Is AI killing craftsmanship?

00:00:00,600 --> 00:00:39,760 [Eric] [upbeat music] Welcome back to the Token Intelligence Show. AI is changing the way we work, and here on the show, John and I try to show you the state of the art, cut through all the noise, and hopefully help you become a leader by acting with wisdom in this new era that we live in. So today, we are going to talk about a subject, two subjects that are near and dear to both of our hearts, and those are craftsmanship and burnout. 00:00:39,760 --> 00:00:40,380 [John] Yeah. 00:00:40,380 --> 00:00:50,880 [Eric] And we're going to start out actually with a post that I found on X that really caught my eye because it comes from a very senior engineer- 00:00:50,880 --> 00:00:51,269 [John] Yeah 00:00:51,269 --> 00:00:56,500 [Eric] ... at a very reputable company, and I'm just gonna read it to start out. 00:00:56,500 --> 00:00:57,920 [John] Let's do it. 00:00:57,920 --> 00:01:11,850 [Eric] So this is, uh, Dylan Mulroy, and we'll link to it in the show notes. He says, "Y'all, I'm not gonna lie, it's way harder to get joy and satisfaction out of building with AI than it was before. 00:01:13,020 --> 00:01:33,970 [Eric] Constantly straddling burnout, being far less immersed in hard problems, constant context switching. I'm tired of the uncertainty of where this is going and how to do it well." This struck me because the way that it's written, you can sense his frustration, right? 00:01:33,970 --> 00:01:34,200 [John] Right. 00:01:34,200 --> 00:01:40,440 [Eric] It almost like, almost like he crafted it, you know, at the end of a long day in a moment of weakness. 00:01:40,440 --> 00:01:40,600 [John] Right. 00:01:41,700 --> 00:01:43,360 [John] Yeah, that's interesting framing. 00:01:43,360 --> 00:01:43,660 [Eric] Mm-hmm. 00:01:43,660 --> 00:01:45,700 [John] I'm not, I'm not gonna lie. 00:01:45,700 --> 00:01:45,760 [Eric] [laughs] 00:01:45,760 --> 00:01:47,590 [John] I had to Google NGL. 00:01:47,590 --> 00:01:47,620 [Eric] Oh, really, NGL? 00:01:47,620 --> 00:01:50,960 [John] Like, I read it a couple times and I was like, "What is NGL again?" 00:01:50,960 --> 00:01:51,580 [Eric] Yeah. Not gonna lie. 00:01:51,580 --> 00:01:54,780 [John] Yeah. Maybe, maybe I need to spend more time... No. 00:01:54,780 --> 00:01:56,230 [Eric] More time [laughs] 00:01:56,230 --> 00:01:56,230 [John] I don't need to spend more time on X. [laughs] 00:01:56,230 --> 00:01:59,010 [Eric] [laughs] I don't know if you need to spend more time on X. 00:01:59,010 --> 00:01:59,080 [John] But- 00:01:59,080 --> 00:02:00,600 [Eric] I spend plenty of time on it for work. 00:02:00,600 --> 00:02:02,220 [John] Yeah. Yeah 00:02:02,220 --> 00:02:17,160 [Eric] But w- tell me about, first of all, do you think that this is becoming more common, this burnout or people who are very, very heavy users of AI stepping back and saying, 00:02:18,400 --> 00:02:20,680 [Eric] "I don't really know if I like this"? 00:02:21,760 --> 00:02:29,120 [John] Yeah, I think so. I have, I have, I just have this visual, like I, I think, I think in visuals sometimes. And I'm imagining, you're a car guy, right? 00:02:29,120 --> 00:02:29,720 [Eric] Mm-hmm. 00:02:29,720 --> 00:02:37,079 [John] So I'm imagining, think about if you had to drive 20 minutes in an old car that really pulled right. 00:02:37,080 --> 00:02:37,380 [Eric] Mm-hmm. 00:02:38,460 --> 00:02:38,880 [John] Like- 00:02:38,880 --> 00:02:38,980 [Eric] Yeah 00:02:38,980 --> 00:02:39,859 [John] ... kind of annoying. 00:02:39,860 --> 00:02:40,299 [Eric] Totally. 00:02:40,300 --> 00:02:46,240 [John] Think about if you're a semi driver in a semi and you had to drive all day in a semi that pulled right. 00:02:46,240 --> 00:02:47,780 [Eric] Yeah. 00:02:47,780 --> 00:02:49,380 [John] That's like my, that's my visual- 00:02:49,380 --> 00:02:49,549 [Eric] Yeah 00:02:49,549 --> 00:02:50,960 [John] ... for driving AI. 00:02:50,960 --> 00:02:51,440 [Eric] Yeah. 00:02:51,440 --> 00:02:55,260 [John] Where it's constantly pulling sometimes left, [laughs] sometimes right. 00:02:55,260 --> 00:02:55,620 [Eric] Yeah. 00:02:55,620 --> 00:03:01,220 [John] And you're constantly pulling it back in the right direction for the thing you want it to do. 00:03:01,220 --> 00:03:01,260 [Eric] Yeah. 00:03:01,260 --> 00:03:05,500 [John] And I think there's something extreme... I- if you're using AI that way, there's something really draining about that. 00:03:05,500 --> 00:03:08,820 [Eric] Okay, great. I have a great story to end on that's a great analogy about that. 00:03:08,820 --> 00:03:08,940 [John] Perfect. 00:03:08,940 --> 00:03:10,500 [Eric] An actual experience like this- 00:03:10,500 --> 00:03:11,260 [John] I figured, yeah 00:03:11,260 --> 00:03:12,520 [Eric] ... from a, from a real car. 00:03:12,520 --> 00:03:13,280 [John] Yeah. 00:03:13,280 --> 00:03:33,070 [Eric] So let's zoom out first, though, because w- we wanna speak to a couple peopl- people here. First, we wanna speak to people like you and I, who use AI very heavily all day long, and who may empathize with the sentiment that Dylan, you know, that Dylan expressed in his, in his X post, right? 00:03:33,070 --> 00:03:33,600 [John] Yep. 00:03:33,600 --> 00:03:45,640 [Eric] Like, "I'm, I'm in this. I'm using it. I'm an advanced user, and I'm... I have questions," right? "I don't know if I like this type of work more than, more than my previous-" 00:03:45,640 --> 00:03:45,710 [John] Mm-hmm 00:03:45,710 --> 00:03:54,859 [Eric] "... you know, mode of work." And then I think there are also a lot of people who have not hit that point yet, and maybe are even in the honeymoon phase of- 00:03:54,860 --> 00:03:55,020 [John] Yeah 00:03:55,020 --> 00:04:07,670 [Eric] ... "Holy cow, this is amazing. This is gonna transform everything. This is so helpful." And then there's probably a, m- you know, people who have used it and maybe they haven't quite seen the power yet. And so I think- 00:04:07,670 --> 00:04:07,670 [John] Right 00:04:07,670 --> 00:04:09,549 [Eric] ... I, I'd love to speak to both of those people 00:04:10,700 --> 00:04:15,520 [Eric] because the f- you and I have been sort of on the edge of where Dylan is. 00:04:15,520 --> 00:04:15,900 [John] Right. 00:04:15,900 --> 00:04:24,400 [Eric] Uh, but then also, I think that it, that people will increasingly experience this as they get deeper and deeper into AI. 00:04:24,400 --> 00:04:24,760 [John] Yeah. 00:04:24,760 --> 00:04:28,060 [Eric] So before we speak to those two types of listeners, 00:04:29,400 --> 00:04:40,530 [Eric] I'd like to zoom out and talk about first principles in terms of craftsmanship and tools. This is actually a topic that goes way, way back to- 00:04:40,530 --> 00:04:40,530 [John] Yeah 00:04:40,530 --> 00:04:42,099 [Eric] ... our very first episode- 00:04:42,100 --> 00:04:42,840 [John] That's right 00:04:42,840 --> 00:04:44,760 [Eric] ... s- uh, s- about six months ago. 00:04:44,760 --> 00:04:45,460 [John] Yep. 00:04:45,460 --> 00:04:47,920 [Eric] Um, almost exactly six months ago, actually- 00:04:47,920 --> 00:04:47,930 [John] Yeah 00:04:47,930 --> 00:04:48,599 [Eric] ... which is crazy. 00:04:48,600 --> 00:04:49,460 [John] Yeah. 00:04:49,460 --> 00:04:49,680 [Eric] Uh, 00:04:50,740 --> 00:05:05,300 [Eric] and it's a s- topic of craftsmanship, right? And so not seeking, um, you know, career advancement or, you know, political advancement in whatever your profession is, but just focusing on being a craftsman. 00:05:05,300 --> 00:05:05,720 [John] Right. 00:05:05,720 --> 00:05:12,979 [Eric] And that is intrinsically rewarding and results in, you know, s- inadvertently joining a community of other people- 00:05:12,980 --> 00:05:13,060 [John] Mm-hmm 00:05:13,060 --> 00:05:14,850 [Eric] ... who share the same value of becoming- 00:05:14,850 --> 00:05:14,850 [John] Right 00:05:14,850 --> 00:05:15,719 [Eric] ... a craftsman. 00:05:15,720 --> 00:05:16,220 [John] Right. 00:05:16,220 --> 00:05:24,280 [Eric] And we've talked on the show about how ultimately that is sort of the most durable form of career advancement, is just becoming very, very good at what you do. 00:05:24,280 --> 00:05:24,460 [John] Right. Right. 00:05:24,460 --> 00:05:39,220 [Eric] Um, and you'll naturally progress in your career. Uh, but what's interesting, I think about Dylan, who is an extremely accomplished software engineer, um, you know, senior level, works on really difficult stuff, 00:05:40,460 --> 00:05:42,550 [Eric] and is a craftsman. 