A Conversation with Workday’s New Chief AI Officer

Speaker A: You have lawful and lawless agents, rogue agents that, uh, don't know the permissions, don't know the underlying rules. You can give them exactly what they need. Rather than, uh, each user having to sort of figure that out. You can then build this more proactive experiences.

Speaker B: Welcome to the Future of Work podcast. That was Joel Hellermark, GM M for sana, and Workday's new Chief AI Officer. Joel sat down with Michelle Dawkins, Workday's VP of Solution Consulting, to define how Workday is drawing the line between lawful and lawless AI and why simply putting an AI assistant at every desk won't deliver the transformation companies expect. You'll learn how to escape the copy paste economy that keeps employees acting as the glue between disconnected systems and how to establish the underlying rails required to move agents from taking rogue actions to operating as proactive teammates. Here's Michelle and Joel.

Speaker C: Joel, thank you so much for joining us. It's lovely to see you in London. Absolutely amazing. You're excited to be here.

Speaker A: I'm missing to be here.

Speaker C: Looking forward to elevate tomorrow?

Speaker A: Yeah, I am indeed.

Speaker C: Are you a fan of the Rolling Stones? Have you noticed we're in the Rolling Stones room in the London office?

Speaker A: Not a massive Rolling Stones fan, unfortunately, but excited to be in the room.

Speaker C: Okay. All right. Um, so we are going to talk about some incredible new research that was just published about AI adoption, the usage of AI in organizations. And I'm really excited to ask you some questions about this, get your perspective. So, uh, we will jump right in, if that's okay.

Speaker A: Of course.

Speaker C: Okay. So, um, if we talk about the difference between adoption and transformation, organizations can often make mistakes around what is the difference between adoption and transformation? Um, we know that just putting an AI assistant at every desk actually isn't going to make the changes that companies are expecting. What is the change that needs to happen within organizations?

Speaker A: I think the first problem that organizations need to solve is genuinely AI adoption. When you get AI adoption, you ride the model improvements. So every time the models get better. If AI is universally embedded in all of your terms, um, your organization will instantly get that sort of intelligence, uh, increase, uh, if you like. But these systems reset the mechanics of how you use them, uh, sort of every 12 months or so now. So we started in the assistant era, and the assistant era was heavily prompt based, so we just put copilots at every desk and people were copy pasting material from any system of record and so on, trying to input it to this copilot to do work for them. What's shifting now and why this sort of pattern is resetting is that we're getting proactive agents and we're getting proactive agents that have context from these apps. So this means we're shifting from co pilots that we're constantly sort of instructing and prompting to get any value out of them to agents that proactively serve us, um, actions and insights before we even ask for that. And as they do that, when they become increasingly embedded into the systems that we use every single day with that context from for example work they can do that much more intelligently. So shifting from reactive copilots to proactive agents.

Speaker C: And that's what's really driving transformation rather than adoption.

Speaker A: Exactly. So the pattern stats that we're seeing is that from a copilot esque interface you could see small sort of 10, 20% productivity increases. But as you're moving towards agents that can do end to end work proactively, um, you can save an order of magnitude more. Um, and that's, that's, that's, that's truly the shift that we're seeing now.

Speaker C: And it's interesting because the research identified, you talked about it, this copy paste into different systems. So the copy paste economy, the research identified that people are losing seven or more hours a week on performing these tasks. That's incredible. It's an incredible loss of productivity. Um, why isn't just adding more AI the answer here?

Speaker A: I think in those cases um, you need um, agents that are fundamentally sort of context aware and that um, to my previous point there are proactive. So ah, basically what we've seen historically is um, across each one of these systems you try to go around and um, the human was sort of doing the context engineering, uh, uh, if you like, trying to at each moment for every single task, uh, sort of provide the right context to the model. And as the systems become embedded and proactive they're uh, assembling all of the right context just in time, uh, uh, to solve the task. But they're also moving from doing single step tasks to multi step tasks. So historically you might have uh, done this to generate a single email or a single document and so on. Now the systems are also becoming better at solving long horizon tasks. So what we're seeing now is that um, uh, the length of the tasks that AI can solve is doubling roughly every seven months currently. So that means we're moving from seven months ago they could um, draft a document, ah, at best to now taking out an end to end uh, process and so gathering the right Context and executing multi step workflows will then you'll get significantly more productivity out of these systems than when you had to sort of for each subtask.

