1 00:00:00,040 --> 00:00:04,160 Oh, I can code 55 times faster. How many hours a day do you code 2 00:00:04,160 --> 00:00:06,320 as a developer? The value was not in the coding. 3 00:00:06,320 --> 00:00:09,640 The value was in finding that solution, in defining that 4 00:00:09,640 --> 00:00:12,600 solution. Junior engineers are more likely 5 00:00:12,720 --> 00:00:15,160 to adopt AI. They are more native to the 6 00:00:15,160 --> 00:00:17,480 technology. They're not too opinionated on 7 00:00:17,480 --> 00:00:19,480 that yet. They're the change agents. 8 00:00:19,600 --> 00:00:22,640 If you cannot experiment on a day-to-day, which means at your 9 00:00:22,640 --> 00:00:24,320 job, it might be actually a good reason. 10 00:00:24,680 --> 00:00:28,280 Just expand your mind, make sure that you you are aware of a 11 00:00:28,280 --> 00:00:31,680 little bit of everything. This is a craft that is changing 12 00:00:31,680 --> 00:00:33,800 super fast. Coding is cheap. 13 00:00:33,920 --> 00:00:37,920 Software is expensive today. I'm joined by Susanna Daniels, 14 00:00:38,000 --> 00:00:41,920 Chief Developer Advisor over at Microsoft, and with close to 30 15 00:00:41,920 --> 00:00:45,160 years of experience, she has more than yours and mine 16 00:00:45,400 --> 00:00:46,920 combined. Potentially. 17 00:00:47,440 --> 00:00:50,720 We talk about the future of software engineering with tools 18 00:00:50,840 --> 00:00:54,720 and a gentic, the things we have nowadays, but more specifically 19 00:00:54,720 --> 00:00:59,000 how you as software engineer can navigate this and much more. 20 00:00:59,440 --> 00:01:07,240 So enjoy, I think a big narrative nowadays and a trend 21 00:01:07,240 --> 00:01:10,480 that really focuses on specifically productivity. 10X 22 00:01:10,480 --> 00:01:13,160 engineer is something that was already a phrase in the past and 23 00:01:13,160 --> 00:01:15,160 now people are saying, oh, it might be a reality. 24 00:01:15,440 --> 00:01:17,720 What is your view on this emphasis on productivity 25 00:01:17,720 --> 00:01:22,360 specifically? So yeah, that's a bit of, that's 26 00:01:22,360 --> 00:01:27,080 an interesting question. Early days when we, when we had 27 00:01:27,160 --> 00:01:33,120 AI, we were talking about mainly code completion and asking an 28 00:01:33,120 --> 00:01:35,800 agent to do some work for you, right? 29 00:01:36,280 --> 00:01:41,920 Task agent, human. Back then the focus was on 30 00:01:41,920 --> 00:01:45,240 productivity because that was what you could achieve, right? 31 00:01:47,080 --> 00:01:50,640 But that marketing was so good that we were talking about, oh, 32 00:01:50,640 --> 00:01:55,560 I can code 55 times faster. How many hours a day do you code 33 00:01:55,560 --> 00:01:58,560 as a developer? How how much is that? 34 00:01:58,560 --> 00:02:01,440 Like 14%? You can say one hour, you can 35 00:02:01,440 --> 00:02:05,280 say 2 hours. So while that is an enormous 36 00:02:05,280 --> 00:02:09,440 benefit, don't get me wrong, the question is, is all there is 37 00:02:09,440 --> 00:02:15,480 productivity? We learn now that with these 38 00:02:15,480 --> 00:02:18,520 agents and these models becoming more powerful and us 39 00:02:18,520 --> 00:02:22,720 understanding better how to operate them that we can also 40 00:02:22,720 --> 00:02:24,200 talk about improving the quality. 41 00:02:24,240 --> 00:02:30,120 We can also improving, talking about improving the outcomes 42 00:02:30,280 --> 00:02:32,240 that we, that we would like to see. 43 00:02:32,880 --> 00:02:36,320 We can talk about instead of productivity, we can talk about 44 00:02:36,560 --> 00:02:41,000 the pace of innovation. So there's a lot more to it. 45 00:02:41,240 --> 00:02:45,240 I think that the productivity narrative has worked really 46 00:02:45,240 --> 00:02:48,440 well, but we, we are way beyond that. 47 00:02:48,440 --> 00:02:52,080 I think productivity is maybe the first level of AI adoption, 48 00:02:52,320 --> 00:02:53,840 right? I'm freeing up some time. 49 00:02:54,200 --> 00:02:59,040 You're going to use the time to learn more, to understand more 50 00:02:59,040 --> 00:03:00,800 about these agents and how they can help you. 51 00:03:01,080 --> 00:03:03,360 And that way you're going to add more and more value. 52 00:03:04,240 --> 00:03:10,200 But yeah, it's a it's a very, very stubborn, reminiscent of of 53 00:03:10,200 --> 00:03:13,080 previous marketing, I think heartache maybe. 54 00:03:13,080 --> 00:03:15,520 But yeah, this isn't it's not all story. 55 00:03:15,680 --> 00:03:18,600 Yeah, No, I like your emphasis on creating value. 56 00:03:18,960 --> 00:03:21,880 That's why I always had my joy. It wasn't necessarily on the the 57 00:03:21,880 --> 00:03:25,720 act of creating software, although I did have a lot of 58 00:03:25,720 --> 00:03:28,360 fulfilment in there. But that part is now I think the 59 00:03:28,360 --> 00:03:30,760 most evolving. The outcomes I'm working 60 00:03:30,760 --> 00:03:33,840 towards, I think those are still what makes me take at least as 61 00:03:33,840 --> 00:03:36,320 engineers specifically. But there are different flavours 62 00:03:36,320 --> 00:03:37,960 of engineers. Some people really like the the 63 00:03:37,960 --> 00:03:40,880 hands on part and they really like and they hold a lot of 64 00:03:40,880 --> 00:03:44,160 passion in the craft of creating code where they say this is 65 00:03:44,160 --> 00:03:46,280 elegant code and we've refactored it and it's not 66 00:03:46,280 --> 00:03:48,440 really nice. That fulfilment part, I think 67 00:03:48,440 --> 00:03:55,280 that's impacted the most so far. Yeah, it's look, if you, if you 68 00:03:55,280 --> 00:04:00,000 look at software development, at least in my opinion, what we 69 00:04:00,000 --> 00:04:03,520 used to do is we would have a problem, a challenge and we 70 00:04:03,520 --> 00:04:07,520 would, would start figuring out what is the algorithm to get to 71 00:04:07,520 --> 00:04:15,400 a solution, AKA the coding the you know, but if you look at 72 00:04:15,400 --> 00:04:18,920 overtime, software development changed a lot, right? 73 00:04:19,079 --> 00:04:24,560 So for instance, gathering the right requirements, that's so 74 00:04:24,560 --> 00:04:27,440 much work. It became a full time job for a 75 00:04:27,440 --> 00:04:31,560 person, right? If you look at other parts of 76 00:04:31,560 --> 00:04:34,600 software development that you know, that became jobs, that 77 00:04:34,600 --> 00:04:39,440 became roles, that became things that other people would do. 78 00:04:40,640 --> 00:04:43,280 So if you look at the essence of software development, a lot of 79 00:04:43,280 --> 00:04:46,760 it was stripped away basically. Why? 80 00:04:46,760 --> 00:04:49,920 Because we're humans. So if we have big things, if we 81 00:04:49,920 --> 00:04:55,520 have big teams, we, we need to figure out the process, we need 82 00:04:56,560 --> 00:05:03,440 handoffs to ensure that we, that, that we, that we keep that 83 00:05:03,440 --> 00:05:08,000 flow going, that we satisfy the requirements of the next person 84 00:05:08,000 --> 00:05:10,160 or the next rule that picks up a task. 85 00:05:12,320 --> 00:05:15,440 Is that really software engineering And not to me when I 86 00:05:15,440 --> 00:05:18,720 started software engineering, what I loved about, about it is 87 00:05:18,720 --> 00:05:22,040 really knowing the problem, knowing who the customers are, 88 00:05:23,600 --> 00:05:26,600 matching that with some ambitions coming down from 89 00:05:26,600 --> 00:05:30,160 business and figuring out how I could make that all work. 90 00:05:30,840 --> 00:05:34,360 And yes, I spent a lot of time coding to do that, but the 91 00:05:34,360 --> 00:05:36,760 failure was not in the coding. The failure was in, in, in 92 00:05:36,760 --> 00:05:40,440 finding that solution, in defining that solution in, in 93 00:05:40,440 --> 00:05:43,200 getting that feedback and, and translating that in some 94 00:05:43,200 --> 00:05:46,360 technical, technical solution for that. 95 00:05:49,120 --> 00:05:54,120 And with, with now with AI, we, we see that we are more involved 96 00:05:54,120 --> 00:05:58,200 in the beginning. This will be increasingly more I 97 00:05:58,200 --> 00:06:02,880 think that the boundary between products and software 98 00:06:02,880 --> 00:06:05,760 engineering will will be non existent at some point. 