Meet Your New Teammate, AI: Asana’s Saket Srivastava

Hey listeners, when it comes to AI, it's hard to decipher what's real, what's hype, and what you need to know. To help make sense of it all, we recommend you listen to the TED AI Show. Host, creative technologist, and AI innovator Bilal Sadu sits down with guests for a deep dive into the thrilling world of tech and its effect on business, education, and more. Each episode will leave you more informed with tools to help you navigate the future of tech. Get ahead of the curve and check out the TED AI Show wherever you get your podcasts.

Stay tuned after today's episode as Shervin and I talk about the changing role of the CIO. How would project management and collaboration be done in the future? Find out on today's episode.

Welcome to Me, Myself, and AI, a podcast on artificial intelligence and business. Each episode, we introduce you to someone innovating with AI. I'm Sam Ransbotham, professor of analytics at Boston College. I'm also the AI and business strategy guest editor at MIT Sloan Management Review.

And I'm Shervin Khodabandeh, senior partner with BCG and one of the leaders of our AI business. Together, MIT SMR and BCG have been researching and publishing on AI since 2017, interviewing hundreds of practitioners and surveying thousands of companies on what it takes to build, deploy, and scale AI capabilities and really transform the way organizations operate.

Hi everyone, welcome to season 10 of Me, Myself, and AI. Today, Sam and I are super happy to be joined by Saket Srivastava, Chief Information Officer at Asana. Saket, welcome to the show.

Maybe I'll start by asking you to describe Asana as a company and then your role within that.

And then your role within that as a Chief Information Officer, I've got to believe AI plays a significant role in what you do and what Asana does.

Wonderful. So AI has been around at some scale at over a decade now, but I would imagine, particularly in the last couple of years with generative AI and all this sort of extra focus and buzz around it, what are you seeing in the marketplace and what is different now versus let's say five years ago?

Well, I'd love to hear some of those examples. What kind of AI initiatives are showing up in your products?

Hey listeners, when it comes to AI, it's hard to decipher what's real, what's hype, and what you need to know. To help make sense of it all, we recommend you listen to the TED AI Show. Host, creative technologist, and AI innovator Bilal Sadu sits down with guests for a deep dive into the thrilling world of tech and its effect on business, education, and more. Each episode will leave you more informed with tools to help you navigate the future of tech. Get ahead of the curve and check out the TED AI Show wherever you get your podcasts.

The theme I'm hearing, by which I'm quite enthused, is you're talking about AI as a coworker, as a teammate, as a pseudo-person. And you're talking about it taking a fair amount of not just predictive work, which is what AI used to do and it does quite well, but really cognitive work. And you know, Sam, contrasting this with like, you know, how this conversation would have gone two, three years ago, is we would have gone right down the graphs. We have a lot of data, it's so much for any human to be able to understand it, we need to predict who will do what when, etc. And I'm not saying those kinds of problems have gone away. I mean, they're quite relevant and we need to calculate risks and probability and propensities and all that kind of stuff which predictive AI does well. But this notion of AI not just as a tool to optimize but as a coworker, as a pseudo-entity with which you can have conversations and adapt and get recommendations and maybe even clarity as you're talking about, that's actually, I think, is quite refreshing. And you know, Sam, we talked about this, I would say in '22, back with our report back then, where we were just suggesting that organizations should think of AI in terms of the roles it would play, not just what it would automate and how fast, but you know, maybe AI as an illuminator or as a recommender. And it's really refreshing to see how generative AI is making that possible, that AI is not just an optimizer but also a coworker.

Well, anecdotally, I could share that in my work at least over the last sort of ten years, I've seen a gradual evolution of the CIO role from running IT services and IT processes to much more a sort of driving the company through innovation, being a partner to the business in terms of that. But I would assume you maybe you've seen this also in your work and the report that probably maybe in the last couple of years there's been really an inflection point even in what CIOs are being asked to do with all of this sort of new technology that's becoming so prevalent and how to control it, how to de-risk it, how to get value from it. What are you seeing?

