Episode Timestamps
0:00 – Introduction & Guest Introduction
3:43 – Jared discusses his interests outside work and the foundation of Daito Design
7:26 – Discussion on the integration of AI in business and UX
11:53 – Challenges and opportunities in UX with AI advancements
19:17 – Future perspectives on UX and AI integration
27:32 – Reflections on technological changes and the UX profession
34:53 – Envisioning the role of UX in a rapidly changing landscape
41:28 – Embracing Change in the AI Landscape
48:00 – Final thoughts & Advice
Join the Conversation Before Everyone Else!
Dive into user research, product strategy, and design with industry leaders. New insights drop every month. Don’t miss out.
Main Insights From The Conversation
“With 90% of the code these days being written by AI, there’s obviously a huge downward pressure on having legions of engineers banging away on keyboards hard-coding out these things, right? And as agents become more and more resilient and independent, less and less engineering is going to be done”.
For Jared Huke, founder and CEO of Daito Design—a UX consultancy specializing in enterprise and industrial transformation projects—this isn’t speculation about a distant future. It’s the present reality reshaping how his Amsterdam-based team designs experiences for energy companies, nuclear facilities, offshore wind operations, and other mission-critical environments where people die and things go boom if you get it wrong.
A physics major from Austin, Texas who became a designer, Jared has spent over 20 years working across China, Vietnam, Thailand, Sweden, and now the Netherlands. His work spans from designing the first enterprise Apple Watch application (though the client’s security requirements prevented Apple from issuing a press release) to building million-dollar applications used by just two engineers—applications that freed up 23 other highly specialized professionals for more valuable work.
Today, Jared is pioneering what he calls “Agent Experience (AX)” design—a fundamental shift from designing screens users manipulate to designing autonomous agents that make decisions, collaborate with other agents, and operate within environments with minimal human supervision. In this conversation, he explains why UX has already changed (not “is changing”), how the automation gradient shapes human-agent collaboration, why requirements gathering and performance evaluation become the critical UX skills, and why we’ll all end up entrepreneurs whether we want to be or not.
His core message? The role of UX professionals hasn’t evaporated—it’s transformed. And those who embrace agent experience design, human performance metrics, and business outcome focus will thrive in a world where screens are increasingly ephemeral and agents do the heavy lifting.
From Physics to Pixels to Mud
Before diving into agent experience, Jared introduces himself with characteristic breadth:
“I’m Jared Huke, founder of Daito Design, which is a UX consultancy focusing on enterprise and industrial transformation projects. We work with a lot of large energy companies as well as other industrial and sort of mission-critical UX design, which is a certain practice”.
“I’m from Austin, Texas originally. I was a physics major who ended up as a designer. So a little bit of left brain and right brain and all that”.
The Digital Detox: Ceramics and Noodles
“Besides all my travels, I do ceramics on the side. I’ve got a couple of young boys that keep me entertained. And yeah, currently living in Amsterdam, which is a nice contrast from Texas, at least in terms of the summers”.
Why ceramics? “I spend a lot of time working with pixels all day long. We’re doing some stuff for the Apple Vision Pro and so very immersive and intense operational experiences. And sometimes it’s nice to just have your hands completely muddy and not be able to even look at your phone for a couple of hours every day just to create an analog baseline for all of the digital stuff that I’m doing”.
“Nowadays they’re calling it digital detox. And I think it’s pretty important for people like us who are almost always in front of computer”.
Squishy Things
“I find myself drawn towards squishy things these days. So a lot of ceramics. I’ve been doing some handpulled lamen noodles for my wife who’s from Shanghai. And I also make tortillas a lot too. So I tend to like some tactility in my detox”.
The Birth of Daito: Contrast and Control
Jared explains how Daito Design came into existence:
“I was working in China. Came back to the US. I worked for a firm that was a sort of a mobile app shop. And I think I can say I brought a mature design practice to that organization that had a pretty close partnership with Apple”.
The Fortune 500 Energy Connection
“We got a lot of introductions with Apple into some of these big Fortune 500 companies in the Houston area—unsurprisingly a lot of energy companies. And so we were dealing with the rise of mobility, mobile application design”.
