1 00:00:00,000 --> 00:00:05,100 Hey, it's Dan Shipper, and this week I'm excited to share a very special episode of 2 00:00:05,100 --> 00:00:05,640 AI & I. 3 00:00:05,860 --> 00:00:09,920 In this episode, we're previewing the release of an experimental new plugin called 4 00:00:09,920 --> 00:00:13,400 Compound Writing, which helps you write with AI the way that we do internally. 5 00:00:14,040 --> 00:00:18,160 It's inspired by Kieran Klassen's Compound Engineering, and it's built by one of our 6 00:00:18,160 --> 00:00:19,180 staff writers, Katie Parrott. 7 00:00:19,360 --> 00:00:24,040 To help talk about this plugin and why Katie made it and how she uses it, Natalia 8 00:00:24,040 --> 00:00:25,940 Quintero is the host of this episode. 9 00:00:26,100 --> 00:00:27,620 She's our head of AI consulting, thing. 10 00:00:28,200 --> 00:00:31,600 And she's been on the podcast a few times, and she's just an incredible interviewer. 11 00:00:31,840 --> 00:00:35,100 So I'm really excited for you to hear Katie and Natalia chat. 12 00:00:49,430 --> 00:00:53,190 Today we're talking to Katie Parrott, staff writer at Ebery. 13 00:00:53,570 --> 00:00:58,730 Katie, if you've been following Ebery for a while now, we'll be very familiar to 14 00:00:58,730 --> 00:00:58,990 you. 15 00:00:59,230 --> 00:01:04,670 She's one of our all star writers, and we're so excited to learn more about her 16 00:01:04,670 --> 00:01:07,430 writing process and her new plugin today. 17 00:01:08,030 --> 00:01:09,230 Hi Katie, how are you? 18 00:01:09,230 --> 00:01:10,250 I'm good. 19 00:01:10,290 --> 00:01:10,870 How are you? 20 00:01:11,830 --> 00:01:12,570 So good. 21 00:01:12,690 --> 00:01:14,330 I am so excited for this conversation. 22 00:01:15,730 --> 00:01:21,270 I am revealing myself as a Katie fangirl on the podcast. 23 00:01:21,770 --> 00:01:23,590 It's a mutual fandom. 24 00:01:25,990 --> 00:01:26,670 Thank you. 25 00:01:27,770 --> 00:01:28,790 Okay, so Katie, 26 00:01:29,710 --> 00:01:34,970 almost exactly two years ago, you wrote one of your most popular pieces for every 27 00:01:34,970 --> 00:01:37,710 and it was about how you had just been laid off. 28 00:01:38,310 --> 00:01:42,910 And also maybe we're in the middle of a mental health crisis, which you've written 29 00:01:42,910 --> 00:01:43,710 publicly about. 30 00:01:44,770 --> 00:01:49,570 And you decided to turn to chat TBT as a career coach. 31 00:01:50,210 --> 00:01:56,310 But at the time you were also an AI skeptic, which might be crazy to think about now 32 00:01:56,310 --> 00:01:57,890 given some of the work that you're doing. 33 00:01:57,910 --> 00:02:00,530 But can you take us back to that period? 34 00:02:01,090 --> 00:02:02,630 What was going on in your career? 35 00:02:03,150 --> 00:02:07,910 And in that moment in time that led you to turn to chat tbt in the first place. 36 00:02:09,170 --> 00:02:14,590 Yeah so I had just been let go from a crypto firm where I was working as an editor 37 00:02:14,590 --> 00:02:20,070 and ghostwriter for the for the team and you know that in like in hindsight like it 38 00:02:20,070 --> 00:02:23,090 was traumatizing at the time but in hindsight it was like the best thing that ever 39 00:02:23,090 --> 00:02:29,070 happened to me because it led me back to every uh and but I was feeling really lost 40 00:02:29,070 --> 00:02:36,030 as you mentioned you know I had tried crypto, I had tried in 41 00:02:36,030 --> 00:02:39,330 -house, I had tried out -of -house, I had tried all these different things and I 42 00:02:39,330 --> 00:02:45,090 felt like I was at the end of what I could do on my own and I really needed an 43 00:02:45,090 --> 00:02:50,330 outside perspective but because I was working on a budget with no clear visibility 44 00:02:50,330 --> 00:02:55,350 into how long I would be unemployed, hiring a real career coach wasn't really an 45 00:02:55,350 --> 00:02:56,150 option for me. 46 00:02:56,150 --> 00:03:00,030 So I had some exposure to chatgbt at that time. 47 00:03:00,590 --> 00:03:07,250 I had actually used it to produce briefs at a previous job as a content marketing 48 00:03:07,250 --> 00:03:07,690 manager. 49 00:03:07,830 --> 00:03:09,790 We had an agency and I realized, 50 00:03:11,030 --> 00:03:15,570 I was like hitting a wall of depression at that job and like could not bring myself 51 00:03:15,570 --> 00:03:17,310 to put words on a page. 52 00:03:17,690 --> 00:03:23,230 But I found that if I used chatgbt to generate some of the repetitive stuff 53 00:03:24,490 --> 00:03:30,090 And it became this assistive tool that I used as a tool for thought, which is a 54 00:03:30,090 --> 00:03:35,470 phrase I'd also seen Dan use around the same time because I was following every at 55 00:03:35,470 --> 00:03:35,990 the time. 56 00:03:36,710 --> 00:03:42,230 And so it was kind of that combination of things where I was like, okay, I can't 57 00:03:42,230 --> 00:03:47,330 afford like $150 an hour, however much a career coach would be, but I can sign up 58 00:03:47,330 --> 00:03:49,830 for a $20 a month ChatGPT subscription. 59 00:03:49,830 --> 00:03:56,490 and get some of that same result of externalizing my thinking and 60 00:03:56,490 --> 00:04:02,570 getting an outside perspective and just holding me accountable for 61 00:04:02,570 --> 00:04:08,270 how I was thinking about my career journey 62 00:04:09,190 --> 00:04:11,630 and challenging me when I would get a little, 63 00:04:12,450 --> 00:04:16,720 like I have a tendency to catastrophize and the worst and it kind of took me out of 64 00:04:16,720 --> 00:04:23,500 myself in that way too and it was just like a really great experience for me 65 00:04:23,500 --> 00:04:29,200 and like kind of and it ultimately led me to every so yeah. 66 00:04:29,700 --> 00:04:36,320 Yeah it's such a cool post that you wrote and you have you just always have such 67 00:04:36,320 --> 00:04:40,960 great zingers in your writing that are just so delightful to read. 68 00:04:41,740 --> 00:04:46,620 I hope when we are sharing this on video we can pull up some of the screenshots of 69 00:04:46,620 --> 00:04:52,720 just what your chat DPT prompts and screenshots look like because it feels vintage 70 00:04:52,720 --> 00:04:56,860 to be looking at chat from two and a half years ago. 71 00:04:57,740 --> 00:05:00,500 But in that post you have this really great quote. 72 00:05:01,320 --> 00:05:04,920 It's a little lengthy, but I'm going to try to read it because I have a question 73 00:05:04,920 --> 00:05:06,040 about it. 74 00:05:06,360 --> 00:05:10,760 So you say, one of my favorite things about writing is that I never know what I 75 00:05:10,760 --> 00:05:12,140 think until I write it down. 76 00:05:13,160 --> 00:05:15,260 But once I write it down, it becomes obvious. 77 00:05:16,300 --> 00:05:18,460 Ultimately, what's prompting ChatTPT is 78 00:05:19,680 --> 00:05:22,120 a form of very specific strategic writing. 79 00:05:22,460 --> 00:05:25,320 Do I think my experiment with ChatTPT added clarity in my thinking? 80 00:05:25,760 --> 00:05:26,380 Yes, it has. 81 00:05:26,760 --> 00:05:29,560 But as with so many of these tools, the real work is still mine. 82 00:05:30,080 --> 00:05:33,940 Clarity, it turns out, doesn't arrive gift wrapped from a digital assistant or even 83 00:05:33,940 --> 00:05:34,580 a human coach. 84 00:05:34,980 --> 00:05:38,460 It's something I had to dig out for myself, question by question and prompt by 85 00:05:38,460 --> 00:05:38,760 prompt. 86 00:05:38,760 --> 00:05:43,060 True clarity, direction, and answers come from the process of engaging with the 87 00:05:43,060 --> 00:05:44,820 prompts and doing the hard work yourself. 88 00:05:45,500 --> 00:05:48,180 AI can help guide the way, but the journey is yours to take. 89 00:05:49,420 --> 00:05:50,760 Just so beautiful. 90 00:05:51,260 --> 00:05:52,400 Just so, so cool. 91 00:05:53,140 --> 00:05:57,220 And, you know, I think one of the things that really resonates with me about your 92 00:05:57,220 --> 00:06:03,020 writing is that this feels so reflective of my experience using AI, where 93 00:06:03,020 --> 00:06:09,600 I am attracted to the possibility 94 00:06:09,600 --> 00:06:13,100 of maybe an easy answer, an easy solve. 