Episode Transcript
[00:00:08] Speaker A: Welcome to Examining, a technology focused podcast that dives deep.
I'm Eric Christiansen.
[00:00:16] Speaker B: And I'm Chris Haunt.
[00:00:21] Speaker C: Okay.
[00:00:21] Speaker A: And welcome to another episode of the Examining podcast, the technology focused podcast that dives deep. I am Erik Christiansen and as usual I am joined by with my co host, Chris Hans. Chris, how are you this morning?
[00:00:35] Speaker B: Good, how are you?
[00:00:36] Speaker A: Terrific. We're getting started early instead of the afternoon, so I'm in much better shape. Today is an exciting episode. We're kind of trying to return.
For several years we've focused on doing more traditional shows, but Chris and I for many years also did interviews. So we're trying to like mix this up a bit. So today we have a, a colleague here or Randall, and I'm just going to let you introduce yourself about your background and who you are and what ARKit is.
[00:01:07] Speaker C: Yeah, thanks, Eric.
Yeah, Randall heard. So I'm a local Calgary entrepreneur and you know, Eric, you and I met recently at one of those Calgary user experience events and what I think really stood out was just this idea of building better tools for humans. How can we come together and make things that are in this time of emerging technology, make things better with, without automating the human elements? So, you know, you asked me how I got here and it's been a long, winding road through big tech and community volunteering and video game culture and human and technology adoption. But the, I think the common thread has always been this connecting line of learning and feedback. So, you know, reaching out to kindred spirits and whether it's an instructional design or product design, and just realizing that I, there's a, there's a lot of really cool things on the horizon for what's happening with emerging tech. And if we're really intentional about how we apply it, then we can make some really great impacts.
[00:02:05] Speaker A: And if we're not intentional, we might have a problem.
[00:02:09] Speaker C: Yeah, the.
There's. It's easy to go down the rabbit hole on the AI doom. Doom Slayer side of it too right now. But I mean, for, for ARKit, to answer your question, you know, what we've really sought out to do with our product is build a genuine tool for visual builders. Because we, we looked at the kind of the world and said, you know what, there's a lot of ways to build things and to ideate, but prompt engineering and writing text prompts seems to be kind of the new way to do it that everybody's convinced is the answer. And we don't think that's true. So we sought out to make something that Had a visual interface. You can just show, not tell, mock up your idea and then turn that into software. And the response has been pretty positive.
[00:02:50] Speaker A: So we'll get into. That's helpful and Chris will ask you a little bit more about ARKit and what it is and how it is from a kind of a position instead standpoint compared to like other UX tools. But how did you get to ARKit? So what is for folks who. Well, before we get into the UX stuff, but like we try to interview, when we interview people here, we like to know kind of the interesting winding roads they take in their career and how they get where they are because that makes it really interesting. And I think people are interested in how people's varied experience kind of comes together to any given point in time. So what was your background into tech before we got to arkit?
[00:03:28] Speaker C: Yeah, careful what you ask for. I guess the very first thing I ever really did in tech was I had the opportunity in Calgary to work the little bespoke video game studio in user testing and beta testing. So we did product testing a long, long time ago on video game design and that just really, really kind of like planted the seed. I was young and had too much responsibility and too much budget under my belt, so quickly became dangerous in that field and at the same time was actually competing in fencing. So there is a way that these threads can act. It was a high performance sport, you know, in a strange sport. It's not a very popular one as you can imagine in western Canada.
But I think Eric, you said you had some experience in fencing. I did.
[00:04:13] Speaker A: I took fencing lessons as a kid. I was never in a competitive aspect, but I, you know, I went through like a pirate phase. So then I think my parents were like, how do we direct this healthily into a sport? So I took, I was put in fencing lessons in that kind of our community.
[00:04:28] Speaker C: See Chris, if you said that you've also done fencing, then that's about 50% of the Calgary fencing community on this call right now.
[00:04:34] Speaker B: I haven't done any fencing, so not yet.
[00:04:38] Speaker C: You've got. Eric can show you some pointers.
But I think going back to your point, like the neat thing was those, those pieces combined. And then I had the opportunity to work with the city of Calgary in their community programming with, with youth and youth music programming and just really started weaving these threads together into this idea of, you know, there's so much that's possible if people are just learning and trying new things and exploring and technology can do a, you know, go a long way to help that exploration or maybe get in the way or convince people that, you know, there's other ways they should think or should do things. And I kind of opted for that first one, that kind of atomic era optimism of, you know, what if we can figure out, you know, psychology and mathematics and economics and we can, we can solve a lot of human woe. So that kept going and then I had the opportunity to work in big tech. So I was working with Apple for a while and that was in mainly in Apple retail. But I also had the chance to work with Apple internally across North America and regards to where I found myself. What was really powerful was, you know, diverse teams all across North America and they, at the end of the day, the teams that were able to test and learn and grow with feedback loops and trying new things and experimenting, whether it was the humans themselves or the tools they were using, always just fostered this really cool sense of almost like wonder. And I just, for lack of a better word, I almost became addicted to that and figuring out how to foster that and grow it in myself and in the communities that I worked with. So that was a long winding road. And then I met our founder, Nate Nagai. I met him at our previous company where we worked together, which is in Calgary.
And he reached out shortly after he left and after I left to say, hey, I think we can build something together. So that was just about two years ago now. And fast forward, we kind of pulled the curtain off of Arkit last December and here we are.
[00:06:32] Speaker A: And you studied psychology you said correct, if I remember in our last conversation.