The First Enterprise Apple Watch App (That Nobody Heard About)
“I actually got—Apple wanted to issue a press release about the first enterprise Apple Watch application, which I had the pleasure of designing and leading on. It was for highly secured work environments. So the client did not allow that to happen”.
Why Start Daito
“I stayed at that place for about four years. Learned a lot about what to do and ultimately decided that I could do what needed to be done better if I was on my own. And so Daito was kind of born from a point of contrast there”.
No Engineering (Until Recently)
“We historically have not done engineering. And so we really focus on the business case and the human experience piece of any technological transformation”.
“Though I think that is changing, and we’re getting more and more involved in the actual—at least experience layer or in some of the actual buildout of the solutions—because to really ensure the fit to purpose and the quality of the experience, we need to have a little bit more control on how it’s being built”.
“So we’ve kind of moved into a little bit more of an ownership position on some of the solutioning with some of our clients. Which has been an interesting transition. And ultimately it’s really cool to be a little closer to the things being done and not sort of handing off to an internal organization that runs off and does some version of whatever it was that you designed”.
The AI Integration Timeline
When asked about AI’s role in Daito’s work, Jared provides historical context:
“Unsurprisingly, AI has been a part of what we’ve been doing. I want to say every project that we’ve worked on has had a ML/AI component for at least the last four or five years, pre-COVID for sure”.
The Inseparable Present
“And in the last two years, AI and any kind of product are just impossible to separate them. And so we’re really seeing the rise of sort of independent and agentic workflows, which is a fascinating area that is changing literally minute-by-minute. There’s new innovations happening in that space at a cycle speed that is just kind of breathtaking”.
“So we’re enjoying that ride and trying to bring the experience piece into this new technological landscape”.
Energy Companies: Invisible Innovation
When told that energy and oil/gas companies don’t seem to do much user research, Jared pushes back:
“I would say that you haven’t seen it. I wouldn’t say that isn’t being done. All of the really cool stuff that I’m working on, I can’t talk about because of very robust NDAs with highly litigious clients”.
Resources and Smart People
“So I can tell you that without a doubt, they are not without resources and they’ve hired a lot of smart people for a very long time. So they’re not in any way ignorant or missing out on what’s going on”.
The Diversity of Energy Work
“Now with the kind of work that these energy companies are doing—we work with nuclear, we work with solar, we work with offshore wind as well as traditional petroleum—you don’t want to disrupt when you’re dealing with these kind of operations, right? People die, things go boom”.
Measure Twice, Cut Once
“You really need to measure twice and cut once, which is part of why our methodology is a bit divergent than a lot of the Silicon Valley approach, which is just sort of—I don’t know what they do, but they have a 94% fail rate with their innovation activities on the whole because of this kind of cavalier attitude”.
“And I think that when you’re working with enterprise and specifically industrial workflows, there is a level of rigor that is required of your user experience work that is being transitioned into how they’re designing AI systems”.
Building Agents for Core Business Functions
Jared reveals what’s really happening inside these energy companies:
“We have several clients right now that are actually building agents for their core business functions. This is how they do the thing that makes them all of the money that they have. And they have been working on this for several years in many different versions of it”.
The Deployment Criteria
“But they’re not deploying these things into production until they can be certain that they’re creating value, that there’s a high level of trust”.
The Car Metaphor
“A lot of the moving parts—there’s the engine, the AI technological engine, which we’re all dazzled by these days, but that doesn’t get you anywhere. You need wheels, you need a car around it. You need seats, you need brakes and gas, and a way to steer and stop and all of these kind of things”.
“And so those pieces of the vehicle that are taking us from point A to point B are where the user experience professionals have an opportunity to guide these things”.
The Million-Dollar, Two-Person Application
When asked about target audiences, Jared shares a revealing example:
“These organizations are the size of small countries, right? Revenue-wise they’re bigger than a lot of countries. So their users are anywhere from complete novice to multiple PhDs trying to do very complicated workflows”.
The 25-to-2 Transformation
“I’ve built over a million-dollar application before that was for 25 engineers to do a job that allowed them to go down to two engineers. So a million-dollar application that is being used by two people because it was freeing up all of this work for other things and other ways. And that was considered a big win for them”.
Different Mindset, Different Metrics
“It’s a very different mindset where you’re not trying to gather sentiment. You’re not trying to manipulate purchase behaviors and things like that. You’re constantly focusing on human performance indicators. You’re focusing on business performance indicators”.