95 00:06:14,080 --> 00:06:19,020 But actually what I constantly find myself doing is learning something that was 96 00:06:19,020 --> 00:06:25,600 always there inside of me, but I had to put some work in and some clarity into to 97 00:06:25,600 --> 00:06:27,520 like actually get it out of me. 98 00:06:28,520 --> 00:06:33,940 Can you tell me a little bit more about what using AI in this very early experience 99 00:06:33,940 --> 00:06:39,340 and this very kind of practical high -value experience, what it taught you about 100 00:06:39,340 --> 00:06:40,240 using AI? 101 00:06:41,220 --> 00:06:47,280 The first thing it taught me is that I can use AI to make change in my life, which 102 00:06:47,280 --> 00:06:50,580 is something that I've continued to carry through over the last couple of years. 103 00:06:51,560 --> 00:06:58,400 I really judge the technology by the impact it's had practically for me, And the 104 00:06:58,400 --> 00:07:05,240 fact that I used the career coach, for example, I remember talking when I got the 105 00:07:05,240 --> 00:07:11,020 invitation from Kate to write what ultimately became working overtime, my column, 106 00:07:11,920 --> 00:07:17,640 I was talking to ChatGPT about whether it was the right step for me to take 107 00:07:19,160 --> 00:07:21,720 because I was kind of on the fence. 108 00:07:21,860 --> 00:07:23,200 It would have been freelancing at the time. 109 00:07:23,320 --> 00:07:27,720 I was kind of on the fence about that versus looking for a full -time opportunity. 110 00:07:29,200 --> 00:07:33,900 And ChachiPT nudged me in the direction of like, well, why not try it in the 111 00:07:33,900 --> 00:07:34,280 meantime? 112 00:07:34,720 --> 00:07:37,540 And I did that and now I'm here. 113 00:07:38,400 --> 00:07:44,380 And throughout my experience with AI, that's really been the thing is that it's 114 00:07:44,380 --> 00:07:49,160 given me both the framework for thinking and obviously it's become the thing I think 115 00:07:49,160 --> 00:07:49,600 about. 116 00:07:50,400 --> 00:07:56,400 But it's really just about the fruits of the labor, so to speak, 117 00:07:57,600 --> 00:08:00,720 and seeing what AI is able to give back to me. 118 00:08:01,060 --> 00:08:06,760 So it sounds like one of the early takeaways that you had in leveraging AI was that 119 00:08:06,760 --> 00:08:10,160 you could really get it to be a valuable tool for you. 120 00:08:11,320 --> 00:08:18,180 And so that reminds me of another post that you wrote that was really popular with 121 00:08:18,180 --> 00:08:18,980 the Every Audience. 122 00:08:18,980 --> 00:08:25,140 and I believe it's titled AI Turned Me Into a Content Agency of One. 123 00:08:26,220 --> 00:08:30,520 And early in this post, you talk about how at the time you were a freelancer, so 124 00:08:30,520 --> 00:08:36,980 this is early 2025, and you realize that you had 125 00:08:36,980 --> 00:08:41,900 overcommitted yourself or potentially overcommitted yourself with the number of 126 00:08:41,900 --> 00:08:44,340 freelance projects that you had agreed to deliver. 127 00:08:45,220 --> 00:08:50,120 And then you go on to list the number of things that you committed to delivering in 128 00:08:50,120 --> 00:08:51,060 a two -week period. 129 00:08:51,700 --> 00:08:58,680 This includes eight blog posts, three ebooks, 24 LinkedIn posts, we're not 130 00:08:58,680 --> 00:09:03,880 even halfway through, 24 x posts, 16 Instagram posts. 131 00:09:04,120 --> 00:09:09,000 I mean, there is just no way I could write this amount of content in a month. 132 00:09:09,560 --> 00:09:12,640 How did you learn to do that? 133 00:09:12,760 --> 00:09:18,600 How did you learn to use AI to generate this amount of content and at the caliber 134 00:09:18,600 --> 00:09:19,740 that I know you deliver? 135 00:09:20,600 --> 00:09:26,440 I think around this time was when I was really discovering, like I wouldn't have 136 00:09:26,440 --> 00:09:30,320 used this language because it hadn't really become a thing yet, but context 137 00:09:30,320 --> 00:09:31,040 engineering. 138 00:09:31,820 --> 00:09:37,580 And what I realized was if I gave the model a really good foundation in terms of, 139 00:09:37,580 --> 00:09:42,820 you know, the brand messaging, product specifics, audience specifics of the 140 00:09:42,820 --> 00:09:44,500 companies that I was working for. 141 00:09:45,420 --> 00:09:50,220 I could shape the AI outputs toward the direction that I wanted. 142 00:09:51,080 --> 00:09:56,360 So it was really a lot of foundational work that once that was set, I could kind of 143 00:09:56,360 --> 00:09:57,880 just iterate through. 144 00:09:58,240 --> 00:10:02,400 And it helped that a lot of the things were content repurposing, which is one of the 145 00:10:02,400 --> 00:10:04,540 first things that people were really excited about. 146 00:10:04,540 --> 00:10:08,500 you know, one of the first products that Every put out was Spiral, which was all 147 00:10:08,500 --> 00:10:12,160 about content repurposing, because that's something that Every was doing a lot. 148 00:10:12,320 --> 00:10:16,200 You know, you have a post, you want to turn it into a LinkedIn post, a blog, you 149 00:10:16,200 --> 00:10:17,800 know, a Twitter post, a blog post. 150 00:10:18,020 --> 00:10:18,420 And so 151 00:10:19,780 --> 00:10:24,340 yeah, it was just like, I realized, I like, I was like, okay, I'll write through all 152 00:10:24,340 --> 00:10:28,260 of the surrounding information that the AI needs to do a good job. 153 00:10:28,820 --> 00:10:35,200 And then when that like, the The more I put in that work up front, the easier the 154 00:10:35,200 --> 00:10:36,620 process got farther down the line. 155 00:10:36,740 --> 00:10:41,260 And that's something that I found with my AI experience overall is just like you 156 00:10:41,260 --> 00:10:46,760 invest, at first it feels really effortful and you're spending a lot of time setting 157 00:10:46,760 --> 00:10:51,420 the scenes, like getting everything lined up, like in terms of your tools and your 158 00:10:51,420 --> 00:10:53,680 integrations and your context documents. 159 00:10:53,780 --> 00:10:58,900 But when you have that environment set up, you can just kind of run and like run 160 00:10:58,900 --> 00:10:59,340 fast. 161 00:10:59,340 --> 00:11:04,720 and that's what I was able to do when I over committed myself that that time. 162 00:11:05,620 --> 00:11:07,520 Which I did again like six months later. 163 00:11:09,420 --> 00:11:14,320 I think this might be a common characteristic of everyone at Emory. 164 00:11:16,480 --> 00:11:23,260 Can we talk a little bit more about what was in those files and in the context and 165 00:11:23,260 --> 00:11:28,120 in the setup that you created to let you fast because I 166 00:11:28,120 --> 00:11:34,920 think I remember some of the context files that you shared with us and that I saw 167 00:11:34,920 --> 00:11:36,480 around that time. 168 00:11:36,840 --> 00:11:38,840 But I'm curious, what does that look like? 169 00:11:38,940 --> 00:11:40,660 Were they folders that you were sharing? 170 00:11:40,960 --> 00:11:42,420 Were they projects on Claude? 171 00:11:42,960 --> 00:11:45,260 Walk me through what some of those things look like. 172 00:11:45,760 --> 00:11:48,480 They weren't projects because projects didn't exist yet. 173 00:11:48,480 --> 00:11:53,640 and I remember at the time, like at the time projects launched, I thought back to 174 00:11:54,090 --> 00:11:57,830 that period I was like I really wish that this had existed when I was doing this 175 00:11:57,830 --> 00:12:01,630 because I had to do a lot of manual copy and pasting like I would have my persistent 176 00:12:01,630 --> 00:12:06,050 documents and I would just in a Google doc and I would copy and paste into the 177 00:12:06,050 --> 00:12:10,510 context windows so that the agent so that it wasn't an agent well I guess maybe it 178 00:12:10,510 --> 00:12:13,810 was an agent I don't know but so that the model would have access to that 179 00:12:13,810 --> 00:12:14,350 information. 