00:05:42,550 --> 00:05:42,840 [John] Mm-hmm. 00:05:42,840 --> 00:05:57,040 [Eric] Right? And what's interesting is that he's using a completely new tool set now f- to do his job, right? B- like, dramatically different than even just a couple of years ago, and so the way that the job is done 00:05:58,140 --> 00:06:13,096 [Eric] has changed. And so I wanted to ask you What, what is crafts-- like, explain craftsmanship in, through that lens of the tools dramatically changing the way that, that something is done. 00:06:14,556 --> 00:06:26,116 [John] Yeah. So specifically I think for, for a senior engineer, like in this case, the, the tools have shifted really dramatically, but maybe even bigger than that. 00:06:26,116 --> 00:06:26,136 [Eric] Mm. 00:06:26,136 --> 00:06:29,395 [John] Let's say that this individual wanted to work for another company. 00:06:29,396 --> 00:06:30,116 [Eric] Mm-hmm. 00:06:30,116 --> 00:06:34,796 [John] Like, even the typical process of you would do like a leak code interview, for example. 00:06:34,796 --> 00:06:35,416 [Eric] Mm. 00:06:35,416 --> 00:06:39,656 [John] Like, there's a lot of questions around that, like which components of that are valuable versus not. 00:06:39,656 --> 00:06:39,816 [Eric] Mm. 00:06:39,816 --> 00:06:45,376 [John] So there's kind of this ... It's not just a day-to-day thing. There's also like an identity thing here- 00:06:45,376 --> 00:06:45,386 [Eric] Mm 00:06:45,386 --> 00:06:51,176 [John] ... where like, I've worked for these companies and I got to work for these companies because I could prove that I could solve these hard problems. 00:06:51,176 --> 00:06:51,576 [Eric] Right. 00:06:51,576 --> 00:07:01,416 [John] I'm not sure my value when at least a growing subset of the hard problems, somebody, uh, AI can solve- 00:07:01,416 --> 00:07:01,496 [Eric] Mm-hmm 00:07:01,496 --> 00:07:04,056 [John] ... as far as think through and apply in the individual- 00:07:04,056 --> 00:07:04,316 [Eric] Mm-hmm 00:07:04,316 --> 00:07:08,375 [John] ... you know, circumstance. So I think that's probably the core. 00:07:08,376 --> 00:07:08,976 [Eric] Mm. 00:07:08,976 --> 00:07:21,396 [John] The core of it for a lot of people. 'Cause you're working in a different way, yes, but you had a, a bit of identity wrapped in, like, I understand this hard problem, I know how to solve it, I can flex it for different companies. 00:07:21,396 --> 00:07:21,786 [Eric] Yep. 00:07:21,786 --> 00:07:27,856 [John] And there's still value in there, like the thinking part that, that got you to be able to learn how to do... Like, there's still value there. 00:07:27,856 --> 00:07:28,336 [Eric] Yep. 00:07:28,336 --> 00:07:31,096 [John] But measuring the value actually just got a lot harder. 00:07:31,096 --> 00:07:31,296 [Eric] Mm. 00:07:31,296 --> 00:07:47,236 [John] If you gave me a room full of n- engineers, and we kind of had a standard, and I say we, um, at least like the Metas and Googles and, and companies like that of the world kind of had this standard of, of how you get these jobs and, and you know, what things you need to know and- 00:07:47,236 --> 00:07:47,516 [Eric] Yep 00:07:47,516 --> 00:07:50,986 [John] ... all the things, and, and that's a little bit broken right now. 00:07:50,986 --> 00:07:50,996 [Eric] Mm. 00:07:50,996 --> 00:07:58,816 [John] And I think it's affecting people's identities a bit in that space. Identities in, in that like, am I any good at this? 00:07:58,816 --> 00:07:59,416 [Eric] Mm. 00:07:59,416 --> 00:07:59,936 [John] Right? 00:07:59,936 --> 00:08:04,456 [Eric] Yeah. Yeah, that's super interesting. Uh, I think that the... 00:08:05,936 --> 00:08:07,846 [Eric] If I think about craftsmanship, 00:08:08,936 --> 00:08:12,455 [Eric] I agree that there is a... 00:08:14,516 --> 00:08:21,476 [Eric] I, I think about it as a joy in going through the difficult process- 00:08:21,476 --> 00:08:21,486 [John] Mm 00:08:21,486 --> 00:08:24,176 [Eric] ... of crafting something great, of making- 00:08:24,176 --> 00:08:24,186 [John] Right 00:08:24,186 --> 00:08:26,356 [Eric] ... something great. To your point, standards- 00:08:26,356 --> 00:08:26,536 [John] Right 00:08:27,716 --> 00:08:28,276 [Eric] ... right? And 00:08:29,356 --> 00:08:33,856 [Eric] the measurement for standards has changed in software engineering, um, 00:08:35,576 --> 00:08:40,596 [Eric] but tool sets have changed historically as well, right? 00:08:40,596 --> 00:08:41,016 [John] Yeah. For sure. 00:08:41,016 --> 00:08:45,656 [Eric] So there's a great example. I asked a guy, an older software engineer, 00:08:46,816 --> 00:09:00,816 [Eric] um... Actually, we were, I was at church, and we just, you know, the service ended, and, you know, you just meet people who you've seen before but maybe haven't talked to, and we happened to be sitting very close, and we started chatting. He asked what I did. You know, I said I work at Vercel. 00:09:00,816 --> 00:09:00,826 [John] Mm-hmm. 00:09:00,826 --> 00:09:08,216 [Eric] He was like, "Oh, I'm in software engineering. I'm familiar with Vercel," et cetera. And he's probably 15 or 20 years older than me. 00:09:08,216 --> 00:09:09,276 [John] Okay. Wow. 00:09:09,276 --> 00:09:11,676 [Eric] Um, and has been a programmer for a very, very long time. 00:09:11,736 --> 00:09:11,946 [John] Right. 00:09:11,946 --> 00:09:28,926 [Eric] And so I was very interested in his perspective. I was like, you know, "You've, a majority of y- a vast majority of your career has been experience without AI. You know, do you use AI in software engineering," et cetera. And he just had a very simple answer, which I love. 00:09:28,926 --> 00:09:28,976 [John] Yeah. 00:09:28,976 --> 00:09:31,696 [Eric] I love sort of what I call like old man wisdom. 00:09:31,696 --> 00:09:32,176 [John] Okay. 00:09:32,176 --> 00:09:36,076 [Eric] He said, "Yeah, I mean, for software engineers, it's kind of like a carpenter. 00:09:37,516 --> 00:09:49,456 [Eric] You know, when the electric circular saw came out, like you could theoretically choose to do your job without it, but no one's going to compete in the marketplace- 00:09:49,456 --> 00:09:49,716 [John] Right. Right 00:09:49,716 --> 00:09:52,396 [Eric] ... if you're not using this really powerful- 00:09:52,396 --> 00:09:52,406 [John] Yeah 00:09:52,406 --> 00:09:53,436 [Eric] ... tool to like do part- 00:09:53,436 --> 00:09:53,446 [John] Right 00:09:53,446 --> 00:09:54,266 [Eric] ... of your job, right?" 00:09:54,266 --> 00:09:54,356 [John] Right. 00:09:54,356 --> 00:09:54,656 [Eric] Um, 00:09:56,136 --> 00:09:59,796 [Eric] and it's not like the other previous tools become completely irrelevant, but there's- 00:09:59,796 --> 00:09:59,805 [John] Right 00:09:59,805 --> 00:10:12,756 [Eric] ... just sort of a new primary way of doing certain things that before were very repetitive and, and the whole process speeds up and, and, you know, can become more precise, and the analogy breaks down at some point with a circular saw. [chuckles] 00:10:12,756 --> 00:10:13,116 [John] Right. 00:10:13,116 --> 00:10:18,956 [Eric] But I thought that was really interesting because the, um... 00:10:20,076 --> 00:10:28,516 [Eric] I think we can over-rotate on the joy of using the tools- 00:10:28,516 --> 00:10:28,916 [John] Right 00:10:28,916 --> 00:10:37,826 [Eric] ... and of letting go of a lot of investment put into getting really good at using a handsaw or a planer- 00:10:37,826 --> 00:10:37,826 [John] Right 00:10:37,826 --> 00:10:39,936 [Eric] ... you know, or other things like that. Um, 00:10:41,236 --> 00:10:41,666 [Eric] and 00:10:42,816 --> 00:10:50,116 [Eric] you know, you put a, you, you... And that is, for a craftsman, part of it is like I l- I mean, people like tools, you know? That could be digital- 00:10:50,116 --> 00:10:50,126 [John] Well- 00:10:50,126 --> 00:10:51,216 [Eric] ... it could be physical. 00:10:51,216 --> 00:10:59,136 [John] And it's not just that. I mean, as a craftsman, people fall in love with some really minute details about tools. 00:10:59,136 --> 00:10:59,656 [Eric] Yes. 00:10:59,656 --> 00:11:04,196 [John] Like, I love this tool because of how it's weighted, like where- 00:11:04,196 --> 00:11:04,206 [Eric] Mm 00:11:04,206 --> 00:11:05,376 [John] ... the balance point is. [chuckles] 00:11:05,376 --> 00:11:05,856 [Eric] Mm-hmm. 00:11:05,856 --> 00:11:10,496 [John] Like, they get into the steel that was made. You know, and people get really deep. 00:11:10,496 --> 00:11:10,966 [Eric] Oh, absolutely. 00:11:10,966 --> 00:11:15,976 [John] And when you completely upend like all the tools they used to use, like I think it's pretty disconcerting. 00:11:15,976 --> 00:11:22,695 [Eric] It's disconcerting, for sure. And I think that there's part of that that's worth lamenting- 00:11:22,696 --> 00:11:23,156 [John] Right 00:11:23,156 --> 00:11:29,556 [Eric] ... because tools are great. You know? There's this, uh, th- there's a another example. There's a, um... 00:11:30,716 --> 00:11:35,396 [Eric] I like to work on cars, I like to work on, um, mountain bikes, and 00:11:36,556 --> 00:11:41,596 [Eric] there's this tool company that makes tools for working on bicycles, and- 00:11:41,596 --> 00:11:42,886 [John] Yeah, you told me about this before, I think 00:11:42,886 --> 00:11:43,936 [Eric] ... Abbey. Yes. 00:11:43,936 --> 00:11:44,056 [John] Yeah. 00:11:44,056 --> 00:11:45,416 [Eric] Abbey, Abbey Bike Tools. 