Nik: Welcome to AI and Design, where we explore how artificial intelligence is reshaping the world of design. I'm Nick Martillero.

Dan: And I'm Dan Staffer, and we're faculty at Carnegie Mowens Human Computer Interaction Institute, the HCI. Each week we break down the latest AI developments, dive deep into topics that matter to designers, and talk with fascinating guests who are right at the intersection of these fields.

Nik: Whether you're a designer working with AI or an AI practitioner interested in design, we're glad you're here.

Dan: And this week we have a special all-guest episode of AI and Design, where we welcome Andrew Hogan and Shane Johnson of Figma to talk about Figma's 2026 AI Report. The report looks at not only how design is changing thanks to AI, but also how AI is changing collaboration among teams. So here's our interview with Andrew and Shane. Okay. Well, this is a very special episode of AI and Design Podcasts because we have for the first time not one but two special guests and that is Andrew Hogan and Shane Johnson from Figma here to talk about the Figma AI report in twenty twenty six. And this is a doubly special episode. Because Shane is our first two timer here. you were our very first guest and now you're back. So welcome Andrew. Welcome back, Shane.

Shane: Thanks, Dan. Thanks, Nick.

Andrew Hogan: Thanks for having us.

Dan: we're excited to talk about this report because there's a lot of crazy stuff in it. So so I mean not crazy, but but interesting stuff in it. we overused the word interesting on this podcast, so we try to we try to limit that. So why don't we start by explaining what the report is, why do you all do it, and this is the second year for it? Or third. Third. Wow, okay.

Andrew Hogan: It's act it's actually the third. Yeah. So we started doing it. I wrote this wacky memo that was that there's so much hype out there about AI. We need some actual data. and shared that memo with Shane. And then internally we all just got behind the idea that we should measure this and we should be able to share across the industry, what are people seeing? What are they feeling? and Shane created this incredible methodology to look at multiple dimensions. so it wasn't just, are they prompting? It's are you seeing your organization change? Are you seeing your own workflows change? Are you seeing your team change? and it was really born out of this idea that people post all kinds of things on social media. What is actually happening? What can we actually measure out there?

Shane: a I think the goal of the report is less to deep dive every year into one specific area, but it's more to look broadly and holistically at how AI is changing the work of product builders, but also the organizational context in which they exist. And by tracking this year over year, what we've seen is a rapid evolution. from what was just a individual productivity story where AI is accelerating this aspect of my workflow or engineers are able to adopt it, but it hasn't really impacted design. what doing this consistently year after year has let us see is how the shape of AI's impact is changing and how it's impacting higher order things. So it's not just productivity anymore. It's how teams are actually working together.

Andrew Hogan: I think one thing

Dan: And that seemed

Andrew Hogan: One one thing that Shane said there is really interesting is that the shift la then twenty twenty-five to twenty twenty six was much bigger than the shift from twenty twenty four to twenty twenty five. And I think we've all felt that. And then to see it come out in the data across all sorts of metrics, whether that's even trainings that are being scheduled, whether that's how you work with your team, I thought that was one of the more interesting things is the acceleration that we've all felt came through in the data.

Dan: Yeah, it seemed like when I was at config with the two of you, last year was like, this is happening. And this year it seems like, okay, this is this is happening. How are we all working with this? And I think the data reflects that, where it's okay, the adoption's gone up and the one of the things that I was really interested in was seeing how the different roles were now acting like the other role and what that meant for people. And the finding I really want to dig into, and maybe we could just do it now, was the idea that the people who were the most excited about and convinced that we need even more design right now were developers. Presumably those are the same developers who are doing more actual design. And so I thought that was a super interesting finding if you can expand on on that.

Shane: I think there's this assumption that design is for designers, right? Or that it's solely the responsibility of those individuals with the title. And it's really not the case. design to some degree has always been a team sport between designers and the folks that they work with. and so like it it kind of makes sense to me that developers are seeing more value in design because it It's not the case that AI can just replace the role of a designer. it makes design a little more accessible for developers. But I think what developers are recognizing is that the process of thinking through business constraints, user constraints, and the technical implementation is fundamental to creating good products. And so I think that's why we're seeing a renewed emphasis on what People might have thought as soft or fuzzy or as aesthetics or the user experience. design is really getting elevated into something that's much bigger than any of those artifacts that designers produced before.