[00:06:36] Speaker C: Yeah, that's right. Let's talk about Meandering Journey. I was in the Haskayan School of Business. I was in Kinesiology, all at the University of Calgary psychology and ended up with the communications and culture degree. And again that kind of common thread being, you know, how do people learn, how do they, how do they learn about things, try new things and connect, connect that to the values that they have and that just, you know, the tool culture loop and McLuhanism for lack of a better word, just really got into me. So I'm very passionate about that.
[00:07:08] Speaker A: Yeah, very cool, very cool.
[00:07:10] Speaker B: Well, maybe why don't you tell us a little bit more about our Arc Net and what makes it different, who it's for.
[00:07:19] Speaker C: Yeah, thanks Chris. Well, and Chris, as I understand you've got some experience in that, in that space as well, I'd love to hear a little bit and help me frame that. Like what are you Seeing in that, in that space of kind of, you know, helping students use AI tools or generative tools in new ways, because that's pretty cutting edge.
[00:07:36] Speaker B: Yeah, well, yeah. So, I mean, right before we started, I mentioned that I was one of the first to allow students to hand in written work using generative AI tools. And honestly, like, I've been even I mentioned this to my students just on Friday.
I, I've been reflecting over the last couple years and I've changed some of my delivery as well. So for instance, in class, I get them to physically write on paper. So I, I bring in the exam booklets there and I just throw them, I just plop them on the, the table. They're like, oh, what's this, like a exam that's happening or what have you. And you know, I get them to actually reflect every class, to give them a little bit of time so that they can jot down some of their, their thoughts. But I describe my brain as being a hard drive. So maybe some of those things aren't really optimizing what's in there. But I think writing, if we just allow this technology to go and, you know, take over that, at what point will we be able to communicate and, you know, convey our thoughts to other people?
So I don't know. I'm, I'm kind of, I, I still like technology. I, I'm a big proponent. But at the same time though, I've been also thinking, like from the generative AI, I don't know if these large language models, they feel like they're plateauing and I don't know if they are going to get any better, even though they've been throwing billions of dollars towards this. But again, for the purposes of the, the courses I teach, like from the business communication side, I think it does a really awesome job. When people are inundated with a few hundred emails a day.
It can certainly expedite the workflow. And again, it's just if you, either you use your brain or you use your brain augmented by AI and I think it's a, in terms of the workflow now what you're doing is critiquing the output and you still need to know, you know, like the, the best practices that way so that you can see whether the, the AI is outputted because it really, it isn't true intelligence either. That's where a lot of, I think the terminology that we're using out there, it's, I think people misperceive things. But anyways, yeah, I'm sure it's interesting from your side. Like, I do think, you know, everything we're going and writing stuff in, in these prompts, and I, I'm looking forward to hearing more about this visual approach.
[00:10:07] Speaker C: Yeah, yeah, thank you. Well, and I think, you know, you kind of, you hit it on the head like the, the experience that we saw, you know, at this point, I mean, it's maybe hyperbolic to say thousands, but it's definitely the high hundreds of interviews that we've been doing in our discovery work over the last year and a half.
And, you know, this thing is how many people in, say, product management or product leadership, software leadership are really excited about, you know, these different tools that you mentioned, but at the same time also just kind of inherently expect to take it with a grain of salt, expect that, you know, it's really cool, but you gotta, you know, hack it this way, or it works until this point, but it's still treated more like a fun product than a genuine tool.
And one of the things. And that's, and that's, I should say, that's not to disparage any of them. They are very great at what they do, but they don't do what we do really well. And that was where, you know, we kind of identified our opportunity with Arket. So at the end of the day, if you are, you know, trying to communicate something, if you're trying to show something.
You mentioned it, Chris, like these, these tools don't have any intelligence to them. And that's not a good or a bad thing. That's just the symptom of the jargon that we're facing at this era of vibe coding and prototyping and AI is going to fix everything in a day. So what we looked at was, you know what, rather than have somebody, somebody argue with a text prompt back and forth just to get their idea across, when, if you zoom out a little bit, it kind of seems a little bit absurd. As soon as you analyze it, you're arguing with a machine interface about your human idea and then taking what the machine is inferring and representing back and saying, hey, this is your idea.
I think, again, not to be too hyperbolic, but in some ways I feel like that's a really violent act against human creativity. And so what we said was, okay, well, you know what? Rather than you've got all these ideas, humans are visual thinkers. We've been that way for thousands of years. And now you have to sit down and force yourself to convert that into text via a keyboard. What if you could use the same model that we already do, which is we think visually, we represent visually, we draw on the whiteboard or the napkin or the sand or whatever it is. And so that's where we started with was this idea that rather than use words, why don't you use visuals and show, don't tell. So with the ARKit interface, the wireframe is the prompt you drag and drop, you mock it all up and then you show it what you want it to build. And that's what it does.
And so the beauty of us being our own software product that is powered by these foundation models, but we're not dependent on any one of them, is that we are able to continue innovating and experimenting with this stuff. But the reality is that a year from now, maybe the tool, the AI products out there, the foundation models, they've changed a lot. But it doesn't really change too much about what we're trying to do. It just makes it easier and better and faster. But our core focus is making it easier to show, not tell, your ideas. And that's been pretty powerful so far.
[00:13:36] Speaker A: So for people who are listening, because we're audio specifically, this is a prototyping tool.
Primarily if you were designing a website or an interface or an app, I suppose it could be used also for any sort of infographic, any really visual process. So it wouldn't have to be like a web interface per se. But some people who've done prototyping and design.
One of the primary desktop and web applications is figma, and you and I spoke a little bit about this before. So how is arkit different than figma? Do they coexist together? Are they in competition? I think for the average individual who's seen prototyping at a distance, they may not need that clarification.
