AI in Design and Research: Revolution or Regression?

John Cleere, Kevin Hawkins at UXDX EMEA. Video: https://youtu.be/er-FKai11Ws

Readable transcript: edited from the recording's captions for readability (fillers and false starts removed, punctuation and section headings added). Wording is the speaker's own. Timestamps are positions in the video. Names marked [?] could not be verified against the audio.

The format and the debaters

[00:00:07] Host: We are about to have a debate about AI, which I'm sure won't get heated, won't get ridiculous, although one of the participants has already said he will turn to physical violence if it doesn't go his way. So I'm really looking forward to moderating this and going home with a very professional black eye at the end of things.

[00:00:23] Host: How this will work: in AhaSlides you will be able to see a poll after each of three questions. For each question we will have two minutes from one of our debaters, two minutes from the other, and then an aggressive 30 seconds to rebut whatever the second person has said. Then you get to decide who won the question. This is a conference, but with winners, and you get to decide who wins. So three questions, you will vote who has won for each of those, they will do a summary, and then at the end you will get to decide on who overall has won the debate. I'd say any questions, but I think it's going to be hard to take questions, so open AhaSlides, hope it works, ask the person next to you if it doesn't.

[00:01:10] Host: It looks like we're good to go, in which case it gives me great pleasure to welcome to the stage two Titans of UXDX, I think we would call them. We have Mr John Cleere to begin with, and Mr Kevin Hawkins. Everyone get out of here, I need space to awkwardly stand between you. Okay gents, how are we feeling? Kind of scary, isn't it?

[00:01:34] John: Yeah, a little bit scary. It's kind of like Terminator One really, the movie, where Kevin is the Terminator with all this machine stuff, and then you're like Sarah Connor in between it all.

[00:01:45] Host: That is who I've always wanted to be.

[00:01:48] John: Stay out of it.

[00:01:49] Host: Okay, I'll stay out of it. Incredible.

[00:01:50] John: And then I'm like the guy who comes back from the future. I cut Kevin off at the pass at this inflection point of AI.

[00:01:59] Host: And if I'm not mistaken you're going to play Johnny B. Goode at the end on a guitar?

[00:02:02] John: No, wrong film. Wrong Back to the Future film.

[00:02:04] Host: Okay, never mind, we'll move on. Right, so you know what's happening. We are going to start with you, John. We're going to give a two-minute intro to how you feel about AI in design and research. Are we heading for a revolution or are we heading for a regression? Kick us off, here we go.

Opening statements

[00:02:23] John: First of all, I'm not a Luddite about this. I'm more like a Cassandra. I'm looking around the corners and seeing what could be coming in the context of design. We call AI artificial intelligence, but I'm wondering, is it intelligent? Is it more in line to call it artificial information processing? That could actually make sense. So if it's not actually intelligent, the first thing we should be doing is thinking about us as designers, that there's a bias there already by calling it intelligent. If we think about it, that's a bias, and if we were to be more honest about it, I just found this only a few days ago from a guy called Stefano Quintarelli, and he said that if we're to be more honest about what it is, we should call it systematic approaches to learning algorithms and machine inferences. That's SALAMI. So for the rest of this thing I'm calling it salami.

[00:03:17] John: What I'd like to say to you guys now is, is salami going to take our creative jobs? Is salami going to be creative? Can salami be creative? I don't really think so. I think it's quite frankly a lot of baloney. But what I'm going to put to you, and I'm going to suggest to you beautiful, lovely people, because you are people and humans and I need you guys to vote for me, is that this is an ultra-processed product, and it could actually be quite carcinogenic when it comes to UX research, to design and to our jobs as well. I'll put to you guys that this salami is really pushing for efficiency over effectiveness, and when we're talking about efficiency, that's a bad thing. What I'm going to say is that it's exploiting over exploring, it's pushing technology over psychology and the human condition, it's also pushing speed over the speed of human thinking, and speed isn't everything, it's not everything good. And it's pushing content over context, and we have to bring that back into the C-suite, because strategy is really important. A really good quote from Jules Goddard: he says strategy is the rare and precious skill of staying one step ahead of the need to be efficient.

[00:04:26] Host: Incredible. Salami. You got anything to beat that, Kevin? Kick us off.

