WEBVTT 1 00:00:18.000 --> 00:00:29.000 We're constantly every week tweaking our own process, but we're also changing our mindset and kind of our own roles as far as how we are shifting as team members how we're working together what our process is like. 2 00:00:29.000 --> 00:00:44.000 And I'm very hopeful for the future as far as where this is all going, and how, how like developers are going to be leverage and designers still in the future but where the job is shifting and kind of the expectations are shifting a little bit. 3 00:00:44.000 --> 00:01:01.000 And so today I actually wanted to talk specifically through that whole change because we're having some questions I know we have some clients who are vibe coding their own apps we had one client that was on a subscription, and they cancelled that subscription after like five years. 4 00:01:01.000 --> 00:01:11.000 And now they've gone and vibe coded their own app. And so there's some cool things happening and it's, it's exciting to see what's happening and where people are going. 5 00:01:11.000 --> 00:01:27.000 But I think there's a really critical role for builders and architects to be in the process and helping clients think through and strategize this and so one thing I thought I wanted to illustrate is kind of kind of what was before and after. 6 00:01:27.000 --> 00:01:35.000 And let me see if I can share this here with you guys. 7 00:01:35.000 --> 00:01:36.000 There you go. 8 00:01:36.000 --> 00:01:41.000 So, let me make this a little bigger. 9 00:01:41.000 --> 00:01:51.000 So, this is kind of how it was right like we used to spend as developers spend mostly like 80% of our time. I'd say like over on this world. 10 00:01:51.000 --> 00:01:55.000 You know, coding, developing right writing code testing things. 11 00:01:55.000 --> 00:02:10.000 And then there was a little bit of like project management, maybe, and there's a little bit of like communication and design but either like a designer was handing you, you know, prototypes, and you were just sitting there and like basically making the prototypes and the requirements doc, you know, check all the boxes. 12 00:02:10.000 --> 00:02:12.000 That is really shifted. 13 00:02:12.000 --> 00:02:21.000 I think now in this kind of new world where like the code has become the 20% right like this is over here on the 20% side. 14 00:02:21.000 --> 00:02:38.000 And now there's this expectation for a, you know, quote unquote developer, which I'm like really thinking these, these are like an AI builder like our team, each of our team members are doing so much more than like writing code like we're doing client communication 15 00:02:38.000 --> 00:02:46.000 design work for prototyping stuff we're shaping things when I say design I'm not thinking like pretty UI and colors. 16 00:02:46.000 --> 00:02:50.000 I'm thinking like real product design of thinking like how does this feature actually fit. 17 00:02:50.000 --> 00:03:00.000 How does it fit with the database design, how does it fit with like the user experience like are we adding, you know, too many features, how can we make it feel simple and easy. 18 00:03:00.000 --> 00:03:09.000 There's obviously the shaping process the, you know, that whole process of actually breadboarding things, trying lots of things fat marker sketches. 19 00:03:09.000 --> 00:03:22.000 And then this other big piece here is the testing and the testing is now become like high expectation because, you know, people can go and do the 20% and this is kind of how I'm explaining it to clients is like we're, we're great if you guys 20 00:03:22.000 --> 00:03:36.000 want to help us do some coding or even pushing a PR or something that's great like do the little 20% piece. What actually needs to happen to make this a really successful product or launch is to do the other 80% which is the management design 21 00:03:36.000 --> 00:03:49.000 that stuff is really that human side. And I don't want to downplay the coding piece because the brilliant part about this is when you work with a really experienced developer like you work with someone on our team. 22 00:03:49.000 --> 00:04:04.000 They are senior developers who are now doing the AI coding as well. And they're doing the other piece so you kind of have like this super AI builder person because you've got someone who understands the code at all of these layers. 