[00:14:22] Speaker C: Yeah, that's great, thank you. Well, I think the thing that stands out to me is if you talk about figma, the first thing that people think of is design is, hey, this is a tool for designers. And figma has done a really good job of building that over the years and building some really good credibility in that community.
I think the reality of that along the way is that it has also become a really kind of heavy duty tool. It can do a lot. You can push every pixel in the right spot, you can get everything looking exactly the way you want to do it and craft this really kind of elaborate experience, which is really great when it's the right tool for the job.
The thing that we identified in a lot of our interviews was that you might have an idea from a customer feedback, say, or a user interview, or maybe you're doing some discovery, experimentation or a B testing or whatever it is. And okay, I want to test this idea. I've got something, I want to see it. I just want to make a quick change here and mock it up.
Teams are facing such pressure in this kind of world of AI is going to fix everything in a day. They're facing such pressure that they look at something like figma and they say, you know what? I need to work, but it's too slow, it would take me too long, would take me too many resources.
I can't pull my designer, I don't want to pull my developer, whoever is responsible for that type of implementation.
I've already mocked it up on my whiteboard. I already have an idea of what I want it to do. And so we said, okay, great, if you want to really get specific with every single pixel, great, you can do that in figma. But if you really just need a disposable MVP prototype to prove something out, mock up an interaction.
We build the whole front end so you can do an infographic. To your point, we've had people throw up, almost like, choose your own adventure condition pass for AI chatbot simulations, logistics operations, transport operations panels.
If you can think it on a whiteboard, you can probably think on it in arca too. So really just making that idea of prototype and interactivity accessible beyond figma, because a product manager, a founder, as someone who's just in a startup, maybe they don't have the expertise or they don't have even the budget to subscribe to, say, a monthly figma membership. And they don't need all that power. What they need is to be able to mock something up and test it quickly. And so that's where we come in.
[00:16:49] Speaker B: So with, with this, like, you know, in terms of, I'm just trying to think, like, what would you upload? You could upload, like even a sketch. You could upload, you know, some files from some sort of, you know, I don't know if you have.
Whether it's infographic or if it's maybe actually a wireframe type of file. You can just upload it in there as well.
[00:17:12] Speaker C: Yeah, yeah, you got it. So I was, I was telling Eric I was getting a little bit cocky the other day. I was doing some testing on a new feature that we just released, and I found an RFP for a local company that was, you know, hey, we're building out this platform, you know, we're accepting Bids. Here's, here's what it needs to do.
And so I took a screenshot from their website and I kind of just kind of one to one, drag and drop to match that screenshot in our tool and then pull the trigger. And in about 15 minutes, I had an interactive prototype with all the layers and all of the content mapping to be able to fill the requirement. Now, I didn't submit it because I didn't go through any of the process to do so, but the idea being that you can reproduce something that exists from a screenshot or from a visual, or you can start from scratch if you know, hey, I've got something I just want to scratch out. Okay, great, do it. And if you're like me and maybe you're scratching something out on a napkin or a piece of paper, it is possible to take a picture of that and throw it in there. And we'll do our best to interpret it as well.
[00:18:17] Speaker B: Cool.
Eric, you want to ask?
[00:18:20] Speaker A: Yeah, I was just kind of curious. So it's interesting to me.
So there's been wireframing tools in the past that are like ultra basic. There's one that looks almost like chalk. It has like a chalk design and it has like one font that looks very Comic Sans. And I have it bookmarked somewhere, but I always forget the title of it. But it's really good for that quick interaction. But then of course, you can't transform it into anything.
Right? And then figma, you can drag and drop boxes, you can simulate what you want a desktop website to look like, but then to refine it, you have to do it all manually. So I get the impression that arkit kind of lives in between. Like if you're not the person doing the finalization of the design, but you have a but maybe because that's not your specialty, and we talked about specialty before the interview and the value and drawbacks to specialties.
So if you're like a product manager, you could draw out the frames, but then have AI generate something that looks more polished with the idea that it would be perhaps handed off to the web developer leader. I mean, I guess the question I'm asking is that are people going from kind of arkit to Figma or are they actually skipping that step? Are they going arkit? Hey, this is really good. This interaction is what we want. Straight to development and design.
[00:19:34] Speaker C: Yeah, I mean, it's funny you say like that. It's kind of, maybe not the best answer, but it's kind of both because we, with our outputs, you can get that interactive HTML and run with it, upload it to a testing platform, put it over and share it with dev, share it directly with a customer.
You can also connect it right up to GitHub. So if you've got a database or other information you want to connect there, that's totally okay to do.
I, you know, the whole objective was how do we make it really easy to keep the momentum of your idea, you know, whatever, whatever that means. And I realize that's kind of fluffy, but you've got an idea, you know, rather than slow it down with text prompts, let's just keep the momentum going. Show, build it, and now keep that running. Maybe it's to show to your developers and communicate that internally. Hey, here's what we're trying to do from a feasibility engineering perspective. Or maybe it's exploration, that's okay too.
And the cool thing is that because it's all already built in an interactive way, the technical feasibility is largely addressed because if it's possible, then chances are it's going to reflect that in the generated prototype.
And if you want to make some tweaks or if you want to make some changes, specifically, we have tools that allow you to effectively arm wave and say, hey, you know what, change this font or animate this this way, or hey, I don't like the way this data is laid out. Can you flip it around this way and it'll regenerate? Just that little portion that you point, it doesn't rebuild the whole thing. You don't have to argue with the AI. It's really meant to be a tool, not a whiteboard marker that talks back.
[00:21:07] Speaker A: It's really interesting because I like the design in Figma, but for me, and you could say the same for Photoshop or actually, I really liked Adobe xd. It's unfortunate that it was deprecated because in some ways I think it had advantages over figma. It was a lot easier to create the animations. It seemed like their process for adding animations and transitions was more straightforward.
