1 00:00:00,040 --> 00:00:03,120 You're not writing enough. Most of engineers that have 2 00:00:03,120 --> 00:00:05,480 something interesting to say do not aspire to write at all. 3 00:00:05,480 --> 00:00:07,920 They're actually perfectly comfortable writing with very 4 00:00:07,920 --> 00:00:11,280 good code, but not very good prose for various reasons. 5 00:00:11,480 --> 00:00:15,720 The sad truth is that if you as a new author write your faults 6 00:00:15,720 --> 00:00:19,160 on something, just nobody cares at all unless you do some kind 7 00:00:19,160 --> 00:00:21,120 of promotion, then nobody's going to ever read it. 8 00:00:21,120 --> 00:00:22,440 So. In your book you also cover 9 00:00:22,440 --> 00:00:24,560 these few things called blog post patterns. 10 00:00:24,560 --> 00:00:27,160 So for some of us who are like beginners in terms of writing 11 00:00:27,320 --> 00:00:29,960 that we could try out. Backhunt is my favorite. 12 00:00:30,080 --> 00:00:32,840 There is a post and that explains how somebody looked for 13 00:00:32,840 --> 00:00:35,120 a bug in the code. It's usually structured like a 14 00:00:35,360 --> 00:00:38,360 detective story a little bit. What is the impact of AI these 15 00:00:38,360 --> 00:00:40,200 days? Could it actually help us in 16 00:00:40,200 --> 00:00:44,240 terms of writing our content? Everyone seen a DPT generated 17 00:00:44,240 --> 00:00:46,240 article? At this moment they are just 18 00:00:46,280 --> 00:00:47,800 pure garbage. They're useless. 19 00:00:48,240 --> 00:00:51,960 Most of them have obvious logical fallacies or great app. 20 00:00:52,080 --> 00:00:55,200 It's something a little different which is using them 21 00:00:55,280 --> 00:00:58,480 as. LLM model or which tool that you 22 00:00:58,480 --> 00:01:00,480 think are good enough for helping us? 23 00:01:00,560 --> 00:01:03,720 To write the easiest answer which I use is Oval. 24 00:01:04,080 --> 00:01:07,120 It's best to just spin up everything that you have access 25 00:01:07,120 --> 00:01:10,320 to and just ask the same question often that there is no 26 00:01:10,320 --> 00:01:12,320 clear winner. How do you actually overcome 27 00:01:12,320 --> 00:01:13,840 this? Kind of like common excuses for 28 00:01:13,840 --> 00:01:16,240 people who do not love writing at all. 29 00:01:16,360 --> 00:01:19,480 The best thing to do to overcome all kinds of excuses is to just. 30 00:01:35,990 --> 00:01:38,430 Hello everyone, welcome back to another new episode of the 31 00:01:38,430 --> 00:01:41,430 Technician podcast. Today I have with me the author 32 00:01:41,430 --> 00:01:44,150 of the book titled Writing for Developers. 33 00:01:44,430 --> 00:01:48,150 His name is Piotr Sarna. So I'm sure when you hear about 34 00:01:48,150 --> 00:01:52,630 this topic, a lot of us want to write, but many of us just 35 00:01:52,630 --> 00:01:56,630 couldn't write for some reasons. So hopefully today Sarna is able 36 00:01:56,630 --> 00:02:00,320 to help us, guide us how we can start it, how we can get started 37 00:02:00,320 --> 00:02:02,720 to write. And welcome to the show, Sarna. 38 00:02:03,280 --> 00:02:05,400 Thanks for inviting me and hello everyone. 39 00:02:06,560 --> 00:02:09,639 Sarna, I always love to invite my guest first to share about 40 00:02:09,639 --> 00:02:12,800 maybe yourself by sharing any career turning points that you 41 00:02:12,800 --> 00:02:15,520 think we all can learn from you. Sure. 42 00:02:15,680 --> 00:02:19,400 The first career turning point was definitely me joining a 43 00:02:19,640 --> 00:02:22,600 company that works on an open source product, which was a 44 00:02:22,600 --> 00:02:29,000 distributed file system, and I stayed in open source for years 45 00:02:29,120 --> 00:02:31,040 later. The next turning point was Cila 46 00:02:31,040 --> 00:02:35,160 DB, which was super high quality open source product. 47 00:02:35,520 --> 00:02:37,800 And that's where I learned really a lot. 48 00:02:38,560 --> 00:02:42,840 But my latest turning point was joining the AI hype, which is 49 00:02:43,200 --> 00:02:46,600 hopefully not just not just a hype, but actually a new era, 50 00:02:47,440 --> 00:02:50,960 which is from a year ago. So these are, yeah, I've 51 00:02:51,000 --> 00:02:53,600 actually have at least three of those points, but those are the 52 00:02:53,840 --> 00:02:57,160 ones that stand out. Yeah, I saw from your profile 53 00:02:57,160 --> 00:02:59,160 you are also active in open source, right? 54 00:02:59,160 --> 00:03:03,200 You are active as a maintainer in the LIP sequel and Cela DB, 55 00:03:03,200 --> 00:03:05,600 right? So tell us, how do you actually 56 00:03:05,600 --> 00:03:08,240 juggle your full time job? Maybe with your open source 57 00:03:08,240 --> 00:03:12,200 maintainer contributions? Oh, that's very easy because 58 00:03:12,680 --> 00:03:16,400 actually both of those are more of honorary titles right now. 59 00:03:16,400 --> 00:03:19,640 I actually don't spend too much time on those anymore. 60 00:03:19,800 --> 00:03:23,200 I do miss them sometimes, but there's only there's so much 61 00:03:23,200 --> 00:03:26,280 time I can spend on them. I do sometimes work on my let's 62 00:03:27,080 --> 00:03:31,520 say weekend projects that are also open source, but mostly I 63 00:03:31,520 --> 00:03:33,400 I'm focused on my full time job at the moment. 64 00:03:34,520 --> 00:03:37,280 Right. And maybe a little bit of the AI 65 00:03:37,280 --> 00:03:40,160 have that you mentioned just sharing what are you up to these 66 00:03:40,160 --> 00:03:41,840 days, right. So what kind of problems are you 67 00:03:41,840 --> 00:03:46,160 solving? So last year, last year, in a 68 00:03:46,160 --> 00:03:49,480 month, I guess, I joined Poolside, which is one of the 69 00:03:49,800 --> 00:03:54,640 companies that works on those largest large language models. 70 00:03:55,840 --> 00:03:59,360 I'm not sure how much I can spell before we release a 71 00:03:59,360 --> 00:04:02,480 product because it's under, you know, 19 layers of Ndas, but 72 00:04:02,480 --> 00:04:07,400 they are working provide well, hopefully state-of-the-art best 73 00:04:07,400 --> 00:04:10,240 models for programming, for helping people to program. 74 00:04:10,240 --> 00:04:13,160 So think per programmer that is really smart. 75 00:04:14,080 --> 00:04:16,560 All right, So fingers crossed that we can see some new cool 76 00:04:16,560 --> 00:04:20,079 things. I feel that every few days or 77 00:04:20,079 --> 00:04:23,440 even not months, right, we will see a lot more new things 78 00:04:23,440 --> 00:04:25,440 happening in the AI, especially with coding, right? 79 00:04:25,680 --> 00:04:28,320 So maybe one day we'll see poolside as well as one of the 80 00:04:28,320 --> 00:04:29,760 solutions. Oh, definitely, yeah. 81 00:04:30,680 --> 00:04:32,800 Yeah. So let's dive in into our topic 82 00:04:32,800 --> 00:04:34,680 for today, right? Writing for developers. 83 00:04:34,760 --> 00:04:37,400 I'm sure many of developers like I mentioned, right? 84 00:04:37,400 --> 00:04:40,240 We all aspire to start writing something, right? 85 00:04:40,240 --> 00:04:43,120 It could be blog post, newsletter, it could be even 86 00:04:43,120 --> 00:04:46,800 like other types of contents. But somehow we've felt that we 87 00:04:46,800 --> 00:04:49,320 are not good enough. Probably our writing is not up 88 00:04:49,320 --> 00:04:51,240 to the mark or things like that, right? 89 00:04:51,440 --> 00:04:54,400 Or we are technical, we don't know how to write better. 90 00:04:54,760 --> 00:04:56,560 So maybe let's start from this, right? 91 00:04:57,120 --> 00:04:59,840 How do you actually see this kind of problem happening? 92 00:04:59,840 --> 00:05:03,760 Maybe outside of your, you know, like your work, right? 93 00:05:03,760 --> 00:05:06,760 Like in the developers in general, how do you see 94 00:05:06,920 --> 00:05:10,040 developers having this challenge of writing I? 95 00:05:10,800 --> 00:05:15,320 Would go even one step further to the negative end and say that 96 00:05:15,320 --> 00:05:18,760 most of engineers that have something interesting to say do 97 00:05:18,760 --> 00:05:22,400 not aspire to write at all. That's the that's the first bit 98 00:05:22,400 --> 00:05:25,000 that they are actually perfectly comfortable writing. 99 00:05:25,000 --> 00:05:29,120 Very good code, but not very good prose or any kind of prose 100 00:05:29,480 --> 00:05:32,200 for various reasons. Everyone has their own. 101 00:05:32,440 --> 00:05:37,080 But the first thing that would help both sides, both engineers 102 00:05:37,080 --> 00:05:40,840 that would aspire to write and those that actually don't aspire 103 00:05:40,840 --> 00:05:43,160 to write but should write because that would be education 104 00:05:43,160 --> 00:05:47,240 of it would be to have writing culture somehow included in the 105 00:05:47,240 --> 00:05:51,360 company they work at. That that's what bootstrapped me 106 00:05:51,720 --> 00:05:53,880 and quite a few of my colleagues as well. 107 00:05:54,840 --> 00:05:57,040 So that would be, yeah, that would be a very good starting 108 00:05:57,040 --> 00:06:00,120 point because most engineers are not alone. 109 00:06:00,120 --> 00:06:02,840 They work at some kind of company, be it startup or the 110 00:06:02,920 --> 00:06:06,400 corporation, but any kind of company would it could have a 111 00:06:06,400 --> 00:06:09,240 right in culture. I find it interesting when you 112 00:06:09,240 --> 00:06:11,480 mentioned many developers actually don't aspire to write. 113 00:06:11,640 --> 00:06:13,200 I guess in a sense that makes sense. 114 00:06:13,200 --> 00:06:16,560 Also for some developers, they just are happy to write code, 115 00:06:16,560 --> 00:06:19,320 write all day long. So you mentioned about writing 116 00:06:19,320 --> 00:06:21,800 culture. I think for some people they may 117 00:06:21,800 --> 00:06:25,160 be unfortunate right enough to actually work in companies that 118 00:06:25,160 --> 00:06:27,800 are all about, you know, actions, just delivering code 119 00:06:27,800 --> 00:06:30,880 features and all that and not spending time to actually write. 