“And you’re trying, like any business, to do more with less, right? Even these larger organizations, they have finite resources. They have every problem in the world. They have every jurisdictional challenge. They have data management problems. They have legacy software issues. They have all of these kind of things”.
Operational Focus
“To craft an experience across a lot of those different challenges keeps you busy. And so the operational space—which is where we’re building tools for people to do work—allows you to have a focus and a maybe a linearity to some extent on how people are going to be using a piece of technology”.
“Relative to, I don’t know, TikTok or whatnot, which is you’re just trying to manipulate behaviors for constant usage at the expense of other physical social relationships and whatever else comes with social media these days”.
Defining Agent Experience (AX)
When asked to demystify “agent experience,” Jared begins with brutal honesty:
UX Is in a State of Flux
“Let’s be honest—UX is in a state of flux. Over COVID we flooded the market with a bunch of let’s say low-training, low-experienced UX professionals that got a certificate online and entered the workforce. And in 2021 they were making $100 grand a year in the US. By 2022, many of them were out of work”.
“And there was a significant downward pressure on how UX professionals were being perceived in terms of usefulness. With automation hitting at around the same time, you end up with this crunch around what is the relevance of a UX professional in modern software construction, in a software build environment”.
Why Coin “Agent Experience”?
“So what I was trying to do with—I don’t know, coining the phrase ‘agent experience’—was really helping UX professionals understand that our role has changed. It’s not changing. It has changed”.
“Now there’s some organizations that are more forward and some that are lagging. And with AI, the spread is quite dramatic from the leaders to the laggards. And so you may find yourself doing business-as-usual UX, but the reality is that you’re going to be doing things that are more for a different user type in this agentic workflow”.
What Are AI Agents?
Jared defines the core concept:
“Building these persistent agents that may have specific roles. They may have even specific personalities and requirements. They may find themselves in disagreement or conflict with other agents from the same tech pool or from external tech pools, and quite possibly working in a hybrid team with multiple humans and multiple agents trying to accomplish work”.
The Key Distinction
“Something that’s important to define about agents in general is that they are tasked with a goal that operates within an environment. It isn’t a bidirectional query-response kind of environment. You set them in motion, and if you design it well, they check in with you from time to time”.
“They know that they can do work. They are making decisions, and those decisions have impacts on these environments that you give them access to”.
Different From Traditional UX
“So that’s pretty different than working through user requirements, aligning against business goals, unifying stakeholders, cranking out some flows, breaking out some screens, getting those hard-coded into an application that doesn’t have a lot of dynamism”.
No Interface (Sometimes)
“I mean, the agentic workflows—if it’s really well designed—doesn’t have an interface other than maybe a conversational interface, right? Occasionally that AI may be able to generate a dynamic interface for you as needed: here’s a graph, here’s a table, here’s a map, or some other output that the agent will be able to generate”.
The Automation Gradient
Jared introduces a critical framework:
“I gave a talk last week in The Hague for the UXPA Netherlands meeting, and one thing that I mentioned was this idea of the automation gradient, right?”.
“So on one hand you have complete automation where the human does nothing. On the other side you have augmentation where you are enabling the human to do more with less”.
Blending Machine and Tool
“And so that blend between the machine and the tool—automation and augmentation—and then designing whatever isn’t augmented and needs to be automated, you need to understand and break down those tasks in a way that an agent can follow, right?”.
“You need to establish your baseline for how things are working today if you’re going to automate them”.
UX Skills for the Agentic Era
Jared outlines what UX professionals bring to agent design:
“So those are some interesting activities that I think a lot of the user researchers—especially the human factors and HCI community—have a lot of the skills that these developers or new ‘vibing developers’ are realizing is it’s hard to automate when you don’t know what all the steps are that you’re trying to automate”.
Designing for Safety and Resilience
“A lot of that designing for the agentic experience is understanding fail-safes and understanding check-ins and reporting and decision moments and things like that”.
“So obviously there’s a lot in there, and it’s a shockingly new field that we’re just trying to unpack how our UX muscles can be flexed in this new paradigm because it’s likely we’re not going to be cranking out screens from a UI perspective in the coming years”.
