Ep 28: A Holistic Approach to AI with Former Xerox CTO and PARC CEO Naresh Shanker

Saam Motamedi: Hi there, and welcome to Enterprise Software Innovators, a show where top technology executives share how they innovate at scale. In each episode, enterprise CIOs share how they've applied exciting new technologies and what they've learned along the way. I'm Saam Motamedi, a General Partner at Greylock Partners.

Evan Reiser: And I'm Evan Reiser, the CEO and Founder of Abnormal Security. Today on the show, we’re bringing you a conversation with Naresh Shanker, former CTO of Xerox and former CEO of PARC. Xerox is a foundational computing and technology company, with over 20,000 employees and multiple spinoff companies operating at the frontier of modern technology. Before running IT and product teams for Xerox, Naresh was formerly the CIO of HP, Palm, and other notable companies. In this conversation, Naresh describes AI’s impact on enterprise software, the importance of data strategy for successful AI applications, and how technology executives can empower teams using AI solutions. Well, Naresh, first of all, it's great to chat with you again. We've known each other for a while, and every time we chat, I'm always impressed with just your personal history of innovation and technology. Maybe to kick us off, do you mind sharing a little bit about some of your experience, maybe give us a quick overview about your career so far?

Naresh Shanker: So I started my career actually in the field of healthcare, and a lot of my work was in diagnostic medicine. And the focus was really around developing products, technologies, services, and the enabling IT architectures to take solutions to market, and envelope it with all of, what I call, health care providers, in terms of doctors, patients, and the nursing care, in the field of diagnostic medicine. So we're talking ultrasound, fetal monitors, patient monitors, echocardiograms, defibrillators, that whole world of diagnostic medicine really focused around mother and child care. So that was one world and then it transitioned over into the biotech, life sciences space. And a lot of the work we did there was really around gas chromatography and liquid chromatography, blood analysis and blood analyzers, that whole world of life sciences; really exciting space. And over time I evolved into the whole world of mobility and consumer computing, developing what I call the first set of mobile phones and interfaces with the webOS operating system. And over time into more commoditized end user computing technologies, you know, data center, traditional infrastructure, and software and application; technology stacks. And then more recently into the whole world of research and development around many different spaces, from IoT and sensing technologies to 3D additive manufacturing, the whole world of advanced material sciences, healthcare, and last but not least, climate technologies. So it's been quite a fun journey and it's been quite a crazy ride as my roles evolved from being a CIO to being a CTO and eventually being a CEO where I headed one of the largest and famous research institutes around the world.

Evan Reiser: I appreciate that. Well, you've had such a storied career leading different technology teams at some of the world's like most innovative companies, right? HP, Xerox. And you've also seen the early years of many of these major, transformative technology shifts. And a lot of those transformative shifts were obvious today, in hindsight, maybe less obvious in the early years. And it feels like AI is also one of those up and coming transformative moments for technology. Some argue it's as big as a shift to cloud software, right? It's a big change in how people build and use technology. How do you think AI is currently impacting the landscape of enterprise software and how are you seeing the world from both your perspective today, but also given that long period in history of being at the forefront of innovation?

Naresh Shanker: Yeah. So the power of AI we know has been fairly transformative, it's matured over the past decade. But what's really exciting about AI is actually the next generation of AI, which we are now calling GenAI. And the beauty about GenAI is that it's expected to be much more advanced, more autonomous in nature, and also much more capable of understanding and adapting to both the human needs and context. And that's what makes generative AI so interesting. Let's talk about really its force in terms of its applicability in the world of enterprise software, and I'm going to touch on a few things that are really interesting and exciting, and not the obvious ones that we've really pursued over the past decade. I think the one space that's pretty interesting is actually the whole world of predictive maintenance, right? If you look at the world of manufacturing, transportation, logistics. The ability of these advanced models to be able to predict failures, right? And when maintenance is required and needed, how do you measure downtime, right? How do you reduce downtime? As well as the maintenance cost that goes into such heavy duty equipment, plant industries, whether it's oil, energy, gas sectors, the world of construction, for detecting material sciences and the advanced material sciences that go into the world of construction. So I think this whole space of predictive maintenance, it really ensures optimal performance and extends the lifespan of a lot of this equipment, and that translates to millions and millions of dollars, capital being saved in this specific type of industry, the whole world of predictive maintenance. So I think that's very, very powerful. I think the second piece that I thought was very interesting was pre and post-COVID, one of the biggest challenges worldwide we were faced with is this whole world of supply chain and manufacturing, because over the last few decades, we have gravitated towards the whole world of what I call centralized manufacturing, and we saw the challenges in a COVID, post-COVID world, the impacts of that. We have massive shortages in parts and components, supplies, and the lead times for manufacturing and gaining access to products and services went from days and weeks to months and years. That really crippled global economies. I think supply chain optimization-, so the first thing was around moving to a much more decentralized manufacturing and supply chain landscape, and then starting to put more sensors in place where the AI actually optimizes these operations in terms of supply demand forecasting, how do you manage inventory levels, optimize your logistics and transportation costs. Tremendous benefits, right, in that compression that we saw in terms of reducing either costs or minimizing wastage and even, overall, improving efficiency. So I think this whole world of supply chain optimization is very, very interesting.