00:11:46,496 --> 00:11:52,756 [Eric] And they're, it's such a great company. Their tagline is, "Precision is our religion." You know, it's Abbey- 00:11:52,756 --> 00:11:52,796 [John] Okay 00:11:52,796 --> 00:11:54,686 [Eric] ... so they have like a monk, "Precision is our religion." 00:11:54,686 --> 00:11:55,716 [John] Hmm. Okay. 00:11:55,716 --> 00:12:01,536 [Eric] And they don't really... It, it's, they just make the best tool possible, and it costs whatever it costs for them- 00:12:01,536 --> 00:12:01,596 [John] Right 00:12:01,596 --> 00:12:02,456 [Eric] ... to get their margin. 00:12:02,456 --> 00:12:02,695 [John] Yeah, yeah. 00:12:02,696 --> 00:12:02,955 [Eric] You know? 00:12:02,956 --> 00:12:03,316 [John] Right. 00:12:03,316 --> 00:12:05,936 [Eric] And so like some of these things are like outrageously- 00:12:05,936 --> 00:12:06,436 [John] Yeah 00:12:06,436 --> 00:12:07,196 [Eric] ... expensive, but- 00:12:07,196 --> 00:12:07,306 [John] Right 00:12:07,306 --> 00:12:16,812 [Eric] ... they literally don't care. And so I remember one year Um, my grandfather gave, gave me a little cash, you know, for Christmas. 00:12:16,812 --> 00:12:17,732 [John] Mm-hmm. 00:12:17,732 --> 00:12:24,032 [Eric] And I decided I was gonna spend 150 bucks on this, like, pretty basic Allen- 00:12:24,032 --> 00:12:25,652 [John] Like tool set or- 00:12:25,652 --> 00:12:26,252 [Eric] No. It's like- 00:12:26,252 --> 00:12:26,802 [John] Just one tool 00:12:26,802 --> 00:12:27,152 [Eric] ... one tool with, like- 00:12:27,152 --> 00:12:27,162 [John] [laughs] 00:12:27,162 --> 00:12:28,382 [Eric] ... a couple of bits, you know? 00:12:28,382 --> 00:12:28,752 [John] Nice. Yeah. 00:12:28,752 --> 00:12:30,472 [Eric] But you mention weight. 00:12:30,472 --> 00:12:30,972 [John] Mm-hmm. 00:12:30,972 --> 00:12:32,582 [Eric] And the, 00:12:33,632 --> 00:12:37,282 [Eric] it, that was immediately the first thing that I noticed, is like- 00:12:37,282 --> 00:12:37,282 [John] Okay 00:12:37,282 --> 00:12:47,692 [Eric] ... it had a very opinionated weight of the tool, and then, like, balance. And so, like, it just felt so different to use it to tighten and, like- 00:12:47,692 --> 00:12:47,722 [John] Mm-hmm 00:12:47,722 --> 00:12:51,632 [Eric] ... the precision level and everything. And it's a heavy tool, like intentionally by design. 00:12:51,632 --> 00:12:52,212 [John] Right. 00:12:52,212 --> 00:12:54,032 [Eric] Um, and 00:12:56,192 --> 00:12:58,872 [Eric] so if I think about, I love using that tool, right? 00:12:58,872 --> 00:12:58,952 [John] Yeah. 00:12:58,952 --> 00:12:59,372 [Eric] Like- 00:12:59,372 --> 00:12:59,382 [John] Yeah 00:12:59,382 --> 00:13:07,172 [Eric] ... I mean, it's almost like, you know, d- having a couple tools like that, it's like, okay, if you have a chance to, like, do something on your bike or fix something, it's like great 'cause you get to use the tool. 00:13:07,172 --> 00:13:07,472 [John] Yeah. 00:13:07,472 --> 00:13:12,382 [Eric] And so I think it's legitimate to lament, like, not using these tools, right? 00:13:12,382 --> 00:13:12,382 [John] Right. 00:13:12,382 --> 00:13:29,432 [Eric] 'Cause the people who created those tools themselves are craftsmen, and there's just sort of an ethos around that, right? And, um, when I think about... Uh, so there's one side of that, right? But then the other side is, well, really, like, tools are a means to an end. 00:13:29,432 --> 00:13:29,652 [John] Yeah. 00:13:29,652 --> 00:13:30,472 [Eric] You're trying to do something. 00:13:30,472 --> 00:13:30,482 [John] Yeah. 00:13:30,482 --> 00:13:41,022 [Eric] You're trying to build something. You're trying to fix something, right? And that's sort of ultimately what it means to be a craftsman, but I absolutely think that there is, and I think a great analogy is actually weight. Like 00:13:42,472 --> 00:13:44,402 [Eric] using AI to 00:13:45,672 --> 00:14:00,552 [Eric] write code, to write w- you know, actual words or copy or prose or other things like that, it makes the t- it makes, like, the physical experience lighter. You know, it's not like holding a heavy tool. It's actually like- 00:14:00,552 --> 00:14:00,562 [John] Yeah 00:14:00,562 --> 00:14:03,072 [Eric] ... you know, it, there's less, like, tactile sensation. 00:14:03,072 --> 00:14:03,222 [John] Yeah. 00:14:03,222 --> 00:14:07,972 [Eric] And th- I think that's a hard adjustment, legi- you know, and for, for good reasons. 00:14:07,972 --> 00:14:08,652 [John] Right. 00:14:08,652 --> 00:14:22,432 [Eric] For good reasons. So let's talk to the person who... Let's talk to Dylan, actually. Uh, you know, he's probably way more equipped to [laughs] to speak on this subject than us and give us advice. 00:14:22,432 --> 00:14:22,512 [John] Yeah. 00:14:22,512 --> 00:14:23,692 [Eric] But let's try to give Dylan advice 00:14:24,852 --> 00:14:28,312 [Eric] because it seems like he's feeling a lot of those things that we talked about. 00:14:28,312 --> 00:14:28,512 [John] Yeah. 00:14:28,512 --> 00:14:29,032 [Eric] Right? 00:14:29,032 --> 00:14:29,052 [John] Yeah. 00:14:29,052 --> 00:14:40,912 [Eric] The craftsmanship, the tool set changing, and some of the joy that he previously experienced, you know, sort of slipping away. So what would you say to Dylan if he were here with us? 00:14:40,912 --> 00:14:47,872 [John] So I actually would say I've, I've experienced all these same things, but it happened about 10 years ago. 00:14:47,872 --> 00:14:48,612 [Eric] Hmm. 00:14:48,612 --> 00:14:50,432 [John] It was when I first became a manager. 00:14:50,432 --> 00:14:50,752 [Eric] Hmm. 00:14:50,752 --> 00:14:52,391 [John] It's so similar. 00:14:52,392 --> 00:14:52,641 [Eric] Wow. 00:14:52,641 --> 00:14:53,592 [John] It's so similar. 00:14:53,592 --> 00:14:53,962 [Eric] That's a great analogy. Yeah. 00:14:53,962 --> 00:14:58,452 [John] 'Cause you're, 'cause you're delegating work, at that time, to another person- 00:14:58,452 --> 00:14:59,192 [Eric] Mm-hmm 00:14:59,192 --> 00:15:05,912 [John] ... who might not have a full understanding of the context that you understand, might not have a full, like, mastery of the same tools that you have. 00:15:05,912 --> 00:15:06,312 [Eric] Mm-hmm. 00:15:06,312 --> 00:15:18,922 [John] And there's such a sharp analogy there, and I think especially, um, for engineers that are at this place right now where they're having to delegate work to this other entity that it's really hard to get your mind around exactly what- 00:15:18,922 --> 00:15:18,922 [Eric] Hmm 00:15:18,922 --> 00:15:21,672 [John] ... does it understand right now, what does it need to underst- all the things. 00:15:22,692 --> 00:15:24,072 [John] It's so similar to that. 00:15:24,072 --> 00:15:26,232 [Eric] Yeah. Yeah. Super interesting. 00:15:27,292 --> 00:15:37,532 [Eric] What... Have you experienced this burnout, like burnout like Dylan has experienced? I mean, he seems to sort of be in a place where he's, at least at, at a minimum, had a bad day, but- 00:15:37,532 --> 00:15:38,002 [John] [laughs] Right 00:15:38,002 --> 00:15:42,812 [Eric] ... you know, is maybe sort of thinking about, "Wow, I mean, like, what, what does this mean for me as an engineer?" 00:15:44,352 --> 00:15:51,232 [John] Yeah. I, I think yes and no. One, in some ways I feel a bit prepared for it because of the management piece. 00:15:51,232 --> 00:15:51,492 [Eric] Mm-hmm. 00:15:51,492 --> 00:15:55,172 [John] Of like, oh, this is, you know, [laughs] this is a lot like managing people. 00:15:55,172 --> 00:15:55,432 [Eric] Yep. 00:15:55,432 --> 00:16:00,852 [John] But a l- it's like managing a new person where you don't under- for whatever reason, you don't... maybe you inherited them from another team. 00:16:00,852 --> 00:16:01,012 [Eric] Mm-hmm. 00:16:01,012 --> 00:16:10,012 [John] You don't know their capabilities. You don't know what context they have. Um, and unfo- and fortunately, it diverges because it's not a human and- 00:16:10,012 --> 00:16:10,282 [Eric] Yep 00:16:10,282 --> 00:16:12,872 [John] ... it's a... And, and we, and nobody knows exactly, like, the right way- 00:16:12,872 --> 00:16:13,572 [Eric] Yep. Yep 00:16:13,572 --> 00:16:23,782 [John] ... um, to do it. Um, [laughs] it diver- and, and then they get the, and then they go through this random training that you don't have any insight to, and then they pop up again, and you're like, "Okay." 00:16:23,782 --> 00:16:23,782 [Eric] Yeah. 00:16:23,782 --> 00:16:29,672 [John] "You seem a little different." [laughs] It's like you have this employee that goes on this retreat, like- 00:16:29,672 --> 00:16:29,702 [Eric] Yeah 00:16:29,702 --> 00:16:32,272 [John] ... a couple times a quarter, and they come back different, and you're like- 00:16:32,272 --> 00:16:32,572 [Eric] Yeah 00:16:32,572 --> 00:16:33,872 [John] ... "How exactly are they d-" 00:16:33,872 --> 00:16:34,342 [Eric] Right. [laughs] 00:16:34,342 --> 00:16:35,492 [John] Like that's the weird part. 00:16:35,492 --> 00:16:37,412 [Eric] [laughs] That's, that's so true. 00:16:38,452 --> 00:16:39,492 [Eric] That's so true. 00:16:39,492 --> 00:16:39,511 [John] [laughs] 00:16:39,512 --> 00:16:39,602 [Eric] Yeah. 00:16:39,602 --> 00:16:39,951 [John] That's the really weird part. 00:16:39,952 --> 00:16:41,212 [Eric] Leadership training. 