180 00:12:14,350 --> 00:12:19,890 And it really came from my experience in the content marketing world building style 181 00:12:19,890 --> 00:12:20,530 guides. 182 00:12:21,370 --> 00:12:27,930 You know, every brand worth its salt has, you know, a source of truth where 183 00:12:27,930 --> 00:12:30,330 they document, okay, who is our audience? 184 00:12:31,150 --> 00:12:32,670 What are their pain points? 185 00:12:33,870 --> 00:12:36,610 How does our product map to those pain points? 186 00:12:37,490 --> 00:12:40,710 What is our competition and what are our differentiators? 187 00:12:41,630 --> 00:12:48,410 These are units of work that I was used to doing as a content manager. 188 00:12:50,510 --> 00:12:52,530 I just started with what I was familiar with. 189 00:12:52,590 --> 00:12:58,170 I started with what I knew and gave that information to the model. 190 00:12:59,550 --> 00:13:01,150 It did a good job. 191 00:13:01,890 --> 00:13:04,630 At that point I was like, okay, this seems to be working. 192 00:13:04,630 --> 00:13:08,050 and then projects came along and I did so suddenly. 193 00:13:08,190 --> 00:13:12,310 I didn't have to copy paste anymore because I could use the project documents and I 194 00:13:12,310 --> 00:13:16,210 could use the custom instructions to sort of set a project up. 195 00:13:16,890 --> 00:13:19,550 And I had a project for every freelance client. 196 00:13:19,730 --> 00:13:26,390 I had a project for every column that I worked on because I was starting to spread 197 00:13:26,390 --> 00:13:29,050 out in terms of what I was doing at every at that time. 198 00:13:29,050 --> 00:13:35,590 But it all comes down to basic rules of writing and knowing who is our audience, 199 00:13:36,550 --> 00:13:40,370 you know, what is our different, like what is, what is our differentiated thing that 200 00:13:40,370 --> 00:13:47,350 we have to say, and just feeding the AI all of that foundational stuff 201 00:13:47,350 --> 00:13:50,730 so that it feels true. 202 00:13:50,950 --> 00:13:52,950 So that it kind of has rails to run on. 203 00:13:53,010 --> 00:13:57,570 That's what I'm always, I always think of is just like creating guardrails and kind 204 00:13:57,570 --> 00:14:00,770 of like a fenced in place for the model to play, 205 00:14:01,930 --> 00:14:08,610 and then you don't have to worry so much about getting something that's completely 206 00:14:08,610 --> 00:14:12,150 wrong because you're kind of pushing it in the right direction. 207 00:14:13,030 --> 00:14:14,850 Right, you've given it a playground. 208 00:14:15,350 --> 00:14:19,490 I know this is something that Mike Taylor on our team cares a lot about, you know, 209 00:14:19,510 --> 00:14:25,330 having really clean context windows and being very selective of what playground 210 00:14:25,330 --> 00:14:26,350 you're giving model. 211 00:14:26,350 --> 00:14:33,250 Knowing what you know now and as a writer, what are the things that you think 212 00:14:33,250 --> 00:14:35,490 have made it so that your projects are so successful? 213 00:14:35,750 --> 00:14:39,550 Is it having a really strong point of view on what good writing is, on what good 214 00:14:39,550 --> 00:14:40,670 content marketing is? 215 00:14:40,870 --> 00:14:44,690 What are those things that are so important to have a strong perspective on? 216 00:14:44,690 --> 00:14:50,450 I think it's less about the 217 00:14:50,450 --> 00:14:55,190 writing at first than it is about 218 00:14:55,190 --> 00:15:01,650 the other kinds of contacts like the 219 00:15:01,650 --> 00:15:07,970 data that you're giving it, the frameworks for 220 00:15:07,970 --> 00:15:12,330 audience personas and who they should be talking to. 221 00:15:12,330 --> 00:15:17,390 Like you need all of that in place before you can even get to stuff like use these 222 00:15:17,390 --> 00:15:19,150 words and avoid these words. 223 00:15:19,450 --> 00:15:25,510 Like prefer this level of reading difficulty and like strike this tone. 224 00:15:25,790 --> 00:15:30,390 Like before you even get to the point where you're sort of messing with that kind of 225 00:15:30,390 --> 00:15:37,190 stuff you need to have those foundational elements kind of set up and then like 226 00:15:37,190 --> 00:15:40,690 you'll we'll be wrestling with this addiction and syntax thing for the rest of 227 00:15:40,690 --> 00:15:41,030 forever. 228 00:15:43,310 --> 00:15:48,810 And when you're referring to the data, what is the data that you would need to to 229 00:15:48,810 --> 00:15:49,510 put in there? 230 00:15:49,830 --> 00:15:51,910 Well, that's the that's the specifics of the piece. 231 00:15:52,170 --> 00:15:57,490 So like something that I feel really strongly about it like as like our role as 232 00:15:57,490 --> 00:16:03,430 humans in this wonderful world of AI and AI writing specifically is that there's 233 00:16:03,430 --> 00:16:10,230 sort of a last mile problem in writing in AI and writing, which is AI has that 234 00:16:10,230 --> 00:16:11,330 knowledge cut off, right? 235 00:16:11,670 --> 00:16:15,630 It's always working with data that's several months out of date. 236 00:16:16,310 --> 00:16:22,290 And our job as the humans is to close that last mile and provide the real world 237 00:16:22,290 --> 00:16:27,630 experiences that AI can't get because they come after the knowledge cut off and it's 238 00:16:27,630 --> 00:16:33,790 also out there happening somewhere in the real world and AI is not in the real 239 00:16:33,790 --> 00:16:35,530 world, the physical world yet. 240 00:16:35,530 --> 00:16:40,150 So when I say research, I'm really talking about the unique or data that it's like 241 00:16:40,150 --> 00:16:44,570 the unique insight so that whether that's like a study that your that your company 242 00:16:44,570 --> 00:16:49,410 has produced or third -party research that you're pulling in which is something that 243 00:16:49,410 --> 00:16:54,290 I love to do in my writing for working overtime it's all of the all of the 244 00:16:55,530 --> 00:17:02,490 constitutive elements that go into the the writing and the model 245 00:17:02,490 --> 00:17:03,630 can then work from that. 246 00:17:03,830 --> 00:17:08,230 But the inputs are really what makes the writing unique. 247 00:17:09,110 --> 00:17:13,070 I think a lot of times when people are writing with AI, they just kind of go with to 248 00:17:13,070 --> 00:17:16,610 it and say, write a blog post about style guides. 249 00:17:16,790 --> 00:17:18,170 Let's just use that as an example. 250 00:17:18,390 --> 00:17:19,990 And they don't give it anything to go on. 251 00:17:20,110 --> 00:17:24,770 They're just like, write a style guide, a guide, a post about style guides based on 252 00:17:24,770 --> 00:17:26,430 what you, the model, already know. 253 00:17:26,910 --> 00:17:30,030 But then what the model already knows as commoditized information. 254 00:17:30,570 --> 00:17:32,290 Like that's not useful or interesting. 255 00:17:32,730 --> 00:17:39,730 So the data is the unique insight, the data point that the model doesn't have 256 00:17:39,730 --> 00:17:45,930 access to yet, the personal experience, all of those things that are working are 257 00:17:45,930 --> 00:17:47,730 ultimately going to make the writing unique. 258 00:17:48,790 --> 00:17:54,970 It all sounds like the documents that you save for what might be a project today or 259 00:17:54,970 --> 00:17:59,170 what might be maybe like a Codex sort of setup is all of the things that can be 260 00:17:59,170 --> 00:18:04,970 codified as like good best practices but then what you bring into every piece it's 261 00:18:04,970 --> 00:18:08,710 like the it's like the quality of the ingredients right so it sounds like what 262 00:18:08,710 --> 00:18:14,110 you're bringing to the piece it's just the the sort of like freshness quality and 263 00:18:14,110 --> 00:18:19,870 unique elements of the piece that you want to highlight and bring out that only you 264 00:18:19,870 --> 00:18:20,610 can really bring. 