AI and the Future of Work: Artificial Intelligence in the Workplace, Business, Ethics, HR, and IT for AI Enthusiasts, Leaders and Academics
00:08:41 · Dan Turchin, Max Wessel
[00:08:41] S1: Some looking in from the outside would look at our roles as product people and say, must be easy because AI does all the work. It writes the code and expects the features and it even recommends what you should be doing more of. That's actually not the case. Talk about what's hard about your job and maybe even what's harder than it was in the past. [00:09:02] S4: There's a great book that was written, I don't know, twenty years ago. That was in the early days of the YouTube moment. And it was about what is creation look like for video in an era where anyone has the tools to build new content. And it was that contrarian position. The author, a woman named Anita Elberse, basically said, in an era where there is completely abundant long tail content, what we're gonna see is gravity to the things that are high quality. And I actually think we lost sight for a while there, that part of the act of product management is around understanding the problem, having the taste and judgment and the accountability to take a decision when it's very easy to generate code. It's very easy to create slap. And so more of the job is what product was always about, which is understanding what real problem are you solving in the world? How do you make something easy and delightful and intuitive for a human that interacts with it? And how do you focus on that very simple interaction making it real?
00:09:02 · Max Wessel, Dan Turchin
[00:09:02] S4: There's a great book that was written, I don't know, twenty years ago. That was in the early days of the YouTube moment. And it was about what is creation look like for video in an era where anyone has the tools to build new content. And it was that contrarian position. The author, a woman named Anita Elberse, basically said, in an era where there is completely abundant long tail content, what we're gonna see is gravity to the things that are high quality. And I actually think we lost sight for a while there, that part of the act of product management is around understanding the problem, having the taste and judgment and the accountability to take a decision when it's very easy to generate code. It's very easy to create slap. And so more of the job is what product was always about, which is understanding what real problem are you solving in the world? How do you make something easy and delightful and intuitive for a human that interacts with it? And how do you focus on that very simple interaction making it real? [00:10:16] S1: One of the things that I thought was most elegant about some of the new interfaces is that it felt like the right amount of UI, like the the the the right amount of conversational AI balance with the right amount of... I don't know if it's a traditional, but classic forms and classic UI. How do you think about the way to build agentic workflows and conversational AI in it... Right point
00:10:16 · Dan Turchin, Max Wessel
[00:10:16] S1: One of the things that I thought was most elegant about some of the new interfaces is that it felt like the right amount of UI, like the the the the right amount of conversational AI balance with the right amount of... I don't know if it's a traditional, but classic forms and classic UI. How do you think about the way to build agentic workflows and conversational AI in it... Right point [00:10:44] S4: in the workflow? So the cool thing about that UI that you saw today, it's it's all generative UI. And so we actually use the interaction with the agents in order to figure out what UI needs to be generated. One of the examples from today, we released a product called Workday Decision Intelligence. Workday Decision Intelligence is an agentic analytics tool. You could chat your way through data, but it's easier for humans to consume data in charts. And when somebody asks for a chart to be created and it comes out the way they want, in decision intelligence, you can save that as a playbook. And the next time you ask the question, when the agent goes through and does the analysis, understands the data relationships, and comes back with an answer, it'll be... Come back in that chart updated. And so we think a lot about how do we connect this incredible alien technology that is AI with all of the tools that we need to translate it into the way we consume information. You said something profound, more profound than I
00:10:44 · Max Wessel, Dan Turchin
[00:10:44] S4: in the workflow? So the cool thing about that UI that you saw today, it's it's all generative UI. And so we actually use the interaction with the agents in order to figure out what UI needs to be generated. One of the examples from today, we released a product called Workday Decision Intelligence. Workday Decision Intelligence is an agentic analytics tool. You could chat your way through data, but it's easier for humans to consume data in charts. And when somebody asks for a chart to be created and it comes out the way they want, in decision intelligence, you can save that as a playbook. And the next time you ask the question, when the agent goes through and does the analysis, understands the data relationships, and comes back with an answer, it'll be... Come back in that chart updated. And so we think a lot about how do we connect this incredible alien technology that is AI with all of the tools that we need to translate it into the way we consume information. You said something profound, more profound than I [00:11:57] S1: think you realized that I think will define kind of the next maybe decade of how humans interact with with machines, the term generative UI. So you kinda tossed it out and then went on. But take a step back. What does that mean at Workday, and how do you think about the role of generative UI?
00:12:14 · Max Wessel
[00:12:14] S4: So generative UI is UI that is created on the fly by an agent. And there are a lot of emerging standards around generative UI, and we're still very early in the process of building out all of the frameworks, all of the tools to be able to give agents the ability to speak in the way we consume information. UI is wonderful. I don't want to chat to my ecommerce sites, because the way they are laid out beautifully with pictures allows me to figure out exactly what I wanna buy, the way that I wanna buy it. It is a experience designed for a human. Getting a code full data dump in a table is not the same. Even though an agent might speak that language, we need that translation layer. And so when it comes to the work that Workday is doing right now, a lot of our focus is how do we build the generative UI components? How do we create the interaction system? How do we coach the agents to use those components to speak to humans? And what you get on the other side is things like forms and charts and workflow tools that feel like the types of systems we are all used to interacting with, but they're created flexibly on the fly by the agents. And so we have a whole team at Workday that focuses on creating those elements for our our Asana agent to use as it interacts with users.
00:13:55 · S2, Dan Turchin, Gerrit Kazmaier
[00:13:55] S2: Finally, we turn to Gerrit Kazmaier, president of product and technology at Workday. Gerrit shares Workday's vision for using AI to restore humanness to work, starting with why nobody comes home proud of answering 50 emails and how companies can decide what AI should do and what should stay human. He also describes the future he hopes for five years from now when more people can spend their days doing work that inspires them. [00:14:25] S1: Talk to us about the... Call it the why behind the what. You introduced a ton of innovation. [00:14:31] S5: Mhmm. [00:14:31] S1: But I think there was a theme beneath some of that innovation. Share with us kind of your philosophy around how you decide what to build.