120 00:06:31,160 --> 00:06:34,480 So tell us what the importance of this writing culture in such 121 00:06:34,480 --> 00:06:37,480 a way that actually it influences you to want to write? 122 00:06:38,040 --> 00:06:43,040 So I think most of the companies would actually benefit from 123 00:06:43,640 --> 00:06:45,640 introducing some kind of writing culture. 124 00:06:45,640 --> 00:06:49,200 Like it's enough to just take a look at some of the top 125 00:06:49,200 --> 00:06:52,000 companies in the world that also have top engineering blogs. 126 00:06:52,000 --> 00:06:55,720 Like when Netflix blog puts something out, it's usually very 127 00:06:55,720 --> 00:06:59,720 interesting and it I think it's still maybe doesn't help them 128 00:06:59,720 --> 00:07:04,920 sell more subscriptions to people who watch TV series, but 129 00:07:04,920 --> 00:07:09,080 it definitely helps gather best talent because engineers write 130 00:07:09,080 --> 00:07:10,960 and read those engineering blogs. 131 00:07:11,640 --> 00:07:15,680 So they could be lured to apply to Netflix if they see an 132 00:07:15,680 --> 00:07:17,760 interesting problem, for instance. 133 00:07:17,760 --> 00:07:22,320 So from company's side, I think it's beneficial even just for 134 00:07:22,560 --> 00:07:25,360 hiring purposes to have an engineering block. 135 00:07:25,640 --> 00:07:29,840 But then it's also a knowledge base, the best kind because it's 136 00:07:29,840 --> 00:07:32,920 usually made of actual real life experience. 137 00:07:33,880 --> 00:07:38,440 And from the developer's side, well, there's so many advantages 138 00:07:38,560 --> 00:07:41,840 for you as an outer. One of the advantages is that 139 00:07:41,840 --> 00:07:46,360 you need to kind of understand what you implemented to write 140 00:07:46,360 --> 00:07:49,680 about it. So it's clarifies what what you 141 00:07:49,680 --> 00:07:51,240 create. Sometimes you can even find a 142 00:07:51,240 --> 00:07:54,400 bug after trying to explain something and figure out that 143 00:07:54,400 --> 00:07:55,720 there's a corner place that you missed. 144 00:07:56,560 --> 00:08:01,720 Another thing would be that perhaps you want to work on your 145 00:08:02,040 --> 00:08:05,200 personal brand even as a developer isn't interested in 146 00:08:05,200 --> 00:08:08,360 becoming famous. It's not exactly about being 147 00:08:08,360 --> 00:08:11,160 famous or not. It's also about meeting other 148 00:08:11,160 --> 00:08:14,720 people from same niche, from the same industry. 149 00:08:15,360 --> 00:08:19,880 Then you can go one step further above blogging and also take 150 00:08:19,880 --> 00:08:24,280 part in conferences, some kind of online meetups, on site 151 00:08:24,280 --> 00:08:27,320 meetups, all kinds of things. So yeah, they're endless. 152 00:08:28,200 --> 00:08:32,360 Or do you need to see if you have to start writing and get 153 00:08:32,600 --> 00:08:35,799 some kind of recognition, because some people are bound to 154 00:08:35,799 --> 00:08:38,720 read it? Right, So maybe can you also 155 00:08:38,720 --> 00:08:41,000 share your personal experience, right? 156 00:08:41,000 --> 00:08:44,000 Because I, I saw that you come from a technical background, 157 00:08:44,000 --> 00:08:45,680 right? And all this super technical 158 00:08:45,680 --> 00:08:48,960 coding experience up to you writing this book, right? 159 00:08:48,960 --> 00:08:52,000 Because writing this book itself is like 1 implementation of 160 00:08:52,000 --> 00:08:54,920 writing for developers, right? So tell us your personal 161 00:08:54,920 --> 00:08:56,480 journey. How do you end up writing this 162 00:08:56,480 --> 00:09:00,040 book? So this particular book is a 163 00:09:00,040 --> 00:09:04,400 continuation of previous book that I also co-authored with 164 00:09:04,960 --> 00:09:08,160 three other people. This time it all started at 165 00:09:08,160 --> 00:09:12,360 Cilla DB, which has a very strong writing culture, blogging 166 00:09:12,360 --> 00:09:15,080 culture. It's strong to the point that 167 00:09:15,080 --> 00:09:19,280 you're pretty much forced to write about something from time 168 00:09:19,280 --> 00:09:22,440 to time, especially if you delivered something interesting 169 00:09:22,480 --> 00:09:25,200 technologically. It's very natural after you 170 00:09:25,200 --> 00:09:29,080 spend some time working at Cilla DB that you even feel that after 171 00:09:29,080 --> 00:09:31,360 you release something interesting, there's going to be 172 00:09:31,360 --> 00:09:33,880 a request for writing A blog post about it. 173 00:09:33,880 --> 00:09:37,760 So you might just as well start even before that, or volunteer. 174 00:09:38,600 --> 00:09:42,400 I think first five or seven or eight blog posts of mine were 175 00:09:42,400 --> 00:09:45,200 released simply as part of this blogging culture. 176 00:09:45,960 --> 00:09:49,040 Then Celadibi also organized the conference. 177 00:09:49,400 --> 00:09:51,400 It was bugged, then called Celadibi Summit. 178 00:09:51,400 --> 00:09:54,920 Now it's diverged into a few online conferences. 179 00:09:54,920 --> 00:09:57,680 Then it was on an Onset conference in San Francisco. 180 00:09:57,760 --> 00:10:02,480 I also spoke about one of these topics from my blog posts. 181 00:10:03,080 --> 00:10:08,320 So that was another thing and and another pretty much result 182 00:10:08,320 --> 00:10:10,240 of me writing blog posts as well. 183 00:10:11,120 --> 00:10:16,080 And then after a few years, I got approached by a publisher, I 184 00:10:16,080 --> 00:10:18,800 think after releasing one of the blog posts of code mildly 185 00:10:18,800 --> 00:10:21,400 popular. So it was also directly, it was 186 00:10:21,400 --> 00:10:26,160 by the maybe blogging and he asked me for if I'm interested 187 00:10:26,160 --> 00:10:30,560 in writing a book about Cila. DB actually wasn't interested in 188 00:10:30,840 --> 00:10:32,520 writing the book. I got intrigued. 189 00:10:32,520 --> 00:10:36,200 So I rerouted him to our marketing and content 190 00:10:36,200 --> 00:10:40,040 department. And from there, Cynthia Michael 191 00:10:40,040 --> 00:10:45,440 outer of this latest book got an idea that we could write a more 192 00:10:45,560 --> 00:10:48,640 generic book about database performance based on CLDB. 193 00:10:49,440 --> 00:10:51,280 What happened? So we released this book. 194 00:10:51,280 --> 00:10:54,360 It's an open book, so we can just go ahead and download it. 195 00:10:54,360 --> 00:10:58,120 It's like an open source book that you can also buy in paper 196 00:10:58,240 --> 00:11:00,520 if you like, or you can just read it online. 197 00:11:01,600 --> 00:11:05,400 And after we released that, it's kind of a slippery slope. 198 00:11:05,400 --> 00:11:09,000 You just want to write more. So when Cynthia suggested that 199 00:11:09,000 --> 00:11:12,520 maybe you write another book and this whole structure of the 200 00:11:12,520 --> 00:11:14,960 latest book was her idea, I just couldn't resist. 201 00:11:14,960 --> 00:11:18,760 So I so I ended up working on it and then we released it last 202 00:11:18,760 --> 00:11:21,560 year. Well, very interesting, right? 203 00:11:21,800 --> 00:11:24,720 I'm actually intrigued when you mentioned Sila DB actually kind 204 00:11:24,720 --> 00:11:26,760 of like forced everyone to write, right? 205 00:11:26,840 --> 00:11:29,080 I understand, you know, just that you mentioned about the 206 00:11:29,680 --> 00:11:33,280 benefits of, you know, writing for the external thing, right? 207 00:11:33,280 --> 00:11:36,480 Maybe for hiring, for the branding, right, for telling 208 00:11:36,480 --> 00:11:38,280 more about the companies and all that aspect. 209 00:11:38,520 --> 00:11:40,960 But do you see also having writing culture actually 210 00:11:40,960 --> 00:11:43,960 impacting engineering culture, maybe in terms of quality of 211 00:11:43,960 --> 00:11:47,600 products, in terms of, you know, best practices or other types of 212 00:11:47,600 --> 00:11:49,760 cultural thing? Is there any relation that you 213 00:11:49,760 --> 00:11:51,720 see when you were working at CILA DP? 214 00:11:52,320 --> 00:11:55,360 Yes, from many angles. Some of it is just actually 215 00:11:55,360 --> 00:11:57,800 finding potential bugs and corner cases. 216 00:11:57,800 --> 00:12:00,440 When you try to explain something for examples and you 217 00:12:00,440 --> 00:12:03,080 see that you obviously missed something, it might come up in 218 00:12:03,080 --> 00:12:05,520 review or you might just find it up yourself. 219 00:12:05,520 --> 00:12:09,320 Kind of like rubber debugging, but on text. 220 00:12:10,480 --> 00:12:14,840 Another thing is that this is a knowledge base and we often just 221 00:12:14,840 --> 00:12:17,720 read that people read A blog post about some feature instead 222 00:12:17,720 --> 00:12:21,240 of trying to find them specifically to code, to 223 00:12:21,280 --> 00:12:23,120 bootstrap themselves some area of code. 224 00:12:23,120 --> 00:12:25,760 Because if it's explained in English, then it's easier to 225 00:12:25,920 --> 00:12:29,320 comprehend potentially. And then there's a whole class 226 00:12:29,360 --> 00:12:33,760 of blog posts about benchmarking and test results. 227 00:12:34,120 --> 00:12:35,680 It's one of our categories in the book. 228 00:12:36,000 --> 00:12:40,560 But the important thing about that is the very benchmarks or 229 00:12:40,560 --> 00:12:42,200 anything that you want to publish. 230 00:12:42,200 --> 00:12:43,880 It's a tremendous amount of work. 231 00:12:43,880 --> 00:12:48,480 So it's it actually makes your product better directly because 232 00:12:48,480 --> 00:12:49,800 you need to write all those benchmarks. 