But that to me seems like it takes an extraordinary amount of time, the mocking up of interaction. And I think about what Steve Krug, the usability icon, said, that you're supposed to test early, like test a drawing and give it to a user to see if they even understand what they're looking at.
So I guess a related question. So now that we have AI and we can kind of augment and make some of the interactivity a lot easier.
And maybe ARKit is part of your answer, but in General, what do you think the future of UX would be? I mean, when I was in grad school, it was still like usability.
You still saw the title usability consultant or usability specialist, you'd still see information architect. Then you saw user experience, researcher, designer. All these titles and job titles are all kind of mixed up.
But do you think that things like researcher, UX designer will be relevant in the AI era? Like, how do you think more broadly as AI gets kind of inserted less as kind of like a standalone chatbot, like Claude feature? And I know that I think you're using anthropic for your, as your frontier model, you said, but as it's kind of inserted into AI specific tools, like what are these jobs going to look like? Are they going to be relevant? Do you think they're just going to be different?
[00:22:48] Speaker C: Yeah, I think as a single voice in this, you know, I take that, take that with a grain of salt too. But I think the thing that stands out to me in the optimistic side of it is as the, as the cost. And I say cost like the mental cost, the cognitive load as these elements have become a little bit maybe reduced by the performance of these tools, making it easier to, you know, say, mock something up or test something or generate work or, you know, you get this new terminology around knowledge workers and how they're able to connect disparate domains together in ways that, you know, are faster than they could before. They're working with unstructured data and combining them in really creative new ways.
I think the thing that stands out to me a lot and what I've seen from the community is that it's getting easier to test and to build and to do some of the things maybe that people always knew about or should have been doing and maybe didn't invest as much in. And one of those areas is in accessible and inclusive design.
[00:23:55] Speaker B: Sure.
[00:23:56] Speaker C: Making that a lot easier now because for the longest time it's been such an afterthought for a lot of organizations to, hey, we'll worry about that when we get to the end and we'll just flip something and away we go.
I don't think those need to be exclusive domains anymore. And so when you asked me about UX what that kind of research looks like, I'm really excited by the perspective and almost like the broadening of possibility that comes in that space because now it's like, oh, these things that before were maybe treated as burdensome because the organization didn't place a huge value on it doesn't need to be as burdensome anymore. It's a lot easier to bring that kind of thinking in to recognize, hey, we know we can build something for a wider audience, which is, you know, always a great thing for a business building technology to try to get their user base out there.
And I think the, the other part of that is that it, in some ways, if it's applied thoughtfully, almost gets people to be more intentional about that human element. So, you know, ux, research, experience, design, I think if anything, those things are going to become all the more important because, you know, Chris, like you mentioned earlier, these are just, you know, there's no intelligence here. This is not something that is experiencing human experience. It can't see anything. Part of the problem we face in the the archive challenge is so interesting is that we're using visual cues and visual inferences and it's with the tool that has no sense of what that is. And so we had to build an entire software around that concept and we've done a really good job. But it also means that, you know, that's not something that can just happen overnight. So I think that to answer your question, UX is not going anywhere anytime soon.
I do think we need to be cautious about some of the ways this can be used in a more negative way, whether it's in dark patterns in an actual UX that's maybe for trickery, or other ways that it can be used to subvert maybe user expectations in a more negative way. I think we need to be mindful of that. And I think at the end of the day, it's going to help us expect better.
Again, going back to the optimist in me, I think that we can have higher expectations and better expectations of what can be done because we know that these tools are out there. And so I think it's okay for us as users and consumers and people that are purchasing and investing our time and money into these products to expect that they respect us, if that makes sense.
[00:26:26] Speaker A: Can I ask a follow up to that? So one of the things that I've always been interested in, and Chris and I have talked about this on the podcast too, and please, Chris, chime in. I'm thinking of examples on the fly. But, you know, technology's introduced and even if it doesn't make a huge splash, it may have a lot a big change in people's behavior. So, like, I kind of straddle between the UX UI world, but information science, right, as a librarian. And so I think about search and search changed, like with Web search, it changed how people think about what they reach for when it comes to information.
Nielsen Norman Group, for instance, has said that they pause it. I shouldn't say that it's proven yet. Their working hypothesis is that AI is really changing how people search for information. Because we've trained people for years to have these really boiled down searches, keywords, keep it simple.
Well, now when you go to search with AI, if you're using a deep research tool, being a bit of a blabbermouth actually works to your favor. It's the opposite.
I think in the product UX world.
The analogy that I'm trying to, or the corollary that I'm trying to bring forward here is, is AI in design tools, do you think going to change how, let's say the product manager talks to the developer or are they going to get better at that?
If you have to describe all the elements in a. Because from what I'm gathering, you draw a user interface. You've shown me before, you can kind of put comments on each section in the user interface that you've mocked up and that helps the AI build it better.
Does the act of doing that commenting, which has almost historically been like you comment in code, you might comment in a document. But is that going to change how people communicate between say, different cross functional teams and a UX environment? I wonder if it's going to change the psychology.
[00:28:20] Speaker C: Yeah, I think, you know, I'm dating myself a little bit here, but I'm old enough to remember the really early days of the Internet, which I. Going back to your point about search. Yeah, right. And there is this, this feeling of, of exploration that, you know, is totally foreign now, but this idea that, you know, you didn't know what was out there on the Internet, you know, maybe your search terms you were figuring out on the fly or, you know, websites and web hosting was kind of the wild west. And so as a result, you know, you kind of crafted this, you know, adventure pack. Okay, these are the searches I'm going to use, these are the websites I like to try to find and everything to aid in that pioneering. And I think that we're kind of at an interesting point now that echoes that sentiment. You know, search for the longest time.