265 00:18:21,770 --> 00:18:26,510 I love that analogy that's a really good one that's That's pretty much exactly it. 266 00:18:26,630 --> 00:18:28,990 You know, the model is the kitchen. 267 00:18:30,170 --> 00:18:30,630 The, you know, 268 00:18:31,630 --> 00:18:33,110 the, you know, 269 00:18:34,010 --> 00:18:40,150 like structure, like the outline is like chopping and the composition is like 270 00:18:40,150 --> 00:18:46,770 boiling, but like you need the individual ingredients to be like fresh and 271 00:18:46,770 --> 00:18:49,850 like good to produce good work. 272 00:18:50,570 --> 00:18:50,850 Yeah. 273 00:18:51,510 --> 00:18:56,270 This is actually something that you referenced very early on in this piece in 2025. 274 00:18:56,550 --> 00:18:58,470 You have this really beautiful quote, 275 00:18:59,730 --> 00:19:03,610 and I'm sorry to read you your great quotes back to you. 276 00:19:03,690 --> 00:19:04,090 It's horrible. 277 00:19:04,570 --> 00:19:10,050 So you write, Two odd years into my AI journey, I have to admit, AI hasn't just 278 00:19:10,050 --> 00:19:12,110 helped me produce content faster. 279 00:19:12,550 --> 00:19:17,110 It has fundamentally changed the scale of what I can do, the limits of what I use to 280 00:19:17,110 --> 00:19:20,470 bump up against time, energy, capacity are way lower. 281 00:19:21,590 --> 00:19:23,650 And it sounds like that's the recipe, right? 282 00:19:23,790 --> 00:19:28,130 It's like you would always be using the same recipe, but now you don't have to worry 283 00:19:28,130 --> 00:19:28,990 so much about the recipe. 284 00:19:29,110 --> 00:19:30,110 The recipe is codified. 285 00:19:30,390 --> 00:19:35,090 Now you can really just come to the table with what is uniquely in your brain. 286 00:19:35,910 --> 00:19:37,650 This is something you also say in the post. 287 00:19:38,010 --> 00:19:42,070 It's become trite to say that AI frees you to focus on the human elements that truly 288 00:19:42,070 --> 00:19:46,130 matter, but AI has freed me to focus on the human elements that truly matter. 289 00:19:47,550 --> 00:19:49,070 I remember that one. 290 00:19:51,210 --> 00:19:54,570 It's magic seeing it from the outside, you know? 291 00:19:54,750 --> 00:20:00,390 Like, I have no idea how you do it, and it's magical every time that you put another 292 00:20:00,390 --> 00:20:01,070 piece together. 293 00:20:01,770 --> 00:20:03,190 It's magical to me too. 294 00:20:03,710 --> 00:20:06,170 Like, that's really all I can say is, like, 295 00:20:07,590 --> 00:20:12,950 the fact, like, working with AI kind of made me fall in love with writing again, 296 00:20:12,950 --> 00:20:16,970 because for a long time writing to me had just felt like such a slog. 297 00:20:17,450 --> 00:20:19,250 It was like so effortful. 298 00:20:19,530 --> 00:20:25,650 I was spending so much like energy and mental capacity just trying to assemble 299 00:20:25,650 --> 00:20:31,750 sentences that I wasn't really paying attention to what the sentences as a whole 300 00:20:31,750 --> 00:20:37,270 said or like how like or like I was just like this is done I'm gonna get it off my 301 00:20:37,270 --> 00:20:43,510 plate it's good enough and now with AI I have like the energy left over to spend 302 00:20:43,510 --> 00:20:46,370 more time wrestling with those harder, bigger picture questions. 303 00:20:47,250 --> 00:20:51,830 And it's just made writing a lot more fun because it's brought back a feeling of 304 00:20:51,830 --> 00:20:58,390 exploration and discovery both with the tool, but then also with my own brain and 305 00:20:58,390 --> 00:21:01,530 what I think and like how I bring that to life. 306 00:21:02,730 --> 00:21:06,470 So I want to turn back a moment to, 307 00:21:07,830 --> 00:21:11,870 you know, there's two things that I admire so much in the way that you use AI 308 00:21:11,870 --> 00:21:12,630 specifically. 309 00:21:14,050 --> 00:21:18,870 And one is obviously the way in which you use it for writing, but the other is the 310 00:21:18,870 --> 00:21:23,650 way in which you use it kind of like as a harness for doing your best work. 311 00:21:24,430 --> 00:21:30,510 And you have this really beautiful piece that you wrote in 312 00:21:31,090 --> 00:21:36,990 about a year ago, so kind of like mid 2025, called AI Solved the Problem that I 313 00:21:36,990 --> 00:21:38,820 couldn't explain to managers. 314 00:21:39,800 --> 00:21:46,180 And in this piece, you talk about how or actually, can you tell us a little bit 315 00:21:46,180 --> 00:21:46,700 about this piece? 316 00:21:46,700 --> 00:21:48,460 Do you remember what you talked about? 317 00:21:48,920 --> 00:21:51,720 You hear a lot about AI tools on this show. 318 00:21:52,160 --> 00:21:55,780 But the one that's truly at the center of your work is your CRM. 319 00:21:56,040 --> 00:22:00,120 It's where all of your customer context lives and where the work either happens or 320 00:22:00,120 --> 00:22:00,740 gets missed. 321 00:22:01,200 --> 00:22:05,600 Adio delivers the work that you never had time to focus on before and more. 322 00:22:05,600 --> 00:22:11,580 It picks up leads at 2am, it catches renewals before they slip, and drafts the 323 00:22:11,580 --> 00:22:13,920 follow -up before you've even thought to send one. 324 00:22:14,240 --> 00:22:19,400 Ask Adio anything about your business from this week's forecast to which deals have 325 00:22:19,400 --> 00:22:21,600 gone quiet and get the answer in seconds. 326 00:22:22,120 --> 00:22:27,800 And because it's extensible with an API and an MCP, your agents work right alongside 327 00:22:27,800 --> 00:22:32,600 your team to build pipeline, advance deals, and grow accounts around the clock. 328 00:22:32,600 --> 00:22:39,520 Try adio for free at adio .com slash every that's adio at o .com slash 329 00:22:39,520 --> 00:22:45,620 every and now back to the episode This is the piece where I talk about my mental 330 00:22:45,620 --> 00:22:49,240 health and my crippling anxiety around email 331 00:22:50,160 --> 00:22:54,520 Yeah, so this is a piece that this is this is like one of the most personal and like 332 00:22:54,520 --> 00:22:58,540 one of my favorite pieces that I wrote because it's about 333 00:22:59,600 --> 00:23:03,900 AI viewed through a different lens that I think we're used to thinking about it. 334 00:23:04,040 --> 00:23:06,520 We think about AI as a productive technology, right? 335 00:23:06,580 --> 00:23:12,520 We value it for its ability to produce output and get things out of our, like just 336 00:23:12,520 --> 00:23:13,760 put stuff out. 337 00:23:14,080 --> 00:23:19,680 And what I found in my personal experience is that AI is just as powerful as a 338 00:23:19,680 --> 00:23:20,880 supportive technology. 339 00:23:21,740 --> 00:23:24,720 So I have bipolar disorder. 340 00:23:24,720 --> 00:23:30,200 I experience spikes of like high energy and honestly frankly I spend a lot more time 341 00:23:30,200 --> 00:23:36,340 in really lows of depression and what I found is that AI kind of reduces friction in 342 00:23:36,340 --> 00:23:39,820 my life in a way that like makes it easier for me to person. 343 00:23:39,920 --> 00:23:45,560 It makes it easier for me to work, it makes it easier for me to you know move 344 00:23:45,560 --> 00:23:47,560 through my day -to -day life. 345 00:23:47,640 --> 00:23:53,860 That's actually become even more true now with agents being able to handle what our 346 00:23:53,860 --> 00:23:56,040 COO, Brandon, calls computer errands. 347 00:23:56,640 --> 00:24:02,380 So I put off making a primary care physician appointment for like... 348 00:24:02,390 --> 00:24:07,530 three years and then I realized I could simply tell Codex to go find primary care 349 00:24:07,530 --> 00:24:13,230 physicians near me that take my insurance and are accepting new patients. 350 00:24:13,450 --> 00:24:16,610 And it went off and found me someone and I booked an appointment. 351 00:24:17,010 --> 00:24:18,890 And that's something that I would never have done. 352 00:24:19,010 --> 00:24:23,030 I would have just kept procrastinating on it infinitely because of the way my brain 353 00:24:23,030 --> 00:24:29,290 works and AI was able to help me over that hump and it's done that over and over 354 00:24:29,290 --> 00:24:29,630 again. 