233 00:12:49,800 --> 00:12:53,160 So you know, have another tool to benchmark your own code. 234 00:12:53,440 --> 00:12:56,840 And with benchmarks, the situation is very tough in 235 00:12:56,840 --> 00:12:58,760 blogging. It's a hard category because 236 00:12:58,920 --> 00:13:02,560 you're going to get accused of biasing the benchmarks in your 237 00:13:02,560 --> 00:13:04,560 favour. It will always happen, but you 238 00:13:04,560 --> 00:13:08,000 can do as much as possible to alleviate that, and that means 239 00:13:08,000 --> 00:13:12,560 improving your code so that's a more direct effect that blog 240 00:13:12,560 --> 00:13:15,520 posts have on products. Yeah. 241 00:13:15,520 --> 00:13:18,280 So I can understand that when you write a lot of things, 242 00:13:18,280 --> 00:13:20,800 right, the knowledge base aspect is actually really important, 243 00:13:20,800 --> 00:13:22,800 right, in engineering, especially as you grow larger, 244 00:13:22,800 --> 00:13:24,720 right? So obviously people want to know 245 00:13:24,720 --> 00:13:26,720 certain stuff. You can just direct them to the 246 00:13:26,840 --> 00:13:30,920 writings and also things like very technical benchmark or 247 00:13:30,920 --> 00:13:33,320 maybe some kind of design docs and things like that definitely 248 00:13:33,320 --> 00:13:36,600 can also improve your own understanding and other people's 249 00:13:36,600 --> 00:13:38,640 understanding as well. So I think thanks for 250 00:13:38,640 --> 00:13:40,640 highlighting that. And obviously, you mentioned 251 00:13:40,880 --> 00:13:43,160 writing itself will take a lot of effort, right? 252 00:13:43,400 --> 00:13:46,840 So some people think, you know, apart from me doing the 253 00:13:46,840 --> 00:13:49,240 implementation, which is by itself is already a lot of 254 00:13:49,240 --> 00:13:52,120 amount of work, I still need to spend a lot of effort to do the 255 00:13:52,120 --> 00:13:54,120 right thing. So obviously, this is one of the 256 00:13:54,120 --> 00:13:57,400 common blockers or common excuse for people not to write. 257 00:13:57,440 --> 00:14:00,200 The other thing is like, yeah, I don't have any other time. 258 00:14:00,440 --> 00:14:03,320 I think I'm not a good writer. So tell us, how do you actually 259 00:14:03,480 --> 00:14:05,800 overcome this? Kind of like common excuses for 260 00:14:05,800 --> 00:14:08,720 people who do not love writing at all. 261 00:14:09,480 --> 00:14:12,880 First of all, if some of the genuinely doesn't like writing, 262 00:14:12,880 --> 00:14:16,160 it's better if they just don't. I'm not going to push so hard 263 00:14:16,160 --> 00:14:18,280 that everyone should write at all cost. 264 00:14:18,800 --> 00:14:23,600 But often those excuses are either being subconsciously lazy 265 00:14:23,600 --> 00:14:29,280 or just scared of the unknown, which kind of goes away after 266 00:14:29,280 --> 00:14:31,560 the first blog post. So the best thing to do to 267 00:14:31,560 --> 00:14:35,720 overcome all kinds of excuses is to just release a single blog 268 00:14:35,720 --> 00:14:38,960 post and see if it's if it really was that bad for all the 269 00:14:38,960 --> 00:14:42,240 reasons of it, or if all those excuses were actually void and 270 00:14:42,360 --> 00:14:45,360 nonsensical. This is what kind of happened to 271 00:14:45,360 --> 00:14:47,520 me. For my first blog post, I didn't 272 00:14:47,520 --> 00:14:51,960 really want to write it. I had very hardcore variant of a 273 00:14:51,960 --> 00:14:54,960 stage fright where I was actually afraid of somebody 274 00:14:54,960 --> 00:14:57,760 reading my piece that I wrote offline. 275 00:14:58,440 --> 00:15:01,640 Which sounds ridiculous but this is just how my brain works 276 00:15:01,640 --> 00:15:04,400 apparently. And only after releasing this 277 00:15:04,400 --> 00:15:07,880 first blog post I realised that it wasn't really but it was just 278 00:15:07,880 --> 00:15:11,040 writing text instead of code and publishing it. 279 00:15:11,200 --> 00:15:13,920 And working in open source is actually very similar because 280 00:15:13,920 --> 00:15:16,840 people also read your code so why not your text? 281 00:15:17,360 --> 00:15:21,560 Second excuse that I didn't have to personally go through because 282 00:15:21,560 --> 00:15:24,600 I don't care that much, but I remember my colleagues having 283 00:15:24,600 --> 00:15:27,800 issues with that was if they aren't English native speakers, 284 00:15:28,240 --> 00:15:31,480 they wondered if grammar mistakes wouldn't be an issue 285 00:15:31,480 --> 00:15:34,920 and they would be ashamed they put something not perfect in 286 00:15:34,920 --> 00:15:37,280 there. There are at least two ways to 287 00:15:37,280 --> 00:15:41,600 define this excuse 1 is that usually some of your company 288 00:15:41,800 --> 00:15:44,200 folks are native English speakers and they could just 289 00:15:44,200 --> 00:15:46,400 review everything. Of course, you can also use 290 00:15:46,400 --> 00:15:49,000 automated tools to do that. And the second one is that a 291 00:15:49,280 --> 00:15:52,720 good rule of thumb is that if online comments about your 292 00:15:52,720 --> 00:15:55,360 article only some grammar mistakes, then it means that the 293 00:15:55,360 --> 00:15:58,840 technical purpose flawless. So congratulations, it's like a 294 00:15:58,920 --> 00:16:02,720 good always love to get only this kind of feedback and things 295 00:16:02,720 --> 00:16:05,320 that they write. Really interesting perspective, 296 00:16:05,320 --> 00:16:06,880 right? So if people need to think about 297 00:16:06,880 --> 00:16:10,120 your grammar aspect of the writing, so that means you're 298 00:16:10,280 --> 00:16:11,960 kind of like the technical aspect is flawless. 299 00:16:11,960 --> 00:16:14,600 So I like that one. So also you mentioned like very 300 00:16:14,600 --> 00:16:17,120 interesting insight, right? So you mentioned writing your 301 00:16:17,120 --> 00:16:19,960 first blog is similar to like writing your first open source 302 00:16:19,960 --> 00:16:22,120 contribution, right? Think of it like this is your 303 00:16:22,120 --> 00:16:25,280 first experience and so you're just inexperienced about it. 304 00:16:25,520 --> 00:16:29,360 So having other people reading, criticizing your writings, I 305 00:16:29,360 --> 00:16:31,760 think is similar, right? And you'll contribute to open 306 00:16:31,760 --> 00:16:34,440 source. So I think let's say people are 307 00:16:34,440 --> 00:16:36,920 interested, so they want to write their first blog post. 308 00:16:37,000 --> 00:16:39,680 So obviously the first blocker is like what to write? 309 00:16:39,800 --> 00:16:42,600 It seems like these days there are so many plenty of 310 00:16:42,600 --> 00:16:46,360 engineering related, technical related blocks already available 311 00:16:46,360 --> 00:16:48,840 online. So how do you help us to 312 00:16:48,840 --> 00:16:51,680 actually find our first blog post topic? 313 00:16:52,720 --> 00:16:56,280 First thing would be to definitely go over whatever 314 00:16:56,560 --> 00:17:01,160 you've just implemented a while ago and look at it. 315 00:17:01,880 --> 00:17:04,000 Try to look at it from the perspective of somebody who 316 00:17:04,000 --> 00:17:06,760 doesn't know all the internals of the company, for instance. 317 00:17:06,800 --> 00:17:10,920 And if it's really that uninteresting because the 318 00:17:10,920 --> 00:17:14,000 feature might be super obvious to you because you already know 319 00:17:14,000 --> 00:17:17,520 all the code base, you know everything, but it might be very 320 00:17:17,520 --> 00:17:20,040 interesting to somebody from the outside. 321 00:17:20,040 --> 00:17:24,400 For instance, I suspect that let's take this Netflix, I think 322 00:17:24,400 --> 00:17:27,240 of blog, for example. I guess it's not that 323 00:17:27,240 --> 00:17:30,240 interesting of an article for an average Netflix developer who 324 00:17:30,240 --> 00:17:31,880 already knows everything about it. 325 00:17:31,880 --> 00:17:35,400 But for the rest of the world, it's actually useful knowledge. 326 00:17:36,720 --> 00:17:40,040 OK. And another really, really 327 00:17:40,040 --> 00:17:44,680 important aspect is that if you think that it's not worth 328 00:17:44,680 --> 00:17:48,080 writing about because it was already written somewhere else, 329 00:17:48,080 --> 00:17:52,360 then that's just false. There are usually 10's and 330 00:17:52,880 --> 00:17:55,280 hundreds of blog posts about the same topic. 331 00:17:55,280 --> 00:17:59,160 So it's not like it's the first person called dips on this topic 332 00:17:59,160 --> 00:18:02,400 and nobody else can write about it because it's the combination 333 00:18:02,400 --> 00:18:07,040 of a topic and your perspective on it that's potentially useful. 334 00:18:07,040 --> 00:18:09,000 That's that might be new information. 335 00:18:09,000 --> 00:18:11,240 If there's actually no new information and you just don't 336 00:18:11,240 --> 00:18:13,640 have anything interesting to say about something, then don't 337 00:18:13,640 --> 00:18:18,040 write about it naturally. But often it's the personal 338 00:18:18,040 --> 00:18:20,520 experience connected to something, implementing 339 00:18:20,520 --> 00:18:23,360 something in the language, trying something out that that 340 00:18:23,360 --> 00:18:26,360 matters. And that makes, yeah, that 341 00:18:26,360 --> 00:18:30,240 produces this unique information that you can put in a post and 342 00:18:30,680 --> 00:18:32,800 release it and it will be interesting to someone. 343 00:18:33,680 --> 00:18:35,720 Yes, I also learned myself right. 344 00:18:35,720 --> 00:18:37,960 So even though you think there are plenty of, you know, 345 00:18:37,960 --> 00:18:40,640 coverage already on a certain particular topic, right, there 346 00:18:40,640 --> 00:18:43,680 are always people behind us who are like just starting maybe in 347 00:18:43,680 --> 00:18:45,960 their career or starting to use that technology. 