What Agent Experience Encompasses
When asked whether AX means agent-to-agent, agent-to-human, or the whole system, Jared clarifies:
“Yeah, I mean, it’s all of it, right? I mean, you can’t have one without the other. So technology is a means for a human to realize let’s say a business outcome or some kind of group or individual goal, right?”.
Goal-Oriented Task Analysis
“One of the tools in the human factors world that I’ve found very useful is this goal-oriented task analysis where you have this goal and you’re breaking it down into what it will take to get there, and then you’re applying technology against that framework”.
“So we have to keep that—we are still humans and we are still in business trying to accomplish these goals. So everything’s ultimately derived from there”.
Designing the Agents Themselves
“Now to your question, we’re designing the agents. So what does the agent do? What is the role of the agent? What does the agent have control over?”.
Tool Builders
“These tools like Gum Loop and Wand and—there’s a lot of these tool builders now where you can crank out your own agent pretty quickly”.
The ChatDev Example
“The one I saw a couple of years ago that was fascinating was ChatDev, where you had multiple ChatGPT 3.5 agents working together to build software, and they could build it faster by being broken into different roles and collaborating than the same technology doing it by itself”.
Collaboration Is Fundamental
“So these agents—and because I don’t know if it’s derived from the large language model or exactly the mechanics of the AI engine that is innately this—but essentially with language, collaboration is a fundamental piece of that. So how we’re designing these collaborations between the humans and the AIs is very relevant”.
Apple’s App Intents: An Agentic API
Jared shares an exciting development:
“How we’re designing their tools that the agents actually use—so Apple, who’s obviously having their challenges with AI, though having a couple trillion dollars in cash, I feel like they’ll figure it out”.
“They had an interesting concept which was ‘app intents.’ And what this means is it is something that would tell Siri how to find out what the weather is, go to the weather app to find out, or to check an email or to book a car—would go to Uber or what have you, right?”.
Not Just an API
“So it’s essentially an API but framed in a way that an agent can understand—that it’s not a hard, just application interface. It’s like an agentic interface. And so the redesign of some of our traditional software where the agent can use it is a very interesting space as well”.
The Foundation of Agent Experience
“And so I think that when I talk about agent experience, it’s the context for which an agent exists inside of an organization, how you manage its goals and performance, and how you are enabling the agent to accomplish those goals”.
“I think at the very simplest, that’s sort of the foundation of what we’re seeing are some activities that UX professionals absolutely can participate in, should even—I would argue compared to engineers or business people—are going to be better at breaking down these tasks”.
Testing the 20%
“And they’re going to be better at testing: is this working as we expected? Like, we automated 80% of our business—that’s great for the 20% of the people left over, how is it working for them, right?”.
“And how are these interactions being designed? Because if you don’t do it intentionally, you get what you get”.
The Collaboration Inversion
The conversation turns to team dynamics:
“It also brings a collaboration issue, because today even if we are practicing Agile or whatnot, even with that—Agile SAFe and so many methods—but in all these methods there is a certain handover where a design sprint hands over to the development sprint”.
“Now it seems the other way around, which means that everything is kind of completely inverted. So it’s like first the human element comes in, which has been tried by the way since quite some time but it was never possible, and then the technology comes in. Or both start hand in hand”.
Jared’s SAFe Agile Critique
“If you’re interested in my critique of SAFe Agile, I did write a book on it that’s available on Daito’s website for free”.
“And the interesting thing about SAFe Agile is that it actually only mentioned the user for the release of SAFe Agile in 2023 in April—was the first time they mentioned design thinking as the end-all-be-all for UX in the SAFe Agile framework, which I find—let’s just say a missed opportunity, to say the least”.
The Rapid Obsolescence of Engineering and UI
Jared doesn’t hold back on where things are headed:
“So I think that—I mean, with 90% of the code these days being written by AI, there’s obviously a huge downward pressure on having legions of engineers banging away on keyboards hard-coding out these things, right?”.
“And as agents become more and more resilient and independent, less and less engineering is going to be done. And I think I thought we had a little more time, but I think that the UI is also falling very quickly”.
The Lovable Moment
“But with things like Lovable and some of these new tools, man, it’s just—it’s one day to the next, whole sections of the economy are just kind of evaporating”.
