You need to have an AI epiphany

00:00:00,880 --> 00:00:20,760 [Eric] AI is changing the way we work, and navigating the flood of hype and new technology isn't easy. On the Token Intelligence Show, we'll teach you how to stay relevant and shape a meaningful career in the midst of today's fast-changing market. Every week, it's a chance for you to slow down, think clearly, and ask the right questions about AI and yourself. 00:00:21,340 --> 00:00:24,560 [John] On today's show, we're going to be talking about AI epiphanies. 00:00:25,500 --> 00:00:43,040 [Eric] John and I cover what an AI epiphany is, why you should have one, and then the questions that you should ask afterwards, like, is AI the most transformative technology in the world? And what kind of future do I envision with AI? I have a pretty strong conviction that 00:00:44,900 --> 00:00:52,020 [Eric] everyone, maybe not everyone, everyone, but most people need to have what I call an AI epiphany. 00:00:52,620 --> 00:00:52,940 [John] Ooh. 00:00:53,240 --> 00:00:54,040 [Eric] You've had these. 00:00:54,480 --> 00:00:54,720 [John] Yeah. 00:00:55,160 --> 00:01:00,400 [Eric] But I wanna hear your definition of an AI epiphany to see if we're thinking about the term the same. 00:01:00,600 --> 00:01:01,040 [John] Okay. 00:01:02,500 --> 00:01:07,400 [John] I think it's, I think it's the moment when you first start u- using AI, or you use AI in a new way- 00:01:07,720 --> 00:01:08,020 [Eric] Mm-hmm 00:01:08,020 --> 00:01:18,040 [John] ... and, and you're blown away because of the, maybe the specific thing it can do, and then maybe a, a realization, or combined with a realization of how many things it can do. 00:01:18,100 --> 00:01:19,740 [Eric] Yeah. I think that's a great definition. 00:01:19,740 --> 00:01:20,240 [John] Is that, are we close? 00:01:20,380 --> 00:01:21,580 [Eric] That's a very good definition. 00:01:21,920 --> 00:01:21,940 [John] Okay. 00:01:21,960 --> 00:01:25,000 [Eric] The, actually, the two terms that came to my mind were depth and breadth. 00:01:25,860 --> 00:01:25,980 [John] Yeah. 00:01:26,040 --> 00:01:26,540 [Eric] Because 00:01:27,620 --> 00:01:30,300 [Eric] it is very cool if you 00:01:31,660 --> 00:01:35,840 [Eric] generate an image and it's, it's like, "Whoa, that was awesome." 00:01:36,060 --> 00:01:36,140 [John] Yeah. 00:01:36,180 --> 00:01:37,780 [Eric] You know, to see that for the first time. 00:01:37,820 --> 00:01:38,060 [John] Yeah. 00:01:39,040 --> 00:01:44,420 [Eric] Knowing, especially if you have experience editing images in Photoshop or other tools like that- 00:01:44,420 --> 00:01:44,500 [John] Right 00:01:44,500 --> 00:01:45,080 [Eric] ... right? It's just like- 00:01:45,180 --> 00:01:45,200 [John] Right 00:01:45,200 --> 00:01:46,520 [Eric] ... this is pretty wild, right? 00:01:47,620 --> 00:01:47,780 [Eric] But 00:01:49,080 --> 00:01:51,680 [Eric] that's v- that's localized to a specific thing. 00:01:51,720 --> 00:01:51,820 [John] Right. 00:01:51,820 --> 00:01:52,760 [Eric] I think the, the, 00:01:53,900 --> 00:02:04,760 [Eric] when you start to realize how many things that could impact or how capable it is to do that level of work across different disciplines is really the big holy cow. 00:02:04,780 --> 00:02:04,880 [John] Right. 00:02:04,880 --> 00:02:05,300 [Eric] And I think 00:02:06,740 --> 00:02:17,080 [Eric] one, one thing that I've seen multiple times is that it can be easy to immediately think about job displacement because of how capable it is across disciplines, right? 