YouTube: https://youtu.be/0o-wn6UnNY0
Karthik Ramakrishnan [00:00:08]
Welcome to people in AI, the podcast exploring the forefront of artificial intelligence. I'm your host, Karthik Ramakrishnan, CEO and co-founder of Armilla AI. Today on the people and AI podcast, we are thrilled to have Dan Chuparkoff, a luminary in the fields of innovation, AI, and the future of teams. Very interesting, given all the conversations we're having about AI and the things that they can do. Today, we're going to explore how the future of teams would look like in the age of AI. Now, with a career spanning over three decades, Dan has held roles at Google, McKinsey and Atlassian and has been instrumental in shaping the mobile and software landscapes that influence billions of users globally. Dan is dedicated to helping teams adapt to the rapid advancements and collaboration, remote work, AI and robotics. Now that's quite the span of things and all exciting. And his insights are rooted in real world experiences with startups, to leading transformations in the world's most prestigious companies. Today, he'll share his unique perspective on the challenges modern teams face and other practical strategies for thriving in an ever evolving technology landscape that is further being transformed by machine learning and artificial intelligence. Join us as we speak with Dan about harnessing the power of change and leading the charge into the future of teamwork. Dan, thanks for joining us.
Dan Chuparkoff [00:01:34]
Karthik, thank you so much. You're too kind. I'm looking forward to the conversation today. There's so much going on. It's the most exciting time for teams and I'm excited to dig into some of the things that teams are going to need to be successful in the coming years. It's going to be a great conversation. Thanks for having me.
Karthik Ramakrishnan [00:01:51]
Now maybe we can start by with your story. How did you get on this journey? What led you to focus so intensely on innovation and team dynamics?
Dan Chuparkoff [00:02:02]
Yeah, I think it started way back. I just happened to be the exact right age. When I first started in my corporate life, it was 1992. I got a job at this company. We didn't have PCs yet. We were using those mainframe computers with the green screen terminals. And about a year and a half into my first job, we got this shipment of PCs. We got PCs for all our desks and nobody had any idea what to do with them.
We didn't know what software to use. We didn't know how to work windows. But computers were always fun for me. I had one at home and so I just helped the people around me and it made me feel good, it made me feel smart, and I helped us with this PC transition. And then five years later, the same thing happened again with the Internet. And then five years after that, we got blackberries in our pocket, and now we could use email and be in meetings at the same time. And then we got cloud, and then we got data science, and then we got remote teams, and then we got AI. For me, this is just like the 7th transition that I've been going through for 30 years, but it's going to be the biggest transition the world's ever seen.
And so it's just always been really fun for me. I'm inherently curious and excited about new things, and so I want to help people through the transition. That's how I got here and why I'm so passionate about it. And we can dig into some of the parts of that, but that's the base story.
Karthik Ramakrishnan [00:03:34]
That's amazing. Would you want to spend some time talking with some of the projects you worked on at Google and McKinsey? And then these are very large companies, and as large companies, I'm sure they're working on some really, really cool, cutting edge stuff. Why transition into what you're doing now? So maybe that piece of the journey, what did you do with these companies?
Dan Chuparkoff [00:03:55]
Right, right. So when I went to McKinsey, McKinsey was going through a little bit of a technical transformation themselves. Like McKinsey is a company that leads digital transformations all over the world. But in inside of McKinsey, they're going through change as well. And so I joined McKinsey to help them figure out how to build software that could solve harder problems than the typical consulting team with Excel and PowerPoint can solve. We started out solving healthcare problems. If we try to look at healthcare data, a year's worth of data for a single healthcare system like a state or a large hospital might have 2 billion rows of data for one year. Now, 2 billion is kind of a lot.
And if you want to look at 20 years worth of data to find trends and patterns that doesn't fit in excel. And so we started using data science to start looking for patterns in data, making sense of it all at scale, in real time. And so I joined McKinsey to help them figure out stuff like that. Once we figured out how to do that for healthcare, we could also do it for manufacturing, we could also do it for agriculture and every transition. So we built the platform inside of McKinsey that solves consulting problems in every practice. And so that was pretty fun. It was great.
Karthik Ramakrishnan [00:05:23]
So transforming the transformers, if you will, that's cool. And how do you contrast that with the type of work at Google, is it similar?