[00:04:31] Kevin: I think this is great that we did this right after the agency talk, because I think AI is just like another agency. Whenever an in-house team loses to an agency there's a big huff and puff online, because oh no, we could have done this in-house. But actually I think the value of any creative person goes up when there's lots of sloppy competition, and so having AI produce tons of C+ material at high speeds makes us all look better. So if AI is so scary and AI is so bad, how are both of those things true at the same time?

[00:05:05] Kevin: People are going to replace us because they're going to use AI. I think we all have heard the adage now, I think 80% of people have said this now, that if you're going to get replaced there'll be people who do our jobs with AI, not AI by itself. But then there's this belief that at some point it'll learn everything we do. We all famously know how Figma launched their AI product and everyone immediately freaked out. We said they Adobe-ified Figma, they're going to take over all of our jobs and charge everyone 7,000 a seat. Then they tried it with Dev Mode. But the reality is, if you're going to defend yourself, I'd rather have competition. I don't want to be in a sea of two people that do amazing work, because you're going to lose the battle convincing our non-design stakeholders that we are the best option. I'd rather have 15,000 competitors and still be considered the best. I'd rather have it produce tons of UI and tons of ideas that fall flat and fail, and have my in-house team and myself look amazing, because we are. So I encourage competition, I encourage tons of people to use AI, and then we'll see who converts the most and who has happy teams.

[00:06:12] Host: I love it. Competition on one hand, old approach, this meat on the other. Let's dive in.

Question one: can AI-driven research build impactful products?

[00:06:16] Host: Question one, I'm going to throw this to you first, Kevin. Can AI-driven research help build impactful products, or is it designed to give superficial conclusions and data? Two minutes.

[00:06:29] Kevin: As you can tell, I'm team AI here. I was told to use funny things to get your votes, so it's team AI, right? I think it's not just superficial conclusions. I think it can be very, very helpful with the aid of humans. I think there needs to be the same quality bar, the same guidance that we just mentioned in the last talk. No agency is going to come in and do a better job without onboarding, without context, without guidance, without a clear goal and definition of success. AI is the same thing. You can judge AI when you give it no context, no conclusions, no goals, no parameters, and say look, it doesn't know how to do anything. And then you can give it the proper onboarding, the proper guidance, the same ones you're supposed to give to all your new hires that you don't do, you give it to the AI, and maybe it actually does a pretty good job. Maybe it does help after all. So let's not judge fish on how well they climb trees. Let's maybe give it a lake and give it a race.

[00:07:29] Host: Amazing, succinct, in just a minute. I love it. John, same question to you.

[00:07:36] John: Sure, two minutes. The first thing I'll talk about is technology and psychology. I think there might be an over-reliance on us now using technology, and that can overshadow, I suppose, the importance of understanding the psychology in UX research. As an example, we can do A/B tests, and they're saying now that we don't need designers to design the interfaces. The AI can design it and then it can pick the best one and then move on to the next one, and then it starts efficiently coming to a conclusion through UX research. I'm afraid that actually could be a lot of content without any real context, and the why someone might be clicking on something or moving something. So we just have to be very careful with that, because this efficiency I don't think is going to work really well.

[00:08:18] John: If you think back to the early 80s when Steve Jobs came out with the first Mac and he said that it can amplify our abilities, that's true, it should be about their abilities. But he did say as well, with Jony Ive, when they brought out the glossy Berry [?] Mac that everyone could have in their house a few years back, they actually put a handle on the top of it. You remember the handle on the top of it? He said the handle on the top of it is so you can throw it out the window, because the computer is not for doing everything. So it's just really, really important in that context that we remember that.

[00:08:52] John: What else do we have? Salami. Like I was saying, we're calling it salami. It relies on strict logic, so it reduces the opportunity to test and research some of the counterintuitive ideas, and this is a bit people forget about. Rory Sutherland, he's the Ogilvy guy, and he said the opposite of a good idea might actually be another good idea. I think we do a lot ourselves, and when we look at some great products that come out, it's easy to, let's say, post-rationalize a good idea. But in our roles in UX, we need to start looking at some of the counterintuitive ideas before they become good ideas, and we need to be able to bring that to the C-suite and give them really good ideas of what that should be. So yeah, I think it's getting interesting, difficult, and we've got to look at those counterintuitive ideas in a logical way that a human can and that AI can't.

[00:09:47] Host: All right, 30 seconds rebuttal of salami and chucking Macs out windows.