Andrew Hogan: I think it's one of those things too where you start to do it and you start to think, I can take I can do part of this other person's job. And then you start doing part of the other person's job, whether that's development or design or product management, and you realize, this person has a really unique set of skills and some unique experience that I I can do part of that. I it can sort of look like it. And then you get behind the curtain and the information architecture doesn't work right, the interactions don't make any sense, you haven't really thought about what you're really trying to accomplish or what you stand for. And then there's some random business, governance problem that you haven't considered either that just sort of showed up in the middle. And all of those things from the outside don't seem like they should be that complex. And then you get into them and you realize how much more valuable and difficult it is. And we've seen that for designers doing development too. it goes every direction.

Dan: Yeah, I feel like I cosplay a engineer sometimes now, but I would never consider myself that because exactly what you were saying, like, wow, some of it as you actually start to get deeper than just some surface level stuff and wanting to do anything that's more complex, you're like, wow, I'm not trained for this, even with the AI. This I can't do this. Or you you soon find out the limits to what you know and don't know, is another way of putting it.

Nik: one of the other big findings that really stuck out to me was that this year it seems like there's a bigger emphasis that people are putting on how AI is changing how the team works together, how collaboration's happening. the stat that we pulled out was 41% say AI meaningfully changes how teams work to together compared to only 7% two years ago. I'm curious if you could speak more about this because I think there was a a point in the report where it said, AI is gonna make you a ten X developer, designer, whatever. it's about you, but your report this year is much more about us the team.

Shane: Yeah, I'm I'm really glad this stuck out to you because this has been something I've long felt. it around the time when Gen AI started to really become integrated into how people are working. I think there was an immediate reaction that this 10X is everybody now. Everybody gets new abilities and can do new things. and I think that really put the emphasis on individuals, unicorn developer designers. And I it is amazing, but I think fundamentally good products are still built by teams of people. They're built by people who have different perspectives who advocate from different perspectives, but who also negotiate conflict together really well. I think it's this friction of Good teamwork.

Andrew Hogan: I think what you're describing, it's funny because in the course of doing this report, Shane was like, I am so tired of hearing about this 10Xing. we're just gonna 10X everybody and then are we gonna get to a million X? And like how does that even work? And I know one of the things that Shane you've been talking about is the prototype is a communication method now, like on par with, writing a document or sending a Slack What does that then do? If you realize that you can now communicate in a higher fidelity, you can then have more of a, a friction, more back and forth, right? And then you end up with, okay, so we're all prototyping together. What skills are now more important? Is it the prototype skill, or is it the ability to have judgment between these different prototypes? And so you end up in these second and third order effects. that I think the data helps you peel apart and what we all feel where you bring this thing to work, you bring this thing to a project, and it used to be that that took weeks and now it takes a day. And then what takes weeks is aligning across these five or six possible directions or a hundred directions. but I can confirm Shane banged the 10X thing up and down and said We need to interrogate this a little bit because this finding is really at odds with a dominant narrative out there.

Shane: Exactly. when we saw that massive shift in people saying that AI is really starting to affect how they work together, I think it's easy to jump to the conclusion that this is a positive thing. What we did this year that was different from past years is that instead of having just open-ended boxes in the survey where people can type a response to an open-ended question like, in what ways has collaboration been impacted? We actually did AI-moderated interviews with 630 people to really dig into these issues. And what we learned from that is that it's not purely a positive story. These this impact to our individual how we work as individuals is starting to ladder up and to collaborate. collaboration and it's really starting to expose the communication and alignment boundaries between teams and between individuals within those teams. And so a lot of the report is focused on What are those problems that are arising in teamwork now? Why isn't 10x collaboration resulting in 10x or 10x why isn't 10x productivity resulting in 10x outcomes? and then what are the organizations who are successfully navigating this doing differently?