99 00:06:06,640 --> 00:06:11,160 Maybe somebody has a role or or you know is is better at doing 1 100 00:06:11,160 --> 00:06:15,320 task than another and and you know keeps the control and, and 101 00:06:15,320 --> 00:06:17,200 has some some responsibilities there. 102 00:06:18,280 --> 00:06:21,800 But it should be seamless. And with that, a developer is 103 00:06:21,800 --> 00:06:25,280 involved, a developer is involved in the outcomes because 104 00:06:25,280 --> 00:06:26,560 that is what we're actually doing. 105 00:06:26,560 --> 00:06:31,400 That algorithm, algorithm that will be solved by AIAI will will 106 00:06:31,400 --> 00:06:34,400 take care of that work for us. And when it doesn't work, we can 107 00:06:34,400 --> 00:06:36,720 still dive into the code and do the nitty gritty. 108 00:06:36,720 --> 00:06:39,840 Some things are maybe way too bleeding edge for an agent to 109 00:06:39,840 --> 00:06:44,960 understand. Could be, but that is then the 110 00:06:44,960 --> 00:06:47,360 more fun work, right? If you ask most developers, we 111 00:06:47,360 --> 00:06:53,080 did internal internal survey and one of the things that came out 112 00:06:53,320 --> 00:06:55,600 out of it is that that's actually the work that 113 00:06:55,600 --> 00:06:59,600 developers enjoy doing and figuring out the the solution, 114 00:06:59,600 --> 00:07:03,400 figuring out the outcomes. So yeah, it is shifting. 115 00:07:03,400 --> 00:07:05,480 It is changing. I think that for the for the 116 00:07:05,480 --> 00:07:08,520 coders that like to code, that's great. 117 00:07:09,040 --> 00:07:12,240 We still have people also writing the Linux kernel, right? 118 00:07:13,000 --> 00:07:16,360 We still have people that worry about, you know, that's writing 119 00:07:16,360 --> 00:07:18,760 assembly. We have people that write COBOL, 120 00:07:19,400 --> 00:07:21,520 that's all fine, that will change. 121 00:07:21,520 --> 00:07:24,600 But yeah, there's something to do for everyone. 122 00:07:24,840 --> 00:07:28,080 I think we're in really exciting times, as you mentioned, where 123 00:07:28,160 --> 00:07:31,120 teams are kind of navigating this newer way of working. 124 00:07:31,640 --> 00:07:34,760 For me, it's probably one of the first times where I've seen this 125 00:07:34,760 --> 00:07:36,880 much impact on what I did as a craft. 126 00:07:37,160 --> 00:07:40,400 I don't have as much years of experience as some other people, 127 00:07:40,400 --> 00:07:42,320 even the people that are listening here, But for me, this 128 00:07:42,320 --> 00:07:44,680 is really cool to experience this part. 129 00:07:45,160 --> 00:07:48,400 There's also fear because as individuals are kind of 130 00:07:48,400 --> 00:07:51,920 amplifying and creating value to a level that is different, that 131 00:07:51,920 --> 00:07:54,160 is evolving. I've also seen organizations 132 00:07:54,160 --> 00:07:57,040 that say, whoa, for the people that are really senior, they 133 00:07:57,040 --> 00:08:01,000 know kind of the boundaries and they know they have quite some 134 00:08:01,000 --> 00:08:03,200 knowledge under their belt to accelerate with this. 135 00:08:03,480 --> 00:08:06,600 I saw a chart where it said that senior positions went up and 136 00:08:06,600 --> 00:08:08,560 then the junior positions actually went down. 137 00:08:09,040 --> 00:08:11,560 And I've also seen a narrative just now I was reading on Reddit 138 00:08:11,720 --> 00:08:14,880 where someone said, well, I have my degree and I have really good 139 00:08:14,880 --> 00:08:18,320 grades, four point OGPAI study from this school, but I've done 140 00:08:18,320 --> 00:08:21,320 X amounts of thousands of applications and I still don't 141 00:08:21,320 --> 00:08:23,720 have a job. That is a little bit concerning 142 00:08:23,720 --> 00:08:27,400 when it comes to new and early in career people actually coming 143 00:08:27,400 --> 00:08:30,160 to the market. I'm curious because you engaged 144 00:08:30,160 --> 00:08:31,800 with a lot of engineering leadership, what your 145 00:08:31,800 --> 00:08:34,760 perspective is on that? Yeah, that's interesting, right. 146 00:08:34,760 --> 00:08:38,440 The research contradicts what we're seeing that what we're 147 00:08:38,440 --> 00:08:45,520 seeing happening slowly. So for and this is maybe also 148 00:08:45,520 --> 00:08:49,680 the positioning about AI. So no judgement there. 149 00:08:49,680 --> 00:08:57,160 I think that a lot of a lot of misunderstanding is that AI can 150 00:08:57,160 --> 00:08:59,280 do the job of a junior developer. 151 00:08:59,280 --> 00:09:02,520 So we don't need the junior developer anymore. 152 00:09:05,320 --> 00:09:07,080 There's there's a few things to that. 153 00:09:08,160 --> 00:09:15,360 I think the most important one is that when you, when you look 154 00:09:15,360 --> 00:09:23,080 at the skill and doing software development at skill, it doesn't 155 00:09:23,080 --> 00:09:26,680 make sense. So you can't only hire senior 156 00:09:26,680 --> 00:09:33,920 engineers, There's not enough of them even like with with the 157 00:09:33,920 --> 00:09:36,920 projections that we see with the number of engineers that that 158 00:09:36,920 --> 00:09:40,400 will be there in the following years, it will not match what is 159 00:09:40,400 --> 00:09:43,880 actually needed. So that's a bit 160 00:09:43,880 --> 00:09:49,040 counterintuitive. I think one other thought is 161 00:09:49,040 --> 00:09:56,800 that when you look at actual AI adoption, research tells us that 162 00:09:56,800 --> 00:10:01,000 junior engineers are more likely to adopt AI. 163 00:10:02,440 --> 00:10:04,360 They are more native to the technology. 164 00:10:04,360 --> 00:10:10,360 They are more able to go beyond the like, what some few as 165 00:10:10,360 --> 00:10:14,480 classic developer work. They're not too opinionated on 166 00:10:14,480 --> 00:10:16,800 that yet. They're, you know, willing to 167 00:10:16,800 --> 00:10:20,960 use AI to do something beyond what is in their original job 168 00:10:20,960 --> 00:10:24,040 description. They're the change agents. 169 00:10:24,040 --> 00:10:28,200 So they're that's also a way to look to look at it. 170 00:10:29,640 --> 00:10:31,360 And I think that makes perfect sense. 171 00:10:31,400 --> 00:10:35,760 And my expectation is that what we'll see happening is that at 172 00:10:35,760 --> 00:10:38,880 some point this trend will be broken and we will actually see 173 00:10:38,880 --> 00:10:42,080 that the junior engineers becoming high demands because 174 00:10:42,080 --> 00:10:47,120 these junior engineers are able to navigate this more perfectly. 175 00:10:48,560 --> 00:10:55,000 They can handle very much the ambiguity that will certainly 176 00:10:56,080 --> 00:10:58,120 certainly exist in software engineering role. 177 00:10:59,560 --> 00:11:06,040 And the senior engineers can be there to as SM ES to codify 178 00:11:06,880 --> 00:11:11,440 their tribal knowledge, to codify best practices so that 179 00:11:11,440 --> 00:11:13,400 agents and humans can leverage that. 180 00:11:13,880 --> 00:11:17,640 And also to train these junior engineers because we, we, we 181 00:11:17,640 --> 00:11:26,320 need steady supply of, of people that will be able to, to become 182 00:11:26,320 --> 00:11:29,080 like, you know, senior software engineers at some point in their 183 00:11:29,080 --> 00:11:33,280 career. Then, then there's also the 184 00:11:33,280 --> 00:11:36,320 profile. So instead of like only looking 185 00:11:36,320 --> 00:11:40,920 at the T shaped engineers, we will also see M shaped engineers 186 00:11:40,920 --> 00:11:46,040 that are, you know, better qualified at two different, 187 00:11:46,200 --> 00:11:49,560 different areas. So yeah, it's an interesting, 188 00:11:49,560 --> 00:11:51,680 interesting thing that is happening. 