Dan Chuparkoff [00:05:33]
So Google was both the same and different from McKinsey. So the first thing is I went to Google to work on mobile related products. So if you're a mobile app developer, you need a bunch of things. You need a Google cloud for phones, you need a database, you need authentication, crash reporting, all of these things. So I worked on those platforms. If you use Android Studio or crashlytics or Firebase, you've used some of those products. But also Google, just like McKinsey did, needs to get better at making software. They need more concurrent projects going on.
At the same time, they need to figure out how to prioritize the work, collaborate as a team globally. And all of those things are things that I've been helping teams with for, you know, 29 years. So I helped Google get better. I helped the teams I worked with at Google become better, at collaborating, better at working remotely like I worked at Google during the COVID era. So many of our coworkers never met each other, and I just helped us do those things while also building the things. If you open your phone and look at the apps on your home screen, all of my software is in every one of those apps.
Karthik Ramakrishnan [00:06:53]
Probably, yeah. And these are global teams, global problems that they're working on, how these experiences shaped your. Your views on what good teamwork looks like. So, tactically, what would your advice be in terms of creating the best team dynamics?
Dan Chuparkoff [00:07:10]
Right. So I think the first thing is people sometimes assume that large teams like Google or Amazon have all these resources and they have infinite everything. And that's actually not true. The teams inside of Google or inside of McKinsey are teams just like yours. My team really had 150 people on it. That's a fairly medium sized startup. We had all the same headcount challenges and all the same resource challenges that every other team faces. One, it's easy to think the grass is always greener at those huge companies, but it's really the same.
They have all the same dynamic challenges that you have. The second thing I would say that I learned from that, especially during the remote work era, is that the. The way we share information is one of the most important things. Teams spend a lot of time in email and in meetings and in chat, exchanging information about the status of products or the strategy or the ideas that we have for what to do next. And making that more efficient is one of the most important things that people can do to make their team better and more successful.
Karthik Ramakrishnan [00:08:26]
How does leadership need to change in this environment where the tactical leadership approaches that may have worked in the past. Obviously, they also need to change. Is there a formula? Here's the three things one needs to do. What would that look like?
Dan Chuparkoff [00:08:44]
Yeah, there is a formula with respect to leadership and decision making. I think what used to be true is there was a small number of decisions that had to be made. And people could group together, have a meeting every two weeks or once a quarter, and they would bring all the propositions to the table and somebody would sit at the head of that table and make the proclamations. They would make all the decisions that they needed, and then the team would scatter and do stuff and they would come back again in a couple of weeks. But the number of decisions that we have to make as people on teams is growing exponentially every day. I have to make dozen decisions at least. And so we can't throttle our decisions to a small number of decision makers anymore. There are too many decisions for that.
And so we need a lot more autonomy on our teams. People need to know what they're in charge of, doing what they're not. One of my favorite things to ask in an interview when I'm interviewing for a job is what am I in charge of? What can I do without telling anyone? And sometimes the answer is, well, nothing. You have to tell us all the things you're working on. And so I don't go work at that place because I'm going to need a little bit of autonomy. It can be small, but you need to have thought about that a little bit. That's a really important thing for teams to think about.
Karthik Ramakrishnan [00:10:17]
You know, that's so interesting because if anything, I think I would have thought it's obvious, right? Like the best team situation is where you empower folks, you leverage the team, you really hire really smart people and you hire really smart people because you want to have them make those decisions. Are you still seeing that? Because the other way sounds very top down, right? Like there's one decision maker and the team runs off and does make executes on those decisions versus this matrix approach or even an autonomous approach where it's the collective decision making which makes the team. You make the best decisions based on that, too. So are you still seeing that being a challenge today? I would have thought that's a done deal. That's how everyone should be working.
Dan Chuparkoff [00:11:03]
It does seem like it should be obvious and should be a lot more common than it is, I think, on highly technical teams filled with a lot of developers that I think is more the case. But on a lot of teams in the world, people still are throttling decisions to a small number of decision makers. And part of that is justified. Part of it is necessary because the teams don't know enough about the strategy to make all the decisions, or they're not spending enough time learning new things to make decisions about those new things effectively. You asked if there was a formula before, and there is. There are five things that make teams more innovative. They need to spend more time learning than they did ten years ago. They need to have a sandbox for experimentation so that they could try new things without any risk.