Dan: I wanna talk about some of those problems but one thing I wanna say before we dig into that is I don't know about you all but I was I was surprised that it was only forty one percent who said that AI has meaningfully changed how their teams work together. Maybe maybe I live in a bubble here but I was shocked that it wasn't sixty, seventy, eighty percent.

Andrew Hogan: So I wanna highlight one really interesting analysis that Shane did. Shane, you should talk about this, your shapes of adoption. because I think the biggest thing is that the organizations are not a monolith. They are wildly different from each other. Even within those organizations, they're different. And I don't think we can underrate that, that the people willing to go tinker and fiddle with the current state of tools is not the same as the median worker or the a median organization. So Shane, maybe I mean, that was an interesting finding that you just came out of the blue. That wasn't we didn't actually have that in the initial what are we going to dig into in this report.

Dan: and this is the four quadrants thing?

Shane: Yes. one of the things that we found really was really interesting is we have all these outcome measures where it's like the degree to which you've been working on AI projects, the degree to which they've been successful, have has you seen ROI on those things? All these things tend to be highly correlated with an alignment between an individual and their adoption and use of AI and their organizational context and the degree to which their organization is emphasizing AI as a priority. And so what we found is that three years ago. around 35 to 40 percent of everyone we sampled fell into this category of nascent. Their organization wasn't really pushing it, and they themselves weren't really finding much value in it, or they had been experimenting with different use cases for AI and their workflows, but nothing has really been sticking or been transformational. Flip to today, That that population of nascent or unaffected has decreased to 18%. So it's roughly halved in two years. But what's doubled in two years is the proportion of people who say that they are aligned with their organization and they're successfully putting AI into practice to get work done. So there has been a pretty remarkable flip in two short years.

Andrew Hogan: Shane, the larger one that I wanna add here is that there's still wide variation and we saw that in the qual too. Dan, your your point, why isn't this higher? I think reflects just how wildly different the work situations really are for for individuals.

Shane: Yeah, and so that's that's the the messy middle right now. And I think when when you say Dan it's like we're you're surprised it's not higher than forty one percent. I think that's more an artifact of where we are. This has been a maturing technology for individual use cases, and we're only just now starting to think of it as a material between organizations or between members of teams.

Dan: Right, 'cause it may be that that I am adopting these tools and using them, but it may not affect the overall process, for example.

Shane: Exactly. And I think as as we see more and more of team conventions, rituals, processes, the more we see those things start to break down and that we're not recognizing 10x outcomes, we're gonna slowly start to address those problems.

Dan: Mm-hmm. What is so get get moving back to that? Like what what i what is is it the collaboration part that is causing the lack of 10x output or more quality or what is it? Because that that's a question that I think everyone is wrestling with. we've been told these tools are amazing, they do all this stuff, and and now I'm a 10x signer, but We're not seeing major shift in output or more features or new or brand new products. We just haven't seen that as much as you would expect.

Shane: I think there's there's a couple of things happening here. One of them is that in many organizations there are pockets of AI first workflows between members of teams, but there's also a lot of legacy products that teams are assigned to. And the question is, is like, how do we start to transform how this team works without completely breaking it? Which is a big question. we heard from a number of folks at config that they have that moonshot group off to the side, building new things, doing cool things, experimenting with their workflows. But their question is, how do I take that and apply it to this completely different context with a completely different set of business concerns around it? the the risk profile is just much higher in those situations.

Andrew Hogan: Wanna add something to that because Dan, there there is evidence of increasing output in lines of code,

Dan: Mm.

Andrew Hogan: apps submitted, the app store numbers are way up, the adoption of apps is not necessarily way up, which is an interesting, supply has increased dramatically, has the demand increased in quite the same way. and so I think there's a difference between outputs, activity, outcomes, and the real issue here is that outcomes come from, teams of people working together often. especially when you think about established companies. And I think what we're one of the reasons this study is interesting is that it reflects a diffusion from individuals able to do things much more quickly out through organizations and products and workflows that shift more slowly. and they're not the same things. and I think it's worth considering what is it that really drives better outcomes? And I would submit that that's from really great teams working together and probably raising the quality bar of how they work together, with these, new tools. But the output evidence seems to actually be up. It's just a question of whether that actually leads to better results in the long run for Those folks.

