Video: Future Forward HR: Middle Managers Are Key Players in HR’s AI Evolution

Neil Jensen: Welcome to Future Forward HR, a Workday series where we explore insights shaping the future of human resources. I'm Neil Jensen, Vice President of HCM Product Strategy. Today, we're diving into HR's AI evolution and the pivotal role of managers in AI's impact. Joining me is Kathy Pham, Workday's Vice President of AI and Machine Learning. Together, we'll share insights for empowering HR and managers in adopting AI. Welcome, Kathy.

Kathy Pham: Thank you. Thanks for having me.

Jensen: It's great to have you here. I'd like to learn a little bit more about you. Can you tell me about your background?

Pham: So I am a trained computer scientist. I have been building machine learning and sentiment analysis models for almost 20 years now. It's been a while. And I think I learned pretty early on the responsibility of taking technology and text, doing some kind of analysis, and then making a prediction on the other side that people then used. And so since then, I've gone to build technologies across the public sector, the private sector, really large-scale healthcare systems, large-scale search systems, large-scale systems of record for taxes, and veterans' access to healthcare and social services. And now I'm here at Workday. I also actually teach a class on building technology that works for people and society, so I think a lot about that topic area of how do we build tech that works and solves problems for different types of people.

Jensen: Your background is really fascinating, and it actually makes me really excited to dig in and talk about this topic of artificial intelligence and the way it's going to impact human resources and, particularly, people managers. I'm curious to know, as you look at your expertise in artificial intelligence and machine learning, what shifts have you noticed in the way that organizations are starting to use this technology?

Pham: Yeah, so many shifts. So the field of AI, as we know it today, many agree it's been around since the 1950s, and there are technologies and ideas that were created then, and there were periods where there was maybe some research. Not a lot was happening. And within maybe the last decade, we've seen changes in the possibilities of compute power, large-scale computing systems, and then the technology, machine learning, AI that's on top of that. It's been the last year that we saw technologies like large language models really come to life. I think what's really exciting about that is to think about which of those problems that we have as managers or as leaders that are really well-positioned to be solved by what large language models are really, really good at, which is summarizing data, which is finding insights from that data, and using that insight to drive really interesting, better processes for the work that we do and that really help make our life easier. What are all those interesting use cases that we can think about that managers especially know deeply so well because they not only take a lot of care in managing the people they work with but also think about the business that they're in as well?

Jensen: I look at these roles today—the human resources executive, the people leader, the people manager—and they're being asked to be so much more efficient and deliver capabilities faster in the enterprise. As you think about this topic of artificial intelligence, what impact do you think that will have on the future of work, where all this goes?

Pham: So I think, right now, we're in a really exciting stage where we have a new piece of technology that's really good at consuming large amounts of data, doing summaries, finding insights, and then perhaps sometimes even generating a new piece of content. And what that means for the future of work, it has this tremendous potential to really think about our business processes, how we do some of the mundane, day-to-day tasks of empowering our employees or thinking about doing performance reviews or thinking about ways to bring in new candidates by job descriptions, etc. And I think we're in a time where it has the power to really amplify the work that we really do and go alongside and really help aid a lot of the work, especially the work that we do in HR, managing people, freeing up our time to do the hanging out and talking to each other and not just being in front of a screen and working through systems but having it now aid the part that it's really good at in the system so that we can have more human interactions.

Jensen: Yeah, I love that. I look at all of the germane and kind of boring stuff that happens in HR and I always think, "Let the machine do that stuff, right? Let's get on with the real business of what we're doing."

Pham: Exactly. Tell me more about that. We just talked about AI and the future of work. Tell me more about how you see that trickle down to managers of different levels.

Jensen: I think it's really getting these amazing digital tools in the hands of people that manage people and help them to do more. As we look at this world of human resources, we know workers leave the organization because they don't think that they have opportunity. And meanwhile, we're creating all of this content, all of these programs aimed at getting workers to see that they can grow and develop inside the enterprise. And it's really getting managers to help connect those dots. I mean, I think of something like what we have here at Workday. We call it Manager Insights Hub. It delivers insights directly to the manager without them needing to do a lot of handholding with data, compiling spreadsheets, pulling things together in the old way where you had to do everything manually. This is AI technology that delivers those insights directly to those managers. And I think it's tools like these, these digital tools that managers need in order to just be so much more effective, not spending the time on the manual, not spending the time assembling the data, but really working with those insights, making decisions, leading their teams, and doing so much more with their time.

Pham: I love that example so much because over time, we've gathered so much information in our systems, right, of documents and memos and ways to advance in your career, and they're all living in these systems that we own. And machine learning/AI is quite good at doing just that, going through all these documents, all this data, all this information, figuring out which pieces might work for different individuals, and matching that with-- let's say you want to surface a document to one of your employees, making that a lot easier. And that's such a great example.

Jensen: I don't know about you, but me personally, I've always struggled with the blank sheet of paper. I will overthink it and overthink it and overthink it before I even get a word down on the paper. And I love just the content generation nature of where this technology goes, because not that it needs to be perfect, but it gives me a starting point that allows me to be more creative, think more about what I'm attempting to write, and get to a place where I can do just so much more. So I love that these tools are coming into Workday and we're going to be giving people those types of capabilities as part of generative AI in our tooling.

Pham: I think you actually touched on something else that's really important, which is this idea of starting from a blank sheet of paper and having a piece of technology that helps you get that first draft. But the first draft is helpful when it's a good first draft. And I think that's where enterprise particularly excels because it's structured data from reliable sources that we can then now train models on that then generate that first draft built on a set of known data sets, memos, documents, whatever it is that we're training on, as opposed to unknown sources. And that allows for the ability to generate that first draft that is so powerful and so different. And there's so much to explore now still in that space of enterprise HR, enterprise finance technology, so I'm really glad you brought that up.