348 00:18:46,240 --> 00:18:48,360 So there are many kind of a spectrum like it could be 349 00:18:48,360 --> 00:18:52,040 beginners content, you know, expert content, like more more 350 00:18:52,040 --> 00:18:54,640 expert content. So definitely there are plenty 351 00:18:54,640 --> 00:18:57,280 of people who could find your writings very useful. 352 00:18:57,520 --> 00:19:00,160 And don't forget, maybe I think it's also for yourself in the 353 00:19:00,160 --> 00:19:02,520 future, right? Maybe what we write today, 354 00:19:02,520 --> 00:19:05,040 right, We may not even remember in the future, right? 355 00:19:05,040 --> 00:19:07,760 So I think it's also a good thing to actually kind of like 356 00:19:07,760 --> 00:19:11,040 remind us what we actually did and what of learnings we got 357 00:19:11,040 --> 00:19:13,960 from there. So I think also you mentioned 358 00:19:14,200 --> 00:19:18,080 past projects that we think we find it maybe like kind of like 359 00:19:18,080 --> 00:19:20,880 normal common things. But sometimes, you know, if you 360 00:19:20,880 --> 00:19:24,320 ask from other people, they think this is super useful and 361 00:19:24,320 --> 00:19:27,160 you may want to write something about it so that other people 362 00:19:27,400 --> 00:19:30,800 can learn from you, right? So you want to say something? 363 00:19:31,240 --> 00:19:33,120 No, I just, I just, I read that definitely. 364 00:19:33,120 --> 00:19:37,960 And the one more thing I could add is that technical blog posts 365 00:19:37,960 --> 00:19:42,040 out of all category blog posts go out of date extremely fast. 366 00:19:42,400 --> 00:19:46,720 It's both dying hyperlinks and the code snippers that don't 367 00:19:46,720 --> 00:19:49,560 compile anymore. So one thing, one aspect of it 368 00:19:49,560 --> 00:19:52,720 is to make sure that your blog posts are up to date from time 369 00:19:52,720 --> 00:19:54,480 to time, just create the old info. 370 00:19:54,480 --> 00:19:57,880 But then that also means that if somebody already wrote about 371 00:19:57,880 --> 00:20:01,520 something, this old blog post might be unusable by this. 372 00:20:01,760 --> 00:20:04,440 So sometimes it's better to just release something that's. 373 00:20:04,960 --> 00:20:06,760 Currently more up to date because it was released 374 00:20:06,760 --> 00:20:09,440 yesterday. Yeah, that's a very good point. 375 00:20:09,440 --> 00:20:12,440 I could remember, you know, let's say if I find a particular 376 00:20:12,440 --> 00:20:15,520 blog post, but it was written in an older version of the library, 377 00:20:15,600 --> 00:20:18,160 sometimes he just couldn't compile all the SDK change 378 00:20:18,160 --> 00:20:20,120 dramatically. So I think that's a very good 379 00:20:20,120 --> 00:20:22,440 pointer as well. So yeah, just don't be afraid to 380 00:20:22,440 --> 00:20:25,000 start writing, you know, whatever topics that you find 381 00:20:25,000 --> 00:20:27,240 interesting and important thing is that you can actually 382 00:20:27,240 --> 00:20:30,600 contribute something, you know, maybe a more authentic kind of 383 00:20:30,600 --> 00:20:34,600 like a thoughts from you. So writing itself, many people 384 00:20:34,600 --> 00:20:38,320 think, you know, it's super easy for those that come natural for 385 00:20:38,320 --> 00:20:40,440 them, right? But for many people, they just 386 00:20:40,440 --> 00:20:43,160 don't know where to start, you know, is it the intro? 387 00:20:43,160 --> 00:20:46,080 Is it the, you know, the skeleton table of contents, 388 00:20:46,080 --> 00:20:48,680 whatever that is? Do you have any tips for us how 389 00:20:48,680 --> 00:20:51,280 to start writing? Maybe even like the first draft, 390 00:20:51,280 --> 00:20:52,680 right? The first raw content. 391 00:20:53,360 --> 00:20:56,480 There are multiple approaches and you need to just find 392 00:20:56,480 --> 00:20:58,400 something that works for you personally. 393 00:20:58,480 --> 00:21:03,320 Synface is an expert in forcing people to outline things and 394 00:21:03,560 --> 00:21:07,640 kick start their writing process. 2 approaches that are 395 00:21:07,680 --> 00:21:12,280 good to try out are as follows. The first one is just start with 396 00:21:12,880 --> 00:21:17,760 writing the headers of the paragraphs that aren't there 397 00:21:17,760 --> 00:21:19,120 yet. Like what would you like to 398 00:21:19,320 --> 00:21:22,120 have? What are your main points of the 399 00:21:22,120 --> 00:21:23,720 post? And then just fill them one by 400 00:21:23,720 --> 00:21:27,520 one, not necessarily in order, and then just reread everything 401 00:21:27,520 --> 00:21:31,240 and see if it makes sense. But then another opposite 402 00:21:31,240 --> 00:21:34,080 approach that may work for some people who are more chaotic 403 00:21:34,080 --> 00:21:36,000 perhaps. I often start like that is to 404 00:21:36,520 --> 00:21:40,600 dump everything you have in super unstructured form and then 405 00:21:41,160 --> 00:21:44,880 fish out the interesting bits and put them into paragraphs and 406 00:21:44,880 --> 00:21:47,080 headers and so on. These structures with headers 407 00:21:47,080 --> 00:21:49,960 and paragraphs is important. Nonetheless, you should end up 408 00:21:49,960 --> 00:21:55,200 doing this for various reasons. One of them is that people will 409 00:21:55,360 --> 00:21:58,280 readers will usually start by scrolling through the article 410 00:21:58,280 --> 00:22:01,920 and parsing headers in half a second before they decide if 411 00:22:01,920 --> 00:22:05,000 they read everything or not. So you should end up with 412 00:22:05,280 --> 00:22:08,480 something structured. The way you end up with 413 00:22:08,480 --> 00:22:11,280 something structured depends how your brain works. 414 00:22:11,280 --> 00:22:14,000 So might be a matter to start with the structure itself. 415 00:22:14,000 --> 00:22:17,440 Might be better to just write to unstructured pages and then 416 00:22:18,240 --> 00:22:20,160 figure them out. Yeah. 417 00:22:20,160 --> 00:22:25,120 So I think The thing is like there's no 100% tried and tested 418 00:22:25,120 --> 00:22:27,640 methodology, right? I think just start whatever 419 00:22:27,640 --> 00:22:29,160 feels natural for us, I guess, right? 420 00:22:29,160 --> 00:22:30,920 It could start with the skeleton and all that. 421 00:22:31,280 --> 00:22:33,640 But I think nevertheless, in your book, you emphasize these 422 00:22:33,640 --> 00:22:37,440 two aspects of writing, the actual draft of their writing, 423 00:22:37,440 --> 00:22:39,640 right? Which is to emphasize on the 424 00:22:39,640 --> 00:22:43,320 goal, like finding out what is the goal of the writing, right? 425 00:22:43,760 --> 00:22:46,400 And then the second aspect is, why do you think your 426 00:22:46,600 --> 00:22:49,880 perspective is actually interesting for this particular 427 00:22:50,040 --> 00:22:51,400 blog post, right? Or writing? 428 00:22:51,720 --> 00:22:54,680 So tell us why these two aspects are really important when you 429 00:22:54,680 --> 00:23:00,120 come up with your writing. So As for getting the draft, 430 00:23:00,520 --> 00:23:04,760 yeah, I mean creating the draft and taking the goal, yeah, it 431 00:23:04,800 --> 00:23:07,440 depends very much on what your goal is. 432 00:23:07,440 --> 00:23:11,320 So if you know what the goal is, then you can pick a process 433 00:23:11,320 --> 00:23:14,080 would work best for you. I mean, it's good if you 434 00:23:14,080 --> 00:23:18,960 consider who is going to who in ideal world would be the reader 435 00:23:19,040 --> 00:23:22,000 of your post, because it might be that you just want to write 436 00:23:22,000 --> 00:23:26,400 it for yourself or for future self, like kind of a personal 437 00:23:26,400 --> 00:23:29,160 diary. Maybe you want the wrote public 438 00:23:29,160 --> 00:23:31,240 to know about your specific approaches. 439 00:23:31,240 --> 00:23:35,120 Maybe you you're targeting a special specific niche for 440 00:23:35,120 --> 00:23:38,440 whatever reason, maybe to advertise some cool project of 441 00:23:38,440 --> 00:23:42,080 yours and maybe to get some feedback on something. 442 00:23:43,080 --> 00:23:46,640 So that really varies between posts. 443 00:23:46,640 --> 00:23:50,000 Some of them are very targeted and some of them are just your 444 00:23:50,000 --> 00:23:54,520 thoughts that you would like to publish whatever reason. 445 00:23:55,240 --> 00:23:59,840 One very specific goal that lots, lots of writers do have in 446 00:23:59,840 --> 00:24:03,320 mind one way or another is optimized specifically for front 447 00:24:03,320 --> 00:24:06,760 page copper news. So this is something you should 448 00:24:06,880 --> 00:24:11,640 also consider whether you do want it to land there or maybe 449 00:24:11,720 --> 00:24:14,840 contrary, you just really don't care if it lands there or not 450 00:24:15,480 --> 00:24:16,800 because it's not that kind of a post. 451 00:24:16,800 --> 00:24:21,560 And once you figure that out and then you can know how to 452 00:24:22,360 --> 00:24:26,800 structure your blog post to fit. Yeah, so I think for some people 453 00:24:26,800 --> 00:24:30,560 they love, you know, ended up in the Hacker News or maybe some 454 00:24:30,560 --> 00:24:32,560 Reddit suppose or whatever that is. 455 00:24:32,840 --> 00:24:35,400 So I think knowing the goal I think is really important, 456 00:24:35,400 --> 00:24:38,160 right, Because that's kind of like narrow down the kind of 457 00:24:38,160 --> 00:24:41,040 like the aspects of your writing or be it the topics, big the 458 00:24:41,040 --> 00:24:44,040 structure, big the content. And also you mentioned about 459 00:24:44,040 --> 00:24:47,480 understanding the kind of like persona, the readers, right, who 460 00:24:47,480 --> 00:24:50,400 are going to read, let's say you're targeting for beginner, 461 00:24:50,880 --> 00:24:53,960 don't kind of like mix beginner aspect and advance all in one 462 00:24:54,040 --> 00:24:57,120 blog post, I guess. So I think talking about writing 463 00:24:57,120 --> 00:25:01,600 these days, it's inevitable people will use AI in some form 464 00:25:01,600 --> 00:25:05,000 of shape to actually help them writing or even including 465 00:25:05,040 --> 00:25:07,280 writing most of most parts of the writing itself. 