23 00:04:04.000 --> 00:04:19.000 So if you're not a designer who's just thinking about design, or a product person who's just thinking about the product, or a QA person who's just QAing stuff. They're actually in the code and they know exactly how it works and they were part of the process of writing the code, not just writing a quick prompt and setting 24 00:04:19.000 --> 00:04:36.000 we're creating so we're also like working on our team of course is how do we go and and create AI tools to help us with the management to help us with design shaping testing etc to make this work every single time because now that these tools 25 00:04:36.000 --> 00:04:45.000 are available and clients are starting to experience it and other founders are starting to experience it. They're getting a taste of how they could build something. 26 00:04:45.000 --> 00:04:55.000 But it's very easy to now like kind of give them all the power tools to go change the code and get access and, you know, give them database access. 27 00:04:55.000 --> 00:05:00.040 But we're seeing the flip side actually made a tweet this morning about this because 28 00:05:04.840 --> 00:05:09.120 And I said here, like, hey, we've been-- 29 00:05:09.120 --> 00:05:12.400 we're experimenting with this, with putting AI tools 30 00:05:12.400 --> 00:05:14.840 in the hands of our clients. 31 00:05:14.840 --> 00:05:16.560 They're able to vibe code an idea quickly. 32 00:05:16.560 --> 00:05:18.980 They can make small tweaks if they need to. 33 00:05:18.980 --> 00:05:20.600 And they started to feel a little more in control 34 00:05:20.600 --> 00:05:23.920 in the ownership of the app that they're investing in. 35 00:05:23.920 --> 00:05:27.240 But funny story, we had a client who 36 00:05:27.240 --> 00:05:31.860 had super-base access and perplexity and changed something. 37 00:05:31.860 --> 00:05:34.400 And the agent was like, hey, I think we should structure it 38 00:05:34.400 --> 00:05:35.160 this way. 39 00:05:35.160 --> 00:05:36.960 And he just said, yeah, sounds good. 40 00:05:36.960 --> 00:05:41.080 And it basically re-built the brand new schema 41 00:05:41.080 --> 00:05:43.040 for super-base, changed a bunch of tables, 42 00:05:43.040 --> 00:05:44.240 changed a bunch of columns. 43 00:05:44.240 --> 00:05:46.080 And he just kept working. 44 00:05:46.080 --> 00:05:48.080 And then only a week or two later realized 45 00:05:48.080 --> 00:05:51.360 that he had basically broken the entire production app. 46 00:05:51.360 --> 00:05:53.240 And so now we're trying to roll that back for him 47 00:05:53.240 --> 00:05:53.800 and help him out. 48 00:05:53.800 --> 00:05:58.600 So it's not perfect, by any means, because you have basically 49 00:05:58.600 --> 00:06:00.080 giving people power tools here. 50 00:06:00.080 --> 00:06:03.000 And they can actually build some really cool stuff. 51 00:06:03.000 --> 00:06:06.160 But they also can break things really, really quickly. 52 00:06:06.160 --> 00:06:08.880 So putting guardrails in place, I think these are things 53 00:06:08.880 --> 00:06:11.560 we're going to keep tweaking as we experiment with clients 54 00:06:11.560 --> 00:06:13.920 being a little bit more involved in the process, 55 00:06:13.920 --> 00:06:17.120 is giving them read-only access, for example, 56 00:06:17.120 --> 00:06:21.680 giving them much closer guardrails of maybe doing things 57 00:06:21.680 --> 00:06:26.400 in a test database, for example, or working 58 00:06:26.400 --> 00:06:29.480 in their own branch, things like that. 59 00:06:29.480 --> 00:06:32.800 So it's also taking a lot to set this process up, 60 00:06:32.800 --> 00:06:35.400 just setting these tools up, getting them connected. 61 00:06:35.400 --> 00:06:37.520 You quickly realize that the coding piece-- 62 00:06:37.520 --> 00:06:39.560 this is kind of what made me think of this-- 63 00:06:39.560 --> 00:06:42.000 the coding piece of building the app, 64 00:06:42.000 --> 00:06:43.800 there's so much management and knowledge 65 00:06:43.800 --> 00:06:48.600 that goes into how Git works, and how deployments work, 66 00:06:48.600 --> 00:06:51.000 and databases work, and servers work, 67 00:06:51.000 --> 00:06:55.080 and how having server side functions versus client side, 68 00:06:55.080 --> 00:06:57.120 and all this stuff that we just kind of know 69 00:06:57.120 --> 00:06:59.480 as we're building things, and just a lot of experience 70 00:06:59.480 --> 00:07:03.520 building apps, you just know stuff is how it works. 71 00:07:03.520 --> 00:07:06.040 And so getting an app, for example, 72 00:07:06.040 --> 00:07:11.200 to run on a client's local machine, it is a task. 73 00:07:11.200 --> 00:07:14.840 And I've been on a couple of these two-hour calls on Zoom, 74 00:07:14.840 --> 00:07:20.840 setting up node, and setting up installing all the pieces 75 00:07:20.840 --> 00:07:22.960 they need to run this stuff. 