years back. And I even
Joshua: Sure, exactly right.
Asha Saxena: though the first twenty years of my life I was executing. I was, a practitioner. but then fifteen years after has been always curiosity. they say, when you are early on in your career, you're learning the hard skills and you get to a place where you know the hard skills. But the magic happens when you start learning the soft skills of leadership, asking the right
Joshua: Yeah, yeah.
Asha Saxena: question are we solving the right problem and I think when you actually get into that zone of figuring out what the big problem you're solving is where the magic starts happening. And I think so for me that's where this whole curiosity of writing the book began 15 years before I actually wrote the book was to really start thinking about are we moving in the right direction? If I'm having this problem, I'm sure other people are having this problem too. I sat as a technologist on the technology side and then I was sitting on the business Side and to
say I don't get this technologies when I'm on the business side, and I don't get the business leaders when I'm on the technology side was a very real problem. And then during the pandemic, when I started writing the book, we all were digital. We I knew that the
Joshua: Sure.
Asha Saxena: the power of the models are gonna just become exponentially high. And so
Joshua: Exponent. Yep. Yep. And
Asha Saxena: it was really even before Chat GPT came out, I was writing the book called The AI Factor, because my whole thesis was on can AI truly impact our businesses and what would that factor be, what would that formula be that would actually, take us to that level of creating that
Joshua: Sure.
Asha Saxena: magic?
Joshua: Yeah. All right. So we've we've talked about this a little bit, but I now I gotta we gotta get into the meat of this. And I'm showing this because I'm directly pulling from this. explain to me what you mean by digital human. You have a very specific definition that you that you use in the book.
Asha Saxena: Yeah. So to me it was really interesting. I wrote the first book, The AI Factor, really for the framework, where is the magic happens and how should businesses think about AI in their business and how should AI the technologies should communicate. But then I the book came out in twenty-three where the whole large language model, the Gen AI era had started. And by twenty four, twenty-three towards the end of thirty-three and beginning of twenty-four, I started writing the second book, the digital human advantage. Because my whole thought was the large language model can only do so much. But what happens when they actually start executing on the work? So the AI agent is a piece of software, how people hear it.
Joshua: That's right.
Asha Saxena: But digital human is actually the role, if you really think about it, because it's really taking a rollover of whatever you give, and that's where the book starts, as you mentioned. In the book, I talk about digital human. This is this so the question is. Isn't that these things what these things can do? in the book I explained the pattern. I talk about that 30 years or so we completely reinvented how we talk to machines. if you really think the evolution has been so fast. I still
Joshua: So wow.
Asha Saxena: remember going to college and working on command line, and then came graphical interface,
Joshua: Mm-hmm.
Asha Saxena: and then came voice and chat. And now machine can actually see and sense and respond like we do at scale. this is incredible,
Joshua: Yeah. Yeah.
Asha Saxena: right? The evolution has been truly incredible. Every single one of those shifts didn't just change the tool. It changed how God to participate in this economy. So when you call something an agent, you start judging it like a software. So it's just a software, it's an agent that can finish the task, how fast, how cheap. So now you're talking really about, is AI giving me value? is how much is it gonna cost me? But when
you call it a digital human, suddenly you have to ask the question, is this a new hire? Because now it is giving you giving it a persona, what's their scope? Who do they answer to? What can they decide on their own? And honestly I think the most AI failures I see inside companies aren't really about the model. They are about the role, you've got the capable I I always say tool is a tool and can do really great work. But if you don't have a job description for an agent or these personas, you're gonna fail because you're really implementing a tool that It's very powerful, but you don't have clear definition of what it's gonna do.
Joshua: when we think about the digital human, it it becomes really interesting because you're now alluding to again a role, a persona that those are tangible things for an entrepreneur, those are tangible things for a for a business owner to to understand when they don't necessarily understand. Why the latest model is so good and why, what's the difference between Kimi and Opus five? they might not know all these nuanced things, but they do know and understand the power of the the sort of human aspects that we've we've had in business for hundred plus years.
Asha Saxena: I I agree with you completely and I think that's why it's important that when we're working with the models we understand are these models and and listen, a lot of the i it just becomes such a fascination for so many of us because we all want to play with these tools and we get so excited about what it can do. But really when it starts coming the conversation starts happening about ROI, it becomes really a challenge. how much impact did it have on your business? And nobody really knows. is because they're really using it for let's find documents. Can it be a HR helper? But really nobody's quantifying. Is it really creating
Joshua: Yeah.
Asha Saxena: the shift you need to create the shift?