They need to share information more effectively and more efficiently because there's more to share next week than there was last week. Then they need to make decisions with more autonomy, but only if you do those first three things. If your team isn't spending any time learning new stuff, if you're not sharing information efficiently, then people won't be able to make decisions with autonomy because they won't have the knowledge or the information necessary for making that stuff. Does that make sense?
Karthik Ramakrishnan [00:12:35]
No, totally. I think that's, you know, it seems, as you said, I think this is what I'm realizing. It seems very obvious, but there are different industries operate differently as well. So technology driven industries, perhaps they're more used to this collective decision making. And I don't sound that as a consensus based approach, but rather individuals empowered to make decisions in their local domains, in their local areas. Right. And other industries, you know, more traditional industries, perhaps, that is. That is still yet to be perceived.
Okay, so let's get to something interesting, something that you can share with us. Where, what's the most challenging, most interesting sort of transformation project that you took on? Like, and why was it interesting?
Dan Chuparkoff [00:13:25]
I think, you know, there's a number. A number of things it's hard to find. Like, the most interesting one, all in one transformation. But I'll first talk about the thing that blew my mind at Google in the first place, and that was the reason that I decided to leap and just do this full time. I was sitting in a meeting, like, I don't know, three years ago at Google, and it was an all hands meeting. There's like 600 people on the call, and that's a meeting that we had every six weeks. So 600 people making Google salaries, sitting on a meeting for 90 minutes, that's an expensive meeting. And most of the time I'm listening.
We're all working remote, so I'm multitasking on my other screen. I'm half paying attention, half not, and I start thinking about, hey, everybody else is doing that too. You know, we're having this really, really expensive meeting, and most of the people are what I call meeting spectators, right? There's a very tiny number of people actually talking, contributing to the information share, and everyone else is consuming it. And that was happening every six weeks for a little bit. And then one day I go into that meeting and there's a Google Meet AI summary, transcribing the meeting as it happens. And the transcription quality was amazing. It was perfect. It was way before ChatGPT. It was back then our thing was still called Lambda, which is now confusing because the, the Facebook one is called that. But we got perfect meeting transcriptions. And the next time this happened, we just didn't go, we just read the transcription after it happened. And we're like, okay, they're talking a little bit about the upcoming performance review process. They're welcoming the new people. There was a tiny blip about the return to office. I want to listen to that.
So I clicked it and just listened to that five minutes. It saved 600 people times 90 minutes worth of time. That's massive. Now I started looking at, okay, how much time do we spend sharing information? And there's two studies. McKinsey did a study that said most knowledge workers spend 68% of their time in meetings, email and chat. Microsoft did a similar study, same thing, 64% of our time. I think those numbers are actually probably low. It's probably 70 or 80%, right? So what that means is every day in the US, we are spending 250 million hour of time just in meetings, email and chat.
And I saw that AI could make that go away. And that was really, really, really interesting to me. And we could talk about more about where I think that's going in ten years or five years. But that was the thing that blew my mind. And so now when I go to transformation places with new clients, the first thing I tackle is the thing AI is best at is language summary, and it's the thing that people spend most time doing. And so if I just mash things together, I can shave 25% of your productivity. I can give you instantly a free extra week every month. And that's mind blowing, massively impactful humans.
Karthik Ramakrishnan [00:16:58]
The reason we have meetings, I think, is because there's something we said about the interpersonal connection that happens when you're hearing someone live versus when you're reading or even reading or listening to it offline, right? I mean, podcasts are great because, for example, why are podcasts so popular? They're offline, but you're getting the stories and you're getting the information that you need, but you're consuming it at your pace. Great. But there's something to be said about live. We still have live events, we still have live meetings, and there's a reason we still do that. How would this tool change that? Are people going to consume information differently given this sort of disconnected, asynchronous consumption of the same information?
Dan Chuparkoff [00:17:45]
Absolutely. So first, people will consume information differently. But second, I don't think in person conversation should disappear completely. I think that would be too far in the opposite direction. But one of the most important things that I prescribe for people that I'm trying to help is to go one week, a month without any meetings, without any email, and without any chat. So just completely communication dark, just do focused work at your house, at the library, in the office, wherever it's most comfortable, but uninterrupted, focused work for one week a month. All the other weeks, keep doing what you're doing, but see if you can go one week just doing work. A lot of people discover they're not actually sure what doing work means if they don't have meetings.