189 00:11:51,680 --> 00:11:57,000 But if you look at everything combined to skill, culture, the 190 00:11:57,160 --> 00:12:02,440 ambiguity and the role that that will be probably what we'll see 191 00:12:02,440 --> 00:12:08,200 happening over the following year years, yeah. 192 00:12:08,920 --> 00:12:12,280 For a lot of my career, I was the person that was early in 193 00:12:12,280 --> 00:12:14,000 career. I haven't been in a position 194 00:12:14,000 --> 00:12:18,440 where I actually engaged with A-Team that is quite younger 195 00:12:18,440 --> 00:12:19,600 than me. I'm now 30. 196 00:12:19,600 --> 00:12:22,480 And I recently had an experience where there was a whole team of 197 00:12:22,480 --> 00:12:24,800 people younger than me. I was like, they actually have a 198 00:12:24,800 --> 00:12:27,480 really good energy and they're experimenting on their own and 199 00:12:27,480 --> 00:12:31,520 they're really wanting with a lot of drive to push this 200 00:12:31,520 --> 00:12:33,960 project forward. And I was like, yeah, like I 201 00:12:33,960 --> 00:12:36,040 used to be like that. I can still see people like 202 00:12:36,040 --> 00:12:37,920 that. I think when you have experience 203 00:12:37,920 --> 00:12:40,760 under your belt, you become a little bit more sceptical, 204 00:12:40,760 --> 00:12:42,680 realistic. Maybe you burned your fingers a 205 00:12:42,680 --> 00:12:46,280 few times like it's normal, but I do like that energy. 206 00:12:46,280 --> 00:12:47,920 I miss that energy. And I feel like if an 207 00:12:47,920 --> 00:12:50,520 organization wouldn't have that, it also doesn't inspire the 208 00:12:50,520 --> 00:12:53,160 people that are there anymore to a level that they should be 209 00:12:53,160 --> 00:12:57,040 inspired, I feel like as well. It's that exactly. 210 00:12:57,040 --> 00:13:03,960 And remember, coding is not only about the tools, it's not only 211 00:13:03,960 --> 00:13:05,880 about the people, it's also about the culture. 212 00:13:06,520 --> 00:13:11,680 And that is also something that's, that's those juniors 213 00:13:11,680 --> 00:13:14,720 bring in these can be your change agents to, to, to get to 214 00:13:14,720 --> 00:13:19,160 that culture. Technology helps us there as 215 00:13:19,160 --> 00:13:21,960 well, right? So if we, if we would see that, 216 00:13:22,680 --> 00:13:24,440 let's imagine you have an A Gentile team. 217 00:13:24,440 --> 00:13:28,040 In my mind, that's three people, lot of customers experimenting 218 00:13:28,040 --> 00:13:31,720 with exactly that number. And that could be like 1:00 PM 219 00:13:31,720 --> 00:13:36,640 and 2:00 juniors. And you know, anecdotally, 220 00:13:37,080 --> 00:13:39,240 people say, oh, that's a nightmare scenario because you 221 00:13:39,240 --> 00:13:41,240 will never know what they will create. 222 00:13:41,480 --> 00:13:44,120 And those agents will will run riot. 223 00:13:44,360 --> 00:13:48,400 And in all their enthusiasm and with their lack of knowledge, 224 00:13:48,760 --> 00:13:52,120 yeah, they might create like something we would think is 225 00:13:52,120 --> 00:13:56,600 suboptimal, subpar or, or whatever negative connotation. 226 00:13:57,360 --> 00:13:59,680 And that doesn't need to be true. 227 00:13:59,680 --> 00:14:03,360 Because if you look at like that model, and again, this is this 228 00:14:03,360 --> 00:14:05,040 is a bigger change, right? This is culture. 229 00:14:07,040 --> 00:14:11,720 Imagine that these SM ES codify their best practices. 230 00:14:11,720 --> 00:14:15,600 We implement platform engineering in the right way. 231 00:14:16,680 --> 00:14:26,920 We have that culture where where collaboration, where creativity, 232 00:14:26,920 --> 00:14:30,480 embracing uncertainty, those principles where, where that 233 00:14:30,560 --> 00:14:33,280 where that is where that is in in in the DNA. 234 00:14:34,760 --> 00:14:38,920 Why would it get off script? Because if they're able to 235 00:14:39,680 --> 00:14:43,680 define the right outcomes and have to write guard feels and 236 00:14:43,680 --> 00:14:47,240 constraints in place, then neither did the human nor the 237 00:14:47,240 --> 00:14:51,920 agent can can, you know, can go off script too much. 238 00:14:52,200 --> 00:14:54,160 You're setting them up for success. 239 00:14:54,520 --> 00:14:56,840 And I think that is a bit of the like the mind shift. 240 00:14:56,960 --> 00:14:58,520 How can it make others successful? 241 00:14:59,200 --> 00:15:04,240 We have a thing in Microsoft and it's it's high culture, high 242 00:15:04,240 --> 00:15:07,160 performance culture. High performance doesn't mean 243 00:15:07,160 --> 00:15:11,240 that one person is super cute and like, you know, wears a cake 244 00:15:11,240 --> 00:15:13,680 all day. That's also the person that 245 00:15:13,680 --> 00:15:19,040 enables others to be successful, that enables others to build the 246 00:15:19,040 --> 00:15:22,000 next great thing. And I think that that is 247 00:15:22,000 --> 00:15:29,120 something that's it just makes sense to to look at it in that 248 00:15:29,120 --> 00:15:32,880 way instead of like, oh, I'm going to replace juniors with 249 00:15:32,880 --> 00:15:36,960 AII think you shouldn't replace necessarily anyone without AI 250 00:15:37,200 --> 00:15:41,200 unless that there's like literally no activities anymore 251 00:15:41,200 --> 00:15:45,120 for a person. But yeah, the strong conventions 252 00:15:45,200 --> 00:15:48,040 and the guard rails and a platform where people can 253 00:15:48,040 --> 00:15:50,840 accelerate, I think those are a lot of the ingredients that 254 00:15:50,840 --> 00:15:54,440 platform engineering brings to the table or also a lot of the 255 00:15:54,440 --> 00:15:57,560 problems that it solves when that is not there, when there's 256 00:15:57,560 --> 00:15:59,880 friction or whether there's different flavours of technology 257 00:15:59,880 --> 00:16:03,240 and kind of wild spread, which really goes hand in hand. 258 00:16:03,240 --> 00:16:05,880 And I feel like platform engineering now should also have 259 00:16:05,880 --> 00:16:09,920 an emphasis on making this work in energetic way of working, 260 00:16:09,920 --> 00:16:11,520 which is really cool, to be honest. 261 00:16:12,240 --> 00:16:16,000 That might be then in this vision where seniors really make 262 00:16:16,000 --> 00:16:19,360 a difference, where they say, OK, these are from what we know, 263 00:16:19,760 --> 00:16:22,320 best practices, strong conventions, differentiating 264 00:16:22,320 --> 00:16:25,440 factors for effectiveness. When we talk about small 265 00:16:25,440 --> 00:16:28,800 incremental teams that deliver value, which can be very cool. 266 00:16:29,200 --> 00:16:31,480 Are there organizations experimenting with this way of 267 00:16:31,480 --> 00:16:34,280 working as you mentioned already, also with regards to 268 00:16:34,280 --> 00:16:37,200 the seniority within teams? Yes, yeah, yes. 269 00:16:37,360 --> 00:16:39,360 So then for people that are early in career and that are 270 00:16:39,360 --> 00:16:41,640 listening, they just have to find those organizations. 271 00:16:42,560 --> 00:16:44,560 Yeah. Look, hang in there. 272 00:16:45,360 --> 00:16:49,640 I think it's, and this is basically advice that's that's 273 00:16:50,320 --> 00:16:53,520 I'm giving to anybody that is, that is new in career. 274 00:16:55,840 --> 00:16:58,520 Just expand your mind. Make sure that you, you are 275 00:16:58,520 --> 00:17:01,080 aware of a little bit of everything. 