Evan Reiser: AI is not a new thing. It's been around for a long time. It's gotten a lot easier and better probably in the last five years. And it's really much more in the spotlight with things like ChatGPT becoming just visible to everyone. Are there examples today that you've seen kind of out in the world where you feel like AI is already having a big impact, but maybe some ways that are hidden to the average person that might be surprising to the average listener?

Naresh Shanker: Yeah, I think AI is going to continuously evolve into the acceleration of GenAI and start marrying this bridge between the whole human condition as well as context. And when you start putting this together, which I think is going to get very real very soon, I think you're going to start looking at things like, for example, its applicability in the field of energy efficiency. So smart cities, smart water, smart transportation, I think these spaces are going to start accelerating at a global level. So that's one specific area that I see in advance because it's going to be a combination of IoT sensing capabilities, computing at the edge, a combination of cybersecurity. And then at the center of it is all about energy consumption and the optimization of improving the human condition and things of that sort. So I think this is going to be a very interesting space that's going to really accelerate. Okay, that is one of the spaces I'm really, really excited about. The second space I think that's going to really advance is expanding and extending healthcare around the world. And this is where, if you look at a combination of hardware and software technologies, both in the IoT spaces, as well as the manufacturability of surgical equipment, diagnostic equipment, soon there has to be the ability to be able to provide not just remote diagnostics, but also be able to perform remote surgeries to such remote parts of the world. And I think this whole space of healthcare is also going to be a pretty exciting space.

Evan Reiser: The way I think about that is almost like you're making the accessibility of knowledge much broader. And for 10 years now, people have been saying, data is the new oil, right? Seems like AI is also going to create the ability for more of that data and knowledge to be utilized and accessed and distributed. How do you think about that? There's no shortage of data, right? I think it's hard to kind of turn that into value for employees, value for customers. Do you see AI as being a bridge there or what are some of the opportunities?

Naresh Shanker: Yeah, so what you've really touched on is, frankly, the future of the culture of Innovation. When you look at the fact that we are now living the whole experience of an absolutely flat world. Where accessibility through technology is now ubiquitous and both data, the ability to learn, to have access to talent, and to have access to ideas-, it really goes back to the heart of what I call The Open Innovation Model. And that's going to become prevalent, it's going to get accelerated, and it's going to become critical in this whole new world in terms of the art of innovation. And how do you transform a lot of these new sciences and capabilities to application oriented experiences that can be very quickly monetized from a time to market and time to value perspective. So I think it's really about this whole open innovation model and that's what the whole world is going to pivot to in a much more accelerated form, which takes advantage of what you just highlighted: the access to information, the access to learning, the access to having a much more progressive talent pool, which is what's going to be required to advance a lot of these technologies in a much more, what I call, inclusive way.

Evan Reiser: So it seems like there's the data assets that different companies have been collecting over the years, right? And whether it's manufacturing data or whatever it is, that seems like those assets become more valuable with AI. Do those data sets become an advantage because you have some differentiated data to feed into your models? Or do those data sets become less valuable because they're more accessible with these general large language models? How does the rise of AI affect data strategy for an organization?

Naresh Shanker: Yeah, I look at it as two worlds. There's going to be data that's going to be, what I call, foundational to these learning-based models that can actually help advance certain capabilities that are much more commoditized. But then there's going to be a whole layer of specialization. That specialization has nuances because there's going to be foreground IP, background IP, meaning intellectual property, right? And then there's going to be all of the policies, regulatory aspects that governs that data, and then there's security and regulatory requirements. All of that has got to get layered. And I think the way we break down this challenge is to just make sure that the data can be, what I call, classified, segregated, compartmentalized in a way that it can actually serve as building blocks around what is truly a commodity to advance a set of capabilities versus what is going to be very specialized to be a differentiator in specific industries to advance specific sciences, right? It is a very complex problem that has to get solved because as you know, like you said, data is everywhere, okay? But I think we've got to start taking a hard look at how to look at this data differently and start creating classifications of this data because this data has to take into account a lot of the complications of boundaries and borders and a lot of the regulatory requirements at a regional and global level. So there's several complexities that come with this and so the folks that get ahead of this are the folks that are going to understand the issues and the implications and the ethics and regulatory concerns that have to be addressed up front and governed and ‘policy-ed’ up front before you unleash the monster. Otherwise, you're going to be fighting trying to address these on the back end. I mean, you're going to be dealing with a lot of delays from a time to market and time to value standpoint to be able to bring these solutions to market and monetize them effectively. So I think there's work that's got to be done up front to be put in place to address this.