Dan: Right, it's all about quality and value. it's very easy now to use these tools to generate something and get into the app store, but if it's not valuable to anyone, then okay, you you put it out there. Congratulations. But

Andrew Hogan: I mean, as an individual, this this is the golden era of side projects. And I have been telling people who are thinking about either switching jobs or finding their own satisfaction, start some side projects, do some things. in your personal life stops you from putting something on your personal computer, even if you're an organization that won't turn the different tools on, has governance around it. This is a golden era to do those things. And that just is not the same thing as working within a business context to create something that fits, what that company's trying to do, what the team's trying to do, the governance structures, all the things, the risk profile Shane talked about. but an inc truly incredible time to do side projects. Really no stopping. the really fantastic.

Dan: let's talk about handoffs. Over half the respondents say they've abandoned linear handoffs. I are the handoffs now Prototypes? Is it code? can you tease that out a little bit? did the what did the qual stuff say about that?

Shane: I don't know if I'd say people are abandoning handoffs in so much as they just don't work anymore. So if I'm a designer, That's a good question. I'd ask the

Dan: did they ever work, really? Right.

Shane: designers in the room, what went wrong with handoffs? Or what was tough?

Dan: I there was always there was always a gap. there was never a hundred percent transfer of knowledge between one step to another. It was practically impossible. And then keeping up with the changing documentation was always also impossible. You would never go back and fix the PRD or fix the wireframes or so what was the the source of truth always became muddled or it was just the final product. And it was like, well that's launched. That's what the source of truth is now. you often needed to have an actual conversation sometimes to convey the intent of what the designer was trying to do. And that was even more hard when you just had static screens. It's well, this thing slides in here and that I mean it got better as you were as it was easier to prototype with things Figma to prototype there to show how things move, but it still wasn't the thing. It was always hard to show things like error states and and edge cases and all kind those kind when this is going on, this is also happening. things that were always very difficult to deliver as part of a handoff.

Shane: But there was a strong desire to, right?

Dan: Sure. I mean they were always supposed to be communication devices, right? I mean, designers were always so dependent on engineers to get the thing built, unless you were one of those rare folks who did both.

Shane: And what we saw last year was roughly ten percent of designers saying that they were using AI to build these fully functional prototypes that completely convey their design intent. We saw that jump to sixty-one percent this year, so almost a six X, right? And it means there's a ton of communication flying around. Because these prototypes are so salient and communicating intent, people feel as if they can get it. But when you start digging into all the things you're talking about, Dan, the the specific states or really specific movements, they're still breaking down in that case. and it's more difficult to discern what the the detail is versus the macro. So we're able to communicate more effectively our intent, but it isn't necessarily completely productive to creating the final thing. I think one of the things that we're seeing with prototypes in particular is they're really easy to create, but they're very difficult to dissect, pull apart, strip down, understand the parts and components, or understand how complete they are.

Dan: That was the one thing that I was gonna mention and the one thing that I really liked That you all showed off at config was that exploded view of the prototype. I was like, yeah, this is this is what I've been waiting for. Just to be able to because right, because otherwise it's well, you can get to the error state, you just have to click click this sequence of things, or I have to make a whole other thing that that has that error state. But if you can do exploded view and see, that's what that's supposed to look like. I got it. I thought that was a really powerful use of the camera. canvas and something that's very hard to replicate without the canvas. You have to the fake button that's put it into the error state, stuff like that.