[00:09:53] Kevin: Okay, I'm going to use your example here for my rebuttal. This handle on a Mac seems random, right? It seems like a really bad idea. It seems maybe like a hallucination of a chatbot. Tons of good ideas, things that ended up being super cool, were originally just really bad ideas that no one understood. I think Jony Ive is famous for being considered the crazy guy at Apple who kept having new ideas. People were like, no one's ever done that, that's bizarre, is that a mistake, you don't really mean that. It kind of feels like the first version of ChatGPT: oh, that's a weird idea, there's no way it intended for that to be the solution. So I don't necessarily think that all ideas that you don't immediately agree with are bad ideas. I think you have to sift through the noise and sift through bad ideas to sometimes get really fabulous ideas. When everyone goes right and you want to go left, maybe you have to try to go left.

[00:10:42] Host: Amazing. I'm going to cut you off, 20 seconds over, mate. I don't know what to tell you. Right, let's see if this works. Who won? Did John win this question, did Kevin win this question? Get voting. Oh no, someone else vote, it's awkward if no one else does. Okay, that's good.

[00:11:05] John: I'm going up against salami here.

[00:11:07] Host: You are going up against salami here. I'm going to give it five more seconds, four more seconds, three more seconds.

[00:11:11] John: Humanity, people!

[00:11:16] Host: You will have your time to rebut. All right, one. Oh, you lost points for suggesting we were humans. Are we? I'm now actually really worried about you for the rest of this talk. Okay, that's one point to Kevin. There we go. We're going to move on to the next question.

Question two: embracing AI across the creative process

[00:11:37] Host: How much should UX designers embrace AI across the creative process? Will it lead to everyone going down the same road? You want to start?

[00:11:47] John: Okay, talking about more efficiency, et cetera. We have an impatience with human intelligence, I think. This seems to be a thing now, and when we AI-design the user interface we say it's great because it's like a shortcut and gets us from A to B. We can get to a final interface very, very fast, and that's great. But the thing with that really is that it's a faster route, but are these actually just really quick, processed, microwavable recipes to get us to some sort of conclusion? And could we actually be nuking some good ideas here? If you think about the value of the creative process: when you were in college you would have had a thesis or you would have done a dissertation, and if you look back at that today you'll read it and go, good Lord, what was I thinking? But the thing is that it was the process of going from A to B, that space in between, that was really important in your development. It also meant that you were collaborating with people, you were reading books that you wouldn't regularly read, you were talking to people you wouldn't regularly talk to. So that process is really, really important.

[00:12:48] John: What else would I say about that? I just have that caution against over-reliance on AI tools, because it's going to lead to a lot of exploitation but not a real area of exploration. If we get into a lecture on the birds and the bees: if we look at bees, they've been around 300 million years, and there's two types of bees in a nest. There's the worker bees. They follow a waggle dance to go out and get the pollen, and then they come back, and they bring more energy back than they expel, and that makes them good working bees. They're very efficient, and if they're not efficient anymore they're kicked out. But then there's 20% of the bees called scout bees, and they explore. They go out and don't follow the waggle dance. They go out and find new areas, new value, in other words, to find new pollen, to find new flowers in different areas. So we have to be very careful that we're not getting into a stage of efficiency where we're just going to start depleting and going towards the exact same thing, our tools going to start mirroring each other, et cetera. We need to be very, very careful that we build up the explore part while AI is really pushing the exploit at the moment.

[00:13:52] Host: Amazing. AI is going to kick out our 20% explorers. An interesting point. Two minutes, take it away.

[00:13:57] Kevin: My favorite topic is this one. All right, so how many of you remember when you used to have very small teams and you used to have no tools, and you used to just do 10 versions of everything? You'd send a client 20 drafts of a logo, 100 mockups, tons and tons of wireframes, and you'd trash most of it, and then somehow we still all make websites look like linear.com. We've done all of this work to get better and better as designers, and there's a website called linear.art [?] which is just a thousand websites that look identical to linear.com, because even with all of our user research, competitive analysis and our senior designers and our agencies, we somehow still came to the same ideas. So if we did it, of course AI sometimes is going to do it. But if we did it and it shipped a thousand times, we could probably catch that much faster if the AI can go and do competitive research in about 30 seconds, when we couldn't do it over the last 12 years.