Once, you know, a lot of paid ads and paid advertising got into that space, it became less about exploration and more about, oh, this person's looking for something, let's get them this, this message with an ulterior motive or an advertising budget behind it as quickly as possible.
And I think that now that were looking at these AI search tools and thinking about unstructured data and access to all this, and you talk about being a blabbermouth can really work in your favor.
The design layer that we really looked at when we were building our tool was if you're whiteboarding and you're really excited about your idea, you're working with your small team and you're whiteboarding your idea, chances are you're not doing that in silence.
You're drawing up on the board and as you're doing it, you're saying, okay, this thing is going to happen here. This is going to go, what do you guys think about this? And so there's this again, that momentum, that energy in that space. And that's what we really wanted to capture with that layer that you mentioned, the commenting layer in our tool.
And so in my mind, it almost connects back to the piece about UX is the human element of how these tools are used and the way that we communicate, if anything, because it's now possible to, you know, build a fully interactive, like end to end prototype in 15 minutes.
You know, that in theory is going to mean that these teams are going to be working together more frequently and more often. And so the interactions, the friction, the feedback loops, whatever it is, those become a lot tighter just by virtue of the fact that, hey, I can get a prototype and work with the team in half an hour rather than three days.
And so I think that acceleration is really exciting because we sought out to not automate the human element of the craft. We just wanted to tighten up these spaces where the machine's already doing the work. So let's just make that faster and make it easier for the humans to have more time with other humans, whether that's more user research or with teams in other departments. So I think the arm waving of the whiteboard and the commentary is not going away anytime soon.
But it also means that the teams that keep that alive are the ones that are going to invest in that type of experience.
If you're a team that is a feature factory and just building on generative AI over time, you're probably not going to be doing a lot of that type of interaction. And if you're a team that values the human experience, you're going to be talking to humans, you'll be doing a lot more research and you'll be working a lot more collaboratively. So I think that's the cultural shift that really excites me and where I see things going well, that's a great.
[00:31:42] Speaker A: Segue to your question, Chris?
[00:31:44] Speaker B: Well, maybe before we get into that, like, I sometimes wonder from the psychology of things, like, I've seen this over the years where, you know, especially with user testing, if you go, as opposed to actually showing them, hey, this is, this is our user interface and this is what it's going to look like. Here's the app and all the interactions.
I find quite often if we show them sketches and show them like, hey, here's a sticky. And this, this, when you press this button, it's going to lead to this.
The people actually feel more engaged in the process, that whatever they're going to contribute is going to make an impact in the overall, you know, design, as opposed to if you show them a polished, you know, interface. And so I sometimes wonder about that with the, sure, we're getting things a lot faster, we can go and, you know, have these prototypes. But then I wonder from the, the user testing, what do you think of that? Like.
[00:32:43] Speaker C: Yeah, I think, to your point, Chris, the thing that's exciting about that, and, you know, a lot of the teams that we've talked to, this is a famous, you know, saying nowadays, but the whole, you know, whether it's anachronistic or not, not anachronistic, the other word for it, but the, you know, if I'd asked people what they wanted, they would have said a faster horse. That whole idea of don't listen to your customers, they don't know what they want, don't listen to your users, they don't know what you want. And I think that's such a terrible way to think about human centric design because it's misconstrued. It gets turned into this, like, don't listen to your users, they don't know what they want. When really what it's saying is don't take their feedback at face value, don't listen to the first thing they say. And I think that the thing that really stands out to me in that space is that, you know, if we, if we say, oh, well, don't listen, they don't know what they want, then we're discrediting the humanity of the craft in the first place. If we flip that around to say, you know what, let's participate and share and build this thing together, whether it's, say, to your example, like, here's a sketch, like, what would you do next? Can they even figure out what they're looking at from these early stages of a design?
There's that human interaction there, and we think that time every tool out there, every product out there is going to say, hey, you'll save time. Hey, you're going to do this. So time is a really weird thing to, to put up there and say, hey, this is what we market around. But I think that what's really powerful is that if you have the resources to be able to invest in more of that human interaction, you see the benefits very quickly. And so whether it's the, hey, press this button and it'll lead to this, they feel more engaged because they just get a chance to see kind of behind the curtain.
And the fact that the organization, the business, whoever it is that's building this is demonstrating that they value human interaction by investing in that type of research in the first place. And I think time and time again, the products that we see out there and the really cool stuff that we see out there, that gets me excited. Inevitably these are teams that are talking to users all the time. They're always doing research, they're always figuring out, hey, is this whether it's UAT or UX studies or experimentation, there's always some form of conversation there because they know that, hey, someone might say they want a faster horse. That doesn't mean, don't listen to them. It means that there's some need there that we have to unpack, experiment with and understand what's happening. And we just really wanted to make tools that would make that whole kind of exploration easier.
I don't know if I want to. I get excited about this, Chris. I might have gone a little bit.
[00:35:24] Speaker B: Yeah, no, no, it's good.
I look at like even one thing that you mentioned too. Even. I just read something recently about like introducing ads and I guess OpenAI Sam Altman is now apparently going to be launching ads in as part of ChatGPT. And so I mean, and then on top of it, apparently he's also creating or investing in something that's similar to neural link. So I, I don't know. Now I wonder who knows. I mean, I would have, I mean they're losing a lot of money, but they're still getting thrown a bunch of money. But now you basically poke the bear, Elon Musk and I don't know what's going to happen with that, but, but in any event, I think going back to what we had in mind for questions. So Amara's law states that people tend to overestimate the impact of technology in the short term and underestimate the long term.