Speaker C: Context Engineer the context point is really interesting. I use a running app at the moment to improve my pace. I'm very slow with my running and when I first started using it it was giving me a pace that was completely unreasonable. And I kept arguing with the AI, I can't run at pace. And then over time it's completely shifted as it's taken in my runs, it's taken in my feedback on how the, my, what my effort was and now it's actually giving me relevant pace based on my data. I just. Yeah, that context piece is really, really important.

Speaker A: Exactly.

Speaker C: Um, I did argue with it a lot in the beginning. So you've talked about um, having context, having end to end processes in the research it identifies this right and wrong approach to AI. So having AI completely separate or having it embed into your systems, what is the crucial difference there in having the layer or having it embedded into your processes?

Speaker A: So there's a few things. The first thing is um, just running on the right rails. And so you have lawful and lawless agents. Um, the lawless agents are largely the paradigm that folks are running uh, to the. It's sort of rogue agents that uh, don't know the permissions, don't know the underlying rules and so on and really struggle to execute tasks that are policy and sort of context aware. When you run it inside of the systems and with that context you can make sure that they're lawful, um, they're executing the tasks that they should be executing. The second thing is that you can bootstrap a lot of this context and you can um, give them exactly what they need to solve the tasks. And so uh, rather than uh, each user continuously having to sort of figure that out, you can then build this more proactive uh, experiences. And so rather than having a reactive experience where you're trying to go in, assemble the right context, put that into the agent to solve the task. The agent can serve you the tasks that it could support you with. Yeah, and so you're moving from this reactive approaches to an agent that is running 24, 7 effectively. Yeah, um, I think about it as you're basically getting an infinite set of you know, 150 +IQ, universally aware, um, um, coworkers in your pocket that you can um, review their work, you can review their suggestions and they're uh, sort of constantly delivering new insights or new actions or New work uh, for you?

Speaker C: Yeah, yeah. That's incredible. Um, building AI that knows the company, knows the context, knows the processes. Sounds simple in theory. What actually makes it hard and what makes it hard about having it sit within the data itself.

Speaker A: The first part is um, building uh, out the rails for this agents um, to run accurately. And uh, this is a very new uh, pattern compared to uh, the previous UI patterns. Historically we were largely uh, building software for humans. Now we're uh, building software for uh, agents. And so you need to deal with the latency requirements of agents, you need to, to deal uh, with how do you gather the right context, how do you create long and short term memory and so on. So we're basically creating a new pattern of software that is built for agents to use instead of humans to use. And so we're uh, building the right policy engines for the agents, we're building the right knowledge engines for the agents, we're building the right memory, uh, systems, uh, uh, for the agents so that they can solve these tasks ah, effectively. Um, so in that sense it's sort of a re engineering moment of how do you make this system super intuitive for agents rather than just super intuitive for humans?

Speaker C: And so if an organization is trying to do that outside of a system where the data sits, where the process sits, where the logic sits, the risk, the compliance, the cost, that all becomes a massive issue.

Speaker A: Exactly. And those systems might not have the deep uh, context of the tasks that you're trying to solve. So I think the domain and knowledge becomes super important here. We sit on the deepest domain knowledge of our tasks, our workflows and we can use that as we create the rails for decisions. But if uh, you're just executing this on an abstract level without having any insight into the specific tasks, that gets really difficult.

Speaker C: Yeah, yeah. You mentioned something earlier, human in the loop. I think it's something that we've talked about for a long time that is something that is still critically important. We are moving more towards autonomous agents that are making decisions, they are taking action, they are running processes, they're supporting and working side by side with humans. How do we keep that human in the loop concept still in place? How do we make sure that we're retaining that?