76 00:07:22.960 --> 00:07:25.640 And you just realize this is when you have a completely 77 00:07:25.640 --> 00:07:28.440 non-developer computer to go set this stuff up the first time. 78 00:07:28.440 --> 00:07:30.400 It's a bit nerdy. 79 00:07:30.400 --> 00:07:32.480 And then, of course, we'll hook it up with codecs, 80 00:07:32.480 --> 00:07:34.760 or hook it up with the Cloud code app, 81 00:07:34.760 --> 00:07:37.880 and then we'll show them how to do branches and all that. 82 00:07:37.880 --> 00:07:41.440 So there's quite a bit of a training step here. 83 00:07:41.440 --> 00:07:43.440 But we were experimenting with it, and we're trying stuff. 84 00:07:43.440 --> 00:07:47.560 But I wanted to kind of just share this thought I had around 85 00:07:47.560 --> 00:07:52.120 this, was that where you can go and spend a lot of time, 86 00:07:52.120 --> 00:07:54.640 where it's an efficient use for our team, 87 00:07:54.640 --> 00:07:57.080 is on the other 80%. 88 00:07:57.080 --> 00:08:00.000 How do we go and get really, really good at this? 89 00:08:00.000 --> 00:08:04.040 And at the same time, AI is now being used across all of these. 90 00:08:04.040 --> 00:08:08.200 So you could really think about AI's kind of all parts of this 91 00:08:08.200 --> 00:08:13.760 is going to be implemented with AI workflows for all of it. 92 00:08:13.760 --> 00:08:16.560 Because that's kind of what makes these projects super 93 00:08:16.560 --> 00:08:19.240 efficient. And it's being able to think really clearly, 94 00:08:19.240 --> 00:08:22.960 and taking a mangled mess of a project. 95 00:08:22.960 --> 00:08:25.760 I mean, a lot of times, we have a project that kind of looks 96 00:08:25.760 --> 00:08:27.200 like this, right? 97 00:08:27.200 --> 00:08:29.600 And we have to take the nest, kind of pull out the strains. 98 00:08:29.600 --> 00:08:32.280 I mean, OK, there's a project that starts here and ends there. 99 00:08:32.280 --> 00:08:34.799 There's actually a project that starts here and ends there. 100 00:08:34.799 --> 00:08:36.520 And there's another one that goes like this. 101 00:08:36.520 --> 00:08:43.679 And that, going from chaotic thinking and figuring out 102 00:08:43.679 --> 00:08:48.080 and pulling the thread apart to these nice, creepy projects 103 00:08:48.080 --> 00:08:49.400 is extremely rewarding. 104 00:08:49.400 --> 00:08:53.360 I think our whole team has really enjoyed that process. 105 00:08:53.360 --> 00:08:54.240 I love this. 106 00:08:54.240 --> 00:08:57.440 It's very therapeutic to get to this point. 107 00:08:57.440 --> 00:09:03.360 And this part is taking a few hours of deep work, 108 00:09:03.360 --> 00:09:06.120 working together, sitting on a Zoom call, 109 00:09:06.120 --> 00:09:08.480 and thinking these problems through, 110 00:09:08.480 --> 00:09:11.800 thinking through how we can bring it really clear, 111 00:09:11.800 --> 00:09:14.080 asking a lot of questions. 112 00:09:14.080 --> 00:09:16.760 And so that's kind of what work is kind of looking like. 113 00:09:16.760 --> 00:09:19.440 Like that, it's not just us staring at terminal 114 00:09:19.440 --> 00:09:22.360 and staring at our ID and coding stuff 115 00:09:22.360 --> 00:09:25.080 over here in the coding world anymore. 116 00:09:25.080 --> 00:09:26.520 It's still there's 100% that. 117 00:09:26.520 --> 00:09:27.960 I mean, I'm being a little generous. 118 00:09:27.960 --> 00:09:29.160 Maybe it's 20%, 30%. 119 00:09:29.160 --> 00:09:30.600 You guys get the idea. 120 00:09:30.600 --> 00:09:34.080 But a lot of our energy is set here. 121 00:09:34.080 --> 00:09:37.960 And this is actually the human piece, which is really fun, 122 00:09:37.960 --> 00:09:38.640 I think. 123 00:09:38.640 --> 00:09:43.640 This whole 80% is all kind of driven by human interaction. 124 00:09:43.640 --> 00:09:46.040 It's us talking to the client, understanding what they need, 125 00:09:46.040 --> 00:09:49.760 understanding their business, and then using what we know 126 00:09:49.760 --> 00:09:52.600 about the code and kind of bringing that into, 127 00:09:52.600 --> 00:09:54.240 and what's possible with code, and bringing that 128 00:09:54.240 --> 00:09:56.640 into each conversation with clients. 129 00:09:56.640 --> 00:09:59.360 It's going really deep on shaping and like me working 130 00:10:05.480 --> 00:10:07.600 this actually affect the business? 131 00:10:07.600 --> 00:10:11.920 So that's really, it's really fun because you get to kind of step away like this other 132 00:10:11.920 --> 00:10:15.920 bigger world outside of just me, you know, writing code. 133 00:10:15.920 --> 00:10:19.280 And now I get to kind of be a part of this other layer of stuff. 