276 00:17:03,040 --> 00:17:06,040 This is a craft that is changing super fast. 277 00:17:07,079 --> 00:17:12,319 It doesn't hurt to know about proper UX, you know, and, and 278 00:17:12,319 --> 00:17:16,599 being able to understand that doesn't, you know, with, with 279 00:17:16,599 --> 00:17:21,480 all of these topics, security, the product management, it, it 280 00:17:21,480 --> 00:17:25,079 helps if you know a bit of that because that is the value that 281 00:17:25,079 --> 00:17:30,040 you will bring extra with it and just don't give up. 282 00:17:31,240 --> 00:17:36,080 I I think and I see that this will change at some point 283 00:17:36,080 --> 00:17:40,600 because economics of skill tell us that this is not a 284 00:17:40,600 --> 00:17:45,640 sustainable model, right? Yeah, a lot of my emphasis I 285 00:17:45,640 --> 00:17:48,680 think also in the impact that I had early in career was the fact 286 00:17:48,680 --> 00:17:51,240 that I'm a good communicator, collaborator and a kind of 287 00:17:51,240 --> 00:17:54,120 bridge gaps. My fear is that, and this was 288 00:17:54,120 --> 00:17:56,800 also just me coming into a project and being the sole 289 00:17:56,800 --> 00:17:59,360 person that needed to build an integration, just get it done. 290 00:17:59,360 --> 00:18:02,040 I was like I I could do it. It was a lot of fun. 291 00:18:02,600 --> 00:18:05,320 I did not interact with as many people within my team because it 292 00:18:05,320 --> 00:18:09,600 was kind of a one man army task. And I feel like that will or 293 00:18:09,600 --> 00:18:11,680 there is the option to do that more and more. 294 00:18:11,960 --> 00:18:15,040 So then there would be less inner operatings with a bigger 295 00:18:15,040 --> 00:18:16,680 team. You already mentioned that 3 296 00:18:16,680 --> 00:18:20,120 might be the the solid number and I see there's a good thing 297 00:18:20,120 --> 00:18:22,240 and a bad thing. I still think that the soft 298 00:18:22,240 --> 00:18:25,600 skills aspect of things collaborating, we will still be 299 00:18:25,880 --> 00:18:28,240 a lot of interaction with people in and of it's own. 300 00:18:29,200 --> 00:18:32,120 But some of the complexity also goes into and genetic work which 301 00:18:32,120 --> 00:18:34,280 is amplified on a different level. 302 00:18:34,880 --> 00:18:38,320 Yeah, if you were to advise people what to focus on, it 303 00:18:38,320 --> 00:18:40,360 could be people that are early in career, it could be existing 304 00:18:40,360 --> 00:18:42,960 software engineers. They're experimenting with 305 00:18:42,960 --> 00:18:46,240 genetic work flows. What should they now looking at 306 00:18:46,240 --> 00:18:48,720 first principles from software engineering focus on with 307 00:18:48,720 --> 00:18:51,600 regards to knowledge from your perspective? 308 00:18:52,800 --> 00:18:56,960 That is a great question in terms of knowledge, of course, 309 00:18:56,960 --> 00:19:00,040 beyond coding itself and understanding these agents. 310 00:19:00,080 --> 00:19:05,160 It's a good name now. Yeah. 311 00:19:08,120 --> 00:19:14,160 What it what it is that I I literally see happening is that 312 00:19:14,360 --> 00:19:22,200 a lot of a lot of principles, a lot of frameworks, best 313 00:19:22,200 --> 00:19:30,600 practices that are older than I am, maybe even they they, they 314 00:19:30,600 --> 00:19:38,000 get a Polish up. There is a lot of principles 315 00:19:38,000 --> 00:19:39,880 that that have been around forever. 316 00:19:39,880 --> 00:19:45,200 There's a lot of best practices and we see that with 317 00:19:49,120 --> 00:19:52,000 specifically the the specification parts, 318 00:19:52,000 --> 00:19:58,480 specifically the beginning of of that software, software 319 00:19:58,480 --> 00:20:03,320 development life cycle. I, I it is those principles I 320 00:20:03,320 --> 00:20:05,800 would. Advice to take another look at 321 00:20:05,960 --> 00:20:10,600 right a, you know, if, if you really understand everything 322 00:20:10,600 --> 00:20:15,760 there is about behavioural driven development, that's a 323 00:20:15,760 --> 00:20:19,880 great one, right. That's valuable understanding a 324 00:20:19,880 --> 00:20:25,880 bit more about systems, understanding more about, I 325 00:20:25,880 --> 00:20:28,760 don't know, the SOLID principles and there's so many, like 326 00:20:28,920 --> 00:20:35,040 there's an enormous list of things that you can just now 327 00:20:35,040 --> 00:20:36,920 acquire in, in terms of knowledge. 328 00:20:36,920 --> 00:20:42,480 It's not difficult to do. But but if, if you understand 329 00:20:42,480 --> 00:20:48,120 these principles, then that will help you massively in, in 330 00:20:49,280 --> 00:20:52,920 defining these specifications and defining these constraints 331 00:20:52,920 --> 00:21:00,680 in making an making an actual software products instead of 332 00:21:00,680 --> 00:21:03,960 codes. My manager has a nice one and I 333 00:21:03,960 --> 00:21:09,680 just have to put it out there. She always says, says coding is 334 00:21:09,680 --> 00:21:14,520 cheap, software is expensive. And that's just the difference 335 00:21:14,520 --> 00:21:17,840 between between just like putting out some codes or being 336 00:21:17,840 --> 00:21:22,560 really intentional about what you're doing, knowing your, your 337 00:21:22,560 --> 00:21:26,200 heritage, knowing, knowing whatever that these researchers 338 00:21:26,200 --> 00:21:31,640 have researched for years, all those best practices, leverage 339 00:21:31,640 --> 00:21:33,800 them. Learn that be, be, be good at 340 00:21:33,800 --> 00:21:37,920 that. Be be the one that says, but 341 00:21:38,280 --> 00:21:42,080 what if we would like, instead of using this method, use 342 00:21:42,080 --> 00:21:46,440 another methodology? You know, I think that's that's, 343 00:21:46,440 --> 00:21:51,360 that's that would be beautiful. Yeah, I'm happy that a lot of 344 00:21:51,360 --> 00:21:54,880 strong principles and like foundational knowledge, knowing 345 00:21:54,880 --> 00:21:59,600 that will still help in the end. We're not really vibe coding on 346 00:21:59,600 --> 00:22:01,840 enterprise skill. We're still creating software 347 00:22:01,840 --> 00:22:05,240 that for the most part is going to be there hopefully for years 348 00:22:05,240 --> 00:22:08,200 on end. Creating is a lot easier. 349 00:22:08,320 --> 00:22:10,640 I feel like creating in the right way we're getting better 350 00:22:10,640 --> 00:22:14,440 and better at, but we will go into a mode where creation is 351 00:22:14,440 --> 00:22:16,880 just there in abundance. So then we have to make sure 352 00:22:16,880 --> 00:22:20,800 that we create the right thing, created the wrong the right way, 353 00:22:20,800 --> 00:22:23,240 and then also has the longevity component to it. 354 00:22:23,960 --> 00:22:26,600 One of the reasons why I really like Go as a programming 355 00:22:26,600 --> 00:22:30,040 language was that the creation process in the end, what you see 356 00:22:30,040 --> 00:22:33,720 in code was quite simple. It's still on version 1 dot 357 00:22:33,840 --> 00:22:35,640 something something. It's probably in the 20s now, 358 00:22:35,640 --> 00:22:38,680 but it's still a one major version, which means I never had 359 00:22:38,680 --> 00:22:40,960 to migrate my code. If there was a new update, it 360 00:22:40,960 --> 00:22:44,200 was kind of seamless. That process was always pretty 361 00:22:44,200 --> 00:22:46,640 simple. There are not many things that I 362 00:22:46,640 --> 00:22:50,280 could do in different ways in Go as I could compare to other 363 00:22:50,280 --> 00:22:53,120 languages, which is why I really enjoyed it now. 364 00:22:53,120 --> 00:22:55,680 It has a strong readability component, which is fantastic 365 00:22:55,680 --> 00:22:57,520 when we're talking about software that is generated 366 00:22:57,520 --> 00:23:01,080 because if you have a language that's easy to read, you will be 367 00:23:01,080 --> 00:23:04,640 reading a lot more I feel like nowadays, which is also funny 368 00:23:04,640 --> 00:23:06,480 when we're talking about programming languages 369 00:23:06,480 --> 00:23:09,640 specifically. I wonder if agentic software 370 00:23:09,640 --> 00:23:12,040 engineering is also going to influence in the end the 371 00:23:12,040 --> 00:23:14,440 language that you choose right now. 372 00:23:14,440 --> 00:23:17,920 I think it will because we are doing a lot of the reviewing 373 00:23:17,920 --> 00:23:21,520 work as humans still with an agent still probably like pre 374 00:23:21,840 --> 00:23:23,920 reviewed, but in the end still as a final pass. 375 00:23:23,920 --> 00:23:26,320 There's humans in the loop. I don't know how that's going to 376 00:23:26,320 --> 00:23:29,840 evolve, which I think is still quite exciting to see how it 377 00:23:29,840 --> 00:23:33,080 plays out. Is this a long intro to asking 378 00:23:33,080 --> 00:23:35,240 the question? Will everything be JavaScript? 