What Remains: Requirements and Evaluation
Despite the automation wave, Jared identifies what stays critical:
“What I do think is very important—and this is at least until the robots really start going on the march, which I think we have at least another couple of weeks on—is the requirements gathering and the evaluation of performance”.
Research Redefined
“So this is where at least the way we define research at Daito is really around the narrowing of the scope, the reframing of the question, the analysis and prioritization of requirements”.
Good Data Gathering
“You need to gather data—for let’s say even if you’re using AI to interpret your user insights, you’re still going to need good gathering, right? You’re going to need good sampling, you’re going to need good documentation”.
“Because the introduction of bias is where software projects go horribly wrong”.
User Researchers Under Pressure
“So I think that the user researcher who had a lot of pressure on them in terms of downward pressure with the automation—I think especially with things like contextual inquiry and in-situ research—is becoming increasingly valuable”.
The $499 MVP Email
Jared shares a sobering moment:
“The UI designer—I mean, we’re working on some products right now and we’re seeing that the trend is towards automation in UI design. I got an email a few weeks ago from somebody who said that they would code me an MVP including UX and UI for $499”.
“Wow. So pretty hard to run an agency if it’s $20 per screen or something like that—or and that’s probably even optimistic, right? It’ll be $20 for the whole app in a couple more weeks”.
Everyone’s an Entrepreneur Now
Jared’s career advice reflects the new reality:
“So I do a lot of mentorship just trying to give back to the community, and some of the things that I’ve been landing on in terms of my advice with people that are having challenges in careers and whatnot is that ultimately we’re all going to end up entrepreneurs whether we want to be or not”.
The UX Advantage
“And if UX professionals are as good as we seem to think we are, shouldn’t we be making better applications than everybody else? Our prompts should be better, our testing should be better, our sampling and requirements gathering should be better”.
SaaS Is Dead
Jared references a provocative statement:
“I feel like with—the way, you know, the CEO of Microsoft said SaaS is dead—I mean, we’re kind of seeing that, right? Where the hard-coded big applications are being transformed by ephemeral, ephemeral dynamic applications that are being designed in the moment they’re needed and evaporating back into the ether once their purpose is served”.
Humans Still Matter
“This will still be used by humans. These will still have business goals associated with it”.
The New UX Mindset
Jared describes the necessary cultural shift:
“And so our discipline in general I think needs to rally religiously around the creation of value for businesses and that the augmented humans are actually getting performance increases”.
“So the measurement of those human interactions are going to be increasingly important”.
From Delight to Performance
“And the delight the users and empathy and some of the design-thinking-derived UX culture I think is going to be rapidly replaced by a much more human-factors-centered, human performance, business performance, efficiencies and measurable outcomes kind of mindset”.
“And I don’t know the implications of all of that, but I can say that there should be a lot”.
Key Takeaways: Embracing the Agentic Future
Jared’s 20+ years across continents, industries, and technological paradigm shifts offer crucial lessons:
1. The role has changed, not is changing
UX professionals must accept that designing screens for humans to manipulate is rapidly becoming obsolete; agent experience is the new paradigm.
2. Agents are goal-oriented, not query-response
Unlike chatbots, agents operate autonomously within environments, making decisions and checking in when designed well.
3. The automation gradient defines the design space
From complete automation (human does nothing) to pure augmentation (enabling humans to do more), UX defines the blend.
4. Requirements gathering and performance evaluation remain critical
While 90% of code gets written by AI and UIs become ephemeral, understanding what to automate and measuring outcomes stays essential.
5. Human factors skills beat engineering for agent design
Breaking down tasks, understanding fail-safes, designing check-ins and decision moments—HCI and human factors practitioners excel here.
6. Contextual inquiry and in-situ research grow more valuable
As automation increases pressure on researchers, ethnographic methods that capture real workflows become differentiated skills.
7. Mission-critical environments demand rigor
Energy companies can’t tolerate Silicon Valley’s 94% innovation fail rate; when people die and things go boom, measure twice, cut once.
8. App Intents represent agentic APIs
Apple’s framework for telling Siri how to use apps points toward redesigning software for agent consumption, not just human interfaces.
9. Multi-agent collaboration is fundamental
ChatDev proved agents in specialized roles collaborating outperform single agents; designing these collaborations is UX work.