Joshua: Yeah. But I would even argue the one question, is it creating this this ROI? But it's also a a there's a lot of concepts in your in your book that get people to that place versus th there's always a question like, Hey, are you driving value? But it's like, okay, if your intent is to drive value, there's probably a a couple tried and true ways to to to get there. And I think you're Your book covers these aspects. Do you agree?
Asha Saxena: I agree, absolutely. And I I think I take them through different stages, to really understand because I think once you understand that agent and digital human and h where and listen, every technology has its own place. But where do
Joshua: That's right.
Asha Saxena: you actually place it, is becomes really, really important. are you using digital human to be a HR assistant or a call center agent or are you using an agent? To automate task and automate the content for your marketing. So it it are two different things. So, how do you use an agent versus a digital human, which is giving a persona? And what are the different kinds of
Joshua: That's right.
Asha Saxena: persona? so in my book I talk about the five personas. I talk about assistant, influencer, companion, digital twin, and advisor. And those are different kinds of personas that a digital human could have. And each one shapes the interface and the autonomy that would follow with these each roles. you don't start with should it have a face, should I call my a assistant Amy or John? mo
Joshua: Great.
Asha Saxena: it's more so about how much should it be allowed to do. s you start
Joshua: Yeah.
Asha Saxena: with what's the job for that digital human or the persona you've given. and the the I would also talk about the interface and the autonomy level are just the downstream decisions once you actually answer the question the important question is the where you start, what's the job?
Joshua: Yeah. Can you can you briefly, just for the the benefit of of listeners, is talk about the different five personas just briefly and help us understand what
Asha Saxena: Yes, absolutely.
Joshua: what what those cover?
Asha Saxena: Yeah, so in the book I talk about child to sage progression. I talk about this whole model of how AI starts with a child stage, is like a new intern, And then human, we as a teacher are teaching this intern how to correct, it makes makes a mistake. And then the adult stage is executing with guardrails. Now we are saying the human now in the loop is acting as a manager because it's building the trust but auditing. And
Joshua: Still watching yeah.
Asha Saxena: Then Sage is where, now the model has evolved, and now you've actually went from being a teacher to this intern, to the manager who's auditing, to being a sage who's actually flagging things before anyone asks. So to me, when you think about the personas of assistant, who's helping you do a job, or an influencer who's helping you create the influence in the company, or as a marketing role, or a companion. who's actually doing the work with you, or a digital twin who can be you to answer the questions, or an advisor or coach who's actually could be again I'm thinking about the HR role or a manager could become a take your advice. Those are different roles personas. Some the biggest persona which we've seen is the call center, who when you call in the call center attendant or agent is answering your questions.
Joshua: Sorry.
Asha Saxena: It used to be so irritating when you call the cult center, but where I've seen actually
Joshua: Tell me about it, yeah.
Asha Saxena: has worked is when you're trying to shop for a skincare. And now you have these personas that actually can help you match your skin tone, which I think is a lot more useful than calling
Joshua: Amazing.
Asha Saxena: for your airlines and getting put on hold and going in a loop.
Joshua: Yeah.
Asha Saxena: those those days and those voice chatbots were super irritating. So how do you evolve that
Joshua: Yeah. I love this because you the way that I I took it from your book is you've got
these you've got multiple frameworks in here, but they're all intertwined, right? Your your child to sage is intertwined with with the different personas. let's s let's shift gears because there's a lot of moving parts here. And I think at least for the executives and and senior leaders listening in, that they're they're trying to figure out, okay, what's my operating model? How how do I put all these pieces together that that make sense so that I can start to go towards working on tasks and projects that actually return value versus just playing with the tools like we were as we call nerding out, how do we how do we think about the operating model? Which I by the way, you you talk about in your book as well.
Asha Saxena: Yes. No, and digital human operating I just call call that operating model in my book digital human operating model. and and
Joshua: Mm-hmm.
Asha Saxena: it's really about when I talk about operating model, because if you think about large language model, you give a prompt, you have a conversation, it's a back and forth. But what a operating model really does, it it tells you how to govern everything. how the decision rights, how it grows over time, who's accountable when it acts. I think that's the difference. And I talk about it also when I talk about child to sage model in the book. It's really not just about working with large language model when you think about or clever prompting, but really thinking about operating model as a infrastructure that is required in your organization.
Joshua: Yeah. I I this is really fascinating because all the folks that talk a bit about AI and cannot say the word AI without governance in the same sentence will love this because it's I I'm such a huge fan of this balance between okay, go forth and innovate and do things beyond your wildest imaginations. But do think about governance. Do think about security and and infrastructure and how do these things play safe so that we're we're not, causing all these other downstream problems.