355 00:24:29,630 --> 00:24:33,890 And so like, you know, using Quora to make my inbox more manageable. 356 00:24:33,890 --> 00:24:38,630 So I'm not confronting a wall of emails, but just the really important ones that 357 00:24:38,630 --> 00:24:40,630 need my human response. 358 00:24:40,850 --> 00:24:45,850 That's completely changed my relationship to my inbox because I don't have to like 359 00:24:45,850 --> 00:24:52,430 go digging and like be like looking for the bomb that is hiding in the, in the, in 360 00:24:52,430 --> 00:24:56,790 the millions of like, uh, things that are just like subscription notices and stuff, 361 00:24:57,050 --> 00:24:59,590 but the things that need human response has come to me. 362 00:25:00,710 --> 00:25:05,070 And now I have an automation set up through Codex, and I have it tell me what's in 363 00:25:05,070 --> 00:25:07,870 my inbox that I need to respond to. 364 00:25:07,990 --> 00:25:12,750 And so my inbox anxiety has been completely resolved. 365 00:25:13,290 --> 00:25:14,990 And I've just seen a lot of other ways 366 00:25:16,270 --> 00:25:23,070 that AI supports me and allows me to work better, not just in terms of 367 00:25:23,070 --> 00:25:26,610 using it for the output, but using it for everything that surrounds the output. 368 00:25:27,530 --> 00:25:27,550 Right? 369 00:25:27,730 --> 00:25:31,710 It's like it's not just the medical appointments, it's the bills, it's the emails, 370 00:25:32,290 --> 00:25:37,870 it's the tasks that, you know, you'll never really get around to doing like around 371 00:25:37,870 --> 00:25:39,550 the house or whatever it might be. 372 00:25:39,970 --> 00:25:46,110 And it does feel to me with AI that I have a little bit more firepower to get to 373 00:25:46,110 --> 00:25:47,050 some of those things. 374 00:25:47,170 --> 00:25:48,050 You'll love this, Katie. 375 00:25:48,230 --> 00:25:55,150 For the first time in my life, I'm filing insurance -free because Codex is filing my 376 00:25:55,150 --> 00:25:56,890 insurance reimbursements. 377 00:25:57,950 --> 00:26:02,530 I would rather pay insurance than how to manually do that. 378 00:26:02,950 --> 00:26:07,330 And it's exciting to have a tool that I feel like can really work in my favor. 379 00:26:08,970 --> 00:26:15,250 In this particular piece, you talk about that and you say, I've been working at 380 00:26:15,250 --> 00:26:19,050 Avery for one year now, and that might not sound remarkable, but it feels like a 381 00:26:19,050 --> 00:26:19,450 milestone. 382 00:26:19,450 --> 00:26:24,030 For me, the proof of AI's value is the fact that I'm still here doing work I'm proud 383 00:26:24,030 --> 00:26:24,930 of. 384 00:26:25,810 --> 00:26:31,130 And it kind of is the greatest measure of success that you get to do more of the 385 00:26:31,130 --> 00:26:34,350 work that you're excited to wake up and do every day. 386 00:26:35,030 --> 00:26:37,130 And to me, that's such a clear AI win. 387 00:26:37,550 --> 00:26:39,430 Can you tell us a little bit more about... 388 00:26:39,430 --> 00:26:46,230 So we started out with you going to ChatTBT over two years ago and 389 00:26:46,230 --> 00:26:47,790 asking it to be your AI coach. 390 00:26:47,790 --> 00:26:52,510 and it's enabled this incredible amount of work that you're doing now. 391 00:26:52,710 --> 00:26:54,350 You've talked about it in some posts. 392 00:26:54,830 --> 00:26:57,630 Now it looks a little bit more like a Codex project, if I believe. 393 00:26:57,990 --> 00:27:01,570 Can you tell us a little bit more about how that has evolved and what that looks 394 00:27:01,570 --> 00:27:02,190 like today? 395 00:27:03,330 --> 00:27:04,550 My writing process? 396 00:27:05,630 --> 00:27:07,970 This is for the kind of career, 397 00:27:08,910 --> 00:27:11,910 like what the equivalent of your career coach looks like today. 398 00:27:12,130 --> 00:27:15,850 Oh my gosh, my career coach has evolved so much. 399 00:27:15,850 --> 00:27:22,550 It is now a project inside of my codecs and it has so much more information than 400 00:27:22,550 --> 00:27:23,630 the first version. 401 00:27:24,270 --> 00:27:31,090 So it has a dossier on me and my role and it has 402 00:27:31,090 --> 00:27:38,030 context around every and our super top secret brand 403 00:27:38,030 --> 00:27:42,690 positioning that I got from our head of marketing, Douglas, 404 00:27:43,850 --> 00:27:47,310 that I was like, okay, if we put, and then what I'm able to do with that is like, 405 00:27:47,310 --> 00:27:49,390 okay, here's the context on me and my role. 406 00:27:49,530 --> 00:27:54,850 Here's the context on every, how do I put those two things together and what should 407 00:27:54,850 --> 00:27:58,870 I prioritize to have the biggest impact that I can? 408 00:27:59,870 --> 00:28:04,250 And so it's helped, it helps me now, now that I'm in a job and not looking, not 409 00:28:04,250 --> 00:28:09,190 meeting the career coach to help me find a job, it helps me orient myself 410 00:28:09,190 --> 00:28:09,990 prioritize. 411 00:28:10,730 --> 00:28:16,370 It does project management for me and now increasingly all of this is happening just 412 00:28:16,370 --> 00:28:18,270 with voice, like with monologue. 413 00:28:18,770 --> 00:28:23,750 I'd monologue like a ton all day, every day and I'll just like talk to my career 414 00:28:23,750 --> 00:28:28,570 coach about like I've got four different priorities like what should I, what should 415 00:28:28,570 --> 00:28:34,270 I, what should I do in what order and it kind of operates as that chief of staff 416 00:28:34,270 --> 00:28:39,310 that kind of like points me in a direction and says this is the highest impact thing 417 00:28:39,310 --> 00:28:44,770 you can do first, this is like maybe something that you can put off until later, and 418 00:28:44,770 --> 00:28:50,890 then that like and it just helps do that kind of that kind of work and it maintains 419 00:28:50,890 --> 00:28:52,570 a Kanban board for me. 420 00:28:52,830 --> 00:28:56,450 I've never been a person that can maintain a Kanban board but now I don't even have 421 00:28:56,450 --> 00:28:56,990 to touch it. 422 00:28:56,990 --> 00:29:02,590 It just lives in this little career coach management project land and I just like 423 00:29:02,590 --> 00:29:06,130 say like, what's the due date on this deliverable? 424 00:29:06,290 --> 00:29:06,950 And it will tell me. 425 00:29:07,870 --> 00:29:10,790 Can you tell us a little bit more about how that's set up? 426 00:29:10,850 --> 00:29:11,850 That sounds incredible. 427 00:29:12,510 --> 00:29:16,270 So like that's the thing is you kind of forget over time as it starts to work, you 428 00:29:16,270 --> 00:29:19,590 just take for granted that it's working and you don't think about what's inside it 429 00:29:19,590 --> 00:29:20,290 so much anymore. 430 00:29:21,150 --> 00:29:25,330 So let me find career coach mode. 431 00:29:25,930 --> 00:29:28,870 Yeah, so it's got evidence is one of the folders. 432 00:29:28,870 --> 00:29:35,110 So, you know, I periodically scrape all the data out of our content management 433 00:29:35,110 --> 00:29:39,670 system into a spreadsheet and like I have that so I know like how my posts are 434 00:29:39,670 --> 00:29:40,150 performing. 435 00:29:40,370 --> 00:29:45,650 It also has like reader feedback and other things like I maintain what I call my 436 00:29:45,650 --> 00:29:49,090 validation folder which is just full of nice things people have said about my work. 437 00:29:49,350 --> 00:29:53,630 That becomes another input because that helps the model understand, okay, what do 438 00:29:53,630 --> 00:29:57,210 people respond positively to in my work that we want to like reinforce. 439 00:29:57,210 --> 00:30:03,190 It has references, like that's where my role positioning is. 440 00:30:04,310 --> 00:30:11,150 It has my OKRs for Q2 and now Q3 because somehow it became 441 00:30:11,150 --> 00:30:15,630 Q3 which should not be legal and some other things like that. 442 00:30:15,770 --> 00:30:18,390 So it just has a ton of these, 443 00:30:19,830 --> 00:30:26,090 again it's all context, it's all documentation that increasingly I don't manually 444 00:30:26,090 --> 00:30:26,650 create. 