00:16:41,212 --> 00:16:41,572 [John] Um- 00:16:41,572 --> 00:16:42,472 [Eric] Um- 00:16:42,472 --> 00:16:43,751 [John] But back to the advice thing- 00:16:43,752 --> 00:16:44,552 [Eric] Yeah 00:16:44,552 --> 00:16:53,832 [John] ... or to answer your question first, I guess. Like, yes, and that, like, I spend a ton of hours, like, how do I do this stuff now? 00:16:53,832 --> 00:16:54,071 [Eric] Mm-hmm. 00:16:54,072 --> 00:16:59,172 [John] Right? 'Cause I feel like I have to understand how to do it to set expectations for other people. 00:16:59,172 --> 00:16:59,932 [Eric] Mm-hmm. 00:16:59,932 --> 00:17:03,232 [John] And, um, it's, it's just good to know how to do it anyways. 00:17:03,232 --> 00:17:03,412 [Eric] Mm-hmm. 00:17:03,412 --> 00:17:10,952 [John] Because there's, there's ways that I can be productive in, in, um, in situations I wouldn't have been able to before- 00:17:10,952 --> 00:17:11,001 [Eric] Yeah 00:17:11,001 --> 00:17:18,932 [John] ... because I can more quickly compile context and then more quickly add value. Whereas before, I'd be like, "Well, like, this is gonna take too long for me to add value." 00:17:18,932 --> 00:17:19,742 [Eric] Yeah, yeah. For sure. 00:17:19,742 --> 00:17:26,392 [John] Um, so there's that. And, and the, and figuring out the balance there is really exhausting. 00:17:26,392 --> 00:17:26,492 [Eric] Yeah. 00:17:26,492 --> 00:17:32,262 [John] There's a bit of burnout of, like, oh, w- I definitely wouldn't have jumped in before. Like, but I think I c- you know? 00:17:32,262 --> 00:17:32,262 [Eric] Mm-hmm. 00:17:32,262 --> 00:17:33,112 [John] But I think I could. 00:17:33,112 --> 00:17:33,472 [Eric] Mm-hmm. 00:17:33,472 --> 00:17:38,712 [John] And then not measuring, um, hard time but also, like, switching cost- 00:17:38,712 --> 00:17:39,332 [Eric] Mm-hmm 00:17:39,332 --> 00:17:47,472 [John] ... is part of that, like, mental equation of, like, should I jump in and switch, and task switch to try to add some value here? Like what's- 00:17:47,472 --> 00:17:47,952 [Eric] Yep 00:17:47,952 --> 00:17:48,682 [John] ... the, the trade-off? 00:17:48,682 --> 00:17:49,052 [Eric] Yeah, yeah. 00:17:49,052 --> 00:17:51,672 [John] And that's exhausting to figure out what the right answer is- 00:17:51,672 --> 00:17:51,842 [Eric] Yeah 00:17:51,842 --> 00:17:52,652 [John] ... right, right now. 00:17:52,652 --> 00:18:05,492 [Eric] Totally agree. I totally agree. I think that the... You know, it was interesting, one of the things that Dylan mentioned in his X post was that there's less time spent immersed in hard problems. 00:18:06,812 --> 00:18:07,272 [John] Yeah. 00:18:07,272 --> 00:18:18,056 [Eric] You know, which is, that's very interesting. It, and the experience of that Is that you're having conversations about hard problems- 00:18:18,056 --> 00:18:18,286 [John] Right 00:18:18,286 --> 00:18:32,126 [Eric] ... with a machine, but you're not necessarily... It's not a, uh, it's not a singular experience of immersing yourself in the problem, going into the valley, and sort of emerging- 00:18:32,126 --> 00:18:32,136 [John] Yeah 00:18:32,136 --> 00:18:34,116 [Eric] ... on the other side, you know, with a solution. 00:18:34,116 --> 00:18:34,436 [John] Yeah. 00:18:34,436 --> 00:18:38,296 [Eric] Which is a very, very different type of experience. 00:18:38,296 --> 00:18:39,776 [John] I think there's one other factor here, too, 00:18:40,796 --> 00:18:47,856 [John] and I think this is a person-to-person thing. If you are an internal processor versus an external processor. 00:18:47,856 --> 00:18:48,316 [Eric] Hmm. 00:18:48,316 --> 00:18:54,656 [John] If you're more of an internal processor, which, you know, maybe Dylan is, I think it's a harder tax on you. 00:18:54,656 --> 00:18:54,946 [Eric] Hmm. 00:18:54,946 --> 00:19:02,766 [John] 'Cause if you're already a little bit of an external processor, you get a bit rewarded working with AI because you, like you and I have talked about. 00:19:02,766 --> 00:19:02,886 [Eric] Mm-hmm. 00:19:02,886 --> 00:19:06,196 [John] You can, you can talk to it for several minutes- 00:19:06,196 --> 00:19:06,256 [Eric] Right 00:19:06,256 --> 00:19:08,536 [John] ... and, like, brain dump, like, a lot of useful context. 00:19:08,536 --> 00:19:09,306 [Eric] Mm-hmm. 00:19:09,306 --> 00:19:11,436 [John] Or maybe for you it j- just flows more easily to, like, type- 00:19:11,436 --> 00:19:11,446 [Eric] Mm-hmm 00:19:11,446 --> 00:19:17,456 [John] ... out a paragraph. Um, but if you're more of an internal processor, I think, you know- 00:19:17,456 --> 00:19:17,626 [Eric] Yeah 00:19:17,626 --> 00:19:21,296 [John] ... I, I think that might, that might be har- it might be harder for those people. 00:19:21,296 --> 00:19:30,916 [Eric] The last point that I'll make that I th- that I've seen contribute to this sort of burnout or questioning of whether you like this new way of working with AI 00:19:32,136 --> 00:19:34,295 [Eric] is that the tool set changes very often. So before- 00:19:34,296 --> 00:19:34,326 [John] Yeah 00:19:34,326 --> 00:19:38,206 [Eric] ... we talked about the tools fundamentally changing, right? 00:19:38,206 --> 00:19:38,396 [John] Right. 00:19:38,396 --> 00:19:41,566 [Eric] So what does craftsmanship mean when the tool set changes, right? 00:19:41,566 --> 00:19:41,576 [John] Yeah. 00:19:41,576 --> 00:19:58,086 [Eric] It's still the same thing because you're using these tools to produce something great. Um, the tools change, but if you think about the circular saw, my dad has a circular saw still that he got from his dad, so this thing is old. You know what I mean? 00:19:58,086 --> 00:19:58,116 [John] Yeah. 00:19:58,116 --> 00:20:02,896 [Eric] At this point it's, you know, gotta be, like, 60 years old, [laughs] and it works great. 00:20:02,896 --> 00:20:03,836 [John] That's awesome. 00:20:03,836 --> 00:20:09,976 [Eric] Uh, and when I compare it to the circular saw that I have, they're not that different- 00:20:09,976 --> 00:20:10,276 [John] Sure 00:20:10,276 --> 00:20:11,936 [Eric] ... fundamentally, right? 00:20:11,936 --> 00:20:15,916 [John] And the economics are interesting, right? Like, how much was the circular saw of that quality- 00:20:15,916 --> 00:20:16,516 [Eric] [laughs] Totally 00:20:16,516 --> 00:20:16,556 [John] ... 60 years ago- 00:20:16,556 --> 00:20:16,806 [Eric] Yeah, yeah 00:20:16,806 --> 00:20:17,956 [John] ... versus today, right? 00:20:17,956 --> 00:20:24,556 [Eric] But the difference, and the reason I bring that analogy up, is, you know, there, of course, have been new tools invented for carpentry- 00:20:24,556 --> 00:20:24,565 [John] Mm-hmm 00:20:24,565 --> 00:20:42,846 [Eric] ... but the circular saw is still a primary tool, and it hasn't changed in, you know, half a century. But one of the tricky things with AI is that not only have the tools fundamentally changed, but there's rapid shift within, within that change once you sort of get to the other side. 00:20:42,846 --> 00:20:42,876 [John] Hmm. 00:20:42,876 --> 00:20:48,376 [Eric] Right? And so just a couple examples of this. Um, new models come out all the time. 00:20:49,496 --> 00:20:57,716 [Eric] Is it going to be better to switch? You know, if you have a workflow built up around a certain model, do you switch, right? Does- 00:20:57,716 --> 00:20:58,016 [John] Right 00:20:58,016 --> 00:21:16,306 [Eric] ... does it make sense to switch, right? Um, there are all sorts of fundamental, like, not fundamental, but there are all sorts of ways that you can change the fundamental behavior of the model that you're using. So you can have skills. You can have system prompts. 00:21:16,306 --> 00:21:16,336 [John] Right. 00:21:16,336 --> 00:21:32,285 [Eric] There are plugins, right? And so that sort of rebaselines the model and gives it more specific capabilities in certain contexts, et cetera. Uh, you know, there's actually still this concept of prompt engineering, uh, which is, you know, one of the oldest- 00:21:32,285 --> 00:21:32,285 [John] Yeah 00:21:32,285 --> 00:21:34,216 [Eric] ... sort of skill concepts- 00:21:34,216 --> 00:21:34,236 [John] Yeah 00:21:34,236 --> 00:21:45,166 [Eric] ... in AI. And of course the interface, the, there are tons of interfaces. You can just go to, you know, chatgpt.com and use it on the web, right? 00:21:45,166 --> 00:21:45,196 [John] Right. 