379 00:23:36,160 --> 00:23:37,800 I hope not. But yeah, what? 380 00:23:37,880 --> 00:23:42,120 How do you see that? Luke, I think that yes, of 381 00:23:42,120 --> 00:23:45,160 course there there there will be some influence here and there 382 00:23:45,160 --> 00:23:49,960 because the agent, if you don't stare at the agent, picks what 383 00:23:49,960 --> 00:23:51,560 the agent will pick. Yeah, right. 384 00:23:52,640 --> 00:23:55,720 On the other hand, who is defining what the agents should 385 00:23:55,720 --> 00:23:59,800 build? It is us humans normally. 386 00:24:00,400 --> 00:24:06,120 Yeah, still so, so I'm, I'm not too, too grim about that. 387 00:24:07,600 --> 00:24:13,080 But one one thing that's this is maybe a bit too hypothetical, 388 00:24:13,080 --> 00:24:21,920 but you could wonder like not if in 10 years or so the the 389 00:24:21,920 --> 00:24:26,560 default programming language will be JavaScript, which we can 390 00:24:26,560 --> 00:24:30,600 argue about, we can argue about this for an hour maybe. 391 00:24:30,920 --> 00:24:34,720 But it is like, what is the future of software even? 392 00:24:35,200 --> 00:24:39,520 And what is the future of software languages at all Like? 393 00:24:40,000 --> 00:24:44,320 Is it too far fetched to think that in, I don't know, a few 394 00:24:44,320 --> 00:24:48,240 years we will create applications using natural 395 00:24:48,240 --> 00:24:51,080 language and that will be compiled. 396 00:24:52,640 --> 00:24:55,400 Who who cares? Like what, how it is compiled, 397 00:24:55,400 --> 00:24:59,600 what language it's using? And so that that's one thing. 398 00:24:59,640 --> 00:25:02,400 And the other thing is what even is an application? 399 00:25:03,440 --> 00:25:09,000 I work with customers that are rethinking their their apps and 400 00:25:09,000 --> 00:25:12,800 they're saying, well, but my customer, yes, sure, they go to 401 00:25:12,800 --> 00:25:18,320 a website right now, but if they are in an in an, in an agent, 402 00:25:19,560 --> 00:25:22,440 they're using, I don't know, something like chit GPT or 403 00:25:22,920 --> 00:25:25,240 clouds or whatever. And they they ask a question 404 00:25:25,240 --> 00:25:29,440 like, Hey, how can I get how, how do I choose? 405 00:25:29,840 --> 00:25:34,320 I don't know life insurance. So why wouldn't it just suggest 406 00:25:34,320 --> 00:25:38,280 me the good life insurance and I just close it right from there? 407 00:25:40,000 --> 00:25:44,000 Or even would it be too far fetched to think that when I 408 00:25:44,000 --> 00:25:47,520 would a life would want a new life insurance that there would 409 00:25:47,520 --> 00:25:53,600 be agents being kicked off to negotiate on my behalf And just 410 00:25:53,640 --> 00:25:57,640 look which one comes back with the right offering or when I 411 00:25:57,640 --> 00:26:03,040 want to do my groceries. So this is super exciting. 412 00:26:03,880 --> 00:26:07,280 But yeah, there there's, there's of course a lot of beliefs about 413 00:26:08,560 --> 00:26:10,960 the way that that we are looking at software development right 414 00:26:10,960 --> 00:26:12,000 now. I think that will change 415 00:26:12,000 --> 00:26:16,200 massively. But but yeah, I on on the 416 00:26:16,920 --> 00:26:20,240 programming language. Yeah, it is of course there was 417 00:26:20,240 --> 00:26:22,960 a preference to the colour purple according to some LLM 418 00:26:22,960 --> 00:26:24,400 right. Oh really, if the people would 419 00:26:24,960 --> 00:26:28,800 build a web app, then it would use the colour purple because 420 00:26:29,200 --> 00:26:33,880 The Color Purple and then some reason, some research, I don't 421 00:26:33,880 --> 00:26:36,840 know. And thus like everything becomes 422 00:26:36,840 --> 00:26:40,440 purple. So if then in the next training 423 00:26:40,440 --> 00:26:42,840 would take the place, then everything would be purple. 424 00:26:42,840 --> 00:26:46,080 Or you know, there's there's a lot of things to to say about 425 00:26:46,080 --> 00:26:48,960 that. I was reading an interview with 426 00:26:50,640 --> 00:26:55,160 it was Reddit and like someone these stated, yeah everything we 427 00:26:55,160 --> 00:26:58,600 say on Reddit becomes becomes the future training data. 428 00:26:59,520 --> 00:27:01,040 Yeah. And I don't know if that's a 429 00:27:01,040 --> 00:27:05,440 good thought or or a horrible thought, depends on, you know, 430 00:27:05,440 --> 00:27:09,720 which subreddit you're on. But yeah, there, there. 431 00:27:09,720 --> 00:27:13,160 There's a lot of implications with that. 432 00:27:14,040 --> 00:27:17,640 Every every so now often I'm on Reddit or I'm on YouTube and I 433 00:27:17,640 --> 00:27:19,680 scroll through the comments and someone makes a comment that 434 00:27:19,680 --> 00:27:21,920 says all of these other comments. 435 00:27:22,120 --> 00:27:24,960 They sound like bots and I'm like I have no clue who I'm 436 00:27:24,960 --> 00:27:28,480 actually talking to nowadays if that is the truth or not. 437 00:27:28,720 --> 00:27:32,640 Is it like with the M dash? So I. 438 00:27:32,920 --> 00:27:35,440 When I see that, I'm like, that's the smell. 439 00:27:35,800 --> 00:27:43,200 But so, yeah. But we think then that that's 440 00:27:43,200 --> 00:27:44,800 that's a thing that an agent does. 441 00:27:44,800 --> 00:27:49,320 But on the other hand, like, I always learned that I used too 442 00:27:49,320 --> 00:27:51,160 many commas. Oh, yeah, yeah. 443 00:27:51,160 --> 00:27:54,000 So what did I do? You removed the commas, yeah. 444 00:27:54,080 --> 00:28:01,400 I moved to M Dash because I that's the way that you know, 445 00:28:01,400 --> 00:28:03,320 that's that's my way of communication. 446 00:28:03,320 --> 00:28:06,280 And then I had a whole period where I was like, yeah, I'm just 447 00:28:06,280 --> 00:28:09,560 going to going to replace all the M dashes with a minus sign 448 00:28:09,560 --> 00:28:13,440 or something maybe. But it's just, yeah, there's no 449 00:28:13,920 --> 00:28:16,200 at some point, no way of telling, I think. 450 00:28:16,560 --> 00:28:19,320 No, I mean, I'm fully there with you actually. 451 00:28:19,720 --> 00:28:21,280 That's so funny. I saw on LinkedIn everyone 452 00:28:21,280 --> 00:28:23,680 people were like, I was doing M dashes before it was cool and 453 00:28:23,680 --> 00:28:25,120 now all my articles are getting rushed. 454 00:28:25,760 --> 00:28:28,560 Like, this is also not a good movement, but yeah, yeah, that's 455 00:28:28,600 --> 00:28:30,960 funny. As that's humans and you know, 456 00:28:30,960 --> 00:28:38,160 the, the, the point of this is as humans, we, we value that 457 00:28:38,160 --> 00:28:42,640 somebody communicates as a, as a, as a natural person, as a 458 00:28:42,640 --> 00:28:44,000 human. Yeah, authentically. 