[00:14:50] Kevin: I think there's ways to use AI in a way that makes you a better designer, and it's something that, again, you shouldn't have over-reliance on. But I'd much rather have my ideas be shut down in the first hour of ideation than to spend two weeks doing the drafts, having things be coded, just to find out that we built the same thing our competitors built, just to find out that we did the same A/B test that they ran last year and bombed. I'd rather know that now at the speed of AI than to learn that at the speed of me, a human, without the ability to scan the entire internet in seconds. So when it comes to expelling bees that are inefficient, I think we may be the bees that are inefficient, but with AI we might be a bit more efficient.

[00:15:34] Host: Amazing. A tool to have, but not be over-reliant on. 30 seconds.

[00:15:41] John: Yeah, I'm trying to think about that. But look, the thing is, as we get more into the exploit area, I think that the AI will take over that a lot more, and that's okay. But the explore area is going to get bigger, because we'll need resilience. Explore is about resilience, and we'll need more resilience as the world moves on. There's more peaks and troughs closer together now than there used to be, in terms of crashes and different things going on, people losing jobs. So explore is going to get bigger, and we're going to need a lot more exploration for processes, for products, for people, humanity, et cetera. So explore is getting bigger while exploit is actually getting smaller, because the machines are doing it all.

[00:16:20] Host: Amazing. Explore is going to get bigger. All right, let's go, everyone. Who won this question?

[00:16:28] Kevin: I'm going to hide my face and look later.

[00:16:30] Host: Oh yeah, at the end I'm going to get you to face that way. I'm going to see the fun thing. Okay, John's in the lead on this one. Five more seconds.

[00:16:40] John: Everyone vote for me, please.

[00:16:43] Host: Three, two, one. Amazing, we are tied 1-1, John, Kevin. Very good. Right, lovely, we move on to our final question.

Question three: will AI steal your job?

[00:16:54] Host: A big one, this. Oh yeah, jobs. Is AI going to steal your job? Can AI tools enhance job security by making designers more efficient, or will it make us all obsolete? Kevin, to begin.

[00:17:12] Kevin: This is going to be a roller coaster of an answer, so bear with me. In the last three jobs I've had, I've had to both conduct layoffs and then been laid off. Very successful part of my career. And so I think efficiency is being forced upon us. I think massive, massive jumps in how we are being measured for value and for success are happening whether we get better as designers or not. So if we are going to have that battle, I'm going to use a bit of a war metaphor here, pardon me, but would I rather be the samurai or would I rather be the machine gun? I think AI is the machine gun. Now, I love a well-crafted, thousands of centuries traditional technique taught father to father to son, but when it comes to winning a battle, that culture and that craft and that tradition matter when it scales. I can be the best designer, but the best designer with no one to design for and no paycheck and no income, are they still a designer, or are they a craft practitioner looking for "open for work" on LinkedIn?

[00:18:15] Kevin: So I think in this battle for keeping our jobs, for being in the game still, you need to pick up the best techniques, the best skills. When something is rushing at you, when someone is trying to break into your house, do you not try to close the door and lock it, or do you just let them in? So I say pick up your tools, learn a new one, and defend yourself.

[00:18:37] Host: Tool and evolution. Oh, a clap, solid. All right, John, two minutes.

[00:18:45] John: Got a clap, damn it. You didn't even have to say anything and you got a clap. So there we go. Okay, we have to look at this in a certain way. First of all, we've had 15 years of zero interest rates. Remember that? Well, some of you couldn't remember a time before there was no interest rates. That's a huge problem, because it meant that VCs loved the cheap money flying around the place, throwing money at companies all over the place. It died just last year, the year before, all that died, and guess what appeared straight away? AI appeared, all of a sudden. So now the VCs are all like, hey, this AI stuff, yeah, you've got to push for AI. So I think it's a bit of a joke, to be quite honest. And what happens then is that you get a lot of cost cutting, and we're using AI to do the cost cutting. You have all these awful companies like McKinsey coming out and they're packaging cost cutting as, this is efficiency, and we're going to do these things for your company and save you loads of money and get rid of a load of people. They're not saying that, but that's actually what they're trying to do. So I think it's a little bit of a joke in that context.