So what do you think the long term impact of AI will be? Let's say in the next five, 10 years. And what are people not talking about?
[00:36:30] Speaker C: Yeah, I think, you know, if I had my generative AI crystal ball I guess, but the, the AI optimist in me, I really like to think about what does it mean for these tools and products to help us move in a really positive direction into that kind of like post scarcity, you know, culture.
[00:36:52] Speaker A: Where.
[00:36:54] Speaker C: These are these large, large models and systems are able to handle things at a scale that, you know, is never been seen before in, you know, civilization.
And, and yet we still also feel the, you know, almost like the trappings of this kind of pre technology capitalism and this world that's saying like, hey, this is how we're making these things happen. You just, your example of, you know, bringing ads to OpenAI is such a great example of this kind of tug of war between the way that the businesses happen and what's possible for humanity. And so I think right now we're going to feel those growing pains, at least for the short term as people try to figure out, okay, what can we do with these things, but also how can we monetize it.
And then I don't necessarily agree with Musk's statement saying that you don't need to save for retirement because AI is going to make that obsolete. I think that's a little too optimistic or maybe unrealistic.
But I do think that there is a lot of potential that we're starting to see the glimpses of for that type of, you know, long term automation. What I think we need to be careful about is using it in a really intentional way because, you know, 10 years from now I'm not interested in AI automating, you know, say music and art creativity. I'm, I'm, if anything I'm more interested in what humans are now able to do with that because there's, you know, there's less resource drain on them in other parts of their life. So I think that that gets me really excited and the, the fun thing that, you know, can come out of all of this is that you've got in theory, more time and resources that weren't dedicated to things that maybe you didn't have to worry about anymore because these tools have managed to, you know, take some of that burden off of society. So if I think about, you know, some of the problems out there in education or healthcare or even unhoused population, there's a lot of areas there that some unstructured data tools can go a long way to helping. We just need to, as a Culture and a society. Invest in that direction.
[00:39:05] Speaker B: Absolutely.
[00:39:08] Speaker C: Again with my meandering answer, Chris. I don't know if that got it for you.
[00:39:11] Speaker B: Yeah, no, I mean, it's.
[00:39:13] Speaker C: What about you? What are you seeing with your students? Like, I'm curious because one of the things that Eric and I connected on when we first met was we were at a talk. And it's easy, I think, sometimes to make students out to be the villains, and they're trying to use these new hacker tools and they're trying to cheat on tests and everything. And I just think that's such a disservice to this generation of innovators. I'm so curious what you're seeing from your side.
[00:39:40] Speaker B: And this is the thing.
I don't know. A lot of people, they take the negative approach to this and just, you know, are basically like, you're saying, like, they're kind of villainizing the, the students. But I mean, from an academic integrity standpoint, I mean, cheating's been around forever, right? It's nothing new. It's just now. And I think what the, the big thing is, you know, people need to redesign and reimagine, you know, what they're actually testing for. I mean, I've, I've had students, you know, that are in Comp Sci what would normally take them, you know, days for an assignment. They put it into Claude and it pumps it out in like five minutes. Well, you know, if I was the Prof. For that course, I think what I would probably do, given, you know, there's some students are doing that, others are still grinding it out over the few days, I would probably change the assessment. And so, you know, knowing full well that there's probably going to be some people using this and now maybe have them explain the underlying code that's been developed or maybe ask them to produce more if, if they can go and pump out in five minutes, what would have taken days. Now maybe I asked them to just, you know, take. Use them full tilt and, you know, produce X, Y and Z, which would have taken, like, who knows, like, like you're saying your, your tool is pumping out things in 15 minutes, which normally, who knows, it might have taken a month.
Right. So, but again, it's.
And keep in mind, too, like, it's like academia goes a little bit slow. There are certain policies and procedures, and I think this goes for anything. Like you, you talked about healthcare, for instance. I mean, even just changing a form in the healthcare system is very difficult.
You, there's a lot of stakeholders there. There's there's rhymes and reasons for it. I mean, there's people's lives are at stake and every day there's people dying because of, you know, just, it's, it's really, it's just like human error. Right. Because of how the system is. And so they're trying to.
That's why like, I mean, I, I just recently, this past fall, I, I was in the hospital and it was funny because the, the nurses, every time, they would always poke and prod and they would check your blood pressure and all this kind of like they would ask you certain questions, but they do that on purpose because from shift to shift and they, you know, they just want to make sure everybody's okay. Right. And so again, there are mechanisms in place. There are reasons. I, I think we need to go and question some of those things given the sheer amount of data that can we be processed, you know, through this. And again, I, I don't know if AI is the best word even. I mean this really what it was is if you go back to like there was machine learning and you know, there's like a certain workflow or what have you, I think this has just taken that to the next level and we're way, like from a AGI, and I don't even know if we want AGI to be honest. Right. But I think we're way, that's like way into the future. But there's, with any technology, there's always going to be impact and, and some pro, some there might be cons. Right. I mean, I still, I'm kind of like you, I'd like to think optimistically, but I, I think what's happening right now, I look at an industry, pretty much everybody is, you know, telling the, the workplace, you know, here, go use AI. But it's not like it's actually cutting down on your time.
They're demanding more. And I mean I look at like Shopify, I think, you know, their whole. With Toby's memo that got leaked, I mean that, that was one of the things where it went viral. But even to make new hires, you have to justify why can't AI do this as opposed to going and hiring X people, X number of people. And so it's, it's become kind of interesting where we're literally demanding more of our, our workplace from our workforce. And so I don't know if it's actually making things like freeing up time like how you're saying, for like creative endeavors or whatever. Like, we're just, we're just doing Way more work because these tools are allowing us to. And that's the expectation from all the managers and everybody.