Evan Reiser: If you can't just dump it all in the data lake, you got to actually have like a strategy.

Naresh Shanker: I wish it was as simple as that.

Evan Reiser: You've been a technology executive leader in a bunch of different companies. And I'm sure like, you know, if you and I were talking 10 years ago, your belief about data strategy is probably different than it is today. Right now that we see some of the opportunities for how data can be used, it probably affects things that maybe were less important, right? The compartmentalization, the segregation, everything's a lot more important in this world that you're describing. What would be your advice to like a new CTO coming into the role, thinking about their data strategy to best set themselves up for the future of AI capabilities? Any advice or guidance or things you think that should be, are more important than they seem or less important than they seem, what would be your kind of pro tips there?

Naresh Shanker: Yeah, so I think at least my observation of the evolution of data is that data started becoming highly proprietary in the beginning, and everything was classified inside the boundaries of the entity, whether it was a public or private sector. It was all, you know, inside the boundaries of these entities. Because it was very unclear at that time what data was actually relevant and mattered. And so the easiest path was to just lock it all up. And the challenge with that is that you tend to be very siloed inwardly focused, and you're not able to think through what are the possibilities of new advancements, new business models, monetization models for market opportunities. So when you start stepping back today, and what I learned through this cycle we went through from public companies to then the newer evolution of companies, a lot more energy went into the classification of data to really understand what data truly mattered. That takes a lot of time, effort, and intelligence to do. But once you do that right and upfront, it's much easier to go down the path of finding new opportunities to monetize and go to market more efficiently and in a much more differentiated way. The biggest challenge I discovered in the past 10, 20 years and in the next 10, 20 years I mean is the folks that are best positioned to understand how to build the right business models, monetization, and go-to-market models are the folks that best understand the data that's relevant to them. And that is having a much deeper understanding of the business and less about technology. And when you understand investors, your customers, your partners, your employees in terms of what really matters, and as they learn and have a deep, deep understanding of the business and the opportunities in terms of how do you bridge and take advantage of what's available in the marketplace with an open mind, I think you can start bringing these worlds together and I think the open innovation model applies not just to traditional companies, but also to new companies. Because it's this open innovation model where you actually, very deeply, understand what's going on in academia, what's going on in research institutes, what's going on in startups, what's going on in other public and private sector entities, right? This broader understanding gets you better exposed to better ideas and opportunities in terms of designing new business models. So I think you've got to first desegregate your data into these classification models and then you've got to start looking at this open innovation model so that you can start thinking through in terms of how do you bring together what's absolutely relevant and what's going to matter, and what you will find out is that a lot less of the data actually matters to design what you need to go forward versus what is really commoditized. That was a huge learning that we discovered going through this process and it takes a lot of effort to actually deep dive into this, to actually develop those models.

Evan Reiser: One thing I've always been impressed with you is your balance of thinking about technology and business and the customer experience, right? People tend to focus only on one of those. And a lot of times when people talk about the impact of AI, they get very focused on efficiency and kind of optimizing the current thing. When I hear you talk about it, right, you're talking about, “hey, here's some of the new customer experience we can create. Here's new business models. Here's new revenue streams”, right? And it's less around, kind of, digitizing the existing business process. That's more about digital transformation. How do you kind of reimagine what's possible, right, as a business and for customers? Are there examples you've seen in the world today where you feel like AI capabilities have opened up new opportunities or use-cases where you feel like, hey, because of how they thought about data and AI, it's unlocked a whole new thing for them that couldn't have happened if they just incrementally improved the old thing.