Asha Saxena: I think governance is the biggest, biggest challenge right now because I I was working with an organization and working on their governance model and I looked at their documentation for the governance they were submitting to their steering committee and to me it looked like a nineteen nineties business analyst document defining,
Joshua: Mm-hmm.
Asha Saxena: the terms and I
Joshua: Right.
Asha Saxena: was just amazed by, the the gap we have to really think about
Joshua: Right.
Asha Saxena: a lot of people are lost about governance. They don't even understand
Joshua: Sorry.
Asha Saxena: how do you navigate this. And I think that's why to me this book was so important because AI is a powerful tool and people are excited, super excited about can what it can do. But also a lot of people are concerned about what how it can harm. there's and there are so many ways you don't even know the harm that can be done. So it is it becomes really critical that you you think about governance seriously. this is a really
Joshua: Yeah.
Asha Saxena: important topic because I started the whole organization because of that. The world leaders in data and AI was started because it was about the passionate leaders, senior
Joshua: Mm-hmm.
Asha Saxena: leaders who come together to talk about responsible AI, ethical AI. it's we are never before are in the place engineers are in the place where we are writing history. We are the one creating this digital framework, the digital, army that The workforce that is going to be working with us, collaborating with us. They will be
Joshua: That's right. Yeah.
Asha Saxena: in our organization sitting next to us, and how do they show up? And we are designing that team that is gonna show up at every work desk. So, how do we really take the responsibility as leaders who actually show up, collaborate? And everybody, it's it's a new place. So I think we all need to figure out how do we really come together to do it. right. And I think that's why these conversations are really, really important. And I'm so glad that you're doing this podcast because this is important that we talk about it. Because not no
Joshua: That's right.
Asha Saxena: nobody has all the answers. We all are in the playground trying to figure out how do we invent and reinvent.
Joshua: Yeah. There's well, there's a lot to to take away from that. the governance conversation is really interesting. And I've had at least two or three other podcasts in this year that were focused on this concept of AI governance. and and one of the best conversation I think I had on this topic was with one of my guests named Fern Halper, who's been in the T DWI space for for a long time. She's been in the field for Probably as long as you have, or more, but she was very adamant about the delineation between the worldview of data governance and AI governance and how those are getting blended together when they absolutely shouldn't be. It was super fascinating. But there is one, this conversation happening specifically on AI governance and what is the new world view look like? But I think the more interesting thing that I I I hear from you is everybody's in in their Ferrari right now. Right? We're all going a hundred miles an hour down the freeway. And but but there are some that some of us that are hyper focused on the technical aptitude, the what the models do, how many parameters they have, what's the the the the sheer possibility of these models on a technical basis. And then there's folks like you that are thinking about, okay, when you're going a hundred miles an hour and you you arrive at your destination fifty miles down the road, what is the world gonna look like? What is the the sort of infrastructure going to be look like? How is the world going to operate with all of these capabilities and all these technical solutions in place? And I I find it Fascinating that we have a good cadre of people across the industry that are, to your
point, talking about it, that they're elevating this conversation to say, hey, look, it's just as important to have long-running queries, long-running prompts with context windows and all this stuff, as it is to understand where's the human, where's the where AI sit? How do we operate? How do we organize? How do we drive value? Because th those are still fundamental concepts that are still true, even with AI.
Asha Saxena: Absolutely. I think I kidded second, third, fourth, that number one you spoke about the governance AI and data. very important that data is governed. It's very important AI is governed. Also, it's really important that we understand where does the role really sit because AI is changing the way we work and and how we are building the system. So the c these conversations are really, really important. And this is happening within the organization. I was talking to the CHR of a very large company. And she was talking about her mandate is about thinking about digital workforce, how digital workforce that includes the people and digital human are going to collaborate and work. So that we have to have new principles and new foundation of how we are going to be running organization structures because it's not going to be about just a role, you're the manager of data intelligence. What is your role? But it's going to be the the capacity as a team. So it's not just how one employee produces, but it's how one employee plus the AI employee. It's the team producing. you must have heard about entrepreneurs who are building companies with just AI agent as employees. So think about the single employee
Joshua: Single employees. Yep.
Asha Saxena: yeah, one human employee with multiple agents.
Joshua: Single humans, yeah. So
Asha Saxena: creating the organization.
Joshua: wow.