Speaker A: So I think we're basically seeing five levels of autonomy. So if you take the self driving cars analogy, we're sort of following that. So first it was this sort of highly reactive uh, systems, uh, maybe like simple auto ah, complete those sorts of tasks and then they became more prompt based and then they're becoming more proactive. And I think as you go towards the higher levels of autonomy, they're going to be more and more sort of policy bound. So we're still at the rate now where humans um, are heavily involved in this sort of final approvals. But instead of um, approving a task that would have taken you one minute to quit, ah, you can um, approve a whole body of work. But once you've been embedded in that loop long enough, the system should get an intuition for when it actually needs you uh, to loop you in. So if you think about self driving cars, it was a quite long era where humans were still sitting in the car correcting it when it did something incorrectly. Um, uh, the autonomous driving system was handing over to the human when it was unsure and so on. And that's largely where we're at today. Uh, and we have a car that drives pretty well but still makes quite a lot of mistakes, hasn't seen all of the edge cases and so on. And now um, we are basically sitting in these autonomous cars correcting them. And then at some point it will have gathered a lot of the policies, a lot of the knowledge from us sitting and correcting it in the, in the car. And at that point it will get increasingly autonomous. So I think we're at like L3 approaching L4 level autonomy, um, ah, for knowledge work. And the final sort of L4 is policy bound. And then if you think about L5 which is really the end state, you have autonomous companies um, that are sort of self improving and so um, you have uh, agents running end to end cycles and then reviewing and improving the processes, uh, as it, as it does that. But I think it's going to be a long journey of sort of L3, L4 getting that right before people will trust autonomous enterprises.

Speaker C: And the trustworthy is really interesting actually because what I was thinking then is I recently um, went in an autonomous car in San Francisco and you still have that desire as a human to take control and drive it yourself. And you're worrying is it doing the right thing? We see that. I mean the research sees that that trust piece is really important. Do you see a change in the way people interacting with AI now where we will become more trustful and not recheck the work over and over and over?

Speaker A: I think at some point uh, it will be quite the opposite where people will lack trust for uh, the human driven systems. Ah, that's certainly how I feel. I feel more safe going into an autonomous car in San Francisco than jumping in with a cab. Driver in London, despite those cab drivers being incredible drivers. Um, if you look at the error rates of the autonomous cars compared to the error rates of the cab drivers, they're um, significantly lower. And that will apply to all work. Right. And I think we're just ah, at the phase where we're starting to make m this shift, but we're certainly seeing that um, humans, um, more and more want to double check everything with an AI system. So you get something from your doctor, you want to send that to an AI system to make sure it was correct. Um, but you wouldn't purely trust the AI system in that case either. Um, so I think that's where you uh, want both, you want the sort of judgment and the intuition, uh, of the humans combined with the AI systems are incredibly unconstrained. Right. They can have all of the world's knowledge, all of your company's knowledge, all of your context. Um, they can run millions of sort of hours of thinking, um, in parallel to solve your task. So naturally they will start doing a lot less errors than humans would.

Speaker C: Yeah, yeah, I get it. You do see the transition coming and the trust. But as you said, the data, the accuracy, the guardrails, all of that needs to be in place to make sure that people do move in that direction.

Speaker A: Exactly.

Speaker C: More confident in the outcomes and um, then increase productivity in that way. So you are Workday's chief AI officer, um, as part of that role you work on bridging the gap between our potential and the return on investment that our customers see. How are you thinking about that? How are you thinking about the value that we drive with AI?

Speaker A: I think we're um, moving into, uh, entirely new sort of category of software. Um, historically we were largely selling the software solutions that you would put in the hands of uh, employees and you would see some productivity gains, ah, uh, as a function of that. And now we're building this, that is the systems that are basically enabling you to create an endless amount of uh, agents that can help you do your work and augment uh, your terms. And um, as we make that transition and we define a new category of software, working incredibly closely with our customers to define, uh, those new patterns. And the results we're seeing are truly staggering. As you move from sort of reactive copilots to proactive agents, people are moving from sitting a few hours a week, um, to doing work that would require months of uh, effort. And so we're incredibly excited about that ROI that we're seeing, uh, through these partnerships. So we love to bring together the Best AI researchers on the planet, the best designers on the planet, with our customers that are sort of pioneering the application of this to define those user experiences um, that uh, augment their teams.