00:02:17,160 --> 00:02:17,260 [John] Yeah. 00:02:17,280 --> 00:02:17,820 [Eric] And I think it's, 00:02:18,900 --> 00:02:23,380 [Eric] we haven't seen that play out on the level that a lot of people thought that it would, but 00:02:24,420 --> 00:02:30,840 [Eric] that sort of one consequence of an epiphany or sort of conclusion from an epiphany is like, "Holy cow, this could take a bunch of jobs because this- 00:02:30,860 --> 00:02:30,880 [John] Right 00:02:30,880 --> 00:02:32,560 [Eric] ... thing is so powerful." 00:02:32,620 --> 00:02:32,720 [John] Yeah. 00:02:34,040 --> 00:02:35,920 [Eric] Do you remember your first epiphany moment? 00:02:35,960 --> 00:02:37,120 [John] My first epiphany moment. 00:02:38,920 --> 00:02:42,920 [John] Okay, this, this ac- I've had a lot. I think, I think I remember the first one. 00:02:44,100 --> 00:02:50,140 [John] I was prepping for a presentation at, at a conference, which I, like, never do. I don't do a lot of those. And 00:02:51,380 --> 00:02:55,440 [John] I was going back and forth, like, what idea am I gonna present? What am I gonna do? So I came up with an idea. 00:02:55,460 --> 00:02:55,960 [Eric] Mm-hmm. 00:02:56,060 --> 00:03:01,280 [John] And I thought it, I think, I think this is 2023-ish, 2023- 00:03:01,340 --> 00:03:01,420 [Eric] Mm-hmm 00:03:01,420 --> 00:03:04,580 [John] ... or so. And I wanted to illustrate, 00:03:05,900 --> 00:03:07,480 [John] for the talk, I wanted to illustrate some things. 00:03:07,500 --> 00:03:08,080 [Eric] Mm-hmm. 00:03:08,080 --> 00:03:12,200 [John] So I used a very early image model, um, 00:03:13,360 --> 00:03:16,700 [John] and generated five or six images 00:03:17,740 --> 00:03:21,200 [John] in sequence where, where I needed some similarity between them. 00:03:21,200 --> 00:03:21,280 [Eric] Mm-hmm. 00:03:21,300 --> 00:03:22,460 [John] And it was a ton of work. 00:03:22,640 --> 00:03:22,820 [Eric] Mm-hmm. 00:03:22,900 --> 00:03:23,900 [John] It was a ton of work, and I was like [laughs]- 00:03:23,940 --> 00:03:25,380 [Eric] That's so much back and forth on that one. 00:03:25,440 --> 00:03:27,500 [John] I could have hired, like, an illustrator or something. 00:03:27,500 --> 00:03:27,920 [Eric] [laughs] 00:03:27,940 --> 00:03:29,460 [John] It was a ton of work, but I did get it to work- 00:03:29,580 --> 00:03:29,780 [Eric] Yeah 00:03:29,780 --> 00:03:33,320 [John] ... and illustrate, like, each of the things I wanted to do. So that was one of the early ones. 00:03:33,500 --> 00:03:34,360 [Eric] Yeah. Yeah, yeah. 00:03:34,360 --> 00:03:34,780 [John] What about you? 00:03:36,400 --> 00:03:47,500 [Eric] I, I, this wasn't an epiphany. The, the first time I got a wow was I was leading a product function, and we had really struggled to prototype really quickly. 00:03:47,520 --> 00:03:47,780 [John] Hmm. Yeah. 00:03:47,780 --> 00:04:02,280 [Eric] You know, as a startup, and so resources are thin and, you know, so it's like, well, do you build all this infrastructure to be able to prototype quickly? And of course, one of the things that AI was good at pretty early on was building a clickable prototype of something that looked- 00:04:02,340 --> 00:04:02,360 [John] Yeah 00:04:02,360 --> 00:04:03,160 [Eric] ... pretty good, right? 00:04:04,260 --> 00:04:05,580 [Eric] And it's only gotten better- 00:04:05,880 --> 00:04:05,940 [John] Yeah 00:04:05,940 --> 00:04:09,960 [Eric] ... since then. That was unbelievable. So we could get ideas in front of customers really quickly. 00:04:09,960 --> 00:04:10,180 [John] Right. 00:04:10,180 --> 00:04:13,840 [Eric] But that was local. When it really hit me 00:04:15,480 --> 00:04:24,460 [Eric] that this was going to... When it... I, I knew from reading and some experience, okay, this is, this is a, a big deal. 00:04:24,700 --> 00:04:24,820 [John] Right. 