Jensen: Yeah, let's dig in a little bit more there because I think there's a barrier here that we need to address. And we're producing insights, we're producing recommendations, we're taking this to this new level of intelligence, but sometimes people don't always believe it. There seems to be what we might refer to as a trust gap. How do you see that playing out? What are ways that you see organizations navigating that?

Pham: Yeah, when I think about building trust, I think about building technology that works to solve problems, right? When we think about technologies that we use day-to-day, the ones that we trust the most are the ones that get us what we need when we need it, the ones that solve the problems for us and not add additional friction, etc. And so I think it's on all of us building the technologies and the managers or the leaders looking to implement the technology in our organizations to really have that curiosity to understand what the technology can do and then how they can best serve the different people in our organization and then help educate, disseminate the tools that way. I actually see it in the same way as I see introducing any new piece of technology into our system, right? Of course it's scary. It's going to disrupt our workforce. It might even disrupt it more than it helps for a little bit because there's a learning curve, and it's okay to recognize that. Change management is hard. We have systems where we bring people in to do change management for us, right? So I think having that care and empathy as the builder of technologies or the people who buy and implement technologies to recognize, "Of course it's hard. Of course it's new, and let's figure out how to bring everyone along." You're like, "This is how it can really help you make your job better," and I think there's an education and curiosity piece there that is a big component at building trust.

Jensen: So we've talked a bunch about this, right? We've got customers that are way out in front, right, experimenting with this, really leading the way. And we've got those on the opposite end that might be more cautious, be more conservative in taking their time with it. What's your advice to help everybody take this idea of AI and really get it into the business?

Pham: Yeah. So whenever there is a new piece of technology, whether it's AI, machine learning, or really anything else, of course it's going to be a lift, right? Change can be scary because especially for us, we work on really critical systems of record for finance, HR, and beyond. And I think you and I had talked about this idea of fostering curiosity. And if you're on the far end of, "Yes, let's bring AI in. We have the infrastructure. We have the talent. We have the resources. We have the [inaudible] to bring it in. Let's figure out what kind of problems we're really solving for," because you also want to buy the right tools to solve the problems that you need. And if we're a bit cautious or worried, it might be because we're in these really mission-critical systems. Maybe we're dealing with healthcare. Maybe we're dealing with social safety nets, whatever it is. Let's also understand what these systems can offer. There might be a version that really works for us. There might be a version that takes into account explainability, that takes into account transparency or interpretability, these terms that are being used in the field to just really highlight what's going on inside the systems.

Pham: Maybe they're taking into consideration security, data privacy, things that we deeply care about here at Workday. We have a whole responsible AI team focusing on that. And they're embedded throughout our design, product, engineering teams to really help build solutions that address folks from across the spectrum, the folks that are really excited about the technology, but folks that maybe, rightfully so, a bit cautious to bring something new into our organizations because we have a lot of care for our employees. We have a lot of cares for our own customers everywhere, and we want to continue to build systems. So I think my idea on thinking about that is to really have that deep sense of curiosity and then foster a sense of being able to explore that curiosity in our organizations.

Jensen: I love that. That notion of curiosity, I think, is just fantastic. And it doesn't have to be curiosity that leads you to go all the way out of the gate. It can be starting small, experimenting, trying different use cases, and just seeing the impact that it has on managers on the workforce.

Pham: Yeah, I mean, it's that idea. You can also see how our employees and the folks in our organizations respond to the systems as they're deployed as well, right? And then we can continue to iterate and change that way as well. It's great for that curiosity to drive how we move forward.

Jensen: I mean, if I'm a manager and I've now got these tools, think of how much better the world gets, right, where I don't have to sit there and agonize about writing an individual performance plan, or I don't have to agonize making sure that I've got the words right in a job I want to post for an opening with my company. What other impacts do you think managers can have?

Pham: Yeah, I think another piece, actually building on that, is-- so the way models often work is we train the models, we develop features or products, like generating these performance reviews, for example, and then as you use the products, we then retrain the models as well. And I actually think that's where managers have a really critical role in understanding the business needs, also the needs of all of our direct reports and the people in our organization on how we just use the tools. The managers are the ones in the weeds using the products, using the tools, and then that informs-- that not just teaches the rest of the organization how to use it, but it also really informs the technology and the models itself to just become better over and over time, right? These models are constantly being retrained, and the people who are using the product, especially managers, play a pivotal role in how these models are trained and used over time.

Jensen: What do you think is that next step, right? We're making progress. These technologies, these capabilities are really starting to filter their way into the workforce. What do you think the next thing is that we can do in order to help our customers, help HR leaders, managers just be more effective with all of this?

Pham: I think the next step for the next few years is that piece that we've been talking about, about pairing the capabilities of the technologies. In AI and ML, we've seen parameter sizes and context windows, all these very technical terms that grow and really fascinate at an incredible pace, to the point where lots of different companies are building these technologies, including us. We're building our own language models. But these models alone themselves are just models. And where managers and customers and partners and all users in the world have a significant part and role, I'm going to use my understanding of my business to help figure out how to build those features and use cases and ideas into tools that then make our lives much more efficient, easier so we can focus on that human interaction parts of work.

Jensen: That's great insight and a great note to end our conversation on. I want to thank my guest, Kathy Pham, for the conversation and joining me on this episode of Future Forward HR. Tune in next time as we explore more topics critical to the future of HR.