[00:44:13] Speaker C: Yeah, and that's. I mean, that's a great point. The. There was a paper I read.
Oh, I can't remember. It was Igor Denisov, bench. I met him at a session last year and he had done some research with the team, demonstrating that the team for code and code reviews. So AI generated code. You know, they were boasting and celebrating internally. And this is across multiple research points that they had reached a, you know, 15 to 20% improvement in their productivity. Hey, you know, hey, this has helped us out in all these ways. But when they actually analyze it, it was the opposite, was almost exact opposite drain on productivity because, you know, it felt good in the moment, as we tend to be as, you know, primates, we like action bias. But then it kind of like scattered out. And when they brought this all together, realized that actually the time that they'd saved it led to, you know, drains elsewhere. Because now there was more review, needed more corrections, needed more work going back and forth to kind of stitch together what they tried to. Tried to create. And I think, you know, your point about students is something that really, really rings true to me is this idea that, you know, if this is a product being used as a supplement or replacement for something, then that's not going to be as effective as, or beneficial as, you know, a product that can also be a genuine tool. Does it actually extend the capability of the person using it? Does it make them possible? It makes something possible for them that wasn't possible before without them really having to think about it. You know, I used the analogy earlier, but I don't want my whiteboard marker to argue with me. And so, you know, I just want to get up there. I forget about the marker, I forget about the whiteboard. I just focus on the idea I'm building. And I think that if you're using some of these tools to replace or in supplement a supplementary relationship with some of these skills, then, you know, that that does a disservice. And I. That's why I use the. The word product versus tool very. I'm pretty harsh about that. Sometimes I feel like I have to argue with something in order to make it work for me. It's not much of a tool.
And that's really what we sought to fix with what we built.
[00:46:30] Speaker A: It almost sounds like your definition would be a tool needs to.
We mentioned when we chatted before about the hammer, it's like affordances. There's only so many ways you can use it. But also it extends people's first principles knowledge.
I think that's what Chris is also alluding to.
But we can always get stuck in the weeds of these bottlenecks. Is the bottleneck your understanding of design, or is the bottleneck the complicated design tool? Is the bottleneck that you don't understand the logic of programming, or is it that you don't know the syntax for rust? And those are very different questions. And I think that sometimes in an assessment context, we evaluate output and details, which are kind of the bottlenecks, but. But they don't necessarily demonstrate first principles. So I guess the optimistic view, Chris, and maybe Kristen, if you agree with me, AI would at least give us the option to focus more on assessment of first principles, on how did you get to this answer rather than how did you get stuck in the weeds of some sort of syntax?
[00:47:43] Speaker B: Well, I mean, I even look at like the, Again, this is where I don't know if AI could even figure out. Well, I mean, again, these large language models and stuff. But let's say, for instance, electric vehicles, right? I mean, to charge an electric vehicle, we don't need to go and make it the same as like where you, you know, you put your nozzle in for the gas. But they decided from, as a design feature to just keep it that way because that's why human beings for how many ever, like, you know, decades now, that's how we've been filling up our cars. I mean, they could have had some other completely different charging mechanism. But again, that's something you have to kind of think about because again, you know, the psychology of us, what we're used to and overcoming that. And that's why even just from a interaction, like, it's funny, I, I honestly, I think we're probably going to develop new, you know, user experience, interaction design kind of approaches. It may take some time, but right now, like, if you look at like, you know, Apple and Google, they release their own wireframes. I mean, you can go and download them. They're on, I believe they're on GitHub.
And so they have basically been telling everybody how to design.
But, and sure, they're, they're obviously some of the brightest people out there, but I think there might be more, there might be other ways to do things. And I mean, I look at like this whole like Apple has launched this Apple glass and then they put that little button there that I don't think that the little like hovering thing, it doesn't Help me much like to get to my note. I could have just done another interaction and get to the notifications or what have you. Right. It's.
I don't know if they fully thought through that whole design interface.
[00:49:29] Speaker C: Yeah, I think, you know, the, the piece you said at the top around, you know, whether it's a hammer or a tool or this idea of, you know, first principles knowledge, I think connecting back, I think it was Ivan Zhao, the founder of Notion, I think he said recently about treating AI like a new industrial material.
And I find that a really interesting perspective because I think that, you know, we talk about, oh, AI will, you know, do this thing or it'll do this thing. And that kind of glosses over the fact that it doesn't do anything on its own.
It's a powerful compute and in the case of LLM, it's, it's unstructured data, that's great, but it's how we work with it, how we work with that as a material. What do we do to bring that forward and you know, kind of figure that out, what are the tools that, that can bring that forward.
And I think that perspective I really enjoy because it kind of removes the almost anthropomorphization of AI as this thing that's going to do all these things on its own. It's like, no, this is a material that we can choose to use. And I mean such as we have a choice given that these foundation models and frontier models are all held by large companies that have the keys and the black box. But I think the intentional and thoughtful application of these to turn them into tools versus toys or slot machines that you can type something in, hope it gets what you want.
I think that is pretty exciting because it almost forces you back into that principles place.
Careful how I talk about the liquid glass design, but I think that there are some really cool things, both from Apple's liquid glass design and Google's material design, that are almost seem like they're anticipating what the next five or 10 years are going to look like. Hey, how do we start planting maybe some of these really early seeds right now so that we can kind of get a sense of how people are going to interact and use this so that two years down the road, three years down the road, when the materials or the technology are at a place where they can do more with it.
They've already done some of that research.