00:04:25,300 --> 00:04:29,320 [Eric] But when that hit me on a personal level, I was migrating my blog- 00:04:30,200 --> 00:04:30,440 [John] Okay 00:04:30,440 --> 00:04:37,160 [Eric] ... from WordPress to a modern, uh, a modern JavaScript framework, Next.js, 00:04:38,300 --> 00:04:44,540 [Eric] on Vercel. I just hosted this blog on WordPress for thir- 13 years, actually. I remember writing a blog post about it. 00:04:45,660 --> 00:04:45,980 [Eric] And- 00:04:46,240 --> 00:04:50,400 [John] There... And then what, what year? Put us, what year are we in where you're doing this migration? 00:04:50,500 --> 00:04:52,960 [Eric] It was last summer. 00:04:53,540 --> 00:04:53,960 [John] Okay. 00:04:53,960 --> 00:04:54,920 [Eric] So not this previous summer- 00:04:54,940 --> 00:04:54,960 [John] Right 00:04:54,960 --> 00:04:55,920 [Eric] ... but the summer before. 00:04:55,960 --> 00:04:57,620 [John] So 13 years prior, all the way to last summer. 00:04:57,680 --> 00:04:59,240 [Eric] Yes, all the way to last summer. So 00:05:00,540 --> 00:05:09,120 [Eric] 2025, the summer of 2025. And I wanted to start writing more on my blog. It had been dormant for a while. 00:05:09,120 --> 00:05:09,140 [John] Mm-hmm. 00:05:09,160 --> 00:05:13,860 [Eric] And I had been frustrated with WordPress for a long time, but, you know, it's like, okay, this works. 00:05:13,900 --> 00:05:14,200 [John] Yeah. 00:05:14,200 --> 00:05:25,660 [Eric] Um, but the... Without going into too much detail, the reason that I hadn't done it is it's a, it's non-trivial for a 13-year-old WordPress press blog because- 00:05:25,780 --> 00:05:25,960 [John] Mm-hmm 00:05:25,960 --> 00:05:30,900 [Eric] ... content had been created over the years and different custom pages had been created- 00:05:30,900 --> 00:05:30,920 [John] Right 00:05:30,920 --> 00:05:34,480 [Eric] ... over the years on significantly different versions of WordPress. 00:05:34,540 --> 00:05:34,660 [John] Right. 00:05:35,100 --> 00:05:40,020 [Eric] And so it creates this problem of long tail edge cases 00:05:41,220 --> 00:05:43,480 [Eric] where it's like, well, you can just export it. It's like, well- 00:05:43,580 --> 00:05:43,840 [John] Yep 00:05:43,840 --> 00:05:48,040 [Eric] ... first of all, the export from WordPress is in a, a atrocious format called XML, 00:05:49,240 --> 00:05:53,980 [Eric] um, that's just not easy, it's not easy to work with. It's not like it's exporting a bunch of Word documents- 00:05:54,040 --> 00:05:54,100 [John] Right 00:05:54,100 --> 00:05:55,680 [Eric] ... or text files or markdown files, right? 00:05:56,920 --> 00:06:05,640 [Eric] And again, the, the change in that over time means that it's just a huge amount of complexity to actually get everything in a standardized format. 00:06:05,700 --> 00:06:05,920 [John] Right. 00:06:06,628 --> 00:06:22,788 [Eric] And I decided to just have Cursor build a plan and iterate through it and just loop through the process until, until the blog was migrated. And it, it happened so quickly. And th- this was before Opus 4.7. 00:06:22,888 --> 00:06:23,308 [John] Right. 00:06:23,368 --> 00:06:31,728 [Eric] Uh, and I was still just unbelievably... I was blown away. It was unbelievably capable and got the whole thing done in a couple of days. Um, 00:06:32,748 --> 00:06:35,228 [Eric] solved all these security issues that the old site had, you know- 00:06:35,288 --> 00:06:35,468 [John] Yeah 00:06:35,468 --> 00:06:36,548 [Eric] ... because of WordPress comments and- 00:06:36,548 --> 00:06:39,548 [John] [laughs] In case somebody wanted to hack, hack your blog. [laughs] 00:06:39,588 --> 00:06:41,908 [Eric] Well, it's like WordPress comments and all these weird- 00:06:42,108 --> 00:06:42,288 [John] Yeah. Yeah 00:06:42,288 --> 00:06:44,028 [Eric] ... I mean, all these weird things, right? Um- 00:06:44,148 --> 00:06:45,048 [John] Yeah 00:06:45,048 --> 00:06:46,608 [Eric] ... that was a huge epiphany moment, right? 