Andrew Hogan: This is a really good example. I mean, it's not that that wasn't needed before, but the second and third order effects of generating prototypes is that you need ways of evaluating them against each other. You need ways of being able to apply your expert judgment to the, the commu essentially the communication artifact that's being shared by, someone who's not an expert in quite the same things that you are. And I think that's part of why this teamwork part feels so early. why it's only, 41% have said it it's it's transformed things, because we are we are sort of finding the bottlenecks and the challenges, it's essentially it is a design process. Okay, so there's a lot more prototypes. So then what does that mean? What do you need to do with that? now there's all sorts of drift and there's it's difficult to fully assess the whole picture and the system. and All of that just indicates to me that we're early in this and early in the adoption of of these tools and it's diffusing the capabilities are diffusing through all the places where you could put them. and that also makes this there's a lot of design challenges to solve here, which then makes, people think that design is more important when they encounter these things.

Nik: we're talking about the use of the canvas here and actually there were some data from folks. I mean, designers of course are spending a lot of time in the canvas, if potentially maybe half of designers that you surveyed were spending their entire time in the canvas. But it was interesting to see the number of developers jumping into the canvas as well. and I'm curious if there's anything from the qualitative data on what they said and then maybe what designers said about working and product managers about working with the canvas as the shared space.

Shane: Yeah, I think someone in the qual mentioned that it's a great highest common denominator for everyone to contribute equally. and that that's always been Figma's hypothesis is that you have this shared space where we can share context. And I think that's really what the canvas is offering to developers. I think it's it's also important to note that that question, we ask them Where they want to spend their time, but we also have a component of it that's about like, do you want to be directly designing things in the canvas or do you want to start with a prompt? And I think for a lot of folks who are working in these collaborative workflows, they they want to be able to work with AI in an environment together to create a a shared sense of reality. If an engineer is just writing lines of code, it may be a reality that they can parse, but it isn't necessarily one that everyone else on the team can understand the intent behind. And so the canvas really represents an opportunity for people to express intent better to one another.

Nik: yeah, maybe as like a quick follow-up on that. I'm curious from the perspective of the survey of respondents and this question of do you want to prompt, do you wanna direct manipulate? Dan and I talk on the show all the time about how you probably need both. Actually we like direct manipulation on a lot of things because there's so many Things where if you generate, you're just like, you got this wrong, and it is it's so much faster for me to go in and just change it. Did you learn anything from the survey respondents about what they want that mix to be or where they where they wanna see some of this going? I know that this was a more of a the state of AI, but I'm curious if you got any forward looking ideas or thoughts from the respondents.

Shane: there's been kind of a narrative of blending roles, shifting roles, collapsing roles, people doing each other's work. And I think that makes a lot of people anxious. And so we jumped to the most extreme mental image of what that might look like in our heads, designer just shipping or making a pull request, or an engineer generating an entire front end. What we heard from people in the qual was that's not really reality. That's that's not what's happening in production environments. What is happening though, is that designers and engineers are able to reduce the amount of tax they put on other people that they work with. So in as a designer, instead of creating an Asana ticket or something for a visual bug, I can go in and fix it myself. That reduces the amount of work that I would have put on an engineer previously, and we just get to do it faster. The same is true for engineers in this context. If they are missing a piece of the system, they'll jump in to a canvas and create it themselves. So it's it's less about doing everyone's job or doing your collaborators' jobs. It's more about knowing enough or being fluent enough in the material that they work in. be able to contribute rather than creating more work, if that makes any sense.

Andrew Hogan: There's also a relationship component, Shane, that I just want to highlight that you mentioned. you don't have to make the ticket that creates work for somebody else. You can do part of that work, which then not only makes things maybe a little bit speedier, but also then makes you able to message that engineer next time about something like a big brainstorm that you want to do together. or, the engineer can or the the PM can say, Here's this prototype that I'm thinking through. And I think in really in good, well functioning organizations with solid communication principles, all of this is an enhancement for the way that you interact together. and then of course there's some others that don't work quite that way where the communication is not great. So it's an amplifier for whatever was already happening. but it is also a way for people who want to maybe do their job better, more successfully, to step in and help and build a relationship with somebody else. And there's all these interesting interpersonal parts to all of this that we're all still working through in the average workplace, let alone students trying to navigate their way through this too, which is also its own, its own thing.