00:21:45,196 --> 00:21:57,976 [Eric] There's interfaces like Codex that are a desktop app, uh, that are all-in-one, right? You can use Clodcode from the command line. You can access AI, very, very powerful AI, through Notion, you know, through their- 00:21:57,976 --> 00:21:58,036 [John] Yeah 00:21:58,036 --> 00:21:58,656 [Eric] ... interface, right? 00:21:58,656 --> 00:21:59,516 [John] Right. 00:21:59,516 --> 00:22:15,376 [Eric] And so there's all these challenges, and I think one of the things that I've noticed myself is that it is very, y- you're co- uh, I mean, quite literally sometimes on a weekly basis, you're evaluating the cost of updating your workflow. 00:22:15,376 --> 00:22:15,386 [John] Yeah. 00:22:15,386 --> 00:22:16,496 [Eric] 'Cause it's not free- 00:22:16,496 --> 00:22:16,505 [John] Exactly 00:22:16,505 --> 00:22:17,756 [Eric] ... to change the way you do things. 00:22:17,756 --> 00:22:18,285 [John] No. Yeah. 00:22:18,285 --> 00:22:21,076 [Eric] And there's no guarantee that it's going to be better. 00:22:21,076 --> 00:22:21,736 [John] Right. 00:22:21,736 --> 00:22:24,016 [Eric] Uh, you know, which is, um, 00:22:25,276 --> 00:22:27,716 [Eric] which is really mentally exhausting, right? 00:22:27,716 --> 00:22:28,706 [John] Yeah. Yeah. 00:22:28,706 --> 00:22:29,336 [Eric] Uh, it's- 00:22:29,336 --> 00:22:30,076 [John] It's, it is. 00:22:30,076 --> 00:22:34,076 [Eric] It reminds me of this, um, we'll try to dig it up and put it in the show notes, but 00:22:36,416 --> 00:22:46,816 [Eric] there was a artist who did a, who created an illustration. I, it was for a major... Well, it was a major publication, so The Atlantic or The New Yorker- 00:22:46,816 --> 00:22:46,826 [John] Mm-hmm 00:22:46,826 --> 00:22:48,546 [Eric] ... or something, you know, something of that nature. 00:22:50,196 --> 00:22:57,925 [Eric] And there was a story about him. He was on, like, version two of Photoshop. 00:22:57,925 --> 00:22:57,935 [John] Uh-huh. 00:22:57,936 --> 00:22:59,626 [Eric] And this is some time ago, you know? 00:22:59,626 --> 00:22:59,925 [John] Mm-hmm. 00:22:59,925 --> 00:23:02,846 [Eric] So it, it was on, like, version, let's just say eight. I don't remember- 00:23:02,846 --> 00:23:02,876 [John] Right 00:23:02,876 --> 00:23:08,996 [Eric] ... the exact numbers, right? But basically, like, had- you know, you, he had to run an older computer to run it. 00:23:08,996 --> 00:23:09,326 [John] Okay. 00:23:09,326 --> 00:23:09,516 [Eric] Right? 00:23:09,516 --> 00:23:09,646 [John] Right. 00:23:09,646 --> 00:23:11,836 [Eric] 'Cause it wasn't supported [laughs] on newer versions- 00:23:11,836 --> 00:23:11,925 [John] Right 00:23:11,925 --> 00:23:13,106 [Eric] ... of the operating system, right? 00:23:14,376 --> 00:23:14,806 [Eric] And 00:23:15,836 --> 00:23:20,056 [Eric] this is a great example of, you know, sort of a craftsman who 00:23:21,356 --> 00:23:26,356 [Eric] is really familiar with the tool and says, like, "I'm just going to use this tool 'cause I'm really good with it," right? 00:23:26,356 --> 00:23:26,506 [John] Mm-hmm. 00:23:26,506 --> 00:23:49,966 [Eric] "And I'm not gonna change what I'm doing." And you know, he was [laughs] successful enough at that, you know, to still produce cover art for a very, very reputable, like, magazine, you know? And so you... I th- I've actually thought about that multiple times 'cause it's like, well, do you just pick a, pick a tool set and workflow and model that works well for you and only very occasionally, you know, switch? 00:23:49,966 --> 00:23:49,966 [John] Yeah. 00:23:49,966 --> 00:23:51,316 [Eric] 'Cause the switching costs can be- 00:23:51,316 --> 00:23:51,326 [John] Yeah 00:23:51,326 --> 00:23:59,876 [Eric] ... can be super high. So I think there are a ton of factors that contribute to that. But let's actually switch and speak to someone who has not experienced this burnout. 00:23:59,876 --> 00:24:00,056 [John] Yeah. 00:24:01,676 --> 00:24:19,712 [Eric] Dylan's a software engineer, and we've said many times on the show that just because of the nature of LLMs working with language and the, the gigantic corpus of publicly available code that all the frontier labs have scraped and processed and, you know, sort of understand, AI has been 00:24:19,712 --> 00:24:31,332 [Eric] Predisposed to solving more difficult problems in software engineering and writing code than in a lot of other industries at a faster rate. But it is coming 00:24:32,352 --> 00:24:44,512 [Eric] with the same force to other industries. So speak to the person who hasn't experienced the burnout yet, and may, may even be thinking, "I don't know. That sounds like I'm not gonna experience that." 00:24:44,512 --> 00:24:45,132 [John] Right. Right. 00:24:46,232 --> 00:24:54,252 [John] I mean, I think for knowledge work, I mean, honestly, you, uh, probably have a good perspective on this because your work kind of like goes between like the tech- 00:24:54,252 --> 00:24:54,352 [Eric] Hmm 00:24:54,352 --> 00:24:58,432 [John] ... and the knowledge work. But I think from my perspective, um, 00:25:00,172 --> 00:25:02,652 [John] I, I do think it'll vary a lot per job. 00:25:02,652 --> 00:25:02,892 [Eric] Yep. 00:25:02,892 --> 00:25:10,652 [John] I think there's a class of jobs where the accuracy is extremely, um, important and/or maybe even regulated. 00:25:10,652 --> 00:25:11,162 [Eric] Mm-hmm. 00:25:11,162 --> 00:25:25,532 [John] Like your, um, medical professional maybe, or some profession where you're... Like, lots of schooling and, and credentials, and the... All the outputs need to be verified and stamped, like, legally. 00:25:25,532 --> 00:25:25,732 [Eric] Yep. 00:25:25,732 --> 00:25:27,692 [John] Or maybe even, like, a, a tax accountant. 00:25:27,692 --> 00:25:27,772 [Eric] Mm-hmm. 00:25:27,772 --> 00:25:28,092 [John] Like... 00:25:29,372 --> 00:25:39,152 [John] So I think for those, um, I think the burnout those positions will face will be spending all of your time doing review. 00:25:39,152 --> 00:25:39,182 [Eric] Hmm. 00:25:39,182 --> 00:25:45,412 [John] Will be really exhausting, and you will m-miss the time actually, like, doing or preparing the work. 00:25:45,412 --> 00:25:45,632 [Eric] Right. 00:25:45,632 --> 00:25:45,652 [John] Um- 00:25:45,652 --> 00:25:48,492 [Eric] Because those are different mental modes, right? If you're not generating- 00:25:48,492 --> 00:25:48,502 [John] Yeah 00:25:48,502 --> 00:25:50,532 [Eric] ... it, then you're just constantly reviewing it. 00:25:50,532 --> 00:25:57,012 [John] Yeah. Yeah, like my, um, father-in-law's a commercial real estate appraiser. 00:25:57,012 --> 00:25:57,652 [Eric] Mm-hmm. 00:25:57,652 --> 00:26:02,252 [John] And for him, he was... I was talking to him about AI a little bit, um, 00:26:03,372 --> 00:26:05,542 [John] and he was just mentioning being able to use it to, 00:26:06,572 --> 00:26:10,032 [John] to review things, but, you know, they have to manually review and sign off- 00:26:10,032 --> 00:26:10,492 [Eric] Mm-hmm. Mm-hmm 00:26:10,492 --> 00:26:22,322 [John] ... and somebody that has the, you know, credentials to do it. Um, but I imagine in, in that type of role, for example, the, um, the velocity is gonna drastically increase for how many, 00:26:23,372 --> 00:26:27,712 [John] you know, reviews one appraiser is, is supposed to do in a week or a month- 00:26:27,712 --> 00:26:27,722 [Eric] Yeah 00:26:27,722 --> 00:26:28,232 [John] ... or something. 00:26:28,232 --> 00:26:28,972 [Eric] Mm-hmm. 00:26:28,972 --> 00:26:29,632 [John] And, um, 00:26:30,732 --> 00:26:36,092 [John] I think maintaining a high quality standard there of just [chuckles] doing, like, review constantly is gonna be- 00:26:36,092 --> 00:26:36,262 [Eric] Right 00:26:36,262 --> 00:26:36,852 [John] ... really hard. 00:26:36,852 --> 00:26:37,932 [Eric] Yeah, yeah. I agree. 00:26:37,932 --> 00:26:38,912 [John] Um- 00:26:38,912 --> 00:26:41,412 [Eric] I agree. I mean, we've seen that in software engineering, right? 00:26:41,412 --> 00:26:42,892 [John] Yeah, with pull request and- 00:26:42,892 --> 00:26:43,792 [Eric] Exactly. 00:26:43,792 --> 00:26:44,192 [John] Yep. 00:26:44,192 --> 00:26:45,932 [Eric] And the 00:26:47,352 --> 00:26:51,392 [Eric] one thing that I've noticed, and there was actually another comment that you pointed out in this, uh- 00:26:51,392 --> 00:26:51,592 [John] Yeah 00:26:51,592 --> 00:26:54,872 [Eric] ... thread about expectations 00:26:55,892 --> 00:26:57,112 [Eric] changing- 00:26:57,112 --> 00:26:57,132 [John] Right 00:26:57,132 --> 00:26:59,651 [Eric] ... from a manager or employer standpoint. 00:26:59,652 --> 00:27:00,372 [John] Mm-hmm. 00:27:00,372 --> 00:27:09,342 [Eric] And the person said, "I, you know, I really don't like the perception that my boss or company has-" 00:27:10,412 --> 00:27:10,461 [John] Right 00:27:10,461 --> 00:27:14,432 [Eric] "... about, like, what should be possible now that we have AI in terms- 00:27:14,432 --> 00:27:14,552 [John] Right 00:27:14,552 --> 00:27:15,382 [Eric] ... of throughput," right? 