00:06:46,648 --> 00:06:46,728 [John] Yeah. That's a big one. 00:06:46,728 --> 00:06:49,348 [Eric] 'Cause it just, it, it seemed mind-boggling 'cause I wasn't... 00:06:50,468 --> 00:06:54,608 [Eric] You know, I didn't have the time to dig in, and I'm not a developer by trade. 00:06:54,648 --> 00:06:54,828 [John] Right. 00:06:54,828 --> 00:06:54,868 [Eric] Right? 00:06:54,908 --> 00:06:55,108 [John] Right. 00:06:55,128 --> 00:06:58,548 [Eric] And so the ability for me to do that was, was incredible. 00:06:59,868 --> 00:07:00,388 [John] That's awesome. 00:07:00,428 --> 00:07:00,628 [Eric] The... 00:07:01,728 --> 00:07:02,368 [Eric] What... Okay, 00:07:03,468 --> 00:07:09,548 [Eric] what was your most recent epiphany, though? Because I think people have this first epiphany, which I think is good, and we'll talk about why- 00:07:09,608 --> 00:07:09,648 [John] Mm-hmm 00:07:09,648 --> 00:07:10,468 [Eric] ... I think that's good. 00:07:12,048 --> 00:07:22,388 [Eric] Uh, but you, you've had a lot of these. I agree. You've probably had more epiphanies than most people I know because you'll text me [laughs] about things that you're very excited about. What's your latest epiphany? 00:07:23,968 --> 00:07:26,488 [John] I have, like, three. I'll try to pick one or two. 00:07:26,888 --> 00:07:27,028 [Eric] Yeah. 00:07:27,248 --> 00:07:33,348 [John] One is a very simple one. There's several new technologies out that are using iMessage as the interface. 00:07:33,468 --> 00:07:34,208 [Eric] Mm-hmm. 00:07:34,308 --> 00:07:44,428 [John] And there's nothing else different. It's just that it's a iMessage, like native blue bubble iMessage on your phone versus a, a, an app or Slack- 00:07:44,708 --> 00:07:44,888 [Eric] Mm-hmm 00:07:44,888 --> 00:07:49,268 [John] ... or something. It makes a difference, and it's, it's weird, and I can't explain it. 00:07:49,268 --> 00:07:50,948 [Eric] It's bringing it to your n- the- 00:07:51,008 --> 00:07:51,428 [John] It feels- 00:07:51,428 --> 00:07:52,628 [Eric] ... familiar channel. 00:07:52,668 --> 00:07:56,068 [John] Yeah, it's familiar, but it feels eas- it feels more frictionless. It feels- 00:07:56,088 --> 00:07:56,088 [Eric] Mm-hmm 00:07:56,088 --> 00:08:02,828 [John] ... easier to use. I find that I want to use it for things. Um, I don't know. I can't fully explain it. So that's a weird one. 00:08:02,868 --> 00:08:03,028 [Eric] Yep. 00:08:03,028 --> 00:08:05,508 [John] 'Cause it's not... there's no feature difference, really. 00:08:05,548 --> 00:08:05,708 [Eric] Yeah. 00:08:05,708 --> 00:08:10,888 [John] It's the same. Uh, the, the other one is I have a migration story, too. Do you remember Evernote? 00:08:11,188 --> 00:08:11,748 [Eric] Oh, yeah. 00:08:11,808 --> 00:08:16,928 [John] Yeah. So Evernote, uh, similar to, like, a Microsoft OneNote, and there's a couple other competitors out there. 00:08:16,987 --> 00:08:17,508 [Eric] Mm-hmm. 00:08:17,528 --> 00:08:18,988 [John] I was on Evernote since 2010. 00:08:19,348 --> 00:08:19,948 [Eric] Mm-hmm. 00:08:20,048 --> 00:08:24,308 [John] And I had ton... I had school notes in there. I had personal notes. 00:08:24,548 --> 00:08:25,168 [Eric] That's a long time. 00:08:25,168 --> 00:08:27,488 [John] I had [laughs] like random things- 00:08:27,728 --> 00:08:27,848 [Eric] Mm-hmm 00:08:27,848 --> 00:08:33,608 [John] ... I wanted to keep. And, and it's just this forever notebook, and you can attach files. And I had a lot of stuff in there- 00:08:33,648 --> 00:08:33,768 [Eric] Mm-hmm 00:08:33,768 --> 00:08:34,628 [John] ... since 2010. 