Yeah, sure. In our research three years ago, we found that people who organized AI functions underneath the CIO tend to underperform places that organized at a higher level. We should ask that question again because I bet what you're saying is that some of that has changed over those years. All this seems like a pressure to do AI things, but one of the things from your State of the IT Leader report was that one in four people felt like they were investing in AI too early. What's behind that? Everything I hear is, oh my gosh, oh my gosh, we got to do much more AI. What's the regret?

Yeah, and I think also part of this, and maybe part of that 25% that are saying maybe we went too soon and too fast, is the fact that this journey that you're talking about, Saket, is not just a technology journey, right? I mean, organizations have to change and have to adapt and roles have to change. And going back to your comment earlier about, you know, AI as a teammate and a coworker, I think that's the big change, right? Whereas in the past, if AI is going to be able to do risk scoring better or propensity calculations better or, you know, personalization of offers and optimization, those kinds of things better, then that just, that is super precise prediction or optimization that basically allows better decision-making and does that which humans cannot do, cannot do effectively. Now you're talking about AI and GenAI as a teammate and a coworker, which means I have a new apprentice next to me and I have to work with it. And if I work well with it, then my job will improve and the apprentice's job will improve too.

Yeah, you know, Shervin, this actually reminds me of how our work back in 2019, which we talked about the producers of AI and the consumers of AI, right? And so the point back then was, look, part of the reason why, I mean, this is five years ago, right? But like, part of the reason companies aren't getting value is because the production and the consumption are not aligned. Because like, people make tools and then the consumption of it requires a fair amount of process reengineering and training and change, and it doesn't happen. Back then, that consumption wasn't being done en masse. That consumption was by few people in the organization, super users, the inventory management group or like the pricing team, but it wasn't every employee. And I think what Saket is saying is in today's paradigm, it is actually most employees, they will have some version of AI in their day-to-day life, whether it is to improve their own proficiency and effectiveness or productivity on daily activities or even to come up with new ideas and innovation.

He said that there's a gap between like the ambition and reality. I would imagine that's also a cause for inaction or maybe some level of paralysis as to like, well, what are we going to do about all this? And then that's where I think leadership has such an important role to play to guide direction and show the path, right?

Shervin, I like the idea you bringing up that, you know, 2019, we thought about this consumption, production, but I think one thing that Saket's talking about is the diversity in the people consuming. You know, this survey was about trying to understand where individual people are. You know, if I look back on what we did before, maybe we were too broad in saying, oh yeah, there's production and consumption, when in fact there's lots of different types of consumption going on. And it looks like the work that you're doing, Saket, is getting at where people are and where they need to be, and understanding that that's, gosh, different for all sorts of people. Maybe just switch a little bit here, you know, go back in my life, I used to work for the Atomic Energy Agency, and we wanted to share information about how people screwed up nuclear reactors so that everybody wouldn't screw them up. That's a long-winded way of saying you have other CIOs that you're working with, and Saket, one of the things I thought was most interesting about your story was how active you were in trying to work with other people and understand how other CIOs are learning, doing things right, doing things wrong. Tell us more about that. How does that work?

Some, that's I think what we think about this show as, you know, we, Shervin and I think about, hey, what would everybody like to know? What are some people doing that other people can learn from? And it sounds like your CIO community is doing a lot of that.

I wanted to ask you your opinion, what do you think is the reason for the gap you talked about in the early part of the conversation, right? The gap between sort of ambition, aspiration, and reality. Like what's fundamental about that gap?

Don't you feel that there is also a double-edged sword here that is because there is so much hype around generative AI, a lot of it is true and a lot of it isn't, or that a lot of other steps are taken or needed to get to that value potential that people might be underestimating. But because of all of this, do you feel like some of the more garden-variety use cases and opportunities with predictive AI that's been tried and true for over a decade are also put on hold because people are saying, well, let's see, let's just wait because maybe there's a silver bullet that's going to come and like fix everything?

Actually, I want to get a little bit of your background. I mean, how did you get interested in this? I know you have a background at GE and IBM and Fujitsu and Square. How did you get interested in all this?

CrowdStrike, as it happened.