00:27:15,382 --> 00:27:16,151 [John] Right. 00:27:16,152 --> 00:27:17,292 [Eric] Which is a really... 00:27:18,972 --> 00:27:27,532 [Eric] I'm interested in your thoughts on that because I, I can see both sides, and I think there are legitimate, there are legitimate viewpoints on both sides, right? 00:27:27,532 --> 00:27:28,372 [John] Mm-hmm. 00:27:28,372 --> 00:27:32,592 [Eric] One of them is that, yes, AI is super powerful, and so, you know, 00:27:33,992 --> 00:27:41,812 [Eric] theoretically you can, like, one person, um, or let's say one unit of labor, if we wanna be clinical- 00:27:41,812 --> 00:27:42,012 [John] Okay 00:27:42,012 --> 00:27:45,132 [Eric] ... can produce, you know, far more output, right? 00:27:45,132 --> 00:27:45,852 [John] Right. 00:27:45,852 --> 00:27:54,122 [Eric] And so, and I think that's a legitimate expectation, right? The other side is that it's not that easy, as we just talked about. The tool set's changing. 00:27:54,122 --> 00:27:54,132 [John] Right. 00:27:54,132 --> 00:28:03,572 [Eric] The models change. It's a complete... You know, there are identity issues wrapped up in that, right? It's not just a matter of like, here's a circular saw. You can, you know, now- [laughs] 00:28:03,572 --> 00:28:11,932 [John] Right. But which, which with that, a circular saw is like a normal saw that, like, moves faster. Like, it's so similar in the, like, practical- 00:28:11,932 --> 00:28:12,432 [Eric] Yeah, yeah, yeah 00:28:12,432 --> 00:28:13,372 [John] ... how, how it behaves- 00:28:13,372 --> 00:28:14,032 [Eric] Mm-hmm 00:28:14,032 --> 00:28:15,711 [John] ... that it's really easy to understand. 00:28:15,712 --> 00:28:15,892 [Eric] Yes. 00:28:15,892 --> 00:28:17,392 [John] And I think that is part of the problem- 00:28:17,392 --> 00:28:17,552 [Eric] Hmm 00:28:17,552 --> 00:28:21,262 [John] ... is we're moving from deterministic, like, XYZ, you can audit- 00:28:21,262 --> 00:28:21,262 [Eric] Yep 00:28:21,262 --> 00:28:24,612 [John] ... exactly, like, what happened to the granular detail- 00:28:24,612 --> 00:28:25,352 [Eric] Yep 00:28:25,352 --> 00:28:31,711 [John] ... into this, like, like, we got from, from input to output, but the middle is, like, pretty mushy at times. 00:28:31,712 --> 00:28:31,852 [Eric] Yeah. 00:28:31,852 --> 00:28:33,352 [John] Like, how did we get there? We don't know. 00:28:33,352 --> 00:28:37,112 [Eric] Yep. Yeah, yeah. I agree. So I, I think that the, 00:28:38,232 --> 00:28:46,772 [Eric] um... This is what I would say to sort of the knowledge worker who maybe hasn't experienced burnout or maybe that even seems unrealistic to them because- 00:28:46,772 --> 00:28:46,802 [John] Right 00:28:46,802 --> 00:28:48,972 [Eric] ... they get a ton of value out of it. 00:28:48,972 --> 00:28:49,551 [John] Mm-hmm. 00:28:49,552 --> 00:29:00,332 [Eric] And I would say the... So first of all, it's, it's likely going to become a problem in a bunch of other disciplines, right? 00:29:00,372 --> 00:29:00,392 [John] Yeah. 00:29:00,392 --> 00:29:00,532 [Eric] So, 00:29:01,752 --> 00:29:06,712 [Eric] you know, we... I'm, I'm... I work on content, right? And we use AI all the time. 00:29:06,712 --> 00:29:06,872 [John] Right. 00:29:07,892 --> 00:29:26,862 [Eric] And it is super powerful, but it's pretty difficult to wield in a productive way. There's a steep learning curve there, and everyone on my team agrees it's, it- it's helpful but actually frustrating in all sorts of new ways that we didn't... weren't frustrated [chuckles] with our jobs- 00:29:26,862 --> 00:29:26,862 [John] Right 00:29:26,862 --> 00:29:27,772 [Eric] ... like previously, right? 00:29:27,772 --> 00:29:28,692 [John] Right. 00:29:28,692 --> 00:29:29,652 [Eric] And, um, 00:29:30,812 --> 00:29:35,132 [Eric] so I think the more, the more that it actually becomes 00:29:36,392 --> 00:29:49,472 [Eric] a core part of the tool set that you use for your core day-to-day work, the more you will experience this, right? The rough edges, the friction, you know, the sort of, um, 00:29:50,652 --> 00:30:06,932 [Eric] the different type of interface where you're not as immersed in the deep problems. You're doing more review. And so the more... Uh, like, if it's not integrated into your core day-to-day, hour-by-hour work, you're probably not going to experience- 00:30:06,932 --> 00:30:07,512 [John] Hmm 00:30:07,512 --> 00:30:09,242 [Eric] ... that as much as someone like Dylan. 00:30:09,242 --> 00:30:09,292 [John] Yeah. 00:30:09,292 --> 00:30:09,612 [Eric] Right? 00:30:09,612 --> 00:30:10,292 [John] Yeah. Sure. 00:30:10,292 --> 00:30:22,868 [Eric] Um, but if it is becoming more integrated, then, uh, it's, you know, I would say it's absolutely something to think about. You know, sort of this burnout and sort of the things changing and even anticipating, like, well- 00:30:22,868 --> 00:30:31,148 [Eric] If there are fundamental changes in the tool set here, how is that going to make me feel about the competency that I have in doing this work a particular way- 00:30:31,148 --> 00:30:31,158 [John] Right 00:30:31,158 --> 00:30:34,268 [Eric] ... that I did, you know, before AI. Um, 00:30:35,368 --> 00:30:35,768 [Eric] and so, 00:30:37,148 --> 00:30:58,148 [Eric] uh, the, the other thing I would say, and we've talked about this a lot on the show, but it, it continues to, um... I don't know if surprise is the right word, but the average person on a $20 plan, um, I don't think is wielding AI to the point... Uh, and they don't have access to enough- 00:30:58,148 --> 00:30:58,158 [John] Mm-hmm 00:30:58,158 --> 00:31:09,188 [Eric] ... power or aren't, you know... They don't, they are not consuming it at a level where they can experience the power enough to integrate it into their day-to-day workflow. 00:31:09,188 --> 00:31:09,888 [John] Hmm. 00:31:09,888 --> 00:31:21,868 [Eric] Right? Um, which is really interesting, right? And so it, it is... I'm becoming less surprised when people are like, "AI is awesome because I can summarize a PowerPoint," right? And it's like- 00:31:21,868 --> 00:31:22,988 [John] Right 00:31:22,988 --> 00:31:26,008 [Eric] ... it... No one wants to go back to a world where you can't do that, right? 00:31:26,008 --> 00:31:26,638 [John] Yeah, sure. 00:31:26,638 --> 00:31:27,778 [Eric] That, it is amazing, right? 00:31:27,778 --> 00:31:27,827 [John] Right. 00:31:27,827 --> 00:31:30,478 [Eric] Like, summarize this PowerPoint for me, and then- 00:31:30,478 --> 00:31:30,487 [John] Right 00:31:30,487 --> 00:31:34,788 [Eric] ... I can go and look at the slides that are, like, really important. Yes, I want this world. This is great. 00:31:34,788 --> 00:31:35,008 [John] Right. 00:31:35,008 --> 00:31:35,148 [Eric] Right? 00:31:36,228 --> 00:31:41,028 [Eric] Um, but that's like s- that is the most basic primitive use case, right? 00:31:41,028 --> 00:31:41,548 [John] Mm-hmm. 00:31:41,548 --> 00:31:48,328 [Eric] And so if that's the primary utility, I can absolutely see how people would be like, "Yeah, I mean, I don't... Why would you get burned out with this?" Right? Um- 00:31:48,328 --> 00:31:49,588 [John] Right. Right. Should be the opposite. 00:31:49,588 --> 00:32:05,968 [Eric] Yeah, it should be the opposite, right? Which is true, and I think that's actually what's a little bit counterintuitive about sort of this, like, craftsmanship AI burnout conversation, right? Is that your initial experience is that it would d- it would, burnout would diminish, right? And we're seeing the opposite, which is really interesting. 00:32:05,968 --> 00:32:06,948 [John] Right. 00:32:06,948 --> 00:32:08,508 [Eric] Okay. Practical advice. 00:32:10,388 --> 00:32:13,168 [Eric] What, what should the person who's experiencing burnout do? 00:32:15,308 --> 00:32:21,848 [John] So I th- I think the, the management, [clears throat] excuse me, I think the management framing is helpful. 00:32:21,848 --> 00:32:22,748 [Eric] Mm-hmm. 00:32:22,748 --> 00:32:32,328 [John] In that... And I actually, just this past week, I thought about this management book put out by Gallup. So Gallup does, like, lots of, like, workplace polls, and they have interesting- 00:32:32,328 --> 00:32:32,357 [Eric] Mm-hmm 00:32:32,357 --> 00:32:33,328 [John] ... content about management. 00:32:33,328 --> 00:32:34,048 [Eric] Mm-hmm. 00:32:34,048 --> 00:32:47,038 [John] This management book, um, came to mind. And one of the principles for the book, uh, there's like five, there's like... I think there's like 10, and, and it's essentially advice to managers of, like, here's how to set your people up to be successful. 00:32:47,038 --> 00:32:47,708 [Eric] Hmm. 00:32:47,708 --> 00:32:54,518 [John] One of the categories for setting your people up to be successful is, do they have the proper resources and materials to do their job? 00:32:54,518 --> 00:32:55,248 [Eric] Mm-hmm. 00:32:55,248 --> 00:32:56,408 [John] Super simple, right? 