466 00:25:07,480 --> 00:25:10,320 So maybe tell us from your point of view, what is the impact of 467 00:25:10,360 --> 00:25:13,360 AI these days? Could it actually help us in 468 00:25:13,360 --> 00:25:16,840 terms of writing our content? Right. 469 00:25:16,840 --> 00:25:21,960 So the impact measured up to this second, in my personal 470 00:25:21,960 --> 00:25:25,840 opinion, is unfortunately quite negative because the direct 471 00:25:25,840 --> 00:25:29,840 impact that I see is absolute influx of garbage articles that 472 00:25:29,840 --> 00:25:34,160 are clearly AI written and you can subconsciously see that they 473 00:25:34,160 --> 00:25:37,880 were just generated without even reading through them. 474 00:25:38,200 --> 00:25:43,120 Most of them have obvious both logical fallacies and sometimes 475 00:25:43,920 --> 00:25:47,480 we're not even covering the topic that it was supposed to 476 00:25:47,480 --> 00:25:51,080 cover. Yeah, everyone's seen GPT 477 00:25:51,080 --> 00:25:54,800 generated article. And at this moment, yeah, they 478 00:25:54,800 --> 00:25:56,880 are just pure garbage. They're useless. 479 00:25:57,320 --> 00:26:00,680 I haven't seen any kind of content that's generated that's 480 00:26:00,680 --> 00:26:02,600 actually high quality that I would have to write. 481 00:26:02,640 --> 00:26:06,000 Maybe we'll end up there one day. 482 00:26:06,120 --> 00:26:11,360 Right now it definitely isn't the case, but what models today 483 00:26:11,520 --> 00:26:13,000 and in the future probably even more. 484 00:26:13,000 --> 00:26:17,880 A really great app is something a little different, which is 485 00:26:17,920 --> 00:26:21,360 using them as reviewers more or less. 486 00:26:22,200 --> 00:26:27,320 And there are also many ways that you can can use AI to help 487 00:26:27,320 --> 00:26:32,360 you make a better blog post from very simple use cases that are 488 00:26:32,520 --> 00:26:34,640 pretty much glorified Google search based. 489 00:26:34,640 --> 00:26:37,600 For instance, synonyms. If you're stuck and you need a 490 00:26:37,600 --> 00:26:40,160 synonym so that you don't repeat yourself too much, but you can't 491 00:26:40,160 --> 00:26:42,840 think of a good one. The limbs are great at that 492 00:26:42,840 --> 00:26:45,280 because they can just suggest how to rephrase a single 493 00:26:45,280 --> 00:26:48,200 sentence so that so that it's you know that's something stuff. 494 00:26:48,200 --> 00:26:50,760 You don't know how to fix it, but hopefully if you see it 495 00:26:50,760 --> 00:26:55,840 generated then you will know. Another thing is that they are 496 00:26:55,920 --> 00:27:00,920 quite good at judging your post. For instance, you can ask an LLM 497 00:27:00,920 --> 00:27:04,880 to point out really logical fallacies that it finds in the 498 00:27:04,880 --> 00:27:07,280 post. Often those fallacies aren't 499 00:27:07,280 --> 00:27:11,160 real and it was just some hallucinated low quality 500 00:27:11,160 --> 00:27:12,840 response. But it's fine because you can 501 00:27:12,840 --> 00:27:14,600 just discard it. You're a human being, you can 502 00:27:14,600 --> 00:27:17,920 just decide. Sometimes it's really is like a 503 00:27:18,600 --> 00:27:20,960 wow moment at all. I really didn't cover that. 504 00:27:22,040 --> 00:27:26,200 It also acts and this is actually a great feature of how 505 00:27:26,720 --> 00:27:30,760 LLM architecture is just implemented as of today is that 506 00:27:30,760 --> 00:27:35,520 the the only context often that this large language model has is 507 00:27:36,200 --> 00:27:38,840 whatever you put in there. So if you're missing something 508 00:27:38,840 --> 00:27:41,840 because you subconsciously assume that everybody knows that 509 00:27:42,000 --> 00:27:45,240 the model doesn't, so it could point out where does this come 510 00:27:45,240 --> 00:27:47,480 from? I don't understand what's the 511 00:27:47,480 --> 00:27:51,520 source of this information. So it's also great for that. 512 00:27:51,520 --> 00:27:54,360 Then there are also simple things you can also. 513 00:27:54,760 --> 00:27:58,360 You can also use AI for more context aware grammar objects. 514 00:27:59,120 --> 00:28:02,440 For instance tenses that you used are coherent and so on. 515 00:28:02,680 --> 00:28:06,440 So for all kinds of reviewing without asking your friends to 516 00:28:06,440 --> 00:28:09,280 review for you. For all those use cases, AI is 517 00:28:09,280 --> 00:28:12,360 great. I already use it and it will 518 00:28:12,360 --> 00:28:15,320 only get better. Yeah, very interesting aspect 519 00:28:15,320 --> 00:28:18,720 that you mentioned, you haven't found any kind of like good AI 520 00:28:18,720 --> 00:28:21,880 written content, right? So I think most important thing 521 00:28:21,880 --> 00:28:24,360 for me also like is the authenticity part of your 522 00:28:24,360 --> 00:28:26,600 writing, right? You, I mean, depends on your 523 00:28:26,840 --> 00:28:29,320 goal especially, right? So if you just want to produce 524 00:28:29,320 --> 00:28:31,640 something and post something under your name, I think maybe 525 00:28:31,640 --> 00:28:34,000 that's a different thing. But let's say you want to post 526 00:28:34,000 --> 00:28:36,880 something that is useful and you are proud of, I guess the 527 00:28:36,880 --> 00:28:39,000 authenticity part actually really matters, right? 528 00:28:39,320 --> 00:28:42,640 So I think AI can help us a lot in terms of what you mentioned, 529 00:28:42,640 --> 00:28:45,800 maybe giving feedback, viewing, improving aspects of your 530 00:28:45,800 --> 00:28:47,560 writings, right? Choosing the words. 531 00:28:47,760 --> 00:28:50,600 Because sometimes, especially if you are not a native English 532 00:28:50,600 --> 00:28:53,040 speaker, right, we don't know the best way to convey a 533 00:28:53,040 --> 00:28:56,200 particular sentence in a more maybe engaging manner or some 534 00:28:56,200 --> 00:28:58,640 different kind of emotion aspect that you want to put to the 535 00:28:58,640 --> 00:29:01,240 readers, right? So I think definitely very, very 536 00:29:01,240 --> 00:29:03,000 useful. And in your book, actually, you 537 00:29:03,000 --> 00:29:07,560 also cover some tips in terms of what kind of prompts you can ask 538 00:29:07,560 --> 00:29:10,360 AI to kind of like improve certain aspects of your writing. 539 00:29:10,640 --> 00:29:13,840 So is there any kind of prompts that to you is something that is 540 00:29:14,040 --> 00:29:17,480 unexpected, Maybe for some listeners here they could also 541 00:29:17,480 --> 00:29:20,720 try it out so that they can actually improve their writing 542 00:29:20,760 --> 00:29:22,760 by using that prompts. Is there such things? 543 00:29:23,400 --> 00:29:28,760 So I would suggest to everyone who already wrote anything, any 544 00:29:28,760 --> 00:29:33,480 kind of blog post, to go through the kinds of prompts, either the 545 00:29:33,480 --> 00:29:36,200 one that I mentioned before or the ones from the book. 546 00:29:36,200 --> 00:29:39,120 They're pretty similar. For instance, the ones that ask 547 00:29:39,120 --> 00:29:45,160 the model to please find any unclear sentences that weren't 548 00:29:45,160 --> 00:29:47,160 explained that their sources unknown. 549 00:29:47,160 --> 00:29:50,640 So maybe it's useful to provide this source and to run all the 550 00:29:50,640 --> 00:29:54,720 articles of yours with this. And you might be really 551 00:29:54,720 --> 00:29:58,040 surprised that it pointed out something quite important and 552 00:29:58,040 --> 00:30:00,840 you should just go and update your old post because something 553 00:30:00,840 --> 00:30:03,880 in there was nothing. You assume because you have some 554 00:30:03,880 --> 00:30:07,120 kind of industry knowledge or maybe knowledge in there, not to 555 00:30:07,120 --> 00:30:09,880 the company that was obvious to you, but it wasn't obvious to 556 00:30:09,880 --> 00:30:13,480 the model didn't work if you this company, obviously. 557 00:30:14,520 --> 00:30:17,240 So maybe I know this can quickly get outdated. 558 00:30:17,240 --> 00:30:19,400 Is there so from your experience, right. 559 00:30:19,400 --> 00:30:23,520 So which LLM model or which tool that you think are kind of like 560 00:30:23,840 --> 00:30:27,520 maybe good enough for helping us write us to write something? 561 00:30:28,480 --> 00:30:33,080 The easiest answer which I use is although it's best to just 562 00:30:33,080 --> 00:30:36,840 spin up everything that you have access to and just ask you the 563 00:30:36,840 --> 00:30:41,760 same question, the same list of things to go through because the 564 00:30:41,760 --> 00:30:46,320 response is very and often it's there is no clear winner at 565 00:30:46,320 --> 00:30:48,160 least. Well, we checked last year and 566 00:30:48,160 --> 00:30:52,480 LLMS are improving daily. But often it was the case that 567 00:30:52,480 --> 00:30:56,320 one of the models answered great about pointing out unclear 568 00:30:56,320 --> 00:30:59,440 sentences, but totally hallucinated, for instance, 569 00:30:59,440 --> 00:31:02,680 logical errors that the contradictions, but while the 570 00:31:02,680 --> 00:31:05,880 other shined of contradictions but didn't have a clue about 571 00:31:05,880 --> 00:31:08,760 those second things. So it's best to just spin up all 572 00:31:08,760 --> 00:31:13,520 of them because then you've got more, more diverse review and 573 00:31:13,640 --> 00:31:17,080 you can just combine everything into a very high quality review, 574 00:31:17,080 --> 00:31:20,880 especially that it's not a problem if the review is sound 575 00:31:20,880 --> 00:31:23,120 right, false, because then you just discard it. 576 00:31:23,120 --> 00:31:26,200 You can just just for yourself if it made any sense or not. 577 00:31:26,560 --> 00:31:31,400 And the real advantage is that while your friends might get a 578 00:31:31,400 --> 00:31:34,320 little sad and offended if you discard their review, LLMS are 579 00:31:34,320 --> 00:31:37,080 totally fine with that. So, so it's actually, it's 580 00:31:37,080 --> 00:31:40,440 actually great to just go with the first round with machines. 