459 00:28:44,200 --> 00:28:49,040 Authentic, human, real. On the other hand, we all love 460 00:28:49,040 --> 00:28:57,520 it when AI does that for us. With with coding, you know, it's 461 00:28:57,520 --> 00:29:00,120 the same story. We still own the code, we are 462 00:29:00,120 --> 00:29:02,440 still responsible. The human is still responsible 463 00:29:02,440 --> 00:29:07,400 for what we're writing. If AI have written 80 or some 464 00:29:07,400 --> 00:29:13,160 customers, even now, 90% of the code, 90%, their engineers are 465 00:29:13,160 --> 00:29:15,760 still responsible. They are still the humans that 466 00:29:15,760 --> 00:29:19,320 are in the loop. We're trying to remove it where 467 00:29:19,320 --> 00:29:24,040 it doesn't make sense if we have enough confidence, but but 468 00:29:24,040 --> 00:29:27,880 still, it's a human that's that's responsible for for what 469 00:29:27,880 --> 00:29:33,280 gets out there. And if I use AI to translate my 470 00:29:33,280 --> 00:29:37,560 ideas and people can actually read it, I think that's way 471 00:29:37,560 --> 00:29:42,000 better than when I spend too much time on writing it myself. 472 00:29:42,480 --> 00:29:46,040 And people read it and think, wow, she's unhinged. 473 00:29:46,040 --> 00:29:48,560 But this doesn't make sense. I don't know. 474 00:29:48,560 --> 00:29:54,520 I stopped halfway, right. So that's, that's always a bit 475 00:29:54,520 --> 00:29:57,160 of the, there's always a train off there. 476 00:29:57,400 --> 00:30:00,080 Yeah, yeah. I had a conversation with Gregor 477 00:30:00,080 --> 00:30:03,600 Hope on the podcast as well and his stance was really OK. 478 00:30:03,600 --> 00:30:07,160 The if the baseline has moved because the baseline is now you 479 00:30:07,160 --> 00:30:10,400 have output with regards to AI, you still have to add value on 480 00:30:10,400 --> 00:30:12,800 top because otherwise anyone can do the same. 481 00:30:12,800 --> 00:30:14,960 Maybe the reasoning or the prompting is slightly different, 482 00:30:14,960 --> 00:30:17,080 the flavour is a little bit different, but if you don't add 483 00:30:17,080 --> 00:30:21,040 your value on top then basically anyone can do it. 484 00:30:21,160 --> 00:30:24,280 Your value is really the value you add, so keep doing that and 485 00:30:24,280 --> 00:30:25,880 otherwise it's useless what you're doing. 486 00:30:27,040 --> 00:30:28,840 And that's an excellent remark, right? 487 00:30:28,880 --> 00:30:35,600 And This is why it doesn't make sense if all people use AI to, 488 00:30:36,160 --> 00:30:38,480 you know, I think the most visible thing is LinkedIn. 489 00:30:39,280 --> 00:30:43,760 If it's like the 19th article that is written by AI and it's 490 00:30:43,760 --> 00:30:47,120 all looks the same and it's like nothing is being added, yes. 491 00:30:47,120 --> 00:30:49,920 What is the purpose? The same you can say for 492 00:30:49,920 --> 00:30:59,360 software, right? And on one end, we, we with AI, 493 00:30:59,360 --> 00:31:02,720 with this technology and with the right culture, we enable 494 00:31:02,720 --> 00:31:05,080 everybody to, to get the most out of it. 495 00:31:05,840 --> 00:31:09,600 Because the the remark of a customer, the idea of, of 496 00:31:09,600 --> 00:31:14,480 somebody in the business, the the, the experience by the PM 497 00:31:14,480 --> 00:31:20,560 and by the engineers that are involved that can, that can 498 00:31:20,560 --> 00:31:25,200 create something great. On the other hand, we also see a 499 00:31:25,200 --> 00:31:28,120 lot of software companies now struggling because they created 500 00:31:28,120 --> 00:31:31,640 the product, but basically it didn't add much. 501 00:31:32,520 --> 00:31:35,760 Think of the example where you have, I'm not going to name a 502 00:31:35,760 --> 00:31:37,960 shame, but you have an open source product. 503 00:31:38,640 --> 00:31:41,840 Monetization with open source has always been challenging. 504 00:31:41,840 --> 00:31:45,960 So one of the scenarios is of course you release the premium 505 00:31:45,960 --> 00:31:49,840 feature and you maintain that and the customer buys that and 506 00:31:49,840 --> 00:31:53,120 with that secure that there flows money into the open source 507 00:31:53,120 --> 00:31:56,600 project. I would wish that everybody 508 00:31:56,600 --> 00:32:00,960 would think that way, but also somebody could just say, well, I 509 00:32:00,960 --> 00:32:06,440 just used the product sheet and the recorded demo and I told AI 510 00:32:06,440 --> 00:32:09,520 built this for me and I got it 90% right. 511 00:32:09,800 --> 00:32:13,960 And I think that's good enough. And of course, I think that 512 00:32:13,960 --> 00:32:20,400 everybody should pay for open source, but but yes, support, 513 00:32:20,400 --> 00:32:23,680 support, support the makers, support the the contributors and 514 00:32:23,680 --> 00:32:28,080 the companies that make the project a life, that invest time 515 00:32:28,080 --> 00:32:30,200 and and resources in it and money. 516 00:32:32,240 --> 00:32:34,360 But yes, you need to add something really special. 517 00:32:34,720 --> 00:32:36,520 Yeah. Otherwise it's not good enough 518 00:32:36,520 --> 00:32:40,320 anymore. I feel like open source is the 519 00:32:40,320 --> 00:32:42,680 value of open source and kind of the monetization model. 520 00:32:43,040 --> 00:32:45,240 It was always challenging. So open source projects really 521 00:32:45,240 --> 00:32:48,600 nailed it and they are thriving. And the other ones that added it 522 00:32:48,600 --> 00:32:54,000 like an add on that one is the, I think impacted the most in 523 00:32:54,000 --> 00:32:57,600 terms of now people being able to kind of copy and and match 524 00:32:57,600 --> 00:32:59,640 things. Which means if that incentive is 525 00:32:59,640 --> 00:33:03,080 gone, I hope we don't get into a cycle where we have less and 526 00:33:03,080 --> 00:33:06,160 less things that are open out for people to use. 527 00:33:06,480 --> 00:33:09,160 Because I do love a lot of my open source software. 528 00:33:09,160 --> 00:33:13,600 And I do think it is, quite frankly, one of the most awesome 529 00:33:13,600 --> 00:33:16,120 things we have about this industry is the fact that people 530 00:33:16,120 --> 00:33:19,600 create, they share with almost nothing in return. 531 00:33:19,600 --> 00:33:23,120 Even the open source projects that have monetised, it's 532 00:33:23,120 --> 00:33:25,960 typically just a fragment of it. It's not the main thing. 533 00:33:25,960 --> 00:33:28,120 And the people that are maintaining it have a second 534 00:33:28,400 --> 00:33:31,360 like a full time day job. This is a side thing for them. 535 00:33:31,920 --> 00:33:36,120 So yeah, I I hope we don't get into a a position where we have 536 00:33:36,120 --> 00:33:38,800 less and less open source, but this is a risk for sure. 537 00:33:39,520 --> 00:33:43,040 Yeah, I, I, I hope the opposite. I hope that if we will see more 538 00:33:43,040 --> 00:33:51,240 open source and, but yeah, it's, it's something to, to that we 539 00:33:51,240 --> 00:33:57,440 need to actively work on. Yeah, the world depends on open 540 00:33:57,440 --> 00:34:01,320 source. It's like that for the 541 00:34:01,320 --> 00:34:04,520 foreseeable future. How many AI she'll build on top 542 00:34:04,520 --> 00:34:07,200 of something and how many platforms she'll build. 543 00:34:07,200 --> 00:34:11,360 Like there's still always that XKCD with that one package 544 00:34:11,360 --> 00:34:15,840 maintainer in Nebraska. But that's, that's the reality, 545 00:34:15,840 --> 00:34:19,480 right? And I, I see a lot of creativity 546 00:34:19,480 --> 00:34:26,600 happening and I think my expectation is that open source 547 00:34:26,639 --> 00:34:29,280 will still be there. It will be, will still be 548 00:34:29,280 --> 00:34:32,360 booming. I don't know about the larger 549 00:34:32,360 --> 00:34:35,760 projects though, because now it's easier to say, oh, well, I 550 00:34:35,760 --> 00:34:38,920 wanted to in another way or, oh, I'm really specific about 551 00:34:38,920 --> 00:34:42,360 something. I don't have the answer to that. 552 00:34:42,360 --> 00:34:44,840 But I think that's a, that's a thing we should solve as an 553 00:34:44,840 --> 00:34:49,360 industry as well. And tell people what open source 554 00:34:49,360 --> 00:34:52,679 is and and why it matters to contribute to like these 555 00:34:52,679 --> 00:34:54,679 projects and and work together. But. 556 00:34:56,040 --> 00:34:59,880 I have one final thought and it goes into the line of I think 557 00:34:59,880 --> 00:35:02,320 tooling that we now have available as software engineers 558 00:35:02,320 --> 00:35:04,800 specifically. I know a lot of people in 559 00:35:04,800 --> 00:35:07,680 start-ups or smaller organizations where tooling is 560 00:35:07,680 --> 00:35:10,880 kind of experimented with. They might fall in love with 561 00:35:11,240 --> 00:35:13,440 Cursor, for example, just to throw one out there. 