And it's funny because the Google, the Google Labs thing of, you know, we run 10 to 15 prototypes every week, that was Kind of our benchmark. We looked at that when we were building ARCA and said if a team wants to do that many prototypes a week, they need a tool to do that. And so that's, that's where we came in. And so I do have frustrations as I'm trying to learn the ins and outs of some of these new UIs. But I. Yeah, I'm. I'm excited for what that could mean for the future, and I just hope that we end up on the optimist side.
[00:52:16] Speaker B: Interesting, for sure.
[00:52:18] Speaker A: Well, I think this is probably. I could probably add more and more to that question, but I think we'll probably move on to our last section of our interview. So at the end of our interviews, what we do is we do a rapid fire question section.
These are not serious questions.
You can opt out if you can't. We don't want to force people to make one. These are kind of either or questions that we have. And it's just. The reason we ask is that it humanizes our guests a little bit more, makes them more relatable in terms of personal preference. And of course, super biased because Chris and I put together these rapid fire questions, and so there's absolutely no UX testing done on this whatsoever. We totally invented them, but we've solidified them down to a list that we think is fun.
So are you ready for the rapid fire question section?
[00:53:06] Speaker C: Yeah, let's do it.
[00:53:08] Speaker A: Okay. Coffee or tea?
[00:53:11] Speaker C: Coffee.
[00:53:12] Speaker A: Mac or PC?
[00:53:15] Speaker C: Mac. PC for gaming. Mac for everything else.
[00:53:18] Speaker A: Okay, fair. IPhone or Android?
[00:53:21] Speaker C: IPhone.
[00:53:22] Speaker A: Standing or sitting desk?
[00:53:24] Speaker C: Standing.
[00:53:25] Speaker A: Okay.
Star wars or Star Trek? You have to be careful with this.
[00:53:33] Speaker C: Both TNG and original series are some of my, my favor. Television I've, I've ever watched. And I mean, Star wars is foundational, so.
Yeah, that, that's. I. I can't, I can't split that. Nice try.
[00:53:46] Speaker A: What's your favorite car?
[00:53:50] Speaker C: My.
My Honda CRV that had 460,000km on it before I got rid of it. That was my favorite car.
[00:53:57] Speaker A: That's good. That's a good endorsement.
[00:53:59] Speaker C: Not. Not a Honda plug, but it was, it was pretty impressive.
[00:54:02] Speaker B: Yeah.
[00:54:02] Speaker C: And in Calgary winters, that was. Yeah, it was great, right?
[00:54:05] Speaker A: Ebook or paper?
[00:54:07] Speaker C: Paper.
[00:54:08] Speaker B: Okay.
[00:54:09] Speaker A: Favorite or open source tech?
I could. Now that could be physical, technically or software.
[00:54:18] Speaker C: I really like what the.
I. I don't know if I have a specific one in mind, but I really like this kind of new era of like note taking apps. Like what we're seeing with some of these. Almost like the next Generation of open office kind of style stuff.
I don't use it personally, but I think it's very cool for what it speaks to in the industry.
Terrible answer. There you go. Eric.
[00:54:42] Speaker A: So not LibreOffice?
No, it's a little clunky. I'm sorry. Corey Doctorow Synchronous Asynchronous or hybrid learning?
[00:54:55] Speaker C: Hybrid.
[00:54:55] Speaker A: Okay, web browser of choice. Now we have a list. Chrome, Edge, Firefox, Safari. But you can pick whatever you want.
[00:55:04] Speaker C: Safari and Safari on Mac and then Firefox or Opera on Windows.
[00:55:11] Speaker A: Oh, interesting. Opera, interesting.
Google Meet, Skype, Zoom or other.
[00:55:17] Speaker C: Google Meet or FaceTime Fair.
[00:55:22] Speaker A: AR or VR or both?
[00:55:26] Speaker C: Oh, both for sure. AR for practical applications. And what's really cool there, Maple Scan is a really neat implementation of AR tech for those Canadian listeners in your audience that you can literally like scan a product and it tells you how Canadian it is when you're at the grocery store or something. Super cool, built by a Canadian guy who had some really cool stuff in the prototyping world. So check that out.
No affiliation with them and VR because I'm a huge Valve geek and so when they announced Half Life Alyx, I I ordered a Valve index immediately and I was done.
[00:56:04] Speaker A: Sold cable or streaming?
[00:56:08] Speaker C: Streaming.
[00:56:09] Speaker A: Okay, who inspires you?
[00:56:14] Speaker C: Trent Reznor is the frontman for Nine Inch Nails and really inspires me in a lot of different ways because I grew up as a grungy punk goth industrial kid, but also because he cut his teeth working the midnight shift at a recording studio so that he'd go in and use the tech for free, went through the 90s like anti everything era and the heroin died and now is scoring Pixar movies. So just what a great career and undying fire of human spirit and creativity.
[00:56:49] Speaker A: Okay, well we can stay with some music then. So this was kind of contingent on that answer. So, favorite jazz group, if you have one.
[00:56:59] Speaker C: I don't know if they're my favorite, but I've been listening to a lot of Comet is coming recently, which is really interesting. I think they're out of the UK kind of brass jazz ensemble. Very strange, very fun.
Favorite drummer, Ooh, Danny Carey from Tool, which I thought was already pretty impressive, but after I watched his performance, I watched one of his performance online and he arranges his drum kit in such a way that he can channel these metaphysical spirits through him to keep his rhythm, but then change his rhythm four or five times or change his time signature in a single song. That just blows my mind.
[00:57:36] Speaker A: Well, those are excellent answers. Well, we really appreciate your time today, Randall, and talking to us about ARKit, but also your views on the future of kind of AI and ux. So thank you so much for taking the time to talk with us on the Examining podcast.
[00:57:50] Speaker B: Awesome. Randall.
[00:57:52] Speaker C: Thank you. Eric thank you. Chris Sam.