[00:19:46] John: In terms of jobs and security, there were two reports out just last month. Notion did a report with the Harvard Business Review, and Goldman Sachs did a report as well about jobs into the future and knowledge workers, and both came to the same conclusion, that there are problems with the workforce in terms of siloed information, bad practices in terms of the processes that we're using at the moment, not new really, I suppose, and the lack of collaboration. They really pushed the lack of collaboration, and they've pushed for that there needs to be a lot more critical thinking, there has to be a lot more creativity, and there has to be a lot more collaboration. Collaboration is the thing that's coming up again and again. Even when we were on the walking tour of Dublin a couple of days ago, I remember we were talking about George Bernard Shaw, and he has a beautiful quote about communication. He said the problem with communication is the illusion that it had ever happened at all. I think we really have to start going back in and looking at communicating with each other when we have things like AI agents coming into the realm of design as well now. What the hell is that? We need to start talking about it, because apparently 2025 is going to be the year of AI agents. You know how far we've come.

[00:20:55] Host: Amazing. Kevin, you're going to have 30 seconds to say a rebuttal on the death of communication.

[00:21:01] Kevin: I know, right? This is great, didn't plan this at all. Communication is my passion. I've traveled to many countries, I've been to 72, I've lived in eight of them. My parents are these crazy travel people and they forced this on me very early. So I've always realized that the ability to communicate is super, super important, to the point where I have a tattoo, and my biggest tattoo on me is a quote in Latin, scientia est potentia, which is knowledge is power. And I think AI enables us to have more knowledge. Do you want it?

[00:21:33] Host: Very cool. All right, we will have one last vote between "it's a tool you need to get on board with or be left behind" or "it will destroy all communication between individuals," kind of. There we go, just for Halloween. Oh, it's going to be close, this one, these strong arguments. Oh, people like locking their doors. No, no, you've had your 30 seconds. Just say "insecure." Four, three, two, one. Just pipped. Kevin, well done.

Closing summaries

[00:22:11] Host: Right, it is not over yet. We have gone through all three of our questions, however you will now have two minutes, no rebuttal, because I want a safe fight. Two minutes to finish and summarize what you have said and why you are right in your position. Go for it.

[00:22:30] Kevin: Okay, fine, all right. Let's go along with this analogy still. We all like our house, and people are trying to attack our house, and I'm going to say we close the door and lock it. Now, this does not mean this is going to work forever. I think when we're talking about AI taking your jobs, we're not talking about the short game, we're talking about the long game. So when I say that you need to pick up the tools and learn AI and learn new skills, I think we eventually need to figure out what is the difference between good design and amazing design, and I think that's everyone's personal journey. The problem is we often don't look around and see other people doing the same journey. We oftentimes make the same mistakes that they did, and it's pretty avoidable sometimes, especially at the company level. You have competitors, you're in an industry. So if you're going to be more aware of everyone, this also allows for more groupthink, which is something you mentioned before.

[00:23:21] Kevin: Now, I think the best thing about this is we still need to be in the room to help the groups think better. AI is going to help with that, and ultimately AI is going to be reliant on us. So if we're going to be doing this, a lot of companies are going to start producing AI as the product, and if you don't know what AI is, how are you going to be the designer that gets kept at the company when the company goes from a tech company to an AI tech company? I think it's going to be the paradigm shift. I think it's going to be the next time we talk about how we design for user experience, which used to be called HCI, which used to be called other things. It's going to be the next thing. So I recommend checking out The Shape of AI, and I suggest that you have the job of shaping AI, because otherwise it'll shape you.

[00:24:05] Host: Very good. John, 30 seconds to sum up. No, two minutes to sum up. Sorry.

[00:24:11] John: Okay. Complexity and creativity first of all. I think creativity is more intricate than straightforward budgeting. It requires advocates in the boardroom. Now, with regards to time for creativity, human intelligence, it takes time. We need more, with regards to the importance of explore and not just exploit, because like I said, the AI is really about exploiting things. In terms of getting the AI to be faster, that removes the idea of critical thinking, et cetera. It's efficiency over effectiveness. Relying solely on efficiency is detrimental to resilience, and that's something that we really need as we go forward.

[00:24:46] John: And then I'll just give some job tips, maybe. I think that's a way to go, because really it's all about jobs. First thing, watch out for the shiny AI companies that are coming out now. I think they'll all move on very, very fast as the landscape changes a lot, so maybe try to avoid them for jobs. I don't think you're going to learn that much. Caution with VC-backed companies, companies who are still going around with CFOs trying to enlarge their TAM, junkies for more funding. I don't think you're going to learn much in those companies either, although we might all be looking for a job in them. And just look for companies that have an environment where they foster, they make their work look like a library and a pub, and they have that kind of communication going on. It's deep work, and then the communication, not like bean bags, but them building communities of practice about these AI agents and things like that. And remember as well, don't get too locked down in this salami, because it's going to make you info-obese, eating all this salami. You don't need to do that, and it's not going to be good for you. Like I said, UX talks a lot about autonomous teams, and we heard a lot of that today, but watch out and try and work with autonomous companies as well that actually are bootstrapped, because it's an enormous sense of wellbeing to be working with companies like that.