Asha Saxena: So think about the power you can have in an organization that actually includes a humans and digital human. How is the world gonna look like when you actually bring those together? We have to really think about which jobs are digital humans are doing and what humans are doing. it's gonna become we're gonna get into the world and that's why I I love the fact that I have this community of world leaders together because the communities are gonna become more important because our roles are changing. people I I teach at Columbia and a lot of my students will ask, how would the job get impacted? because we are going to have new roles in the organization. And I always say is that you have to start upgrade yourself. I was running a board readiness with EY and we I had public board members in the room who were asking me, how do we show up? And I say the same answer that I give to a graduate student, I gave it to the board members Is that you need to upgrade because now
Joshua: Mm. That's right.
Asha Saxena: it's not just about getting your board deck and reading the board deck, but now within the framework, you can use the LLM to understand the market and competitive analysis. You should be more educated about how you show up in the boardroom and what questions are you asking. And I have to tell you,
Joshua: So true.
Asha Saxena: super smart board members, super highly accomplished, and their first question was we don't even know what questions to ask. How what should we be asking in the board? boardroom. And so I think those are the important conversations to have because they are l on this journey of becoming a lifelong learner. AI has made all of us students of life again. we have to become students to learn about this technology and know how to implement and how to be curious.
Joshua: That's right.
Asha Saxena: I I love that you were having this conversation with me. You were curious about my journey. You were curious about the questions
Joshua: Very.
Asha Saxena: to ask. And I think that is the most important skill we need to have students, professionals, an accomplished leader to go back and start asking good questions to figure out are we in the right direction? Should we be giving this these tasks to AI agent or to humans? relation I I work with another organization, CPG, and the CIO said we have mandated that our organization anything that is touching a relationship or a real real trade off is going to be recommended only for for humans, so we're also not going to use a AI avatar for any human reflection. So we're not gonna use it
Joshua: Interesting.
Asha Saxena: for skincare products. We don't want AI avatar or models, be re representing what humans really care about. So really thinking
Joshua: So
Asha Saxena: about what is what are the values that you embody in an organization.
Joshua: Yeah. It's so fascinating.
Asha Saxena: And where do you want to create that distinction between human and digital humans?
Joshua: Yeah. the two things that come off the top of my head. The the first thing is how humbling AI is, right? I laugh at this every probably once a week, that there are board members who are humble enough who are able to say complete
Asha Saxena: I don't know.
Joshua: complete change. Yes, I don't know. I'm gonna have to, learn. I watched this amazing workshop with someone who I I really admire, Ally K. Miller, who interviewed Mark Cuban. And Mark Cuban started rattling off all this AI that he's using and and how he's using it. And I just sat there dumbfounded. I was like, this guy gets it. This guy's billionaire. he he gets it. He understands the the the technical side enough, to understand how it drives value, how it actually makes a difference in the he talked about cost plus drugs, which is a is a a his effort, and how it's impacting. And I was just like, God, there there's hope in the world because there there are a new world view of of leaders. So everybody from your your senior manager up to your board members that really have to put their nose to the grindstone, they have to understand what they're talking about, they have to understand how it applies to them. The models, the
framework, these are all tools for them to to deploy and to tap into. But they're new. you just wrote about it less than a couple of years ago. And so th this idea that that they everybody's gonna have to go back to school and and do their do their work to to catch up, is is really fascinating to me. I really love
Asha Saxena: You I have
Joshua: this.
Asha Saxena: I I have to say that isn't it true that when you think about AI, it is absorbing a lot of volume, from from our task, right? So what it is giving us, our best people get to spend more time on the parts that actually need a human. Not that judgment quietly disappears from the process because nobody's taking anymore. It's really giving that time and space for humans to actually evolve. for us to become better humans. And really the whole premise of collaboration when we talk about it, and I talk about it in the book is humans being more moral agency, creativity, real empathy, accountability. That now lies with humans. AI brings speed, scale, recall, consistency. But scaling judgment means more of the human strength shows up in more places. I did the whole TEDx on human upgrade. I think this is such a great time we are living. I'm so excited that I'm
Joshua: Yeah, yeah.
Asha Saxena: still in the workforce after thirty five years because I am having the best time of my life being an engineer, being a leader, because we get to create. We get to actually show up and we get to evolve. We get to actually think about Mark Cuban not as a technologist, can now play with this tool and build things and learn things and use AI to be more productive. No, so we do have this powerful tool, but I keep going back to governance because I think it is such an important topic that you touched. and I talk about it all the time, is because it too truly is about if we don't put create the control plane within our software, our technology, to truly understand who's accountable. Are you consistently monitoring the models? Are you understanding where where the agency stands? It becomes really really important that we think about this as engineers, as architects, to start putting the governance, in the in the AI framework.