562 00:35:13,680 --> 00:35:16,080 And if they then move to a bigger organization, they will 563 00:35:16,080 --> 00:35:17,920 no longer have access to the tool that they love. 564 00:35:17,960 --> 00:35:20,880 So now they use whatever is in our organization, whether that 565 00:35:20,880 --> 00:35:24,640 be something open, AI related, clawed, or get a copilot doesn't 566 00:35:24,640 --> 00:35:26,240 really matter. In the end. 567 00:35:26,240 --> 00:35:28,800 They have to be more flexible when they switch jobs to also 568 00:35:28,800 --> 00:35:32,680 switch tools, which is pretty funny because I love my tooling. 569 00:35:32,680 --> 00:35:36,400 I love mastering tools because then I feel like I can leverage 570 00:35:36,400 --> 00:35:38,920 that to achieve my outcomes faster. 571 00:35:39,040 --> 00:35:43,400 It was also in the IDEI love my hotkeys, my bindings to make 572 00:35:43,400 --> 00:35:45,520 sure that when I actually execute code, I do that 573 00:35:45,520 --> 00:35:48,320 efficiently because like to be effective in what I do. 574 00:35:48,680 --> 00:35:52,240 So letting go of a tool and then mastering another one, it's not 575 00:35:52,240 --> 00:35:53,600 really something I look forward to. 576 00:35:53,920 --> 00:35:56,840 There are a lot of, I feel like, underlying principles when we're 577 00:35:56,840 --> 00:35:59,560 talking about agentic development, the way you prompt, 578 00:35:59,600 --> 00:36:04,840 the way you context manage, how you kind of parallelize agents 579 00:36:04,840 --> 00:36:06,880 or how you navigate certain structures. 580 00:36:08,520 --> 00:36:11,280 We're still learning and we're still evolving also with how we 581 00:36:11,280 --> 00:36:14,280 structure context nowadays. There was this discussion on 582 00:36:14,280 --> 00:36:17,720 skills and agents, but I'm curious from your perspective, 583 00:36:17,720 --> 00:36:21,440 when you advise engineering leadership on the tooling they 584 00:36:21,440 --> 00:36:24,040 make available for their department, for their teams, for 585 00:36:24,040 --> 00:36:27,480 their organization, what do you make sure that is there? 586 00:36:27,880 --> 00:36:30,120 Because I'm quite curious from the people that are listening, 587 00:36:30,800 --> 00:36:33,520 they might not have the tooling they want, but it might be 588 00:36:33,520 --> 00:36:35,040 effective for what they need to do. 589 00:36:35,280 --> 00:36:36,560 So I'm curious to hear how you think. 590 00:36:37,760 --> 00:36:39,120 Yeah, that's, that's a great one. 591 00:36:39,200 --> 00:36:50,000 I'm a lot of the, a lot of the tools, to be very blunt here, 592 00:36:51,000 --> 00:36:57,520 they're not so much different. Can you achieve 80% the same 593 00:36:57,520 --> 00:37:00,640 thing with all of these tools? Probably, most likely. 594 00:37:01,000 --> 00:37:03,200 I didn't do the research on it, so don't pin me down. 595 00:37:03,880 --> 00:37:08,000 But look in, in general, these tools are, are different, but 596 00:37:08,040 --> 00:37:13,720 but not that different. One thing that that's, and this 597 00:37:13,720 --> 00:37:19,200 is regardless of the technology, we cannot just ship a tool and 598 00:37:19,200 --> 00:37:22,080 give it to a developer, especially not when it's highly 599 00:37:22,080 --> 00:37:26,720 sophisticated like an AI coding assistant or an an agent for 600 00:37:26,720 --> 00:37:33,960 doing anything basically. So what I advise these leaders 601 00:37:34,320 --> 00:37:36,920 is to at least enable the developers. 602 00:37:38,440 --> 00:37:41,120 And with enablement, I do not mean like, oh, this is how you 603 00:37:41,120 --> 00:37:42,840 get the license and this is where you find it. 604 00:37:43,000 --> 00:37:45,520 Although that's a really good start, don't get me wrong. 605 00:37:47,200 --> 00:37:49,560 And already better than like maybe like half of the 606 00:37:49,560 --> 00:37:52,160 companies, you know, it's like usually it's just like, here's a 607 00:37:52,160 --> 00:37:53,360 license. Good luck. 608 00:37:55,640 --> 00:37:58,280 You need to train on that. You need to understand like 609 00:37:58,920 --> 00:38:01,640 where do I put constraints? What are these constraints? 610 00:38:01,640 --> 00:38:04,880 Like this is a lot of concepts to grasp agents. 611 00:38:05,360 --> 00:38:08,800 Oh, I can write my own agent. You have to know how. 612 00:38:08,800 --> 00:38:10,840 You have to know where. You need to understand the 613 00:38:10,840 --> 00:38:14,320 specifics of these tools and how they process this information. 614 00:38:15,080 --> 00:38:17,360 Luckily, we have more and more standards, right? 615 00:38:18,200 --> 00:38:21,680 This is why we as an industry trying to figure out like, OK, 616 00:38:22,080 --> 00:38:23,800 where do we put these instructions? 617 00:38:25,080 --> 00:38:28,040 Let's adopt these skills. Let's look at certain protocols 618 00:38:28,040 --> 00:38:30,720 that help us. MCP is a great example of a 619 00:38:31,720 --> 00:38:35,960 highly fast moving, fast based industry that competes with each 620 00:38:35,960 --> 00:38:39,200 other to their life. OK, we need to settle on this. 621 00:38:39,200 --> 00:38:41,360 We need to find a, find a standard for that. 622 00:38:41,640 --> 00:38:43,840 Whatever you think of MCP, it's a standard, right? 623 00:38:44,000 --> 00:38:48,600 Yeah, but people need to learn what is MCP, how does that work? 624 00:38:48,600 --> 00:38:50,800 How do we build that securely? How do we skill that? 625 00:38:50,800 --> 00:38:56,000 What can't we do with MCP? Looking at the different rules? 626 00:38:56,720 --> 00:39:00,240 If you are a software developer, that's really great. 627 00:39:00,240 --> 00:39:01,760 There's tons of information out there. 628 00:39:02,040 --> 00:39:06,240 Even then, we need to teach you to understand how you can 629 00:39:06,240 --> 00:39:09,120 leverage the tool. But we also have data 630 00:39:09,120 --> 00:39:14,760 scientists, we have test engineers, we have PMS, we have, 631 00:39:14,920 --> 00:39:17,760 there's so many people that are involved in that whole coding 632 00:39:17,760 --> 00:39:21,200 thing and they all need to be enabled and understand like how 633 00:39:21,200 --> 00:39:24,680 they can get the best out of the tools that they use. 634 00:39:26,440 --> 00:39:31,720 Also about the models, what happens when I pick one LLM over 635 00:39:31,720 --> 00:39:34,720 another LLM? Sharing best practices 636 00:39:34,720 --> 00:39:39,120 internally, making sure that you set up an inner source culture. 637 00:39:39,520 --> 00:39:43,040 There's a whole, you did not need to do that all at one time, 638 00:39:46,280 --> 00:39:49,640 but there, there's like this whole trajectory where you can 639 00:39:49,640 --> 00:39:53,760 just mature and, and for each phase, you can add a little bit 640 00:39:53,840 --> 00:39:57,480 to it to ensure that you're successful with the tools. 641 00:39:57,480 --> 00:40:00,560 And then when you need to switch the tool and you're like, oh, 642 00:40:00,840 --> 00:40:03,320 this is like, it feels uncomfortable because it's 643 00:40:03,320 --> 00:40:07,960 different or it's in another IDE even, or it doesn't even have 644 00:40:07,960 --> 00:40:10,800 it's own IDE or whatever. It's ACLI suddenly. 