Naresh Shanker: Yeah. So, I mean, let's take the world of autonomous engineering. It's amazing what a lot of these new companies have done with data. All of our cars today, the intelligence of our experience and how we interact with the machine, all of that data is now available for automotive companies to design the right experience and the right capabilities that are very quickly going into the next generation automotive manufacturers. So you're going to hear about cars that are actually designing a significant number of safety systems in their windshields to detect stress and strain, and things of that sort, right? I think it's very powerful because I think the whole world of automotive, which is also applying to aviation and aeronautical systems, is going to take care of passengers in a very, very advanced and productive way like never before. And so I think it's very, very powerful. I think that the whole world has the same thing with transportation and logistics. How do you optimize at a global level the shipping of goods and services? And how do you optimize in real time what is the right path globally,  air, ship, rail, to get things to you in the fastest time possible at the lowest cost? And how do you consolidate shipping lines? How do you consolidate transportation lines, rail, auto? I mean, it's amazing. So I think this whole world of autonomous engineering, very advanced, and it's going to continue to accelerate. But guess what? It's consuming a lot of personal data, a lot of human data. And before you know it, the next generation of cars coming out by the same alternative manufacturers already know the kinds of user experiences you're looking for. Okay, so I find that whole world of autonomous engineering very powerful because it's applicable across so many industries.

Evan Reiser: I just find this so just exciting as like a human that when you think about today, like a lot of these contributions that people make, especially like an enterprise software business, right? Whether you're building software or you're doing different types of analytics, today requires a lot of knowledge and technical skills. Like even answering a question, a sales forecasting question, you have some expertise and you have some business acumen, you understand Salesforce or whatever your tool is, you can imagine a future with the AI co-pilots, the number of people that can now perform that job and do an even better job than what we can do today, that gets broader. And so now more people can do more things and just the accessibility for people to access data, make big decisions, abstract their work to things that may be less mundane and more interesting and exciting. It's an exciting future for work, I suppose.

Naresh Shanker: So what's interesting is that today people have highly specialized roles and highly cross-functional roles and you can envision a day when a lot of these roles are going to be interchangeable because the data and the AI is going to basically power up people to be very, very intelligent cross-functionally and not have to be necessarily so specialized. So I think the future is going to be fungible intelligence and transferable intelligence. Yes, there will always be a need for certain specialization, but, eventually, it's that specialization that translates to fungibility. And the fungibility of experience translates to making broader and better decisions faster and much more economically over time. So I think there's going to be this shift, which is going to be so interesting in terms of how roles get developed in the future and how people get trained and educated in the future.

Evan Reiser: Yeah, you can almost imagine, I don't know when the time arises, but some future world where it's almost like every employee has a bachelor's degree in every subject. They may not have all the wisdom and experience and judgment, but at least the foundational knowledge and information can be accessed very easily. And like if you have an entire global workforce and civilization with those abilities accessible to them over time, you can just go do more great things.

Naresh Shanker: I think all degrees will have some element of data sciences, right?

Evan Reiser: They're going to have to, yeah.

Naresh Shanker: I mean, data science is going to be the foundation for all of these degrees and the ability to discern patterns and to be able to drive decisions and outcomes is going to be so different. And that makes it super exciting, which means technology is going to get very, very advanced very soon, at an accelerated pace.

Evan Reiser: I mean, that's kind of the nature of technology we've seen for a while, especially in the last, even six months, right? There's more advancement in the last six months in AI than maybe the 18 months before that. And so we’re on this exponential curve and the future's coming very, very fast.

Naresh Shanker: It's almost like the future is yesterday.

Evan Reiser: Yes, I feel like that quite regularly. You're talking about the transformation of business models, products, services, and what kind of suppresses innovation sometimes, kind of this incremental, like, hey, we're going to do the old thing a little bit better. And some of the biggest jumps, right, so people kind of reinvent or they really think about, “hey, what is a better way to go do this from scratch?” You've been in some just, you know, companies that are iconic for innovation, right? Thinking outside the box. And there's many technologies that are complex today that would not be around without some of the work and innovation of some of the companies you've been a part of. I'd love to hear any thoughts you have from a kind of cultural leadership perspective, right? As you're kind of leading teams, right, there's behaviors you want to reinforce to encourage people to think outside the box, to experiment with new things. Any kind of guidance or principles you'd share about what you've seen be effective in creating innovation, you know, culturally with inside teams?