581 00:31:41,600 --> 00:31:43,640 Very interesting aspect. Yeah, sometimes I also use 582 00:31:43,640 --> 00:31:46,680 different models, different tools to actually ask about the 583 00:31:46,680 --> 00:31:49,800 same thing, right, and pick out the best things that they, you 584 00:31:49,800 --> 00:31:52,080 know, could suggest, right. So I think it's still the same 585 00:31:52,080 --> 00:31:55,680 case, I guess, as you mentioned. So I think one other aspect 586 00:31:55,680 --> 00:31:58,320 working with AI, right, especially writing, you know, 587 00:31:58,320 --> 00:32:01,080 maybe a blog post, longer form content. 588 00:32:01,600 --> 00:32:04,560 Do you advise us to do more like a chunking approach when we 589 00:32:04,560 --> 00:32:06,800 finish a certain paragraphs or sections? 590 00:32:06,800 --> 00:32:09,800 Or do you suggest us to do in, you know, like after we finish 591 00:32:09,800 --> 00:32:11,960 the writing? So any kind of tips how should 592 00:32:11,960 --> 00:32:14,880 we work together with AI, right? So is there any kind of like 593 00:32:14,920 --> 00:32:18,840 best practice or maybe suggested practice that you find very 594 00:32:18,840 --> 00:32:22,640 useful for yourself? My answer might have been 595 00:32:22,640 --> 00:32:26,320 different last year when it released, but now the context 596 00:32:26,320 --> 00:32:29,000 window of all these models is large enough to restrict your 597 00:32:29,000 --> 00:32:32,320 whole post, so you can iterate and it's the. 598 00:32:32,520 --> 00:32:36,480 Usually the paragraphs should be quite self-contained, or at 599 00:32:36,480 --> 00:32:40,880 least progressively. It makes some sense so that next 600 00:32:40,880 --> 00:32:44,360 paragraph wasn't depend on something from the future, which 601 00:32:44,360 --> 00:32:48,120 you can also ask LLM to check. By the way, it's quite, it's 602 00:32:48,120 --> 00:32:51,840 quite with the depth. I would much rather just write 603 00:32:51,840 --> 00:32:53,600 the whole thing and then go for the review. 604 00:32:53,600 --> 00:32:57,000 But that is just how I work. On the other hand, if you are 605 00:32:57,000 --> 00:33:01,400 stuck writing and you get this writers blog that I don't really 606 00:33:01,400 --> 00:33:04,920 have so I can't speak from experience, but let's say that 607 00:33:04,920 --> 00:33:07,720 you do have it. You can just copy paste whatever 608 00:33:07,720 --> 00:33:09,640 you have right now and ask for a nudge. 609 00:33:10,160 --> 00:33:13,000 Like how to continue this? What would you latch upon next? 610 00:33:13,000 --> 00:33:15,920 Not actual content, but the idea, the suggestion. 611 00:33:16,680 --> 00:33:18,960 What would be a good continuation of this? 612 00:33:19,080 --> 00:33:22,320 I think because I'm stuck. Yeah, very interesting. 613 00:33:22,320 --> 00:33:26,120 So definitely writer's block is I assume almost everyone will 614 00:33:26,120 --> 00:33:27,920 have this kind of like problem, right? 615 00:33:27,920 --> 00:33:30,160 So one day you're just stuck like you don't know what next 616 00:33:30,160 --> 00:33:32,720 idea to write. So definitely AI could help 617 00:33:32,760 --> 00:33:35,160 giving you some prompts, you know, like, OK, maybe you can 618 00:33:35,160 --> 00:33:38,280 talk about ABCD and maybe from there we can start picking up 619 00:33:38,320 --> 00:33:42,800 again our so hot flow, right? So in your book you also cover 620 00:33:42,800 --> 00:33:45,720 this few things called blog post patterns, right? 621 00:33:45,720 --> 00:33:47,080 I think there are a couple of them. 622 00:33:47,240 --> 00:33:50,240 I find this very useful. So for some of us who are like 623 00:33:50,240 --> 00:33:53,320 beginners in terms of writing, but also narrow down in terms of 624 00:33:53,320 --> 00:33:56,600 like what patterns of writing, you know, like the post itself 625 00:33:56,840 --> 00:34:00,040 that we could try out. So maybe from your view, what 626 00:34:00,040 --> 00:34:02,520 are some of your favourites that maybe you can share to us here 627 00:34:02,520 --> 00:34:05,080 and why are they good to start with? 628 00:34:07,080 --> 00:34:12,639 First effect that I first of all I have and I also heard from our 629 00:34:12,679 --> 00:34:16,679 readers that they also have is that after you read about all 630 00:34:16,679 --> 00:34:19,600 those patterns that we figured out or cover most of the 631 00:34:20,000 --> 00:34:23,960 blogosphere, you can't help but qualify everything you see 632 00:34:24,760 --> 00:34:29,040 subconsciously and recognize this that oh, this is this is 633 00:34:29,040 --> 00:34:32,239 backhand. There are some there are some 634 00:34:32,239 --> 00:34:34,400 blog posts that don't fit well into the patterns that we 635 00:34:34,400 --> 00:34:37,199 listed. We just picked some that we 636 00:34:37,199 --> 00:34:41,400 fought together, cover most of the most of the content out 637 00:34:41,400 --> 00:34:44,159 there. Backhand is the first pattern we 638 00:34:44,159 --> 00:34:47,159 explained. And it's also my favorite 639 00:34:47,560 --> 00:34:52,719 because just because most of my favorite blog posts happen to be 640 00:34:52,719 --> 00:34:56,960 from this category. So I can talk about this one and 641 00:34:56,960 --> 00:35:01,400 this one first, It's quite self descriptive, but the bug hunt is 642 00:35:01,400 --> 00:35:05,560 about it's a post that explains how somebody looked for a bug in 643 00:35:05,560 --> 00:35:08,760 the code. It's usually structured like a 644 00:35:09,520 --> 00:35:13,600 detective story a little bit. Usually at the during the first 645 00:35:13,600 --> 00:35:16,560 paragraphs, it's building tension and you don't yet know 646 00:35:16,560 --> 00:35:20,000 as as the author, you don't even know what's happening. 647 00:35:20,000 --> 00:35:21,800 You're you're trying different things. 648 00:35:21,840 --> 00:35:26,200 And then there is this cathartic paragraph that says we found it 649 00:35:26,320 --> 00:35:31,480 and it explains what it was, why some of the previous 650 00:35:31,480 --> 00:35:35,320 investigations didn't make sense and why this one was valid and 651 00:35:35,640 --> 00:35:42,560 hopefully how to also fix it. This is my favorite blog blog 652 00:35:42,560 --> 00:35:46,040 post category, mostly because it's super educational for me, 653 00:35:46,040 --> 00:35:50,240 because usually along the way you learn very interesting 654 00:35:50,320 --> 00:35:54,360 techniques, how to deal with common issues, let's say. 655 00:35:54,360 --> 00:35:58,840 So if if somebody just sees the effect of a bug that something 656 00:35:58,840 --> 00:36:02,360 bad happens, for instance, the program goes out of memory. 657 00:36:02,920 --> 00:36:06,240 They usually describe trying out a few techniques, how to check 658 00:36:06,240 --> 00:36:10,920 what happens, memory profiling, even debug printing and so on. 659 00:36:10,920 --> 00:36:15,600 And one of those works. And then whenever in the future 660 00:36:15,600 --> 00:36:19,040 you see a similar issue, it's easy to just remember that they 661 00:36:19,040 --> 00:36:21,040 read the blog post about a similar problem. 662 00:36:21,040 --> 00:36:23,920 So let's go back to it and try out the same things. 663 00:36:23,920 --> 00:36:27,080 And maybe you hit the same after the same investigation, maybe 664 00:36:27,080 --> 00:36:29,000 not, but maybe it was a different issue that was also 665 00:36:29,000 --> 00:36:31,760 kind of described in there. This, This is why it's my 666 00:36:31,880 --> 00:36:34,480 favorite pattern because it's deeply educational. 667 00:36:34,480 --> 00:36:39,720 Almost exclusively. It's you always gain some unique 668 00:36:39,720 --> 00:36:41,960 knowledge after reading this kind of blog post. 669 00:36:43,760 --> 00:36:45,680 Yeah. So I could remember sometimes 670 00:36:45,680 --> 00:36:48,720 when I face a particular challenging box, so to speak, 671 00:36:48,720 --> 00:36:50,400 right? So maybe most likely it's the 672 00:36:50,400 --> 00:36:53,840 external related box, right? Be it the library that you use 673 00:36:53,840 --> 00:36:56,120 or certain internals of the libraries that somehow just 674 00:36:56,120 --> 00:36:59,640 doesn't work with your use case. So finding out this kind of like 675 00:36:59,640 --> 00:37:01,880 Buckhand blog post is super useful. 676 00:37:02,520 --> 00:37:04,600 It's kind of like safe saving your day, right? 677 00:37:04,600 --> 00:37:07,960 So I think Buckhand could be one pattern that you actually 678 00:37:07,960 --> 00:37:11,360 explore to write something that is interesting for people to 679 00:37:11,360 --> 00:37:12,560 learn. And it's like what you 680 00:37:12,560 --> 00:37:14,800 mentioned, right? Super educative when other 681 00:37:14,800 --> 00:37:19,360 people also face the same bug. So other aspects of pattern that 682 00:37:19,360 --> 00:37:22,520 you mentioned is about, you know, covering a certain trends, 683 00:37:22,520 --> 00:37:24,280 right? So these days especially, right, 684 00:37:24,480 --> 00:37:29,080 you have so many things to talk about AI, the maybe AI replacing 685 00:37:29,080 --> 00:37:32,280 developers or maybe the new super cool model that just got 686 00:37:32,280 --> 00:37:35,360 released, right? So is this something that also 687 00:37:35,360 --> 00:37:38,480 people I don't like? For some people, they like to 688 00:37:38,480 --> 00:37:41,120 cover certain trends, right? But for me, I find it quite 689 00:37:41,120 --> 00:37:44,440 exhausting because you have to always keep up, right? 