Dan: Yeah, we'll we'll talk about that in a minute too. I've heard from a lot of engineers that say that, they love that designers can go and do the P threes and P fours that they never wanted to do in the first place because it was all this fiddly crap that you you care about and It's and designers are like, great, I get to do that fiddly crap to brown that button or, fix that spacing that I've always wanted to do. then it frees the engineers up to be working on the stuff that they really want to do, the P zeros, the P ons, the big new features, all those kinds of things. As so I've I've heard it as a a net positive on both sides. that removing the tax of my god, here comes the designer to ask me to to to to to do this thing. I'm I'm deep in this other thing. And so yeah, heard a lot of positive about that level of of role overlap.

Nik: Yeah, that idea of the tax that you have when you make a request or do it, submit a ticket or ask someone for something. I mean, that's my my inner computer support supported cooperative work, researchers going off being man, I wonder if anyone's studying that. And if anyone's tracking any data, because I think we we focus so much about say the use of AI as well, this is gonna make our product better, or it's somehow gonna make our productivity better. but you gotta define productivity, but is it just work output? But actually this idea of if you can reduce friction, and that's really where things are are are happening and where it's really helping, and then that from that reduced friction, hopefully you get better products and better teamwork. Now I'm thinking who's tracking that? Can you track that? because yeah, as Dan said, we we have heard that people talk about these kinds of things. And I would be really interested to see if that's where maybe one of the first impacts of a well-functioning team that uses AI and incorporates AI into their processes actually sees benefit.

Shane: I think also to build on this idea a little bit, one of the other things that we're hearing designers and engineers start to take on more is thinking through the overall system itself. A lot of designers are really reinforcing the importance of design systems now. It's not as if they weren't important before, but I think a lot of the way in which a AI is impacting collaboration is really exposing the gaps in the system, the gaps in the organization, and the gaps in communication. And so rather than focusing on screens, we're focusing on systems. And what these things do, they're they're not just organizing UI components. Microsoft actually had a great report on this. They called it orchestrating relationships. Because that's that's really what it's doing at this point. It's it's the the transactional nature of the relationship between a designer and an engineer can be boiled down to some degree by the outputs of the designer and the inputs of the engineer. so what these systems are really doing is mediating that relationship much more effectively in a way that designers can contribute to higher order system needs while engineers are still getting what they need. Yeah.

Andrew Hogan: mean the other part to that is enabling your teammates, right? Like custom plugins, these are these are things that Figma's thinking a lot about because it's very clear that when you can enable other people to do something that's on brand, some motion system, then you extend the impact significantly, you accelerate everybody. And I think it makes the people who can see these things, Shane said, the system who and and I mean that in the most meta macro way of thinking about it, the system of work too, and I think he does as well. When you can see it, then you can see the opportunities to change it. You can see the opportunities to support it. And that is one of those things that seems like it'll become immensely more valuable moving forward because it helps to create coherence with everything else. It helps to accelerate the overall work, which I think would likely lead without divergence, would I think would likely lead to better work between each other. so I mean Nick, I don't I don't know. Obviously we're thinking a lot about these things. I think there's great opportunity to study them more.

Dan: I was gonna say that this has been in ye olden times, meaning like three years ago, this was the role of design ops, right? It was to try to help scale the design organization and increase its impact. And now AI can do that in a way that we couldn't even fantasize about three years ago where it was well, yeah, if you have a design system and you've got all the context and design tools and the system understands how all the pieces get put together and what the intent behind each piece is and about research that has been done about each different part. I mean that stuff is is incredible. And if you can bake it into the system, it is such a powerful tool that no matter who is using it, designer engineers or Bob and HR to make a new intranet web page.

Andrew Hogan: the trap of course is thinking that you don't have to have someone who's an expert in those things to do it well or to make those decisions. and I think what many folks are finding in organizations is that they still there's a high degree of domain expertise that's needed in either intranet design or, potentially in whatever that field is that the intranet page is about or whatever the internal employees it doesn't remove any of the judgment or need to understand. it probably just helps you ask better questions of the people who have that expertise. and I think the thing folks will find is that you still need the expertise from somewhere. but I think it does make it easier to get input, get a judgment back.