445 00:30:26,970 --> 00:30:31,010 I how they I created for me like through an interview process like I'm I'm always 446 00:30:31,010 --> 00:30:36,810 having AI interview me Alex Duffy gave me that tip early on in my time at every and 447 00:30:36,810 --> 00:30:39,890 it's just like completely changed how I work with AI because I'll just be like ask 448 00:30:39,890 --> 00:30:44,310 me questions to draw out what I think about this and that's how I was able to pull 449 00:30:44,310 --> 00:30:49,470 together the career coaches like I just said I want to create a career coach you 450 00:30:49,470 --> 00:30:53,170 know this is what I wanted to do and then I started just showing it a bunch of 451 00:30:53,170 --> 00:30:58,190 context documents like all of the data from the performances and like my validation 452 00:30:58,190 --> 00:30:58,670 folder. 453 00:30:59,030 --> 00:31:03,630 And we just talked until we got to a place where we had a system that worked the way 454 00:31:03,630 --> 00:31:06,590 that I kind of like had envisioned in my mind. 455 00:31:07,690 --> 00:31:14,290 So in two years, which is, you know, a lot of time in AI time, but in two years, you 456 00:31:14,290 --> 00:31:15,030 went from 457 00:31:16,610 --> 00:31:20,990 prompting a simple prompt on chat TBT, you know, how can you help me as a career 458 00:31:20,990 --> 00:31:27,690 coach to having a fully autonomous codec setup that maintains your to -do list, 459 00:31:28,430 --> 00:31:34,730 your OKRs, it keeps you on track with a Can -bin board and at this point your 460 00:31:34,730 --> 00:31:37,170 codecs is maintaining it for you. 461 00:31:38,470 --> 00:31:38,970 Yes. 462 00:31:38,970 --> 00:31:39,890 Is that insane? 463 00:31:40,510 --> 00:31:41,890 So I mean, this is amazing, right? 464 00:31:42,170 --> 00:31:48,810 I think one of the things that you're uniquely great at is using AI for kind of like 465 00:31:48,810 --> 00:31:55,350 powering your career harness and also your ability to write really, really great 466 00:31:55,350 --> 00:31:55,810 pieces. 467 00:31:56,510 --> 00:31:59,210 And so going back to the writing component, 468 00:32:00,230 --> 00:32:05,490 you know, we talked a little bit about some of the how you've created AI to 469 00:32:05,490 --> 00:32:09,390 basically kind of like hold the recipe for you, like how you think about good 470 00:32:09,390 --> 00:32:13,810 writing, and then the elements that you bring to each writing session so that it can 471 00:32:13,810 --> 00:32:15,430 just allow you to do your best work. 472 00:32:15,430 --> 00:32:21,390 And now you are, 473 00:32:22,310 --> 00:32:27,430 I think by the time that this podcast airs, we will have published your compound 474 00:32:27,430 --> 00:32:29,510 writing plugin. 475 00:32:29,990 --> 00:32:36,830 Can you tell us a little bit about what compounding is as a concept and then about 476 00:32:36,830 --> 00:32:38,350 the writing plugin? 477 00:32:38,990 --> 00:32:45,830 Yeah, so compounding is the idea that every piece of feedback 478 00:32:45,830 --> 00:32:50,270 that you give to AI should feed back into the system to improve the next output. 479 00:32:51,070 --> 00:32:55,510 So obviously for folks that follow every they'll be familiar with this from Kieran 480 00:32:55,510 --> 00:32:57,810 Klassen's compound engineering plugin. 481 00:32:58,210 --> 00:33:02,090 He was really the one to pioneer this idea that you should only have to give a piece 482 00:33:02,090 --> 00:33:08,490 of feedback once, but if you codify it the right way, it will be there forever and 483 00:33:08,490 --> 00:33:11,570 the next time you're in that same situation, the AI is going to make a better 484 00:33:11,570 --> 00:33:11,990 decision. 485 00:33:13,170 --> 00:33:17,270 And so the Compound Writing plugin, it started, I literally just forked. 486 00:33:17,870 --> 00:33:22,770 Forked is a word that I know now because I wouldn't have known if I weren't working 487 00:33:22,770 --> 00:33:29,750 with AI, but I pulled the compound engineering plugin down and talked to 488 00:33:29,750 --> 00:33:36,150 Claude about it said, I want to take this compound engineering plugin and adapt it 489 00:33:36,150 --> 00:33:36,650 for writing. 490 00:33:37,190 --> 00:33:40,550 Because in certain ways, writing is very similar to code. 491 00:33:41,210 --> 00:33:46,950 Like the compound engineering plugin has brainstorming, ideating, planning, 492 00:33:48,430 --> 00:33:50,310 working, and then reviewing. 493 00:33:50,750 --> 00:33:57,530 And very similarly, in writing, you have brainstorming, outlining, drafting, 494 00:33:58,070 --> 00:33:58,510 reviewing. 495 00:33:58,510 --> 00:34:03,450 and it's kind of a similar kind of flow, but there are also differences. 496 00:34:03,750 --> 00:34:06,210 Like obviously, I don't need my, 497 00:34:07,110 --> 00:34:10,570 you know, for writing you're not looking at things like 498 00:34:11,430 --> 00:34:14,990 coding conventions, you're looking for things like structure, voice, 499 00:34:17,070 --> 00:34:20,110 evidence, and like all of these kinds of different things. 500 00:34:20,250 --> 00:34:24,130 So it's a different, like there's a different, it's a different container, but like 501 00:34:24,130 --> 00:34:25,850 it just made sense to me. 502 00:34:25,850 --> 00:34:31,310 So like again, I just talked to Claude and said, you know, I want I want these steps 503 00:34:31,310 --> 00:34:38,070 in the process I want you know, I want an outline to look like this I want a draft 504 00:34:38,070 --> 00:34:41,950 to be done this way and like these are the kinds of review and then like I want two 505 00:34:41,950 --> 00:34:45,930 different kinds of Reviews like a substantive edit That's like the big picture 506 00:34:45,930 --> 00:34:49,410 structure in the argument and then a line at it that looks at the individual writing 507 00:34:49,410 --> 00:34:54,710 and then a final pass that like looks at like just the final details that you want 508 00:34:54,710 --> 00:34:55,890 something to be publication ready. 509 00:34:56,090 --> 00:35:01,410 And I was able to pull that together just like, I always like all of my best AI 510 00:35:01,410 --> 00:35:04,690 ideas, I've ripped off from someone else on the every team. 511 00:35:05,130 --> 00:35:10,670 So it was very much compound engineering plus Dan's agent native architecture guide 512 00:35:10,670 --> 00:35:13,110 that kind of combined together to create Compound Writing. 513 00:35:13,470 --> 00:35:17,810 And now that's what I use all the time for every piece that I write. 514 00:35:17,950 --> 00:35:19,810 I'm no longer in a chat window. 515 00:35:19,810 --> 00:35:24,970 I'm no longer in a project on Claude .ai or chadgbt .com. 516 00:35:25,150 --> 00:35:31,910 I am like in the Claude desktop app or the chatgbt desktop app working with 517 00:35:31,910 --> 00:35:37,230 this system that has all that again has the recipe of how I want the drafting 518 00:35:37,230 --> 00:35:38,310 process to go. 519 00:35:38,810 --> 00:35:45,710 And then it has the, the context files about this, but like still the, many of 520 00:35:45,710 --> 00:35:49,490 the same context files that used to live in the project in terms of like a style 521 00:35:49,490 --> 00:35:50,770 guide, example pieces, 522 00:35:51,790 --> 00:35:57,210 and then again I just bring the fresh information into that ecosystem and 523 00:35:58,050 --> 00:36:00,290 a piece comes out of it one way or another. 524 00:36:01,750 --> 00:36:03,910 Somehow a piece appears. 525 00:36:05,150 --> 00:36:08,950 I'm so excited to use it when it comes available. 526 00:36:09,810 --> 00:36:13,790 I hope to be an alpha tester of it when you're ready to share it. 527 00:36:14,170 --> 00:36:14,790 Oh, it's on the GitHub. 528 00:36:15,170 --> 00:36:17,170 You can get it now if you're on the ever GitHub. 529 00:36:18,110 --> 00:36:18,250 Dang. 530 00:36:18,610 --> 00:36:18,890 Okay, great. 