Joshua: Yeah, it's the the governance question will be interesting. We're gonna have to bring you back on a year from now and see how it's evolved from that just on that topic alone because it's it's it's really wild. I wanna go back to the this other tangible idea and see if we can flush this out because I'm I'm maybe I'm selfish in this regard, but I'm knee deep in this. so I I've gone and built an agentic organization. I've got 30 plus agents that are going out doing different things. My travel, my expense reports, handling emails and Slack and all this great stuff. And I'm super in admiration mode, holy smokes, this is possible. it goes and and does all these things without me asking. And then the things I do ask, it goes out and and helps me. And I'm it it's a continual growth thing, but Where my head goes is I'm fairly technically inclined, I'd like to believe. And so by building this, I now have a new world view of what's possible. But I sit back and I imagine what if the entire organization had their own agentic organization, right? Now you l let's just do a hundred-person startup, hundred
Asha Saxena: Yes.
Joshua: humans. We're gonna have to start to say, humans to be very delineate. we have a hundred people and each of them have a near-infinite agentic organization that they can deploy to anything that is needed to be done. What does that world look like?
Asha Saxena: my God, that's the world we are entering right now, Josh, because Yes we are.
Joshua: It's we're just jumping into it right now. Yep.
Asha Saxena: And that's why I wrote the whole book because the world needs to understand that it's not as I said earlier, it's not going to be is it digital or is it human? We're gonna be meshing this whole world that you wouldn't know is the work getting done by my digital agent, digital human, or is it being done by a real human? But what really I think matters is as you said, I wanna go back to what you spoke about, that you build these agents to do your work. When you talk about organizations, can organization actually focus on What kind of work are they really? If they're looking for only the headcount saving and ex and automating these existing processes, or they're actually trying to do more? Is it that
Joshua: Mm-hmm.
Asha Saxena: automate my travel, automate my marketing? in the book I call that that it's really the trap because you fall into the optimizer quadrant. I talk about the four
Joshua: That's right.
Asha Saxena: quadrant in my in my book, the power quadrant. And if the result is just new work, faster judgment cycles, things you've used to have to just skip because
Joshua: Yeah, exactly right.
Asha Saxena: there wasn't enough human band bandwidth, now that's capacity. The different outcome I would talk about is that you have to actually be looking at is it more I can do with AI? Can I help create
Joshua: That's right.
Asha Saxena: new products? Can I help create reinvent my business model? So when you start thinking about not just optimizing, optimizing my work, manual work labor, but really thinking about I call it in in in the book I I have the four extender, optimizer, I talk about innovator and multiplier. is where you start now in re innovating new products, new ideas, new business models. Multiplier is when you actually reinvent your own bus whole business, reinvent the whole process. When you say, you know what, this is how I go to market for my products, I want to reinvent completely. Netflix was a great example when we had Blockbuster, Netflix actually said, I want to truly understand my customers and create a queue so I don't have to have brick and mortar. I can can now understand what my supply and demand should look like. Now now you're reinventing, Rightshare, the whole Uber, industry came from
thinking what do my customers really need and can I create supply chain without having vehicles? So can you actually use AI to really reinvent? So I would say just not sticking on the trap of just automation, but really thinking about which quadrant you you could be. to get the most value becomes really important.
Joshua: Yeah. I I I agree with this. I just on a a personal journey, I I I was always enamored by when AI because I've been like you've been doing this for twenty plus years and and came from data science machine learning. But when AI came, I was like, okay, the prompting is cool. All right. I I could get excited about but when agents came and they could do stuff, I was like, All right, I'm done. I'm gone. You no one's gonna
Asha Saxena: Yes.
Joshua: hear from me for years. But what what I found interesting from a very personal aspect. And I've talked to a lot of people when they come up with the argument that AI is going to take their jobs. As humbly as I can, which is hard sometimes because I start giggling because they're not willing to take in the whole picture. Most people that I talk to are behind or feel behind a year at least. So what agents have done for me in the last three to six months has got me caught up. And and I think there's
Asha Saxena: Okay.
Joshua: this this human tangible perspective slash reality that is playing out that when people start going after agents, yes, I would say the North Star is exactly where you're headed, is let's let's get to reinvention, let's relook at the things, the assumptions that we thought were were true and challenge them. But you need mind space to do that, you need
Asha Saxena: That's true.
Joshua: space To to stop and think and not worry, my God, I have 2,000 emails in my inbox. And
Asha Saxena: I agree.