Speaker C: So you're really looking at outcomes based, an outcomes based approach when it comes to defining the value that the AI that we're building is delivering. Do you look at it, uh, is it productivity? Is it um, access to data? Are there specific metrics that you have that you use to define what success looks like?

Speaker A: I think um, um there's a tendency to try to um, over measure and ah, I think uh, given the current trend lines, the list of our issues will be the ROI of AI. The issues will be do we re engineer our processes around it, do we adapt quickly enough to adopt it and so on. And why this shift is very different compared to previous shifts is that historically we used to rely on a couple of years where we could adopt these technologies and then um, a decade of sort of harvesting the value of adopting it. Now this adoption to harvesting cycles is basically yearly. We spent last year trying to just adopt these copilots and then this year we're trying to adapt to agents and so on. So we're moving from an organization that does these transformations every decade or so to having to fundamentally re engineer how we work every single year. And so what we're really focused on is how do we um, together with our customers um, really define new ways of working that can re engineer how they work every single year. But if you're stuck defining uh, your pilot and your ROI criteria, you'll be stuck doing that for the old paradigm. Um, so I believe a lot in sort of AI maximalism if you like, try everything, run it in parallel, um, and um, over invest in adoption. I think for leaders today they'd much rather have been over rotated than underrotated. And I think the risk with um, sort of AI minimalism in defining a few sets of projects, running a pilot, defining the ROI and based on that sort of defining whether you invest more is that you'll be massively under, under rotating. And so what we see is most successful is uh, companies that are exploring, that are testing um, uh a lot of different approaches and then doubling down on the projects that, that work that are delivering.

Speaker C: Yeah, it's interesting. A lot of the customers I talk to, um, it's building creativity inside their own companies as well. So giving people the opportunity to explore, come up with their ideas and then see what would be adopted, you know, at a global level or a company level is that Something you're seeing as well. Organizations really going down that route.

Speaker A: Exactly. So Ethan Mollick, who's um, you know, exceptional researcher in this field, talks about sort of the lab, the crowd and the, and the leadership. And um, your implementation really has to hit all through. Um, so first starting with the leadership, you know, every, all hands, every performance review, it's sort of universally embedded in, in, in how you, how you work and um, the, the, the leader of the company needs to be very AGI paled basically. If the leader is, is. It's not really going to trickle down from there. If the leader is sort of an AI minimalist that's taking a few sets of projects and wants to measure the ROI of every single initiative and so on, it's going to hinder a lot of that. So first the leader needs to be very um, sort of AGI pilled if you like. Uh, the second is then the lab. So that's sort of the central AI team that sets the principles, evaluates the tools, just makes AI adoption internally very easy. So can set up a lot of that internal AI tooling that a lot of other companies uh, or a lot of other teams can rely on. And then finally the crowd. And what works really well when you invest in the crowd is to find the folks in each function that have the domain and knowledge and are really sort of um, um curious and um, explore the new AI models uh, a lot. So if you combine that, you basically have ambassadors in every single time, every single team who knows the craft, knows the models and is applying it there. Um, you have the lab that is um, overseeing sort of making sure all of the teams have the right tooling and so on. And then um, you have the leaders that is sort of constantly reinforcing this um, in every aspect of the process, process. And if you get all of those three right, those companies tend to be most successful in their AI implementations.

Speaker C: Joel, we were talking about Swedish Midsummer, um, a great time of year and that uh, quite often there's a long holiday during that time. You mentioned that you tend to use that to do a bit more research or learn some new things. What are you planning for this summer?

Speaker A: I think this is a very uh, very fun era of home robotics. Uh, so I think this summer what I've been working a bit about on this, creating a uh, sort of home security system, um, and you can have this sort of always all McLeod Bots, uh, ah, over ceilings, no bad guys sort of entered the building. And so I think something like that would be a fun project for the summer.

Speaker C: This goes way beyond a robot vacuum. I like it. Thanks. I know you've got a lot going on over the next few days, but I really enjoyed the conversation. Thank you for joining the podcast.

Speaker A: Thank you so much for having me.