531 00:36:19,070 --> 00:36:20,450 I will definitely have to try it. 532 00:36:20,650 --> 00:36:24,710 One of the things that I find so exciting about a great plugin. 533 00:36:25,210 --> 00:36:32,130 A great plugin to me is like an expert instruction manual for how to do something 534 00:36:32,130 --> 00:36:33,610 you have no business doing. 535 00:36:34,890 --> 00:36:36,950 And so for me, for compound engineering, 536 00:36:38,470 --> 00:36:44,090 I am not an engineer by trade, and yet I spend my weekend building different 537 00:36:44,090 --> 00:36:49,550 products and products that I rely on now every day. 538 00:36:50,570 --> 00:36:57,050 And what compound engineering allows me to do is to build a product using the logic 539 00:36:57,050 --> 00:37:01,690 and as practices that a great engineer and actually like a great team of engineers 540 00:37:01,690 --> 00:37:06,850 would use for different steps in the kind of like product build process. 541 00:37:07,130 --> 00:37:10,210 Obviously it begins with something like a PRD or a plan. 542 00:37:10,870 --> 00:37:14,210 Things are scoped really well, there's security element, obviously there's 543 00:37:14,210 --> 00:37:19,370 compounding at the end, but this is a process that if you are not an engineer or an 544 00:37:19,370 --> 00:37:22,830 expert at building products, it's just totally foreign to you, right? 545 00:37:22,970 --> 00:37:26,590 So in the same way that when you were describing the things that you would 546 00:37:26,590 --> 00:37:33,410 intuitively create as context files for your cloud projects over a year or two 547 00:37:33,410 --> 00:37:33,950 ago. 548 00:37:34,750 --> 00:37:41,310 Now it's like I can access Katie's brain recipe with this Compound Writing 549 00:37:41,310 --> 00:37:48,170 plugin where I can both use the tool and the tool 550 00:37:48,730 --> 00:37:50,270 can help me do better work. 551 00:37:50,390 --> 00:37:54,330 But I think what's most exciting about a great plugin to me is that it can teach you 552 00:37:54,330 --> 00:37:56,310 how to do that kind of work, right? 553 00:37:56,470 --> 00:38:01,350 So it can teach you how to be a great writer because it presents the steps, the 554 00:38:01,350 --> 00:38:05,110 options, and the different directions that you might take at any step. 555 00:38:05,570 --> 00:38:08,310 Do you find that for writing in particular? 556 00:38:08,410 --> 00:38:13,750 Because it's not so deterministic in the way that code is, that it's a little bit 557 00:38:13,750 --> 00:38:17,670 more flavored, like it's more of a reflection of how you write, or can it be kind of 558 00:38:17,670 --> 00:38:21,170 like a generic tool that other people can use, even if they're not professional 559 00:38:21,170 --> 00:38:21,550 writers? 560 00:38:22,890 --> 00:38:26,270 I definitely think people can use it who are not professional writers and it may 561 00:38:26,270 --> 00:38:32,850 even be better for them because they don't have strong pre -existing opinions about 562 00:38:32,850 --> 00:38:38,930 the writing and so they're going to be less inclined to nitpick and they'll just be 563 00:38:38,930 --> 00:38:43,510 excited that they can produce something that they maybe hadn't been able to produce 564 00:38:43,510 --> 00:38:44,070 before. 565 00:38:45,890 --> 00:38:46,290 But, 566 00:38:48,370 --> 00:38:51,990 and like, I think that, and like, honestly, I think, and I wrote about this in a 567 00:38:51,990 --> 00:38:55,990 piece for working overtime, like writing with AI is harder than you think. 568 00:38:56,710 --> 00:38:58,090 It's made me a better writer. 569 00:38:58,470 --> 00:39:03,070 Like the analogy that I use for the Compound Writing plugin is that it's a gym. 570 00:39:04,110 --> 00:39:09,250 It's a workout system to like make you a better writer because most of the, so like, 571 00:39:09,370 --> 00:39:12,370 for example, in the editing step, I built in a lot of these reviewers. 572 00:39:12,370 --> 00:39:15,210 One of my favorites is Vonnegut. 573 00:39:15,610 --> 00:39:18,790 So Kurt Vonnegut had these like eight elements of story structure. 574 00:39:19,510 --> 00:39:22,110 So it's things like start as close to the end as possible. 575 00:39:23,410 --> 00:39:25,810 You know, every sentence has to earn its place. 576 00:39:26,350 --> 00:39:30,170 Give the reader a character they can root for. 577 00:39:30,550 --> 00:39:32,950 And I just turned that into a skill in the plugin. 578 00:39:33,410 --> 00:39:35,630 And now you can base it, you can borrow this framework. 579 00:39:35,850 --> 00:39:40,350 It's not that like Kurt Vonnegut, the spirit of Kurt Vonnegut is reviewing your 580 00:39:40,350 --> 00:39:44,490 work, but he has this framework that I think is really useful for a lot of writers 581 00:39:44,490 --> 00:39:47,910 that you can kind of distill and turn into a scale. 582 00:39:48,050 --> 00:39:51,790 And I've done that with several different writers that I admire. 583 00:39:51,970 --> 00:39:53,210 So another one I have is Hitchcock, 584 00:39:54,390 --> 00:39:58,730 which is based on Hitchcock's principle of suspense. 585 00:39:58,970 --> 00:40:04,810 So he has this famous quote about a bomb under the table where like if there's a 586 00:40:04,810 --> 00:40:08,330 bomb under the table and it explodes, that's a one moment of surprise. 587 00:40:08,650 --> 00:40:12,850 But if you have people sitting around the table and you show the audience the bomb 588 00:40:12,850 --> 00:40:17,210 under the table and say this bomb will go off in five minutes, now you have five 589 00:40:17,210 --> 00:40:19,290 minutes of suspense. 590 00:40:19,530 --> 00:40:24,330 And so I turned that into a skill that assesses, okay, does this, does this make the 591 00:40:24,330 --> 00:40:25,090 reader lean in? 592 00:40:25,250 --> 00:40:30,590 Does this make the reader like anxious and curious to see what will happen next? 593 00:40:30,630 --> 00:40:33,930 And that becomes another lens that you can apply to, to writing. 594 00:40:34,330 --> 00:40:38,850 And so I get a lot of value out of that as an experienced writer even because like 595 00:40:38,850 --> 00:40:43,770 I'm not always able to like consider all of those different lenses but now I can 596 00:40:43,770 --> 00:40:49,810 just sort of like invoke Vonnegut, invoke Hitchcock, invoke Sorkin, invoke Sedaris 597 00:40:49,810 --> 00:40:56,310 for humor and get those kinds of perspectives on my writing that are 598 00:40:56,310 --> 00:41:01,610 informed by these really brilliant storytellers that I admire. 599 00:41:02,450 --> 00:41:03,610 It's such a gift. 600 00:41:03,710 --> 00:41:05,670 It's so cool that you've done this. 601 00:41:06,290 --> 00:41:12,370 I mean, I could totally see how, you know, presented with the opinions of Sorkin and 602 00:41:12,370 --> 00:41:13,850 Zadaris and Vonnegut. 603 00:41:14,630 --> 00:41:16,770 Someone might say, oh, I really like Vonnegut. 604 00:41:16,910 --> 00:41:19,130 Like, that's a framework that really resonates with me. 605 00:41:19,170 --> 00:41:23,010 And maybe that is closer to the writing style that I aspire to have and might 606 00:41:23,010 --> 00:41:23,630 pursue. 607 00:41:24,250 --> 00:41:25,610 Or equally helpful. 608 00:41:26,470 --> 00:41:31,170 I don't like the Hitchcock, you know, sort of like method of writing or sort of like 609 00:41:31,170 --> 00:41:32,010 this leaning in. 610 00:41:32,190 --> 00:41:35,850 And that's what actually would enable you to develop your own taste, which I think 611 00:41:35,850 --> 00:41:40,170 is really what any ambitious AI user wants to do, right? 612 00:41:40,270 --> 00:41:43,870 It's like we want to do things that we have no business doing, right? 613 00:41:44,030 --> 00:41:48,050 Building products that we have, that we wouldn't have been able to build, you know, 614 00:41:48,190 --> 00:41:51,270 10 years ago or even five years ago, right? 