690 00:37:44,880 --> 00:37:48,000 So tell us maybe from your perspective, how how effective 691 00:37:48,000 --> 00:37:53,240 this kind of pose is? First of all, this, this may be 692 00:37:53,240 --> 00:37:56,160 a tiny bit depressing, but it also would be motivating. 693 00:37:56,160 --> 00:37:59,360 But the fact is that force on trends, the pattern that we 694 00:37:59,360 --> 00:38:02,760 have, that's how we called it and kind of an expert level 695 00:38:02,760 --> 00:38:10,040 pattern that you need to to advance to after writing blog 696 00:38:10,040 --> 00:38:15,640 posts from other categories. Because the sad truth is that if 697 00:38:15,640 --> 00:38:20,480 you as a new author write your thoughts on some top, there's a 698 00:38:20,480 --> 00:38:24,280 very high chance nobody cares at all unless you're given a tiny 699 00:38:24,280 --> 00:38:26,600 bit recognizable in your niche, whatever it is. 700 00:38:26,960 --> 00:38:30,360 Then people might think, Oh, I would like to know what this guy 701 00:38:30,360 --> 00:38:32,560 thinks about about a certain topic. 702 00:38:32,680 --> 00:38:36,760 But if you, if it's your first steps, there's, it might be 703 00:38:36,760 --> 00:38:38,960 discouraging because you write something that is deeply 704 00:38:38,960 --> 00:38:41,480 personal to you, but nobody reads it because they just 705 00:38:41,480 --> 00:38:43,520 scroll through it and didn't find anything uniquely 706 00:38:43,520 --> 00:38:46,280 interesting. So this is this is something 707 00:38:46,280 --> 00:38:50,360 that we recommend not as the first thing to pick when when 708 00:38:50,360 --> 00:38:53,840 writing, because even our examples that we went through 709 00:38:53,840 --> 00:38:57,640 are pretty much only from well established bloggers that you 710 00:38:57,640 --> 00:39:02,880 already recognized. And often it's their name, their 711 00:39:02,880 --> 00:39:07,360 name combined with the title that makes people actually read 712 00:39:07,360 --> 00:39:10,720 it, not just the title, because the title is already hyped. 713 00:39:10,720 --> 00:39:15,400 So, so yeah, this is an interesting observation that 714 00:39:15,400 --> 00:39:21,560 this is not a beginner friendly pattern, but it it is 715 00:39:21,560 --> 00:39:25,480 interesting because usually those bloggers are already 716 00:39:25,480 --> 00:39:27,240 experienced. So they're also good at 717 00:39:27,600 --> 00:39:30,840 condensing knowledge into a few minutes of reading. 718 00:39:31,800 --> 00:39:36,960 And also you can you can see that these patterns have 719 00:39:38,080 --> 00:39:41,760 naturally become serious. For instance, somebody tried out 720 00:39:41,760 --> 00:39:44,360 the language and then five years later they write the 721 00:39:44,360 --> 00:39:47,800 continuation of this story, going over of all new features 722 00:39:47,800 --> 00:39:51,440 and so on. So it is very nice to just go on 723 00:39:51,440 --> 00:39:54,880 and read for half a day about the whole story, starting from 724 00:39:54,880 --> 00:39:58,800 1996 to today. Right, wow, I think that that's 725 00:39:58,800 --> 00:40:01,520 very unique insight, right. So I I thought that you could 726 00:40:01,680 --> 00:40:03,960 actually just cover any trend, but I think it's more like an 727 00:40:03,960 --> 00:40:06,040 advanced kind of like a pattern, right. 728 00:40:06,040 --> 00:40:08,680 So maybe you should start with some other patterns first, you 729 00:40:08,680 --> 00:40:12,040 know, maybe that is beginners friendly rather than catching up 730 00:40:12,040 --> 00:40:13,800 with all the trends. So. 731 00:40:14,640 --> 00:40:18,120 I think 1 aspect after we right finish it is is always good to 732 00:40:18,120 --> 00:40:20,560 actually publish it, right? So don't forget to publish it. 733 00:40:21,320 --> 00:40:24,440 So publishing itself, it's not enough, I think, right? 734 00:40:24,440 --> 00:40:27,680 So I think you would need to promote it, market it somehow. 735 00:40:28,240 --> 00:40:31,280 Some people these days, you know, maybe publish on Medium 736 00:40:31,280 --> 00:40:33,080 and sub stack. Maybe it gets recommended by the 737 00:40:33,080 --> 00:40:35,240 platform. But other people might post on 738 00:40:35,240 --> 00:40:37,920 their personal websites and things like that and think 739 00:40:37,920 --> 00:40:42,800 that's just it, right? So maybe tell tell us why we 740 00:40:42,800 --> 00:40:47,680 should do promotion of our, you know, writing and how can it be 741 00:40:47,680 --> 00:40:52,840 beneficial for us? So the obvious thing is that 742 00:40:53,640 --> 00:40:56,520 unless you do some kind of promotion, you're new, then 743 00:40:56,520 --> 00:40:58,840 nobody's going to ever read it because they wouldn't find it 744 00:40:58,840 --> 00:41:03,280 because we've got enough enough content being published every 745 00:41:03,280 --> 00:41:04,760 day. You might get lucky at London 746 00:41:04,760 --> 00:41:06,920 Hacker News if something's very interesting. 747 00:41:06,920 --> 00:41:10,240 But again, landing on Hacker News usually comes from some 748 00:41:10,240 --> 00:41:12,720 kind of promotion. Somebody must have posted it in 749 00:41:12,720 --> 00:41:14,360 the first place. If it was you, it was also 750 00:41:14,360 --> 00:41:17,680 promotion. So first of all, there are 751 00:41:17,680 --> 00:41:22,000 multiple, like technically speaking, there are multiple 752 00:41:22,360 --> 00:41:24,760 ways you can publish A blog post. 753 00:41:24,840 --> 00:41:29,520 The easiest one, if it's just your technical company blog and 754 00:41:29,520 --> 00:41:33,200 it happens to be well maintained, high quality block 755 00:41:33,200 --> 00:41:36,680 anyway, for instance, is a great example of that. 756 00:41:37,080 --> 00:41:41,280 And then it's super easy because usually someone else takes care 757 00:41:41,280 --> 00:41:43,200 of everything. You just need to provide the 758 00:41:43,920 --> 00:41:47,280 content and then it's edited, it's put out there, it's 759 00:41:47,280 --> 00:41:51,960 promoted by the company accounts and everywhere and then your job 760 00:41:51,960 --> 00:41:54,920 is done. If you decide to self publish, 761 00:41:54,920 --> 00:41:57,440 then obviously you also need to put in some more work. 762 00:41:57,560 --> 00:42:01,680 Figure out things that Cynthia is expert about and she wrote 763 00:42:01,680 --> 00:42:05,520 all wrote all of her knowledge condensed into these chapters 764 00:42:05,520 --> 00:42:08,840 like how to make sure that people can actually find it, how 765 00:42:08,840 --> 00:42:12,000 to submit it to all kind of aggregators, Hacker News, 766 00:42:12,000 --> 00:42:14,800 lobsters and all the other kinds of blogs. 767 00:42:16,000 --> 00:42:19,880 There is also a whole bunch of companies that specializes in 768 00:42:20,440 --> 00:42:25,280 reposting not original content. So they are absolutely fine. 769 00:42:25,480 --> 00:42:28,080 And actually it's their whole business to just take your blog, 770 00:42:28,440 --> 00:42:31,040 say somewhere in the footer that is actually not original. 771 00:42:31,040 --> 00:42:34,000 It comes from here, but repost it on their side. 772 00:42:35,000 --> 00:42:40,240 So there are, yeah, a whole bunch of tips and tricks how to 773 00:42:40,440 --> 00:42:45,400 make your blog was visible to the to the larger objects. 774 00:42:47,240 --> 00:42:48,440 Right. I think those are some great 775 00:42:48,440 --> 00:42:51,120 tips, definitely, right. But what if let's say I am a 776 00:42:51,120 --> 00:42:53,920 person, I don't have any help maybe from the company, I don't 777 00:42:53,920 --> 00:42:59,080 have anyone with great networks. I'm just like a normal developer 778 00:42:59,080 --> 00:43:01,680 with small social media following probably right. 779 00:43:02,600 --> 00:43:05,960 So what tips would you give us? Would you recommend to actually 780 00:43:06,600 --> 00:43:09,560 start promoting you know our own content without you know such 781 00:43:09,560 --> 00:43:11,880 help? I was first of all lucky enough 782 00:43:11,880 --> 00:43:16,200 that my first few articles were just piggybacked on on the on 783 00:43:16,200 --> 00:43:19,440 the technical blog post that that my company had. 784 00:43:20,480 --> 00:43:25,680 But still, especially after joining an even earlier stage 785 00:43:25,680 --> 00:43:32,560 startup, I broke my promise to never have a social media 786 00:43:32,560 --> 00:43:35,240 account. And I started using bug and 787 00:43:35,240 --> 00:43:38,120 Twitter for promoting just because everyone else was doing 788 00:43:38,120 --> 00:43:41,320 it because we really needed the company to get more broadly 789 00:43:41,320 --> 00:43:43,720 known. But it does work exactly the 790 00:43:43,720 --> 00:43:47,680 same way with blog posts as well, your personal posts and 791 00:43:48,120 --> 00:43:52,200 well, to bootstrap yourself, assuming that you have some kind 792 00:43:52,200 --> 00:43:57,360 of niche that you, that you were writing about a certain language 793 00:43:57,360 --> 00:44:01,200 or I don't know, databases, web programming, whatever else. 794 00:44:01,880 --> 00:44:06,720 If you just start engaging with the community, even reviewing 795 00:44:06,720 --> 00:44:10,280 someone else's posts, commenting about something, then you start 796 00:44:10,280 --> 00:44:14,720 getting those smaller circles. And then if you publish a post 797 00:44:14,720 --> 00:44:18,080 that is interesting to them, then hopefully some of them will 798 00:44:18,080 --> 00:44:21,200 share it and you'll wrote in your circle a little bit and you 799 00:44:21,200 --> 00:44:24,840 just go from there. Right. 