Joshua: I don't have any. I I has all my Slack messages. Have I seen them? Have I responded to them? Am I up to date on all my projects? I'm still not up to date on all my projects, being very transparent. and there's a lot of projects that I want to help drive that I I still need bandwidth. I still and I I like to consider myself a high performer as a human. And so now having this agent piece that there is. Do you agree with this premise that before agents really start to take off, that all the call it baggage, if you will, has to be rectified. Whether you are
Asha Saxena: I agree.
Joshua: bold enough to be like, screw it, I'm not dealing with it, throw it in the garbage, or let me clean it up real quick and then let me move on to because I I've spent that time cleaning up. And now I'm just at the gate to where I can start to to fire on all cylinders. And and I'm you can tell I'm all jazzed up about it, right?
Asha Saxena: Yes, exactly. No, I think i i it's a a great point and observation what you made. Yes, we all are behind our task because we have millions of emails coming and billions of work tasks that we are behind. So it's great because it helps make us more productive. it's creating the productivity for us. But I also wanna say that there has to be a balance. somebody can get stuck in just
Joshua: Minusha, yeah.
Asha Saxena: Cleaning up and not thinking about the future, how do you create the balance that yes, clean up but also think about how you can reinvent? Clean up
Joshua: Yeah. Yep.
Asha Saxena: but also see what are the new ways you can continue progressing. So I think having the balance becomes really important when you have a tool like this. But you're absolutely right. It the the tool helps us become more productive. But where are we being more more productive? Is it some s people get
Joshua: Yeah. Totally agree.
Asha Saxena: so excited about it that it becomes a a a a toy for them. They're now building things that may not land land them to being productive because now they're spending a lot more time on just playing with the tool. So
Joshua: Exactly.
Asha Saxena: how do you really create that balance that yes, be more productive and use that time to launch so that it can get ahead?
Joshua: Yeah. I this was my my AI gut check. the moment when You get to a point where I don't have to look at emails, I don't have to look at Slack, I don't have to do all this minutia work that I used to to keep busy. I go to a task list, it's already gotten all the stuff that I need to do and it's sorted ranked to the highest priority, and now I'm starting to chew off that stuff. That gut check I think has got to happen to everybody because
Asha Saxena: Well absolutely.
Joshua: it has to be a point at which AI now takes away any sort of excuse that you have to work on, minute stuff. And you're almost gonna be forced to to go after big, bold and innovative ideas. hopefully you get there with, excitement and passion, but it's
Asha Saxena: Yeah, exactly. So let the AI do worry about your email so now you have a little extra time that you can take that little time to decompose, You can take that little time to or have a strategy meeting with your team, So really use
Joshua: Yeah, yeah.
Asha Saxena: that time effectively.
Joshua: So I I wanna shift gears now because Do you do you wanna share how you would get started? How do if you're interacting with other executives or or leaders, w what is it that you do or say or get them to implement when they're just getting started? When they just have this new world view of a digital human. That they they have a whole bunch of new frameworks in their head that they they want to, understand and implement. Where where do you get started? Where do where do you advise them?
Asha Saxena: So I always say that You create, of course you need the data layer, you need the infrastructure layer, but really think about the business problem. You want to start with something that's a quick win for you. Because if you don't show a quick win, it will be very hard for you to create the buy-in. people don't trust. And
Joshua: That's right.
Asha Saxena: technology is something that if you first you get out of
the gate and it fails, then it'd be very hard to get going. So I always say is that start with a business problem that is of value and also start with something that can create create a quick win for you. It's not gonna take, long implementation cycle, but something which is quick to show results.
Joshua: some some value to it. Yep. is there any good examples where you'll you you see a business problem come to you and you're like, yeah, but that's not that's not one that we would implement with AI in in some regard.
Asha Saxena: Always. really, a lot of time because you're so fascinated about AI, you wanna solve every single problem with AI. But that's not true.
Joshua: Right? The the the hammer and
Asha Saxena: you really wanna what you wanna think about is that what is the problem? Where can I implement a AI tool or a digital human in my book as I talk about? that is bounded repeatable, first draft language is sorting, flagging, the person can handle the real judgment call. but really you're thinking about where AI can actually come in and truly create that impact. now also th there's another caveat here because a lot of people how you define AI, anything that is analytics, that is anything that is data engineering, that is anything that is a predictive problem, everything is lumped into AI now. or
Joshua: That's right.