134 00:10:19.280 --> 00:10:22.960 And so there's a, this is something I would encourage all developers right now to be 135 00:10:22.960 --> 00:10:28.960 focusing heavily on all of this human side, the process, getting super clear, knowing 136 00:10:28.960 --> 00:10:34.440 how to breadboard and shape and that whole process of getting things to be super concrete. 137 00:10:34.440 --> 00:10:41.000 Because that skill, while you can like, you know, just blur it out and use, you know, a 138 00:10:41.000 --> 00:10:45.800 voice, you just do like a voice memo to an agent and have it kind of like, bring your 139 00:10:45.800 --> 00:10:47.860 thoughts a little bit more clear. 140 00:10:47.860 --> 00:10:53.440 That part, you can, is still taking a human piece to get really, really good at it. 141 00:10:53.440 --> 00:10:58.140 And not just making it like, you blurted out a bunch of requirements to me, and I just 142 00:10:58.140 --> 00:11:01.320 like organize them in an outline because that's kind of what AI does. 143 00:11:01.320 --> 00:11:05.520 It's actually like distilling it from like, oh, it's not this like blob. 144 00:11:05.520 --> 00:11:08.000 It's really just these two or three things. 145 00:11:08.000 --> 00:11:13.260 And so many projects or so many conversations we've had with customers recently, you know, 146 00:11:13.260 --> 00:11:17.060 we have a founder who gets on and just is like, you know, on this, I want that. 147 00:11:17.060 --> 00:11:18.060 Hey, I've thought about this. 148 00:11:18.060 --> 00:11:19.300 I had this thought in the shower. 149 00:11:19.300 --> 00:11:23.580 I want to like, this is the new thing we need to do, take all of the ideas and then we 150 00:11:23.580 --> 00:11:27.420 go sit for a few hours and we distill them down to like, hey, I think it's this, like 151 00:11:27.420 --> 00:11:29.620 I think this is really the problem. 152 00:11:29.620 --> 00:11:32.140 And this is really the outcome that person needs. 153 00:11:32.140 --> 00:11:36.300 And that is very, that is like where there's so much value there. 154 00:11:36.300 --> 00:11:40.860 So when we're building things, if you're a freelancer, if you're a builder, obviously 155 00:11:40.860 --> 00:11:42.620 you can go and sell your hours. 156 00:11:42.620 --> 00:11:47.320 And I think that part of it is, you know, the hours of coding are becoming less valuable. 157 00:11:47.320 --> 00:11:53.220 So people can get a little scared of like, what does this mean for my job, my future career. 158 00:11:53.220 --> 00:11:57.820 But what is coming, if you can continue to make yourself incredibly valuable and bring 159 00:11:57.820 --> 00:12:00.340 value to the process. 160 00:12:00.340 --> 00:12:01.980 And that's all happening on this side. 161 00:12:01.980 --> 00:12:05.340 It's not that like, you know, you're just going to go and like, you're doing a whole 162 00:12:05.340 --> 00:12:09.500 lot less work and you can, you know, because this, if you just do this, like you're just 163 00:12:09.500 --> 00:12:12.380 going to get, you're going to get through all your work a lot faster. 164 00:12:12.380 --> 00:12:16.060 And if you're charged, charging hourly, that's like, you're going to basically make less 165 00:12:16.060 --> 00:12:17.460 money. 166 00:12:17.460 --> 00:12:21.980 But if you're keeping yourself very valuable on this side, but I think there's this idea 167 00:12:21.980 --> 00:12:26.020 like, I kind of even in the last few weeks had a little bit of a shift, because I was 168 00:12:26.020 --> 00:12:31.060 thinking like, all of these pieces were really fitting into this little thin layer of, you 169 00:12:31.060 --> 00:12:34.420 know, that happens before the coding, or maybe I had thought I was like 50 50. 170 00:12:34.420 --> 00:12:39.340 I know I've drawn somewhere here and probably drawn this 100 times, but our little diagram 171 00:12:39.340 --> 00:12:41.780 of this like, there's this sort of this uphill piece. 172 00:12:41.780 --> 00:12:42.940 And then we go through the coding. 173 00:12:42.940 --> 00:12:46.140 So I kind of was thinking is kind of 50 50. 174 00:12:46.140 --> 00:12:49.420 But the more I'm getting into it, I was like, this is, if you looked at our like a cross 175 00:12:49.420 --> 00:12:54.260 section of our time, like in toggle for a particular project, I honestly think this 176 00:12:54.260 --> 00:12:58.780 is really where so much of the work is being done in the process. 