800 00:44:25,160 --> 00:44:28,120 So yeah, definitely. I also learned like long form 801 00:44:28,120 --> 00:44:31,280 content, be it writing or maybe podcast, it's a kind of like a 802 00:44:31,280 --> 00:44:34,600 longer term thing, right? You cannot just expect other 803 00:44:34,600 --> 00:44:37,440 people to find the content that you produce, right? 804 00:44:38,000 --> 00:44:39,440 First of all, yeah, you need to produce more. 805 00:44:39,440 --> 00:44:41,120 I guess. It's not just one time thing or 806 00:44:41,120 --> 00:44:44,560 so that you promote and you expect others to find you, but 807 00:44:44,560 --> 00:44:47,320 also, yeah, because starting from social media, even though 808 00:44:47,320 --> 00:44:51,200 you have small circles right in your social media, just post it. 809 00:44:51,200 --> 00:44:54,520 Hopefully somebody shares and you know, organically, you know, 810 00:44:54,520 --> 00:44:57,560 your posts can be found by others if let's say it's it's 811 00:44:57,560 --> 00:44:59,680 found to be useful for some people. 812 00:45:00,680 --> 00:45:03,520 So I think 1 aspect that you mentioned in the very beginning 813 00:45:03,520 --> 00:45:05,960 and also in your book, right, you mentioned that blog posts 814 00:45:05,960 --> 00:45:09,080 could be a beginning of something that is bigger, right? 815 00:45:09,360 --> 00:45:13,440 So the most common aspects that normally people experience is 816 00:45:13,720 --> 00:45:16,800 being invited to write a book, just like in your case, right? 817 00:45:17,080 --> 00:45:20,320 Or maybe giving a talk in the conference, if let's say, the 818 00:45:20,320 --> 00:45:24,000 topic itself can be super useful for a particular conference. 819 00:45:24,360 --> 00:45:26,760 So tell us this experience because I'm sure some, some of 820 00:45:26,760 --> 00:45:31,000 us also would like to know what, what benefits can it bring us 821 00:45:31,360 --> 00:45:33,400 just by writing, for example, blog post. 822 00:45:34,880 --> 00:45:38,280 Yeah. First of all, my experience was 823 00:45:38,280 --> 00:45:41,320 again, a little bit specific because I was also pretty much 824 00:45:41,320 --> 00:45:46,360 forced to go and present on my first first conference live. 825 00:45:47,360 --> 00:45:51,360 But the thing with blog post is that first of all, that you, you 826 00:45:51,360 --> 00:45:55,240 kind of get noticed by people who look for either speakers or 827 00:45:55,240 --> 00:45:59,680 outdoors or whatever, because this is exactly, you're exactly 828 00:45:59,680 --> 00:46:01,520 the target. You're somebody who posted 829 00:46:01,520 --> 00:46:05,320 something potentially interesting so that you, you 830 00:46:05,320 --> 00:46:07,240 have something interesting to say about this topic. 831 00:46:07,240 --> 00:46:10,520 So you're, you're a great fit for writing a book about it or 832 00:46:10,520 --> 00:46:14,920 speaking about it. The very nice thing about 833 00:46:15,320 --> 00:46:20,000 especially technical blog posts is that sometimes you can just 834 00:46:20,320 --> 00:46:24,080 not even improve it into a into a slide deck. 835 00:46:24,120 --> 00:46:27,720 You can just translate it into a slide deck for a 20 minute 836 00:46:27,720 --> 00:46:31,520 presentation and then present it online without changing too many 837 00:46:31,520 --> 00:46:35,520 words really. Because if your structure is 838 00:46:35,720 --> 00:46:38,880 quite right and you've got this paragraphs, you've got some 839 00:46:40,320 --> 00:46:43,800 either benchmark results or maybe cold snippets, whatever 840 00:46:43,800 --> 00:46:47,680 else, those are slides. You can just often take them 841 00:46:47,720 --> 00:46:49,920 past this. You just, you just need to press 842 00:46:49,920 --> 00:46:53,000 the screenshot button a few times and you've got a whole 843 00:46:53,000 --> 00:46:57,360 deck. If this blog post was even 844 00:46:57,360 --> 00:46:59,880 moderately popular, then you know that there's interest in 845 00:46:59,880 --> 00:47:02,560 the topic. So we can just you can just take 846 00:47:02,560 --> 00:47:05,120 it and present it in the other way because you're simply 847 00:47:05,120 --> 00:47:08,920 covering more people. Because some people prefer to 848 00:47:09,440 --> 00:47:14,640 just put your presentation in a in another browser tab and 849 00:47:14,640 --> 00:47:16,880 listen to it. Some prefer to attend 850 00:47:16,880 --> 00:47:20,320 personality to a conference and then talk to you after your 851 00:47:20,320 --> 00:47:23,400 talk. So you just take the same 852 00:47:23,400 --> 00:47:24,880 content. The hard part is done. 853 00:47:24,880 --> 00:47:27,720 You already already investigated everything. 854 00:47:27,720 --> 00:47:29,080 You know what you're talking about. 855 00:47:29,160 --> 00:47:31,520 And you just present it in another way so that more people 856 00:47:31,520 --> 00:47:35,440 can access it, which is just yeah, that's just a good idea. 857 00:47:35,440 --> 00:47:39,320 It's you're already more than halfway through after the block 858 00:47:39,320 --> 00:47:42,360 is out. Yes. 859 00:47:42,520 --> 00:47:45,240 So I find if you got this opportunity being offered, you 860 00:47:45,240 --> 00:47:48,400 know, invited into a conference or writing a book that kind of 861 00:47:48,400 --> 00:47:51,200 like tells, oh, you have reached a certain different level, 862 00:47:51,200 --> 00:47:52,880 right. So from where you started in the 863 00:47:52,880 --> 00:47:55,080 beginning when you started writing the blog post? 864 00:47:55,520 --> 00:47:59,000 So Sana, is there any other thing that you want to convey 865 00:47:59,000 --> 00:48:00,640 here? I know that we have covered and 866 00:48:00,640 --> 00:48:03,040 discussed a lot of things. Is there any other things that 867 00:48:03,040 --> 00:48:06,400 you want to give us listeners here about aspects of, you know, 868 00:48:06,400 --> 00:48:09,960 writing as a developer? Is there any other things that 869 00:48:09,960 --> 00:48:15,160 you want to cover today? Or I could just repeat great 870 00:48:15,160 --> 00:48:19,840 words from Ryan Country, who also quoted another person who 871 00:48:19,960 --> 00:48:22,400 quoted another person. These words are you're not 872 00:48:22,400 --> 00:48:24,160 writing enough. And I love this quote because 873 00:48:24,160 --> 00:48:29,120 it's shortened to the point you should really consider sharing 874 00:48:29,120 --> 00:48:32,320 more because there's bounds to be a few things that you could 875 00:48:32,320 --> 00:48:35,960 have described and put out there so that people can learn from 876 00:48:35,960 --> 00:48:39,240 it. So just, yeah, I, I just 877 00:48:39,240 --> 00:48:42,240 encourage everyone to go and do it because there's no downside 878 00:48:42,240 --> 00:48:46,560 except spending some time. I love it, right. 879 00:48:46,600 --> 00:48:49,800 So definitely we are all we are all not writing enough, be it 880 00:48:49,800 --> 00:48:52,200 you know at work, be it personal thing. 881 00:48:52,200 --> 00:48:54,280 Sometimes it could be your journal or diary, right? 882 00:48:54,480 --> 00:48:57,440 Just to distill your thoughts, right or things that you are 883 00:48:58,160 --> 00:49:00,240 going through. So I think definitely writing 884 00:49:00,240 --> 00:49:03,960 more is something very useful. So Sana, as we reach the end of 885 00:49:03,960 --> 00:49:07,120 our conversation, I, I only have one last question that I'd like 886 00:49:07,120 --> 00:49:09,080 to ask you. I call this question the the 887 00:49:09,080 --> 00:49:11,240 three technical leadership wisdom. 888 00:49:11,600 --> 00:49:14,200 So if you can think, if you can think of them just like advice 889 00:49:14,200 --> 00:49:16,760 that you want to give to us. So anything that you want to 890 00:49:16,760 --> 00:49:19,800 share as part of your wisdom? Yeah. 891 00:49:19,800 --> 00:49:23,600 So chronologically from my career, first one would be 892 00:49:23,720 --> 00:49:27,800 engaging, open source, it really based off the second one would 893 00:49:27,800 --> 00:49:32,760 be to write more and the third one to create a culture around 894 00:49:32,760 --> 00:49:34,560 you that encourages people to write more. 895 00:49:36,720 --> 00:49:38,720 Wow, lovely. So it's all about contribution, 896 00:49:38,720 --> 00:49:40,880 it's all about writing. So thank you for sharing those. 897 00:49:41,400 --> 00:49:45,280 So for people who would love to maybe learn more from you right 898 00:49:45,280 --> 00:49:47,760 about this writing aspect or they want to follow your online 899 00:49:47,760 --> 00:49:50,120 resources, is there a place where they can find you online? 900 00:49:51,600 --> 00:49:55,920 The best place to learn about first of all the book and the 901 00:49:56,080 --> 00:50:00,480 continuation of the book is a web page to write that dot blog. 902 00:50:01,000 --> 00:50:07,640 This is kind of an aggregator where we go through new blog 903 00:50:07,640 --> 00:50:10,600 posts every month. Like we pick a few that are 904 00:50:10,600 --> 00:50:14,360 popular and categorize them with those patterns and shortly 905 00:50:14,360 --> 00:50:16,320 summarize why they're interested. 906 00:50:16,320 --> 00:50:21,520 So this is the kind of blog post aggregator and promotion for the 907 00:50:21,520 --> 00:50:26,160 book that you can go and see. All right, thank you. 908 00:50:26,160 --> 00:50:28,640 I'll put that in the show notes. So thank you again for your time 909 00:50:28,640 --> 00:50:31,400 today, Sana. So I really love what you shared 910 00:50:31,400 --> 00:50:35,480 just now and especially the invitation for us to write more, 911 00:50:35,720 --> 00:50:37,840 I guess because we are all not writing enough. 912 00:50:38,160 --> 00:50:39,640 Yeah. Thanks for the opportunity to 913 00:50:39,640 --> 00:50:42,000 first. It was great.