00:32:56,408 --> 00:32:57,168 [Eric] Mm-hmm. 00:32:57,168 --> 00:33:01,908 [John] And I was thinking about that. I was like, "Well, that's super applicable for, for AI." 00:33:01,908 --> 00:33:02,448 [Eric] Hmm. 00:33:02,448 --> 00:33:02,577 [John] Like, 00:33:03,728 --> 00:33:03,908 [John] so 00:33:05,008 --> 00:33:11,588 [John] I guess the point is, 'cause we talk about, with AI, like, we talk about context, and we talk about tool use and, and this, like, long, you know, 00:33:12,668 --> 00:33:21,468 [John] long number of things that's all revolving around, like, resources and materials. But I got there... The point of this is I got there 'cause I was thinking about, like, a manager, right? 00:33:21,468 --> 00:33:22,048 [Eric] Hmm. Yeah. 00:33:22,048 --> 00:33:23,538 [John] So I think there's a bunch of things 00:33:24,748 --> 00:33:28,208 [John] framing yourself as a manager, and go read management books, honestly. 00:33:28,208 --> 00:33:28,438 [Eric] Hmm. 00:33:28,438 --> 00:33:35,148 [John] 'Cause the concepts, most of the concepts work. I think it's important to distinguish these aren't humans, and they do work differently than humans. 00:33:35,148 --> 00:33:36,108 [Eric] Right. Right. 00:33:36,108 --> 00:33:39,328 [John] But there's a bunch of these concepts that work. Um, 00:33:40,368 --> 00:33:46,088 [John] and I'll do... Like, another one is, is metrics, right? Like OKRs- 00:33:46,088 --> 00:33:46,608 [Eric] Yep 00:33:46,608 --> 00:33:50,818 [John] ... um, or, or standards or templates. I mean, templates, checklists, like templates- 00:33:50,818 --> 00:33:50,818 [Eric] Yeah 00:33:50,818 --> 00:33:51,708 [John] ... and checklists, right? 00:33:51,708 --> 00:33:52,068 [Eric] Yeah, yeah. 00:33:52,068 --> 00:33:53,408 [John] They're around forever- 00:33:53,408 --> 00:33:53,648 [Eric] Yep 00:33:53,648 --> 00:33:55,488 [John] ... um, but are actually immensely helpful- 00:33:55,488 --> 00:33:55,928 [Eric] Yep 00:33:55,928 --> 00:33:57,428 [John] ... um, to validate, like, 00:33:58,448 --> 00:34:11,288 [John] just using a checklist. Like, we literally worked on this yesterday. We're having a... 'Cause it's, it is a real struggle to get an entire team on the same page with, like, how we're gonna use AI and get the entire team's quality up to, like, the same standard. 00:34:11,288 --> 00:34:11,468 [Eric] Yep. 00:34:11,468 --> 00:34:21,008 [John] Like, struggle for me, I think struggle for a lot of people. Well, and there's just a practical one that emerged the other day. It's like, we should have a checklist for this particular thing, and it was like 20 things long, and- 00:34:21,008 --> 00:34:21,668 [Eric] Mm-hmm 00:34:21,668 --> 00:34:27,528 [John] ... when humans were trying to review the AI, like, output, like, 20 things was too much to try to remember- 00:34:27,528 --> 00:34:27,588 [Eric] Mm-hmm 00:34:27,588 --> 00:34:28,468 [John] ... obviously. 00:34:28,468 --> 00:34:28,928 [Eric] Mm-hmm. 00:34:28,928 --> 00:34:35,548 [John] So that's, that's a really practical one of, like, here's our checklist. Um, if you're more in coding, it's evals, like- 00:34:35,548 --> 00:34:35,738 [Eric] Mm-hmm 00:34:35,738 --> 00:34:42,088 [John] ... and, and you can do it. The AI has, um, evaluations that you can build into your AI agents and stuff. 00:34:42,088 --> 00:34:42,158 [Eric] Yep. 00:34:42,158 --> 00:34:46,188 [John] But just practically, like you can use a checklist with, with whatever AI that you use- 00:34:46,188 --> 00:34:46,848 [Eric] Yep 00:34:46,848 --> 00:34:48,328 [John] ... and, um, can be really helpful. 00:34:48,328 --> 00:34:57,568 [Eric] Yeah. When I think about management, you know, as a, as a manager and having had both really good and really, you know, really bad managers in the past- 00:34:59,328 --> 00:34:59,338 [John] Right 00:34:59,338 --> 00:35:02,728 [Eric] ... really good managers say the same thing over and over and over again. 00:35:02,728 --> 00:35:03,608 [John] Uh-huh. 00:35:03,608 --> 00:35:06,928 [Eric] And because they just return to first principles- 00:35:06,928 --> 00:35:06,958 [John] Right 00:35:06,958 --> 00:35:09,588 [Eric] ... and priorities with their team over and over again. 00:35:09,588 --> 00:35:10,028 [John] Right. 00:35:10,028 --> 00:35:10,508 [Eric] Right? Um, 00:35:11,788 --> 00:35:14,528 [Eric] constantly changing things. I mean, I do think that there is 00:35:15,808 --> 00:35:26,908 [Eric] v- a lot of value as a manager in not, i- in, um, in envi- in creating an envi- a team environment that does not reward or foster complacency. 00:35:26,908 --> 00:35:27,488 [John] Right. 00:35:27,488 --> 00:35:36,628 [Eric] Right? But in terms of the core values of what the team's working on and the core metrics that we're driving towards and other things like that, it requires constant repetition- 00:35:36,628 --> 00:35:36,728 [John] Right 00:35:36,728 --> 00:35:38,307 [Eric] ... and reminding. [laughs] Like- 00:35:38,308 --> 00:35:38,968 [John] Yeah 00:35:38,968 --> 00:35:39,238 [Eric] ... yeah. 00:35:39,238 --> 00:35:39,248 [John] Yeah. 00:35:39,248 --> 00:35:42,567 [Eric] That is very similar to working with AI, right? Like, uh- 00:35:42,568 --> 00:35:42,588 [John] Yeah 00:35:42,588 --> 00:35:44,478 [Eric] ... no, we're trying to do this thing. [laughs] 00:35:44,478 --> 00:35:54,128 [John] Well, and just like in managing a team, when you find yourself saying it a bunch of, a bunch of times, like, figure out how to integrate it into the work and into the workflow. 00:35:54,128 --> 00:35:54,358 [Eric] Mm-hmm. Yeah. 00:35:54,358 --> 00:36:02,728 [John] 'Cause before, like, like imagine you were in charge of a, a team cleaning a restaurant kitchen. Like, it's like on the back of the door is this checklist. Like initial- 00:36:02,728 --> 00:36:02,758 [Eric] Yeah 00:36:02,758 --> 00:36:05,328 [John] ... the checklist before you leave. Like there's things like that- 00:36:05,328 --> 00:36:06,008 [Eric] Yeah, yeah, yeah 00:36:06,008 --> 00:36:10,068 [John] ... that it's easy to get so sucked into, to the current, like everything's changing- 00:36:10,068 --> 00:36:10,708 [Eric] Mm-hmm 00:36:10,708 --> 00:36:13,358 [John] ... that, um, that you can do, that actually it works. It makes it better. 00:36:13,358 --> 00:36:31,376 [Eric] Yeah, yeah. No, it's great. Uh, I, I had a couple of thoughts here as we were discussing this before we hit record. One of them is that- A- and we mentioned this a little bit earlier when we were talking about tools changing, but craftsmanship was never about the tools. 00:36:31,376 --> 00:36:32,236 [John] Yeah. Yeah. 00:36:32,236 --> 00:36:40,736 [Eric] Uh, and the way that I think about that is, uh, there are a couple specific points here. 00:36:42,456 --> 00:36:46,416 [Eric] One, the old tools are still available, so if you like using them- 00:36:46,416 --> 00:36:46,426 [John] Yeah 00:36:46,426 --> 00:36:48,106 [Eric] ... you can still use them, right? 00:36:48,106 --> 00:36:48,636 [John] Yeah. Good point. 00:36:48,636 --> 00:36:49,116 [Eric] I mean- 00:36:49,116 --> 00:36:49,696 [John] Yeah 00:36:49,696 --> 00:36:56,376 [Eric] ... you can go write code by hand. You can go generate a marketing campaign on a- 00:36:56,376 --> 00:36:57,016 [John] By clicking 00:36:57,016 --> 00:36:57,606 [Eric] ... blank page. 00:36:57,606 --> 00:36:58,286 [John] [laughs] Yeah. 00:36:58,286 --> 00:36:58,556 [Eric] Yes. [laughs] 00:36:58,556 --> 00:36:59,656 [John] You can use your mouse still. 00:36:59,656 --> 00:37:00,696 [Eric] You can u- Yeah. 00:37:00,696 --> 00:37:01,496 [John] Like, yeah. [laughs] 00:37:01,496 --> 00:37:04,716 [Eric] [laughs] You feel free to touch your trackpad. 00:37:04,716 --> 00:37:04,936 [John] [laughs] Right. 00:37:04,936 --> 00:37:05,556 [Eric] Um, 00:37:06,816 --> 00:37:08,196 [Eric] you can still do those things. 00:37:08,196 --> 00:37:08,536 [John] Mm-hmm. 00:37:08,536 --> 00:37:15,556 [Eric] And so I do think that there can be over-rotation on, you know, lamenting that the old days are gone, right? 00:37:15,556 --> 00:37:16,276 [John] Right. 00:37:16,276 --> 00:37:25,906 [Eric] Um, but you can still do it, and I would actually even go a step further and say that intentionally doing things the old way actually is very healthy, 00:37:27,076 --> 00:37:34,316 [Eric] right? To keep yourself sharp. We've talked about this on the show m- multiple times before. Um, 00:37:35,856 --> 00:37:49,756 [Eric] but I think i- it's a balance of accepting that there is a new tool that is going to be ubiquitous in some professions that already is completely ubiquitous in the way that the job is, is done, right? Or, or- 00:37:49,756 --> 00:37:50,036 [John] Right 00:37:50,036 --> 00:38:00,006 [Eric] ... com- ubiquitous is a tool that everyone uses to perform their job, right? And that's spreading to more and more professions. Um, but you can still work on your craft, right? 00:38:00,006 --> 00:38:00,095 [John] Right. 