So the first thing is, you're absolutely right. Emotional connection, being heard, feeling valued, seeing smiles, all those things are deeply important. And so I don't think they should go away completely. But I think maybe spending 80% of our time just getting smiles and feeling heard is too much. And if we could just throttle that back a little bit by taking 25% of our time and making sure it's not spent in meetings, in email and chat, then I think you instantly become 25% more productive. Regardless of how big your time is, 25% more productive is massive.
Karthik Ramakrishnan [00:19:29]
Excellent. One comment there. I actually listened to this talk between Farhan, who's heads up engineering at Shopify, a good friend of mine, and the CTO of stripe. And in the discussion, that is what they were saying they have these weeks where there are no meetings and it's focused coding time for the engineering teams. That's all. You just are heads down, undistracted. And I think they have different terms of how they refer to those weeks. I think Shopify calls it studio week, essentially our studio time, and that you just spend your time, just don't work.
Now, the other thing I think about is the newer generation. So if you're new in your career and you're just starting off, I learned so much by just being in the room and observing. Yes. I wasn't saying much in the meeting, but I was listening to how the decisions were being made. What were the criteria that were being used to analyze a situation and then arriving at that decision, or file the objections to a decision or the discussion, the debate that happens if you're not again in the room and consuming it offline, do you still get that same effect and be folks younger in their career, how do they learn?
Dan Chuparkoff [00:20:47]
Yeah, that's a super important question, and I think we'll have to dig into that a lot more. The first thing is we're around the same age. I don't know. And I learned in that way as well. Sitting in meetings and seeing how my boss made arguments or people I looked up to, how did they persuade others? All of that stuff is super important, and it is how I learned. I would challenge a little bit whether that still happens as much when you're all on Zoom, when meetings are all remote and everybody is just a picture in a window. I don't know that you absorb enough or as much as we used to when we were all around a conference room table. But the second thing is, I think we still need really strong mentorship programs.
You're not going to learn as much by osmosis. You have to learn by deliberate mentorship relationships, deliberate shadowing. A lot more. Like in engineering, we have the notion of pair programming where you might work on something with somebody else that's more senior, or certainly there are peer reviews where you do something and somebody else looks at it. If youre not an engineer, those practices dont exist and they need to. If you're a brand new marketing person working on stuff, there should be peer reviews and peer marketing, and those things should exist and well need to make them if we want to survive this world of scattering.
Karthik Ramakrishnan [00:22:24]
Fair enough. Yeah, I think that from there I would go into now that AI is coming into the fray, I think you've given us a great example of how AI can be leveraged, at least to say productivity and efficiency in our current way of operating. But let's extrapolate, ten years from now, and agentic work being done by AI, more and more is offloaded and will be offloaded to these systems. And I have a theory of these agents, I would say of these agents being synchronized machines. So you have different agents for different things. Some will end up writing code themselves, and I think your code production is done that way. Your security systems basically doing the work that a human would do, reviewing security issues that come through in the IT systems marketing teams, getting agents, making decisions on what marketing directions to go to. I think that's where we'll end up.
A, what does the future of work look like for humans in the enterprise? And b, again, how does teamwork get impacted in these situations where there's more asynchronicity now? And B, as these agents take on more and more of the work, there's more information and more decision making. That's asynchronous. So what does the future of teamwork and interplay between humans look like? Are there less humans now? Are there more humans, and what are the humans doing, and how do they then collaborate in this new world?
Dan Chuparkoff [00:24:00]
Yes. So I think the first thing is as we start to live in an AI agent world where we have agents that are doing specific tasks, I, as a person, might be controlling 27 agents that are all doing parts of the work for me, and that, I think, will make me dramatically more productive at that point in time. One, I have an AI that's starting to get to know me, what I think about stuff I'm working on. I'm asking questions all the time to my AI assistant. I'm getting it to help articulate my thoughts, draft documents for me. I will have a master agent AI that knows me and my thoughts pretty intimately. And then I'll also have these worker AI's that are doing tasks for me. And my job as a knowledge worker in that world is to watch all my bots and make sure they're doing the right stuff, look for problems, assess the situation that already happens.
But we call those people DevOps people, right? You're just sitting in a room, and you're looking at Splunk, and every once in a while, Splunk goes, hey, this might be a problem. Can you look at it? And the person jumps in and looks at that service and figures out what's going wrong, and they restart the service, and they undo the stuff, roll it back, and start again. That's what I think the world looks like for all of us. We just become people that are monitoring our servers. And the world for 15 years, 22 years, has built an infrastructure that helps people monitor and survive. A world of working agents. It already exists. It'll just look like that weather. We'll be nurses doing that. We'll be looking at Splunk and watching all our patients and how they're doing, and we'll intervene when we need to. Or construction people watching all the construction robots. Every job, I think, becomes what a DevOps job is. Now, I think.