The final vote

[00:25:59] Host: Amazing. Before we dive into the final vote, can we just have a round of applause for two incredibly eloquently put across, and very politely put across, sides of a debate. This is not what you promised me at all. I love it. All right, I think we are going to head on to the final. So gentlemen, I would like you to come to the front of the stage and not look at the screen behind us. That way, fine. We'll have everyone start voting. It will tell me anyway. Start voting. And in the audience who are here, I mean you can do it at home but we won't hear it, a little bit of a drum roll please as we go. Five, four, three, two. This is a terrible drum roll. I'm going to start from five again. Better. Four, three, two, one. With less than 5%, Kevin is the winner. Look at these numbers. That is basically the same.

[00:26:59] John: You regret it [?].

[00:27:01] Host: Congratulations. Thank you very good, amazing, fantastic. Oh, I get to ask you questions apparently. See, I knew what I was doing when it was a normal format. You can take your seats again, or each other's, whichever you'd like. Oh, there are no unanswered questions. We are asking questions? We're not asking questions? Okay, you can type. Okay, let's type some questions. Real human hands. Ask ChatGPT what you should ask for our amazing debaters.

Q&A

[00:27:30] Host: All right, so for anyone who wasn't in the room for the last panel, this is going to seem very odd, and if you were, this is still very odd. Will AI replace sausage rolls?

[00:27:44] John: Me? Sorry.

[00:27:45] Host: It's a real question. Yeah, it's a very serious question. You were talking about it being salami anyway.

[00:27:49] John: So this is the kind of question I might ask, what's it called, ChatGPT, because I only use it for satire. So I don't know. Is AI in satire? No, probably not.

[00:28:02] Kevin: I mean, we already have vegan sausages, right? So why can't we have AI sausages? Representation for all types of sausages.

[00:28:12] Host: I love it. Okay, it's going well. Right, brilliant. So you've both obviously fiercely fought on either side of this as a debate, but is there a happy medium? Can we make the best of both worlds, AI that's useful but doesn't replace us?

[00:28:29] Kevin: Yeah, I'm going to quote one of the previous speakers, who's like, B2B software is great for adoption because they have to use it. So when the company buys AI tools, you go to work and it's there. We're going to be using it, so you might as well figure out how to do your job still when they change the software to AI tool software.

[00:28:45] Host: Sure. All right, fantastic. There's another great question. So I guess you've mentioned you use ChatGPT for satire, John. Are there any aspects of your professional life in which you use AI?

[00:28:56] John: Of course, I use it all the time. We're just being a little polarized up here, but yeah, no, AI is very important. It's great for when you stare at the blank page, just in a very simple way, if you're working on a design or you're just writing. Writing is still a huge thing, like I was saying earlier on about when you're writing your diploma and dissertation or whatever, but I'd hate to think that people are now using GPT to write that, because you don't experience the writing. I can't spell any of those things, but writing is really important because it gave you clarity of thought. So I think we have to make sure that we keep writing and things like that, because GPT is great just to get you started, it's a good kickstart, but it's a process, writing, not an end state.

[00:29:41] Host: I get it, yeah. AI, how are you using it, Kevin?

[00:29:43] Kevin: We're actually much closer in opinion than it seems.

[00:29:47] Host: That doesn't make for good content, though.

[00:29:50] Kevin: I already won, so it's fine. But genuinely, I agree. It's a good starting point. I think my whole point is not to get faster to done. It gets me faster to V2. I think a lot of us spend time in jobs where we're doing an MVP, and secretly the CTO has planned to make that the final version since day one. So if we can get to their finished plan five times faster, we can now get to the V2 faster as well. So I love to just get past the busy work, the V1s. I don't want to keep shipping V1s and calling it done just because we're out of time. So I do agree, it's just the beginning.

[00:30:28] Host: Amazing. That is all we have time for, so we have a massive round of applause once again for Kevin and John.