615 00:41:51,970 --> 00:41:54,010 In compelling ways that feel 616 00:41:54,850 --> 00:42:00,970 authentic and and of a caliber that like we would be proud of but it's actually 617 00:42:00,970 --> 00:42:05,550 really difficult to do high -quality work and it takes a lot of effort there's 618 00:42:05,550 --> 00:42:12,450 really kind of no shortcut but with a plug -in like the Compound Writing plug -in it 619 00:42:12,450 --> 00:42:16,430 just gives you something to react to it gives you so much more to react to that you 620 00:42:16,430 --> 00:42:21,490 can either agree or disagree with and that can make you a better writer so much 621 00:42:21,490 --> 00:42:21,970 faster. 622 00:42:22,270 --> 00:42:24,790 So I'm personally so excited to be using that tool. 623 00:42:25,410 --> 00:42:27,390 Thank you for making the plugin. 624 00:42:27,810 --> 00:42:30,050 It is a gift, especially coming from you. 625 00:42:31,250 --> 00:42:36,410 And I think it really informs this new chapter that we're in AI, which is, you know, 626 00:42:36,510 --> 00:42:42,410 we're getting expertise that is codified that allows more people to do more great 627 00:42:42,410 --> 00:42:44,410 things, which is just so exciting. 628 00:42:45,110 --> 00:42:48,890 Yeah, and I'm benefiting from that on my end. 629 00:42:49,190 --> 00:42:52,810 You know, like I also use compound engineering to do things that I have no business 630 00:42:52,810 --> 00:42:53,250 doing. 631 00:42:53,870 --> 00:42:58,470 Sometimes that gets me in trouble, but then AI gets me out of trouble as well. 632 00:42:59,570 --> 00:43:04,390 So that's like, yeah, just like I wrote my like my one of my most recent pieces for 633 00:43:04,390 --> 00:43:08,090 working overtime was about how I accidentally coded a security risk into a product 634 00:43:08,090 --> 00:43:09,870 that I had built. 635 00:43:10,010 --> 00:43:14,930 I had built an MCP into this app I made called Tastemaker that actually lets you 636 00:43:14,930 --> 00:43:20,290 pull clips from different writing you admire and it distills that into a style 637 00:43:20,290 --> 00:43:20,650 guide. 638 00:43:20,830 --> 00:43:24,950 It's like okay this is like this is what you admire and appreciate in writing that 639 00:43:24,950 --> 00:43:29,110 you might want to inform your own pieces and I built an MCP because I was like I 640 00:43:29,110 --> 00:43:34,390 want to be able to like add new clips and like pull my style guide down without 641 00:43:34,390 --> 00:43:39,010 actually having to go to the app just while I'm in Codex and like or Claude and like 642 00:43:39,010 --> 00:43:43,610 I built it and I don't think I had the most up -to -date compound engineering plugin 643 00:43:43,610 --> 00:43:49,390 that had more robust security features and Claude was just like I'll build that for 644 00:43:49,390 --> 00:43:54,590 you no problem and then like I asked Codex to review it and Codex was like there are 645 00:43:54,590 --> 00:44:00,290 five security vulnerabilities here so AI got me in the trouble and then AI got me 646 00:44:00,290 --> 00:44:04,930 out of the trouble, but like just the fact that I'm here like building these kinds 647 00:44:04,930 --> 00:44:09,570 of products, like I built my website, like that's crazy. 648 00:44:09,790 --> 00:44:13,530 Like it's crazy that I'm able to do that, but all I have to do is like borrow 649 00:44:13,530 --> 00:44:17,190 Kieran's brain and I can do it. 650 00:44:19,020 --> 00:44:22,900 Well, I borrow Kieran's brain and I'm excited to be borrowing your brain. 651 00:44:23,160 --> 00:44:23,660 So thank you. 652 00:44:23,740 --> 00:44:25,440 Thank you for putting it out there. 653 00:44:26,500 --> 00:44:31,580 Katie, you've been on this journey for really over two years now, you know, all the 654 00:44:31,580 --> 00:44:37,620 way from starting to use Chat DPT as a career coach to now having a 655 00:44:37,620 --> 00:44:44,480 codecs, not just helping you run your work day to day, but now having built your own 656 00:44:44,480 --> 00:44:45,940 custom writing plugin. 657 00:44:46,380 --> 00:44:49,220 It's such an incredible journey that you've been on. 658 00:44:49,840 --> 00:44:55,560 And in a few months, you'll be speaking at the every conference thesis on November 659 00:44:55,560 --> 00:45:01,160 fifth at Pioneer Works in front of so many executives and leaders in the AI space. 660 00:45:01,780 --> 00:45:04,200 What is your thesis on AI today? 661 00:45:04,920 --> 00:45:11,600 My thesis on AI today is that education and access are going to 662 00:45:11,600 --> 00:45:14,580 matter more than they ever have before. 663 00:45:15,560 --> 00:45:20,160 So something I feel very strongly about is the distribution of opportunity. 664 00:45:20,160 --> 00:45:23,860 I am based in Columbus, Ohio. 665 00:45:24,520 --> 00:45:25,740 I grew up in the Midwest. 666 00:45:27,860 --> 00:45:33,080 I haven't really spent time in New York or SF or Austin and all these big tech hubs. 667 00:45:33,620 --> 00:45:36,960 And I believe, but I believe really strongly that this technology should be in the 668 00:45:36,960 --> 00:45:38,600 hands of as many people as possible. 669 00:45:39,140 --> 00:45:42,940 But in order for that to happen, you have to create the right conditions. 670 00:45:43,120 --> 00:45:47,580 So I'm very fortunate in the fact that I have connections. 671 00:45:47,580 --> 00:45:52,680 Like, I have access to really great thinkers, and like, I can learn from Kieran, I 672 00:45:52,680 --> 00:45:54,420 can learn from Dan, I can learn from you. 673 00:45:55,320 --> 00:46:01,580 But all of that has been facilitated by the fact that I have a certain amount of 674 00:46:01,580 --> 00:46:02,680 economic comfort. 675 00:46:02,960 --> 00:46:08,980 I have a certain amount of free time and budget to experiment with these tools and 676 00:46:08,980 --> 00:46:13,040 like get a subscription and vibe code until 2 a .m. 677 00:46:13,040 --> 00:46:14,480 if that's what's interesting to me. 678 00:46:14,480 --> 00:46:20,840 And I think that what I want to challenge the AI community with is thinking about 679 00:46:20,840 --> 00:46:24,960 how we can expand those kinds of opportunities to more people and more kinds of 680 00:46:24,960 --> 00:46:26,460 people in more different places. 681 00:46:26,800 --> 00:46:32,920 Because AI does have this huge multiplicative capacity to end this compounding 682 00:46:32,920 --> 00:46:33,920 capacity. 683 00:46:34,340 --> 00:46:40,180 But the risk there is that value is going to compound to the small subset of people 684 00:46:40,180 --> 00:46:41,320 that happen to be early. 685 00:46:41,320 --> 00:46:46,320 And I just think that a much more compelling vision for an AI future is one where 686 00:46:46,320 --> 00:46:48,300 everyone is able to come along for the ride. 687 00:46:48,980 --> 00:46:51,900 I aspire to that version of the future. 688 00:46:52,200 --> 00:46:53,120 Thank you so much, Katie. 689 00:46:53,480 --> 00:46:55,640 It's been such a treat to connect today. 690 00:46:55,860 --> 00:47:00,320 So excited to have the public release, to see the public release of the compound 691 00:47:00,320 --> 00:47:01,000 writing plugin. 692 00:47:01,560 --> 00:47:04,080 And thank you for your time. 693 00:47:04,460 --> 00:47:05,400 Thanks for having me. 694 00:47:05,480 --> 00:47:05,900 This was fun. 695 00:47:19,290 --> 00:47:23,930 eye and I why because this show is the epitome of awesomeness it's like finding a 696 00:47:23,930 --> 00:47:27,790 treasure chest in your backyard but instead of gold it's filled with pure 697 00:47:27,790 --> 00:47:33,450 unadulterated knowledge bombs about chat GPT every episode is a roller coaster of 698 00:47:33,450 --> 00:47:38,330 emotions insights and laughter that will leave you on the edge of your seat craving 699 00:47:38,330 --> 00:47:43,170 for more it's not just a show it's a journey into the future with Dan Shipper as the 700 00:47:43,170 --> 00:47:49,770 captain of the spaceship so do yourself a favor hit Like, smash subscribe, and strap 701 00:47:49,770 --> 00:47:50,750 in for the ride of your life. 702 00:47:51,210 --> 00:47:56,090 And now, without any further ado, let me just say Dan, I'm absolutely, hopelessly in 703 00:47:56,090 --> 00:47:56,550 love with you.