Dan: last year's report asked whether roles and titles and responsibilities are gonna be a thing of the past. do you think based on this data, you think designer, developer, PM are these gonna survive as job titles? Or are we kinda converging on product builder?

Shane: Personally, I think we're converging on a collective aptitude for product building, but I don't think it negates the need for expertise for a lot of the same reasons Andrew's talking about. I and I think people are maybe confusing the two and feeling their own roles and responsibility being stretched into new areas. Andrew and I were talking about hard skills versus soft skills. prior to this, because there's long been this downplaying of the role of soft skills or even saying things communication, collaboration are soft skills, I think is patronizing in some way, and that the real value exists in technical ability. But I think comparison of those two things is just wrong. from the start because hard skills are usually thought of as technical abilities and those are technical aptitudes but there's also an aptitude for the application of those skills and I think that's where a lot of soft skills or collaboration tends to fall is this is the application of our collective technical abilities towards that outcome. And just to come back to one other point, I think When everybody can create anything, the real bottleneck and what our data is showing is that decision making is that bottleneck. It was a big CGI boondoggle, to some degree. It was a great book, great movie. But these it was the

Dan: Is that the one with the tiger in the boat and

Shane: tiger in the boat, and they spent like Two or three X, the actual production budget, just in post, because the team, the producers, the directors, deferred everything until later, right? They said, we'll fix it in post, we'll fix it in CGI. And that that that was the prime example of ballooning budgets really killing a promising project because of this deferral of decision making. I worry that the same will be true for organizations. They can produce more, but it's really just more churn for an inefficient decision making process.

Dan: We we have heard that before that this idea that the actual decision making process, some people are treating it like a bug when it's actually was a feature. It actually was a good thing that people weren't launching everything. deliberating and taking time and working on things together was actually net positive and we are just finding that out now.

Andrew Hogan: I mean you find new new people who have input or authority over something. You find new expertise you didn't know existed. I don't know if I would go so far as to say all decision making processes are net positives. I think I do think there's something to the act of finding those things and the bottleneck moving. I wanna go back to this thing. this idea of identity, titles and identity. And I absolutely think there's a group of people who want to be product builders, who want to think about themselves as product builders, and they really identify with that. But there were designers before there were screens, or before screens were as big a part of our lives. I find it really hard to believe that there won't be designers after that. and the developers at one point called themselves computer programmers. even the idea of software developers, software engineers rose as well. So these things shift and morph and change based on our collective taste and what's considered valuable. but I find it interesting that

Dan: Right, the the the Yeah, the the first computers were people. that's what they called themselves. They were the computers. And they were mostly women too, who were doing the computation.

Andrew Hogan: Yeah. So these things always change, but the title of designer has continued. People identify themselves that way and they identify their skills that way. it seems unlikely that that will not become a thing going forward. it just seems surprising to me if that if that's where we end up.

Dan: What do you think has caused the big shifts that we've been talking about from last year to this year? I know one of the things that you mentioned in the report is around training, that AI training doubled basically to fifty four percent in the last year. Is that contributing to it? Is it the is that the models are getting better? The tools are getting better? what what is making this or is it just time and and hype going into this?

Shane: So I can I can tell you what I think is causing it. I don't have any conclusive causal relationships. Your point about training, we did see the the idea of formal top down training jump this year significantly. I think that's more in reaction to what the what new products have entered the market. What we saw last year, which was March of 2025, was remarkable jump in the agents being a a key technology. I think along with that, what we saw was a large jump among engineers. And so what I what I think is really happening is that AI first impacted engineers, but last year the impact on designers was largely the same as 2024. So it's more of a a a case of differential adoption between individual functions within a team. What we saw this year now is that AI has become much more useful to designers in their core workflows and what they produce and the new abilities that it's given them. so that to me is is really what's causing that. Now companies are drafting off of a lot of the organic adoption they're seeing, they're concerned that this isn't going to be homogenous across their teams. It's it's just going to be in little pockets here and there. and so that was more why I think we saw training come out so strongly this year.