177 00:12:58.780 --> 00:13:03.620 And a lot of we're like the real movement in the process is happening here. 178 00:13:03.620 --> 00:13:06.780 And then we'll go off and we'll like code or we do a spike and we learn something and 179 00:13:06.780 --> 00:13:08.780 we'll come back. 180 00:13:08.780 --> 00:13:12.940 That part is it's been interesting to see how that affects it. 181 00:13:12.940 --> 00:13:18.740 So we'll like end it there, I just, this is sort of an incur, hopefully encouraging everybody, 182 00:13:18.740 --> 00:13:26.220 not, you know, putting, you know, scaring anyone in this new kind of reality that we're 183 00:13:26.220 --> 00:13:27.740 stepping into. 184 00:13:27.740 --> 00:13:33.820 But just essentially just keep leaning in like be be okay to be expanded and challenged 185 00:13:33.820 --> 00:13:39.860 and kind of embrace this new way of like, hey, I can actually bring a lot more value. 186 00:13:39.860 --> 00:13:45.100 But you kind of come in as sort of like the technology expert into each of these different 187 00:13:45.100 --> 00:13:52.340 spaces and the different parts of the process as sort of the coding tech expert. 188 00:13:52.340 --> 00:13:55.620 And a lot of people say like, hey, yeah, we're getting a promotion with AI and now like, 189 00:13:55.620 --> 00:13:58.500 you know, the AI is doing the coding and you're kind of senior management. 190 00:13:58.500 --> 00:14:01.580 I don't really know if that's it, obviously we're working very much. 191 00:14:01.580 --> 00:14:05.260 It's a lot more back and forth like that. 192 00:14:05.260 --> 00:14:08.820 But just lean into that process would be my advice. 193 00:14:08.820 --> 00:14:11.580 And there's there's so much opportunity. 194 00:14:11.580 --> 00:14:19.020 I show this, you know, diagram of like the people who are like building things that are, 195 00:14:19.020 --> 00:14:23.260 you know, actually using code for this is just so, so tiny. 196 00:14:23.260 --> 00:14:27.940 There's a great graphic here and I'll pull it in and this will be the last thing I touch 197 00:14:27.940 --> 00:14:28.940 on. 198 00:14:28.940 --> 00:14:34.580 But if you guys have seen this, this has been shared a lot on, on X, but each dot is 3.2 199 00:14:34.580 --> 00:14:40.780 million people, 60, let's look at the stats down here, so 84 people 34% have never even 200 00:14:40.780 --> 00:14:49.180 touched AI 1.3 billion have a free chat bot user 0.3% have actually paid more than $20 201 00:14:49.180 --> 00:14:58.020 for AI and 0.04% have actually used some sort of coding tool for AI. 202 00:14:58.020 --> 00:15:00.020 And I don't remember exactly what this is from April. 203 00:15:17.000 --> 00:15:27.000 We are talking to some customers now that are doing things on paper that are using PDFs and printing things and faxing stuff and we came in with like a form builder. 204 00:15:27.000 --> 00:15:38.000 And it's like, this is the future, like we can do forms and we can have someone submit a form here and we can process it this way, and it's going to like, it's going to affect hundreds of thousands of people with this tool. 205 00:15:38.000 --> 00:15:56.000 And there's industries out there that are just so like ripe for helping and bringing expertise to so don't, don't be sucked into the bubble of X and social media, and kind of seeing all the things that could like, oh, like, I don't know if we're going to be needed anymore in this space 206 00:15:56.000 --> 00:16:10.000 like this is building things and building things with quality is not going away, the craftsmanship of what we're building, having to, but you can't just rely on the 20% of coding and be like, I coded it and I handed it back. 207 00:16:10.000 --> 00:16:13.000 That's, that does that part is going away. 208 00:16:13.000 --> 00:16:28.000 This is there's this expectation, especially around testing, especially around thinking through like fully formed features, and not just hey this requirement or bug came in I passed a code and it's fixed it's like, did we really test it that we tried on all the devices 209 00:16:28.000 --> 00:16:38.000 and I think that I put the app on my phone in my hand and I push the buttons and feel it, you know, it didn't feel fast or it felt clunky. 210 00:16:38.000 --> 00:16:53.000 That's where you're going to bring a lot more value to it is kind of getting the sort of the finished shipable piece out into the world so we'll open up for some conversation and I'll probably go and stop the stream here to keep it short. 211 00:16:53.000 --> 00:16:55.400 was helpful and I will see you guys on the next one.