Karthik Ramakrishnan [00:26:09]
Interesting and so this goes back to something I've been playing with for a few years now. So probably should write this up somewhere. But I think what we are headed towards is self driving organization, right? Just like today, humans drive a car, but the car takes on more and more these decisions and takes on everything about the environment, makes the decisions of how it should get from point a to point b. And the human's job is to say, I need to get to point b, get me there in the fastest, safest possible manner. And in an organization, a similar thing is what will happen. So you've got these individual agents in these different departments, but you have a master agent that then is coordinating between those two. And asynchronicity may not be an issue because the communication that needs to happen between teams now happens between these agents. And they can actually make decisions much faster, react much more quickly.
And if they're a representation of how a human would have made the decision, because learnt how those things are, okay, we can now move faster, right? And all the humans have to do is set well, I want to maximize profits. That's it, right? That has dangers in and of itself, because we go through the nail or the widget making robot that then goes super efficient at making widgets and then unintended consequences. But without getting into that piece, I think that there's a world that's we are headed towards in terms of self driving organizations. But again, I still wonder what the role of the humans are and how many humans are there. Are we looking at the future? How is the future of work going to change entirely now?
Dan Chuparkoff [00:27:52]
Yes, so in lots of ways, for sure. But the first thing I'll address is how many humans are there working question. Because I think that's an important one, that a lot of people imagine. A world where robots and algorithms are doing our work for us, so there's nothing for us to do. I believe that there's demand for infinite work. Like if you hired an intern for the summer and they came in and did all the stuff you were planning on doing for the summer, you wouldn't just take the summer off, you would do all the other stuff that you didn't have time to do. And if you hired five interns for the summer, or ten interns, or 20 interns for the summer, you still wouldn't take the summer off because your customers keep wanting more and more from you. You could solve problems a little bit better, you could do slightly higher quality stuff.
You could produce a little bit more, a little bit faster. There is demand for infinite work. We never run out of stuff people want, because we all want infinite things. So first, I think we never run out of work. Second, when algorithms are doing this work for us, when we have these AI agents that are putting our business on autopilot, we'll have two issues. The first is, AI is trained on the existing world, so it doesn't know what we wish the world was like. So if we want a world that keeps doing the same stuff, then AI will be great at that. But one of the reasons bias is so challenging in AI is because we don't want a world like it is already.
We want a world like we wished that it was. And that's not in the training data. There are no examples of the unbiased world that isn't filled with hate. And so AI can't ever be trained in that way. So people need to pull the AI toward a place that AI can't see. And that's what I think our role as humans will be. Does that make sense?
Karthik Ramakrishnan [00:30:13]
That makes a lot of sense to me, because then the role of the humans is to ensure that the AI's are headed in a direction that we want it to go towards. And so ultimately, what you're saying is the human's infinite potential for creativity, thinking outside the box versus a machine today thinks inside the box is the ultimate definition of the machine today. Right, right. But do you think, but the one challenge, I challenge myself in that thinking, because once these machines get really, really good and we have better reasoning, that's being put in. So what's missing in which is today, but the latest version of GD 40 or 40, it has really good reasoning power? And if that's here today, in ten years from now, would a machine not be able to say, well, actually, humans, here's where we should be going. Are they going to take on this creative work too? Do you think we'll get to a world like that?
Dan Chuparkoff [00:31:15]
I think the definition of the word reasoning, I think is important in understanding the answer to that question. I think that when many people evaluate the power of AI's reasoning capabilities, what they will mean is it appears to reason at a level of human. When presented with challenges that require reasoning, however, it's only going to be able to reason scenarios that it saw somewhere else when presented with a brand new problem in which there were no examples previously. I don't know technically that there is a way to make it good at things like that. Now it will almost be good at things like that, because it'll see reasoning in this industry is kind of similar to the reasoning I need in this new industry. And so I can just infer that I can solve the problem in that similar way. AI is good at pattern recognition, not at problem solving. In if a problem keeps happening, then it's not a problem, it's a pattern, it's a process.