645 00:40:12,200 --> 00:40:15,840 Look, it's the. Basically these LL Ms. and then 646 00:40:15,840 --> 00:40:22,240 this is of course talking, you know, working at Microsoft, we 647 00:40:22,320 --> 00:40:27,000 we can just switch the LL Ms. to whatever LLM that you use of 648 00:40:27,000 --> 00:40:29,000 course. And the most popular ones are in 649 00:40:29,000 --> 00:40:32,880 there and will be probably more. If the one is not in there, your 650 00:40:32,880 --> 00:40:36,200 organization could edit even, right, to make sure that you can 651 00:40:36,200 --> 00:40:41,160 access that. And this will be further 652 00:40:41,160 --> 00:40:45,360 commoditised. We already see that, like what 653 00:40:45,360 --> 00:40:47,960 we now do with GitHub. We're our platform, we're 654 00:40:47,960 --> 00:40:52,840 welcoming. So if you want to use Clot as an 655 00:40:52,840 --> 00:40:55,680 agent, sure. You use Clot as an agent. 656 00:40:55,680 --> 00:40:58,480 If your org enables that, that's really great. 657 00:40:58,600 --> 00:41:02,440 Then you only have another bit of a different interface, sure, 658 00:41:02,440 --> 00:41:06,040 but your agent is still there, uses the LLM that you're used 659 00:41:06,040 --> 00:41:08,080 to. It's the agent that you're used 660 00:41:08,080 --> 00:41:11,320 to. And, and I think that that's, 661 00:41:11,320 --> 00:41:15,320 yeah, that commoditisation will will go on and I think that 662 00:41:15,320 --> 00:41:19,360 everybody will benefit of that. But yeah, you can't just give it 663 00:41:19,360 --> 00:41:23,080 all to a developer. That's that's no, that's, that's 664 00:41:23,080 --> 00:41:26,960 not a driving adoption. Yeah, I like, I like that you 665 00:41:26,960 --> 00:41:30,080 say that an arbitrary percentage, but most tools right 666 00:41:30,080 --> 00:41:31,520 now, they kind of do the same thing. 667 00:41:31,880 --> 00:41:34,560 And The funny thing is, I thought of this and yeah, if 668 00:41:34,560 --> 00:41:36,960 they don't do the same thing, then within a few weeks, within 669 00:41:36,960 --> 00:41:39,240 a few months, they will still do the same thing because people 670 00:41:39,240 --> 00:41:41,480 look at each other. If my tool has something that is 671 00:41:41,480 --> 00:41:43,880 very special and other tools are out there, they are definitely 672 00:41:43,880 --> 00:41:47,360 going to also adopt something very similar to achieve the same 673 00:41:47,360 --> 00:41:49,080 outcome. It just makes sense. 674 00:41:49,520 --> 00:41:52,400 It's good that we have competition because as consumers 675 00:41:52,880 --> 00:41:55,680 we get the best products in the end that really I enjoy. 676 00:41:55,960 --> 00:41:58,880 Yeah, and, and me too and I love what's happening. 677 00:41:58,920 --> 00:42:01,840 I, I love looking at what's what's happening with, with 678 00:42:01,840 --> 00:42:06,600 competition and of course I'm ready to, you know, and I follow 679 00:42:06,600 --> 00:42:10,000 these discussions and people showing off and on YouTube and 680 00:42:10,000 --> 00:42:13,960 on TikTok. It's just by looking at how they 681 00:42:13,960 --> 00:42:16,680 are using a product and how they're talking about these 682 00:42:16,680 --> 00:42:21,280 products and helps me understand what is what is important. 683 00:42:21,360 --> 00:42:25,800 And there's some nice innovations that are being put 684 00:42:25,800 --> 00:42:29,640 out there. Not everything I would say, 685 00:42:29,640 --> 00:42:32,520 well, that's that's a nice implementation or that doesn't 686 00:42:32,520 --> 00:42:38,280 really sound enterprise to me. But yeah, in the end, 687 00:42:38,280 --> 00:42:42,200 competition is what makes software good and great and 688 00:42:44,080 --> 00:42:47,280 makes it inspirational to be in this field at this at this time. 689 00:42:47,840 --> 00:42:50,800 For the people that are listening and they are not 690 00:42:50,800 --> 00:42:54,960 enabled to a degree where you say this is kind of the tooling, 691 00:42:54,960 --> 00:42:58,480 that you should at least have at least one of the tools, do you 692 00:42:58,480 --> 00:43:01,680 feel like that warrants them changing the organization that 693 00:43:01,680 --> 00:43:05,400 they are in? Because I think I do almost if 694 00:43:05,400 --> 00:43:07,360 it was easy to just switch organisations. 695 00:43:07,840 --> 00:43:10,840 Disregarding that, I feel like now things are evolving and 696 00:43:10,840 --> 00:43:13,720 accelerating so much in our craft that if you cannot 697 00:43:13,720 --> 00:43:17,560 experiment on a day-to-day, which means at your job, it 698 00:43:17,560 --> 00:43:19,760 might be actually a good reason. Yeah, totally. 699 00:43:19,840 --> 00:43:31,280 If you if you look at at organizations and looking at 700 00:43:31,280 --> 00:43:39,080 their, their strategies on on retraining and retaining talent, 701 00:43:39,280 --> 00:43:46,560 there's a lot to do. Having a good culture, having a 702 00:43:47,920 --> 00:43:53,520 mature tools, which goes from your day-to-day software 703 00:43:53,520 --> 00:43:58,400 development to how easily can I obtain a platform? 704 00:43:58,400 --> 00:44:01,440 How can I skill? How can I make these changes 705 00:44:01,520 --> 00:44:03,720 during all your platform engineering implementation? 706 00:44:04,400 --> 00:44:07,080 Yeah, there will be, there will be choice, There will be 707 00:44:07,080 --> 00:44:11,920 choices. And you know, the more reason 708 00:44:12,080 --> 00:44:18,720 for organisations to think this true really, I'm certain that we 709 00:44:18,720 --> 00:44:21,480 will see more and more platform engineering conversations 710 00:44:21,480 --> 00:44:23,800 popping up. I don't know if we did a great 711 00:44:23,800 --> 00:44:25,520 job at defining platform engineering. 712 00:44:25,520 --> 00:44:28,680 I think we crammed way too much in there with a focus on 713 00:44:28,680 --> 00:44:31,840 engineering. So maybe it's time for another 714 00:44:31,840 --> 00:44:34,200 iteration on on on that approach. 715 00:44:35,120 --> 00:44:37,400 But there's so much in there and we need that. 716 00:44:37,400 --> 00:44:40,440 We need, we need that for the agents that help your customer. 717 00:44:40,600 --> 00:44:43,120 We need that for the agents that help us write codes. 718 00:44:43,360 --> 00:44:47,000 We need that for the agents that deploy our stuff into production 719 00:44:47,240 --> 00:44:49,880 and manage that and observe that for us. 720 00:44:50,320 --> 00:44:56,160 We, we need all of that. People that are listening or 721 00:44:56,160 --> 00:45:01,000 watching might think, hey, yeah, but shouldn't it always have 722 00:45:01,000 --> 00:45:02,640 been that way? Yes. 723 00:45:03,080 --> 00:45:10,240 But when you slap a high onto some project like that, chances 724 00:45:10,240 --> 00:45:13,400 that you get the budget for it and you get the resources for it 725 00:45:13,400 --> 00:45:14,920 are a bit higher as well. Yeah. 726 00:45:15,120 --> 00:45:18,360 So do yourself a favour, implement platform engineering, 727 00:45:18,360 --> 00:45:22,760 make sure that you have the best culture look, be creative with 728 00:45:22,760 --> 00:45:28,200 with your collaboration model, your operating model experiments 729 00:45:28,200 --> 00:45:30,520 a bit with that. Make sure that you celebrate 730 00:45:30,520 --> 00:45:35,640 your success, that people are enabled and encouraged to use AI 731 00:45:36,320 --> 00:45:40,600 and to fill. And I think if you do that 732 00:45:40,600 --> 00:45:45,000 right, you'll be solid and you're you're doing probably 733 00:45:45,000 --> 00:45:48,200 better than some other company and you will attract the talent. 734 00:45:48,640 --> 00:45:51,400 Yeah, beautiful. Thank you so much for coming on, 735 00:45:51,400 --> 00:45:52,600 Susanna. This was a blast. 736 00:45:52,960 --> 00:45:54,280 Yeah, thank you. It was so fun. 737 00:45:55,040 --> 00:45:57,160 Cool. If you're still here with us 738 00:45:57,160 --> 00:45:58,880 listening, let me know in the comments section what you 739 00:45:58,880 --> 00:46:01,160 thought of this episode and we'll see you in the next one.