Naresh Shanker: So I'm going to touch on a handful of things that may or may not be obvious, but at least have been a set of guiding principles that I have used with a lot of my peers, colleagues, friends, as we evolved and built companies and transformed them over time. I think the first one I had alluded to earlier is that I'm a strong proponent of the open innovation model. I mean, these models that involve very, very deep partnerships with external organizations, startups, academia, will become more and more prevalent. And businesses will start tapping into this global pool of ideas and expertise to drive innovation, and I find this is one of the most powerful models to help accelerate the whole innovation mindset. Another set of principles we've applied is always doing more with less. So this whole notion of applying agile and lean approaches and principles, super critical. Because in the world we live in, which in the past has been very, very software-driven, but as we go forward, I believe strongly we're going to see more deep tech. Where there's going to be a combination of sensing technologies coupled with very strong AI, GenAI capabilities all on the look by software, right, and software experiences. That pivot that's going to happen is going to be very critical, so putting in place the right frameworks that can take advantage of both these agile and neat frameworks that can take advantage of both hardware and software ecosystems so that we can iterate quickly, learn from failures and adapt to these changing market conditions globally is going to be super critical. And then the last one that comes to mind that I found very valuable was having an organization, team, and workforce that's continuously learning and upscaling its capabilities, they thrive on that. And to stay competitive, to stay relevant and to be able to really apply yourself with an open mind, especially which is what innovation is all about, you have to make continuous learning and upscaling an absolute priority. So lifelong learning has to be in the lifeblood of innovators.

Evan Reiser: Before we go, I want to have a quick lightning round and we’re just looking for your one tweet sound bites for these. So how do you think a company should measure the success of a CIO or CTO?

Naresh Shanker: I would say that the measurement system for a CIO should be no different than the CEO, the CFO, the COO, or the CTO. Because at the end of the day, what your investors and shareholders care about is how do customers measure your success? And at a very fundamental level, it really is about, are you producing the greatest products with the greatest experiences for your customers? And whether it's measuring net promoter scores, lifetime value, there are so many amazing metrics in terms of brand loyalty and things like that. So I think that's the first thing. The customer's gotta be front and center to the extent you can meet a customer experience front and center and have a greater than 90% retention rate of your customers, right? I mean, it's phenomenal. So I think customer-centricity is super critical.

Evan Reiser: What advice do you have today, or that you wish you could have given yourself back then?

Naresh Shanker: I think the first thing is to really surround yourself with a group of trusted advisors early in your career, and even when you're moving from company-to-company. And this could comprise of people that are friends, both professional, they could be personal in nature, but a trusted group of advisors that have broad diverse experiences that can actually help you navigate through different challenges over time. And people that bring an outside-in view of the world and a much more open view of the world in terms of how we solve the problems. And so that's important. So I call it having your own personal board of directors to guide you. The second thing is team matters. Bringing together a team, right people, right place, right time matters. And the collaborative nature of the team is so important. You may not hire the brightest person on the planet if they can only operate in an isolated fashion. And so I think bringing together a team that gels and that brings together broader views and expertise that can really cross-functionally play as a team is very important. Early in my career, I focused a lot on bringing in highly specialized people in different roles. Extremely hard to put the jigsaw together. So I think getting a team together up front with the same cultural values and norms, mission, vision in mind, and people that are trusting, people that know how to work well together, and that are smart, I mean, really matters. Team matters.

Evan Reiser: This may be more of the personal side, but is there a book you've read recently that's had an impact on you, right, or your leadership? And if so, I'd love to hear which book so I can go buy it.

Naresh Shanker: The book that comes to mind for me more recently that I just completed is a book called From Strength to Strength and it's by Arthur Brooks. I think it's a beautiful book and it really explores the concept of finding purpose and meaning and fulfillment in life. Actually, how do you transition from your first stage of life as a professional, very focused in career and growth and opportunities, to the second phase of life, which is really about your own personal growth and spiritualism and having a fulfilling life, leveraging your strengths and virtues from your prior experiences. So it's really that pivoted shift from your professional life and everything that you lived and breathe professionally and what mattered then to more enhancing your personal life, that whole shift. Great book and I think it's an enjoyable read.

Evan Reiser: Naresh, thank you so much for joining us today. As always, I really enjoyed speaking with you and super excited for us to chat again soon.

Naresh Shanker: Good to see you, Evan, and thank you so much for your time. Appreciate the opportunity. Bye now.

Evan Reiser: That was Naresh Shanker, former CTO of Xerox and former CEO of PARC.

‍

Saam Motamedi: Thanks for listening to the Enterprise Software Innovators podcast. I'm Saam Motamedi, a General Partner at Greylock Partners.

‍

Evan Reiser: And I'm Evan Reiser, the CEO and Founder of Abnormal Security. Please be sure to subscribe so you never miss an episode. You can find more great lessons from technology leaders and other enterprise software experts at enterprisesoftware.blog.

‍

Saam Motamedi: This show is produced by Luke Reiser and Josh Meer. See you next time.

‍