10. SaaS is being replaced by ephemeral experiences
Hard-coded applications give way to dynamic experiences designed in the moment and evaporating after serving their purpose.
11. Everyone becomes an entrepreneur
With $499 MVPs and automation eliminating traditional roles, UX professionals must leverage their advantages—better prompts, testing, requirements.
12. From delight to performance
Design-thinking-derived empathy culture transforms into human-factors-centered business performance and measurable outcomes focus.
Final Thoughts: Digital Mud and Agent Futures
Jared’s journey from physics major to designer to ceramic artist who makes handpulled lamen noodles reveals something essential about thriving in technological upheaval.
When pixels dominate your days and agents orchestrate your work, tactility grounds you. Muddy hands that can’t touch phones for hours create the analog baseline needed to design the digital experiences that users—and increasingly, agents—rely upon.
The million-dollar, two-person application captures the essence of what agent experience design enables. Not manipulating sentiment or purchase behaviors, but freeing 23 specialized engineers for higher-value work by automating complex workflows with surgical precision. That’s human performance increase. That’s measurable business value. That’s the new UX.
The automation gradient forces honest conversations about what should be automated versus augmented. Complete automation means humans do nothing; pure augmentation means humans do more with less. Somewhere on that spectrum lies the optimal blend for every workflow, and UX professionals trained in goal-oriented task analysis, fail-safe design, and human factors are better positioned than engineers or business people to find it.
The $499 MVP email isn’t just anecdotal evidence—it’s a wake-up call. When entire applications with UX and UI can be coded for less than one hour’s billable rate, running traditional agencies becomes untenable. But Jared’s response isn’t despair—it’s entrepreneurship.
If UX professionals are as good as we claim, our prompts should be better, our testing should be better, our requirements gathering should be better. We should be building superior applications with AI assistance, not competing on screen-cranking speed.
Apple’s app intents framework hints at the future. Not APIs for developers to hard-code against, but agentic interfaces that tell autonomous agents how to accomplish tasks across software ecosystems. Redesigning traditional software for agent consumption—not just human manipulation—opens entirely new design spaces.
ChatDev proved that agents collaborating in specialized roles outperform monolithic single agents. Designing these multi-agent teams—defining roles, managing conflicts between agents from different tech pools, orchestrating human-agent hybrid teams—requires UX thinking scaled beyond individual user journeys to organizational performance systems.
Energy companies building agents for core business functions that make them all their money work on these systems for years before production deployment. Because when things go boom, you don’t get second chances. This rigor—understanding every failure mode, designing every check-in and decision moment, establishing trust through consistent performance—is where UX professionals add value that raw AI engines can’t provide alone.
The CEO of Microsoft declaring “SaaS is dead” reflects the ephemeral future Jared sees. Applications designed in the moment they’re needed, serving their purpose, then evaporating back into the ether. But humans still use these experiences. Business goals still drive requirements. And someone needs to ensure augmented humans actually gain performance rather than just creating 20% of workers left confused by the 80% that got automated.
The cultural shift from empathy and delight to human performance and measurable outcomes won’t satisfy those who entered UX to “make the world better through design thinking”. But for practitioners like Jared working in industrial environments where rigor trumps vibes and business performance determines survival, this evolution feels natural.
Whether we’re ready or not, agent experience is here. The role has changed. And those who embrace requirements gathering, performance evaluation, contextual inquiry, human factors rigor, and entrepreneurial mindsets will thrive.
The rest? Well, they might consider learning ceramics. There’s something deeply satisfying about creating with muddy hands when the digital world moves faster than you can track.
Thank you for reading!
If Jared’s insights on agent experience design, the automation gradient, and the future of UX resonated with you, share this article with designers and researchers navigating AI’s transformation of their field.
Have questions about agentic workflows or industrial UX? Connect with us at hi@uxarmy.com
Special thanks to Jared Huke for his pioneering work defining agent experience and for sharing hard-won lessons from building mission-critical systems across energy, nuclear, and industrial sectors.
And thank you to all of you for being part of the User Insights community.
⚡ This podcast is brought to you by UXArmy, an all-in-one UX research tool.
Join the Conversation Before Everyone Else!
Dive into user research, product strategy, and design with industry leaders. New insights drop every month. Don’t miss out.