Andrew Hogan: one hundred percent agree. I do think this is productization of AI capabilities and that getting stronger and the the actual design of things to do design and development with AI. I think it's you're seeing that play out. and then that's then driving the the jump that we saw. So just a finer point on what Shane said, designing agentic products increased dramatically in the years before. Now we see agentic products coming into play, which then causes new bottlenecks and new things to form. It's linked relationship between these things. paired with a healthy dose of hype. Let's not ignore that. let's not ignore that that is you you're being told you need to upskill on the ninety percent of respondents said they need to learn to work well with AI for the future of their career. the it's an important skill for them to have. That then itself breeds, I'm gonna learn how to use these tools more effectively so that I can tell people I can use these tools effectively. so I think all of these things are interlinked and interrelated.

Dan: That is an excellent transition into our f final self serving question here for Nick and I, which is, yeah, so ninety percent of people saying that it is essential to their future careers. Nick and I are putting together our syllabuses for this fall. what do we tell Our design students starting their design schooling this fall about what is it to learn and what to not bother learning. What what are we what are we teaching them that works with that ninety percent?

Shane: I'm gonna say this in a self serving way because I am a researcher is to do research. I still stand by the belief that context is the number one thing to optimize for in the future. And what we're seeing in our data is reality is changing faster than we can track it. And so your ability to gather context on reality is the differentiating thing that you can do as a professional in this space. building an aptitude and Collecting context about people, synthesizing what it means for the company you're working for or the problem you're trying to solve, and then creating a shared sense of context with the people that you're working with. I think those things are fundamental to what it means to be a designer, engineer, product manager, or whatever in the future. and designers have always done. research themselves. They they've recognized this need to understand the constraints of the environment they're operating in. yeah, so I think the takeaway should be get better at research, get better at understanding or get better at collecting signal that matters,

Andrew Hogan: I love that. And I think that the skill of looking at a hundred different prototypes or something like that and then making some degree of sense of it. when you yourself maybe didn't actually make those things, using the context that you have might actually become really essential. And then can you bring five other people who aren't involved in it? Can you bring them along and facilitate some way of them participating? Because I just see this world where you're over overwhelmed is the wrong word. You have so many options. How do you find your way through it? What are the steps to even get there? that seems like a world of absolute abundance that is happening now that seems like it'll be critical to navigating your way through in the future. and maybe there's, ways you can write applications to help you do that. Maybe there's ways that you can use your, shared context creation Shane just mentioned to help you do that. But that situation seems like it'll become more and more critical in the future.

Dan: training judgment is definitely up there, right? When when when the cost to build is is is very low and very fat and and the speed is very fast, you're left with judgment. And yes.

Andrew Hogan: on on things you didn't make, on things that you didn't are given to you also very rarely end up in zero to one situations too. So there's a okay, we got this thing and here's what we could do with it. Where do you go?

Dan: And then talking to people because yeah, the technology is changing. Unfortunately, people do not. we change very slowly over over tens of thousands of years, alas. but maybe even hundreds of thousands of years. I'm not sure. When do we when do we start walking upright? two hundred thousand years ago? Something like that.

Andrew Hogan: Deep questions.

Dan: But deep deep questions for a different podcast than this one. But Andrew Shane, thank you for coming on the podcast. Great to have you both here to talk us through this. Look forward to seeing next year's report. Of course, we'll be linking to this year's report in the show notes. And yeah, any other place Our listeners can find you or find out more about this?

Andrew Hogan: I love LinkedIn and I'm on LinkedIn all the time posting non AI stuff. So happy to be there. I'm trying to bring Shane into that. Shane, is LinkedIn the right place for you?

Dan: Ha ha ha.

Shane: It is. the DMs. If you if you want my candid takes, shoot me a DM.

Dan: All right. Well great. Thank you both. Thank you, listeners, for joining us here on AI and Design. We will be back next week with more AI, more design, and we'll see you then.