Asha Saxena: is it Gen AI, or is it a agent? so how do you define AI becomes really tricky in every organization and how people are having a conversation becomes really interesting. But I I I I agree with you. Is really thinking about is it a predictive problem? Do we want to do a forecasting in Prediction, or is it an age in which we are automating a work workforce, or is it actually using digital human personas? So all different categories can fall into AI, but these are different problems. And
Joshua: Yeah. Yeah.
Asha Saxena: how do you really distinguish? So I think the first step is that what problem do we want to solve? I talk about that first. Is that is that a true problem that the business wants to solve? Number two, I talk about is the data layer ready and do you have the enough data? So that you're not waiting forever to get the data ready. And number three becomes that understanding which category of that AI you're going to be using. And can that be solved? And you always
Joshua: Yeah. Yeah. There there's a
Asha Saxena: want to start, or you also say start with what you got. don't worry about what I need to go buy immediately. What you got, and can you use that to create the quick win?
Joshua: Yep. Yeah, like most executives I know are great problem solvers. they don't you don't get to where you are as a VP, S VP without being able to solve problems. I think the skill is being able to see the problems and then understanding conceptually and then in the future tangibly How AI is deployed to that benefit. Because the ability to
Asha Saxena: Exactly.
Joshua: solve a problem is huge. We we all know, business is about solving problems. However, to your point, if I don't need to use AI, if it's simple, hey, I just need to build some queries, I need to build a machine learning model, if I need to do some data management, stuff that that problem might not. might be solved without the use of AI. However, an argument could be made that, hey, what if I instead of doing these things to solve this problem, I I train my agents to do that. My agent can can write a query, my agent can build a predictive model, my agent can do XYZ. And so there's this the skill is understanding the art of the possible to the point that You can deploy your resources the most intelligent way possible, right?
Asha Saxena: Exactly. Exactly. It's it's the balance between understanding my resource allocation because AI can get very expensive. I have leaders who I'm talking to and they're like debating is like, is it truly saving us money? I'm spending more time, more money, more resources onto deploying AI. What is the value I'm getting? So there's a lot of questions around it, so very, very
Joshua: Yeah. Yeah.
Asha Saxena: important to understand what can I solve this problem with, I and I love that because I think that That gives us a starting point of understanding what can AI do and what it cannot do and what it should
Joshua: Yeah. Yeah.
Asha Saxena: do and what it should not do.
Joshua: Yeah. I spend all my time and my role at Alteryx, just the art of the possible. What what's possible, what could be done, what probably is not a great idea to to go down, what are useful useful mechanisms, use cases, that sort of thing to go down.
Asha Saxena: Exactly.
Joshua: all right. L let's get into it. I have a little fun segment I like to do on this podcast called Just Simple Rapid Fire. and it it meant to sorta get you on your toes, but also quick answers off the cuff. Are you are you okay with this? You good? All right. I
Asha Saxena: I'm ready, yes, let's do it.
Joshua: I just like to build up the tension and make it, really really
Asha Saxena: What?
Joshua: pressure. so I've got a list here, but I I want your your quick thoughts But you are free to expand on them if you feel like it. So the first one we talked about do digital human, do we start with digital human or artificial an AI agent?
Asha Saxena: to me a digital human because agent describes software. So to me agent is just a software, but digital human is a layer above where it actually has a role. And I think it's important to have a role to describe what it's going to do and what the output will be.
Joshua: Perfect. So digital human, what's the best first role for that digital human?
Asha Saxena: I would say assistant. You really want something that is low risk but has a clear ROI.
Joshua: What is one permission that you would not give it permission on on should never receive on day one? a permission that you would not give it on day one.
Asha Saxena: But yeah. You should not give the permission the ability to act without a human
checking the first output. You cannot let AI run without human in the loop.
Joshua: Yeah. Yep. one sign that you see when when implementing this that it deserves more authority, that you're you're at a point where you're like, hey, it's proven to me that it can do it. I need to give it more authority to have drag more value.
Asha Saxena: Only when the escalation rate drops below the threshold you are setting going in. So if
Joshua: Mm.
Asha Saxena: you set this is the threshold and it goes below that, then it's ready for promotion.
Joshua: Love it. That's so tangible. one sign that you you look at and you're like, yeah, no, this is we need to take away permissions. We need to take away and reduce this. What what would you think about that?
Asha Saxena: When it's catching people off guard instead of catching things they missed themselves.
Joshua: There you go. That's good one. one artifact that every digital human, digital coworker needs.
Asha Saxena: Everyone every digital coworker should have a role charter. in my book, I spoke about it in this interview. I think the role charter is really important because it's important to know what is the job description, what is the decision rights, who's the named owner. So I think that's really, really important.