There's a process for when a self driving car drives down the street and it sees a tree that fell, it knows that's a pattern I've seen before. I need to stop. I need to look for an alternate navigation. That's not problem solving, that's pattern recognition. And there will be times when that is sufficient. But when presented with a brand new thing that's never been encountered before, people are good at that. And that's where people will have to interject and look for ways to solve the brand new problems in the world. And that's where creativity comes in.
And science and AI will accelerate our ability to be great problem solvers and great discoveries, specialists and great creators, but it can't replace that stuff completely. Those are inherently impossible to do with training data.
Karthik Ramakrishnan [00:33:32]
So the way these AI or machine learning machines are architected, they're based on past information. I think to your point, we can extend them further, the current architecture, to maybe reason a bit better and make those leaps of decision making in unknown situations, but there may be limits to it. And I think while I agree with that, there may be different architectures that we think about, that these machines can be infused, and maybe that's a totally different paradigm of how these machines are built and trained and how they operate. We're not there yet. Hopefully there's a breakthrough. But when that happens, I think we'll have another transformation that we had to think about.
Dan Chuparkoff [00:34:19]
Yeah, yeah, exactly. And certainly the focus of our discussion right now has been these AI agents, but there's a pending robot revolution right behind the AI one. So we'll have things, physical agents walking around helping us do things as well. And then there will be another one after that, and another one after that, and then a space race or whatever. There will be a continuous stair step of revolutions coming. And so our job is to figure out how to adapt to the new things we thought of as humans. And as long as we keep thinking of new stuff, then there will still be work for us to adapt all our existing structures. We're in this constant race of thinking of a new thing and then updating all our existing structures and then thinking of the next new thing and updating the structures again. That's where the work is. As long as we keep thinking of new stuff, there will be adapting. To do that.
Karthik Ramakrishnan [00:35:25]
I can 100% get behind. In the work that you're doing today, where the companies are today. What is the framework that you put in place for them to move with this coming world? Maybe a basic question first, are they even seeing this? Do they even recognize that this is coming?
Dan Chuparkoff [00:35:45]
I think that's one of the first questions that I ask people when I'm about to work with them is, if you have a knob, on one end there's a number one and on the other end there's the number ten. Where are you on the AI journey? Are you at ten? It's all you think about. You'd want to go full speed ahead, or are you on one where you're like, this is a hype cycle. I'm going to wait for the dust to settle. And wherever you are on that transition impacts the ways that I help you. If you're all the way on eight or nine or a ten on your AI journey, then I'm going to help you with specific things. For wrestling the information in your organization, using AI tools to manage that information, translate to other languages in real time, reach the other 94% of the people in the world that don't speak English as their native language. Those kinds of things are the things that are in the near term, things you could do today with existing tools, effectively and efficiently.
On the other end of the dial, people are like, this is a hype cycle. I don't even know what AI really means. Is it just the chat bot on our website or is it agents that talk to my customers without me involved? For those teams, I help them figure out how to learn a little bit more about what AI even is and how to try experiments so that they do things safely without risk. How to share more information about AI so that people know what's being worked on and what's not, how to add more autonomy to their teams. Those are the things that people that are not quite ready for full on AI, they need to focus on those four fundamentals first.
Karthik Ramakrishnan [00:37:35]
Got it. And if they're not ready, or if they are ready, how do you see their transformation journey from that point onwards?
Dan Chuparkoff [00:37:45]
I think, first of all, a lot of teams are already going through a transformation like this and they call it a digital transformation. And digital transformation means a lot of different things to a lot of different people. And that's why most studies show that most digital transformations fail. McKinsey, Deloitte, all of them show that digital transformations fail at like 70 or 80%, which is massive. That's because the word digital isn't specific enough. It means a lot of different transformations. It means a new mobile app, a new IT system. It means hiring more engineers.
It means a lot of things. The AI transformation will look like that as well. So the first thing I talk to people about is stop calling it an AI transformation. Let's talk about agent based research. Let's talk about conversational customer service. Let's talk about real time translation. All of those things will impact different people on different timelines. We start breaking AI into the different component parts that it that it is so that you're tackling one thing at a time.
When you're ready for it, all of your people should be asking Claude or Gemini or OpenAI some questions. Sometimes it's just the new search. Everyone should be trying that a little bit right now. But when customers are ready to do a little bit more than that, I can help them figure out what the next step looks like.