00:38:00,096 --> 00:38:11,356 [Eric] And I think I'm seeing more and more that people who still choose to work on their craft the old way actually become better and better at using AI and can find more joy in it. Um- 00:38:11,356 --> 00:38:11,756 [John] Yeah, yeah. 00:38:11,756 --> 00:38:22,796 [Eric] So that's one thing, is that there's... I- it was never about the tools, and it's not like the old tools have been taken away, and I think that that's a, that's a really important thing for us to remember. Um, 00:38:24,176 --> 00:38:25,776 [Eric] the other, uh, 00:38:26,936 --> 00:38:31,366 [Eric] the other thought that I had was around the 00:38:32,815 --> 00:38:35,356 [Eric] invention of new types of work. 00:38:36,516 --> 00:38:37,186 [John] Yeah. 00:38:37,186 --> 00:38:37,756 [Eric] And so 00:38:38,896 --> 00:38:47,196 [Eric] there's a... It's, it's very easy for us. We like to distill reality, right? So we distill reality into- 00:38:47,196 --> 00:38:47,536 [John] Yeah 00:38:47,536 --> 00:38:48,736 [Eric] ... like a, a map, right? 00:38:48,736 --> 00:38:50,686 [John] Are you saying the map is not the [laughs] territory? 00:38:50,686 --> 00:38:57,376 [Eric] The map is not the territory. [laughs] We, we distill reality down into a format that's easy for us to consume and make sense of, right? 00:38:57,376 --> 00:38:58,536 [John] Yeah. 00:38:58,536 --> 00:39:01,656 [Eric] Um, but we lose fidelity as that happens, right? 00:39:01,656 --> 00:39:02,036 [John] Right. 00:39:02,036 --> 00:39:19,116 [Eric] And so I think as we consider... Uh, when we struggle with this, to, to actually be even more precise. When we struggle with this new world of AI or the fact that our job is changing or, you know, are losing some of our identity because we were really good at this certain thing with these certain tools- 00:39:19,116 --> 00:39:19,225 [John] Right 00:39:19,225 --> 00:39:20,326 [Eric] ... and now all of that has changed, 00:39:21,516 --> 00:39:23,796 [Eric] it's really easy to forget that, um, 00:39:25,356 --> 00:39:31,036 [Eric] we are... Like, new ways of doing things are already being created, right? 00:39:31,036 --> 00:39:31,196 [John] Yeah. 00:39:32,336 --> 00:39:43,086 [Eric] Software development is, is changing as a profession, and so the definition of what it means to be a software developer is changing, right? It's not that the old way of doing things- 00:39:43,086 --> 00:39:43,216 [John] Right 00:39:43,216 --> 00:39:45,576 [Eric] ... completely goes away and that skill set's not valuable- 00:39:45,576 --> 00:39:45,586 [John] Right 00:39:45,586 --> 00:39:46,876 [Eric] ... and all those sorts of things, right? 00:39:48,076 --> 00:39:57,356 [Eric] But I think there's a lot of excitement in saying that, well, actually, like, this, this entire new category of jobs is being invented, right? 00:39:57,356 --> 00:39:57,856 [John] Yeah. 00:39:57,856 --> 00:39:58,236 [Eric] And 00:39:59,496 --> 00:40:19,496 [Eric] that, I think that's a very exciting thing. If we go back to the circular saw example, right? Um, it sort of o- it, there's a... It's an open-ended, very future-looking question about, okay, like, it used to take you, let's just say, six months to build this structure, right? 00:40:19,496 --> 00:40:19,836 [John] Yeah. 00:40:19,836 --> 00:40:28,186 [Eric] What if you could build it in three months, right? Would you build more of them? Would you think about new types of structures to build, right? 00:40:28,186 --> 00:40:28,256 [John] Right. 00:40:28,256 --> 00:40:49,095 [Eric] New materials come out, right? And so there's a, there's an immense amount of opportunity there, and I think that what I would... Like, a very practical thing, like mindset that I would encourage people to adopt is that there is a craftsmanship in learning what that new job is and how to do it, right? 00:40:49,096 --> 00:40:49,976 [John] Right. 00:40:49,976 --> 00:40:55,316 [Eric] Um, which is cha- you know, that's a challenge, right? Because, because of all the tools and all the change, especially- 00:40:55,316 --> 00:40:55,416 [John] Right 00:40:55,416 --> 00:40:56,376 [Eric] ... the rate of change in AI. 00:40:56,376 --> 00:40:57,085 [John] Right. 00:40:57,085 --> 00:40:58,266 [Eric] But I think that 00:40:59,556 --> 00:41:12,216 [Eric] I would adopt a less binary mindset of, like, I don't know if I n- I like this new world, right? It... Really, the exciting thing is I actually have agency to shape what these new, what the new role looks like. 00:41:12,216 --> 00:41:12,396 [John] Right. 00:41:12,396 --> 00:41:16,836 [Eric] You know, and sort of even define the way that I do things and define what it means to be a craftsman, you know- 00:41:16,836 --> 00:41:16,846 [John] Right 00:41:16,846 --> 00:41:17,376 [Eric] ... in this new world. 00:41:18,476 --> 00:41:20,585 [John] I think, I think one last thing that 00:41:21,716 --> 00:41:46,896 [John] does come up, come up a lot, especially when I talk to leaders or executives, um... So this, I think this may be a little bit less true of an, of an engineer or individual contributor, but especially when I talk to leaders and executives, they're... Most of them, when they're honest, they're like, "There's probably two or three decisions I make a year that make the biggest impact." 00:41:46,896 --> 00:41:47,476 [Eric] Mm. 00:41:47,476 --> 00:41:48,456 [John] And, like, most of- 00:41:48,456 --> 00:41:48,816 [Eric] Yeah 00:41:48,816 --> 00:41:52,076 [John] ... most of the decisions we make are pretty much incon- inconsequential. 00:41:53,256 --> 00:41:54,286 [Eric] Mm-hmm. 00:41:54,286 --> 00:41:56,876 [John] Um, so I mean that in a encouraging way- 00:41:56,876 --> 00:41:57,776 [Eric] [laughs] Okay. 00:41:57,776 --> 00:41:58,456 [John] Of- 00:41:58,456 --> 00:41:58,876 [Eric] Explain that 00:41:58,876 --> 00:42:03,576 [John] ... like, because it's only, like, a couple of things. Like, one, we're like, "Oh, I hope I don't miss them," [laughs] right? 00:42:03,576 --> 00:42:03,896 [Eric] Yeah, yeah. 00:42:03,896 --> 00:42:04,936 [John] Which is fair. 00:42:04,936 --> 00:42:05,136 [Eric] Sure. Yeah. 00:42:05,136 --> 00:42:06,806 [John] Um, but the other piece of it is, 00:42:08,196 --> 00:42:16,516 [John] in some ways, it should take a little bit of the, the pressure off as far as, as far as zooming out a bit. 00:42:16,516 --> 00:42:16,756 [Eric] Mm. 00:42:16,756 --> 00:42:23,876 [John] And I think we're so focused on output, speed, maybe even efficiency with our new tooling- 00:42:23,876 --> 00:42:24,636 [Eric] Mm-hmm 00:42:24,636 --> 00:42:29,136 [John] ... to, to realize that, like, I don't think the math changed there. 00:42:29,136 --> 00:42:29,206 [Eric] Yes. 00:42:29,206 --> 00:42:36,256 [John] Pre-AI, it was, like, just a couple of, like, key decisions that got us, like, post AI. I think that math is the same. 00:42:36,256 --> 00:42:36,736 [Eric] Mm. 00:42:36,736 --> 00:42:37,736 [John] And therefore- 00:42:37,736 --> 00:42:37,746 [Eric] Yep 00:42:37,746 --> 00:42:48,296 [John] ... you have the capability of being, like, 5X, 10X, whatever X more productive. But I don't know that, like, at the highest level of reality, that the math changed there- 00:42:48,296 --> 00:42:48,406 [Eric] Yes 00:42:48,406 --> 00:42:55,576 [John] ... as far as, like, we can actually still only do two or three things really well or only have two or three really high impact things that we do. 00:42:55,576 --> 00:42:56,296 [Eric] Mm. 00:42:56,296 --> 00:43:03,856 [John] Um, and that, maybe that changes. I don't know. But at least from talking to people, I, I haven't gotten the sense that that has fundamentally changed. 00:43:03,856 --> 00:43:05,036 [Eric] I completely agree with that. 00:43:06,196 --> 00:43:12,786 [Eric] The ability to contextualize... This is gonna sound very AI, but- 00:43:12,786 --> 00:43:13,176 [John] [laughs] 00:43:13,176 --> 00:43:21,716 [Eric] ... as a human, take in a bunch of disparate information, contextualize it, and make a really good strategic decision- 00:43:21,716 --> 00:43:22,076 [John] Right 00:43:22,076 --> 00:43:26,696 [Eric] ... is, uh, is huge. Um- 00:43:26,696 --> 00:43:26,736 [John] Yeah 00:43:26,736 --> 00:43:28,976 [Eric] ... and I completely agree with that, and I think that 00:43:30,096 --> 00:43:31,056 [Eric] AI can, 00:43:32,476 --> 00:43:34,835 [Eric] I think, help gather information as- 00:43:34,836 --> 00:43:34,866 [John] Yeah 00:43:34,866 --> 00:43:35,936 [Eric] ... part of that process. 00:43:35,936 --> 00:43:36,656 [John] Mm-hmm. 00:43:36,656 --> 00:43:44,256 [Eric] But, um, but I think you're right. Retaining your ability to do that is, is critical, and I don't think that math has changed. 00:43:44,256 --> 00:43:44,376 [John] Yeah. 00:43:45,416 --> 00:43:58,476 [Eric] All right. Well, thank you for joining us on the Token Intelligence Show, and we will catch you on the next one. [outro music]