Joshua: Yeah. Huge. I I this is probably in in my conversation previously around the agentic work. Telling my agents what their role is in minute detail, where they hand off, what they do, what they don't do, what they perfect and get good at has been an the exponential game I I really love it. All right, last one. One belief about digital labor that you think is gonna change this coming year within the next twelve months.
Asha Saxena: I think I think it's autonomy. Autonomy is something you grant once instead of something you keep re earning, So I think that would be something that would be evolving.
Joshua: Got it. So we've talked about a ton of stuff. By the way, good job. Your rapid fire was pretty pretty good. Some some folks have gone off of deep dives and all that and they they lost the plot, but good job. what was what is the one idea that you as we're wrapping up here that you you'd really love to just post it note to people's heads and and take away from this from this podcast?
Asha Saxena: We are entering a new era where digital humans and humans will be working together to create the output. So focusing on the output, understanding the collaboration, understanding the role of digital human and AI in your world. It is a powerful tool. Take advantage of this AI factor in your world, in your business, to create that digital human advantage so that you can continue progressing and really using the human capacity for. Something bigger and better that could be for you or for your organization. I think we're in such a great time to be able to take advantage of these technologies and to create that space like you spoke about, to find that time to be more productive and be able to create bigger, better things. So it's truly the time for human upgrade. And every single professional, no matter what age, should be curious, hungry to learn and keep evolving.
Joshua: Yeah. So where where can people follow you if they they like what you had to say today and they want to learn more?
Asha Saxena: LinkedIn is the best place for me to for everyone to
Joshua: LinkedIn.
Asha Saxena: follow me. And if you need to reach out to me, LinkedIn is the best place to reach out to me. and of course I do keynote speeches, so I'm around
Joshua: Mm-hmm.
Asha Saxena: I'm traveling around the globe. any of the conferences you'll get to meet me. Say hello to me, stop by and say hello to me. and I also
Joshua: Love it.
Asha Saxena: have ashasaxina.com where a lot of the material I talk about is on the website.
Joshua: Perfect. If we if we go to a meet you at a an event, I'm first in line. I I need to say hi. the the last thing
Asha Saxena: that's very sweet. Absolutely. I would love that.
Joshua: I'll I'll leave. You can you get this on Amazon and everywhere?
Asha Saxena: Amazon, yeah, absolutely. You can get it on Amazon pretty quickly. Kindle,
Joshua: Yeah.
Asha Saxena: Amazon, har hardcover, whatever you like. the book Digital Human Advantage on Amazon. And also my first book, The AI Factory, is also available on Amazon.
Joshua: Yeah. Yeah, I I truly recommend this. first of all, it's not five thousand pages. It's it's digestible. But it's one of these books that I will say as I'm going through it, it I'm going to have to come back to it many times over. So hopefully by the time you see me next, it's got worn corners and doesn't look so brand new. But the the the concepts in here are stuff you have to sit with. And I am regurgitating and then I'm thinking through it okay, what what about this? As I'm building out my agentic org, I come back to it and look at it and be like, Huh, yeah, I either got to the same place that you got or you helped me to get there way faster by by doing it. So really tangible stuff and and kudos for for getting this out with that, thank you so much, Asha, for for spending your time with us. I know you're busy and you're you've got a million other things going on, so for you to carve out time with me is truly appreciated. Thank you so much. My pleasure.
Asha Saxena: I love it. Thank you so much. Really appreciate it.
210 Outro: One of the things that really stuck out with me from this conversation was Asha's distinction between an AI agent and a digital human. At first, that might sound like semantics, but when you treat AI like software, you just plug it in and expect it to work. When you treat it like a new hire, or what she calls a digital human, you naturally ask much better questions. What is this role actually responsible for? What decisions is it allowed to make on day one? And what has to be escalated to a person? I also really liked her framework of moving AI from a child child. To an adult, to a sage. Because anyone who's experimented with agents knows they don't show up on day one ready to run the company. At the beginning, they're a lot like an eager intern. Your job isn't just to prompt them, it's to teach them, audit their work, and gradually let them earn more autonomy as they prove you can trust their output. And of course, there's her warning about the optimizer trap. It's so easy to use AI just to clear out your inboxes or do the exact same work a little bit faster while missing the bigger opportunity to build something new. Maybe that's a good question to leave you with today. If the AI tools you're using right now were a new team member you just hired, would they actually know what their job description is? And what would it take for them to earn a promotion? Thanks to you all for listening to Alter Everything. You can find us on YouTube and anywhere podcasts are found. I'm Michael Kisick. See you next time.