Karthik Ramakrishnan [00:39:21]
Okay. Yeah, I like that. Which is don't see or don't think of AI as monolithic thing, but rather break it into your area or these different areas and how could it be applied. And it's different answers. So it's not, again, one AI, but many different things. That is super interesting. I think looking ahead, there's the one thing about tactically, how do you implement this technology from a leadership standpoint, what personal habits or mindsets are crucial for these companies to go on that journey?
Dan Chuparkoff [00:40:01]
I think the first thing is we talked a little bit earlier in the conversation about how engineering teams somehow have a little bit more autonomy. They understand at Shopify what deep work looks like. They spend a week just coding, they take a break from meetings and all that. Helpful. One of the reasons that is possible on engineering teams is that, like, the chief operating officer of the company doesn't know how to program in Node JS. So, like, the node JS developer can do whatever they want. No one looks at it, right. The VP of whatever isn't involved in that level of decision making because it's a technical thing that they don't know how to answer.
Right. So they focus on different kinds of decision making, strategy and customer targeted and stuff like that. So the first thing leaders should think about is whether this separation exists. Right? Should you have some AI specialists on your team that are helping you make the right decisions about AI? Yes is probably the answer. If you're trying to make all the decisions about whether you use AI or not, or whether you use Claude or you use Gemini, you might not be equipped to make those decisions. And so start talking about decision making on your team and how you support smart decision making with technical specialists that are digging into the details. That's, I think, the most important thing for people because a lot of senior leaders just think, I've been here the longest, I should be able to decide stuff. But you might be the person playing around with the technology the least. And so it's time to give some of that decision making across your team.
Karthik Ramakrishnan [00:41:50]
I feel like we've come full circle from where we started about decentralized decision making and empowering teams to make those choices because you're restricted by what you know and what you don't know.
Dan Chuparkoff [00:42:05]
Right.
Karthik Ramakrishnan [00:42:06]
Excellent, Dan. This has been awesome. What would you like to talk about? Something that we did not cover and you wish we had, that you think the audience needs to hear from you?
Dan Chuparkoff [00:42:17]
I think the thing, I touched on it briefly, but I think it's an under discussed benefit of AI that I think we could dramatically improve. Right now there are about 1.6 billion people in the world that speak English natively. There's another like six and something billion people that don't speak it or speak it, but it's not their most comfortable language. AI can instantly democratize access to information in the world, right? There's what, 5.8 billion people that can't read English, and so they're blocked from learning, they're blocked from collaborating with others, and they're forced to learn English so that they can participate in the business world. And that's messed up. AI translation is good enough now that it can democratize access to your products, to the textbooks in the world, to the YouTube videos, to the learning, to the collaboration. People should be able to collaborate in the language that's most comfortable for them. And there is no technical blocker anymore to that happening.
So if you're a team and you're trying to collaborate globally or you're building things for people in the world globally, don't forget that there's like eight point something billion people in the world. And so maybe try to reach all of them with the wisdom that you have to share and the products and services that you're building. That's, I think, the most important thing to me.
Karthik Ramakrishnan [00:43:59]
That's amazing, because as we go into more and more remote work, we're having folks across the globe, we're not restricted by geography anymore, so. And we shouldn't be restricted by language anymore.
Dan Chuparkoff [00:44:12]
Right?
Karthik Ramakrishnan [00:44:13]
And that just expands the base of employees that you can have beyond just the english speaking world. I think that's a phenomenal takeaway. Thank you for sharing that, Dan.
This has been amazing. Where can people learn more about your work? The best way to engage with you or your content online?
Dan Chuparkoff [00:44:32]
Yeah, so I'm pretty active on LinkedIn. That's probably my most active platform. I'm a pretty easy guy to search on the Internet. My last name is Chu, c h u Park P a r k off o f f. You google me, you will find me. There aren't many of us in the world, so that's the best place. Connect with me on LinkedIn.
I'll keep sharing things I'm available to do, work with teams. I speak at conferences all the time, so let me know how we can help. This has been a wonderful discussion. Thank you, Karthik, for having having me. I hope I'm helpful for the audience.
Karthik Ramakrishnan [00:45:08]
Amazing. Dan, this has been awesome. Again, we've never touched on future work and I think you just opened up a whole new thread of conversation. Obviously it's top of mind for everyone. So thank you for coming on and sharing your insights and your work. But brilliant look to connecting on LinkedIn and keeping on top of everything that you have to share.