Saket, we'd like to end our episodes with a series of rapid-fire questions. So these are intended to be just the first thing that comes to your mind. You can answer quickly, short answers are fine, preferred even. Okay. What do you say is the biggest opportunity for AI right now to really make it more mainstream, have it become that teammate that we're talking about, be someone that a human can naturally look at as a teammate and lean on for advice and for action?

Watching machines work. What was the first career you wanted?

When do we have too much AI?

What's one thing you wish that AI could do that it can't currently do?

Hey Shervin, that was pretty interesting talking to Saket. I mean, one thing he kept talking about was the evolving role of the CIO. I'm thinking back on the days of the mainframe printers with green screen computers and overnight reports. This is a very different world for a CIO from 30, 40 years ago, but even different world from just two or three years ago with the transition to remote work or whatever. Yeah, you talk to a lot of CIOs, what are you seeing today?

What I'm seeing in many companies is leading CIOs are taking a much more forceful and vocal ownership of the technology agenda. And they're doing this in a number of ways. One is with AI and now generative AI, they're becoming the leaders in the organization that helps everyone make sense of this, right? With the proliferation of all the tools and technologies and all that, and all of the safety and cybersecurity and responsible AI concerns that go around that, right? That's one role we've seen CIOs play. But even thinking beyond that, I think what CIOs are increasingly doing and should be doing is taking the view of evolving technology landscape and helping align that with the corporate vision of the company and bringing all these things together. A role I see a lot of CIOs play is become really almost a fiduciary in sort of figuring out in the soup of technology what are some of the no-regret moves you make now, what do you wait for, and how do you still advance the business agenda without having a hundred different solutions in a hundred different places and a massive amount of tech debt that then down the line they need to bring together.

Yeah, that seems like a huge problem because, and I think we see that with the advent of generative too, that everybody's looking to do something, do something quickly. It's changing quickly, and if you're not careful, you get a dozen different standards that evolve over the last two years and then suddenly you got dozens of standards out there, and that's not a standard.

Yeah, that's for sure. The other thing that resonated with me was how they're using AI and this notion of, you know, he called it a teammate in the context of the platform and how AI and generative AI are working to take a fair amount of thinking and cognitive work load away from humans, right? Because when I think about AI ten years ago, five years ago, even when we wrote our first paper together back in 2019, it was a tool. It was a tool that would predict better or optimize better and do things that humans could not do, right? I mean, the argument was billions of data points, all this complexity, all these variables, too much for people, how in the world is a human able to do that? That enter AI can do it. That's still going on and that's still very valuable. We're now talking about, and that's what Saket was talking about, and we're seeing this theme come up in a lot of our conversations, is that look, AI is actually beginning to do that which humans are doing as well or sometimes better. And that is some things with reasoning, some things with scheduling and organizing, some things with just cognitive functions in terms of writing and summarizing and editing and those kinds of things. And of course, that's great, it's still a tool, but the profound implications on workforce, if you think about it like three, four, five years from now, what skills are you going to need to hire, how many, and for what kind of roles, and how would those roles evolve? I think that's a profound decision that many companies are beginning to tackle. And across all functions, right? Like whether it's customer servicing or marketing or underwriting or whatever, it's of course a question of, okay, how can AI help you do stuff better and with more accuracy and more efficiency? But then if you fast forward three, four, five years from now, there is no future where generative AI is not playing a very significant role in all of those functions. So then who do you hire, how many, in what roles, and how does the nature of work change?

Next time, we'll speak with Alexandra Salah, senior director of artificial intelligence and data science at Shutterstock. Please join us. Thanks for listening to Me, Myself, and AI. We believe, like you, that the conversation about AI implementation doesn't start and stop with this podcast. That's why we've created a group on LinkedIn specifically for listeners like you. It's called AI for Leaders, and if you join us, you can chat with show creators and hosts, ask your own questions, share your insights, and gain access to valuable resources about AI implementation from MIT SMR and BCG. You can access it by visiting MIT SMR.

That explores cutting-edge research at the intersection of business and tech. Check out MIT Sloan Management Review's podcast from the MIT Computer Science and AI Laboratory. They offer an understandable view of what's around the corner, from how to spot AI hype to what keeps cybersecurity experts awake at night. Subscribe to the MIT Sloan Management Review podcast and stay ahead of the curve.