00:08:35,688 --> 00:08:44,228 [John] Um, I decided... Oh, they got bought. Evernote got bought by private equity, and they have jacked up the price every year- 00:08:44,528 --> 00:08:44,748 [Eric] Mm-hmm 00:08:44,748 --> 00:08:49,168 [John] ... the last, like, five years, maybe four years. I was like, "I'm done with this." 00:08:49,168 --> 00:08:49,268 [Eric] [laughs] 00:08:49,288 --> 00:08:52,828 [John] Like, I, I don't actively use it much. It's more of storage, like- 00:08:52,848 --> 00:08:52,908 [Eric] Mm-hmm 00:08:52,908 --> 00:09:00,848 [John] ... for me. So I did the same thing you did and moved everything off of that to something called Obsidian, which a lot of people are using now. 00:09:00,908 --> 00:09:01,108 [Eric] Mm-hmm. 00:09:01,208 --> 00:09:04,828 [John] It... partially 'cause it was... I was frustrated, and I was like, "I'm, I'm tired of, like- 00:09:05,108 --> 00:09:05,148 [Eric] Yeah 00:09:05,148 --> 00:09:07,348 [John] ... my cost double..." even though it's not that much money. 00:09:07,968 --> 00:09:08,808 [Eric] Mm-hmm. 00:09:08,808 --> 00:09:16,948 [John] Um, and it was the same, it was the same experience. It was crazy. Like, it just worked. I, I tried to do it before and got frustrated because of the weird format file thing- 00:09:16,968 --> 00:09:16,988 [Eric] Yep 00:09:16,988 --> 00:09:17,228 [John] ... and I just- 00:09:17,248 --> 00:09:18,468 [Eric] Totally. All the exports- 00:09:18,488 --> 00:09:18,768 [John] Mm-hmm 00:09:18,768 --> 00:09:19,628 [Eric] ... and all the edge cases, too. 00:09:19,688 --> 00:09:27,748 [John] Yeah. So that was the other one that... And I did that this... earlier this year. And th- there's probably a lot more, but those- 00:09:27,788 --> 00:09:28,008 [Eric] Yeah, yeah 00:09:28,008 --> 00:09:28,768 [John] ... I stop at two. [laughs] 00:09:28,788 --> 00:09:31,988 [Eric] Yeah, yeah. My most recent one actually relates to the show. 00:09:32,048 --> 00:09:32,428 [John] Yeah. 00:09:32,428 --> 00:09:33,748 [Eric] So we've gone through different iterations. 00:09:33,788 --> 00:09:36,568 [John] I'll tag onto your recent one, too, 'cause I got to experience a little bit, too. 00:09:36,608 --> 00:09:37,068 [Eric] Oh, yeah, yeah- 00:09:37,108 --> 00:09:37,188 [John] Yeah 00:09:37,188 --> 00:09:38,308 [Eric] ... 'cause we kinda worked on this together. 00:09:38,348 --> 00:09:39,108 [John] Yeah. 00:09:39,108 --> 00:09:47,048 [Eric] The... we've gone through different iterations of the way that we do video for the show, right? So do we edit in the service that we use online- 00:09:47,128 --> 00:09:47,208 [John] Right 00:09:47,208 --> 00:09:59,988 [Eric] ... that sort of records into, which is like a web-based interface, and you can edit there. And we wanted to do a couple things, update the graphics for the show and the intro and clean things up a little bit and grow up a little bit as a show- 00:10:00,188 --> 00:10:00,248 [John] Right 00:10:00,248 --> 00:10:02,228 [Eric] ... now that we're closing in on, you know, 40 episodes. 00:10:02,608 --> 00:10:03,328 [John] Yeah. 00:10:03,428 --> 00:10:03,708 [Eric] And 00:10:04,988 --> 00:10:13,328 [Eric] we actually... I remember messaging you and saying, "Should we just..." We have a friend who's like a video... He does video for a living. It's like, "Should we just pay him to just do a good job on this- 00:10:13,328 --> 00:10:13,348 [John] Yeah, we talked about that 00:10:13,348 --> 00:10:14,228 [Eric] ... and build a template and- 00:10:14,268 --> 00:10:14,348 [John] Yeah 00:10:14,348 --> 00:10:19,228 [Eric] ... you know, whatever?" And you said, "Well, I actually have seen people..." Like, GPT launched Astra- 00:10:19,408 --> 00:10:19,688 [John] Mm-hmm 00:10:19,688 --> 00:10:23,367 [Eric] ... very recently, like, you know, less than two weeks ago. Which- 00:10:23,388 --> 00:10:25,848 [John] Which I guess is GPT-6. I just- 00:10:26,168 --> 00:10:26,307 [Eric] Yeah 00:10:26,307 --> 00:10:28,247 [John] ... that... [laughs] I didn't even know it was called that until recently. 00:10:28,307 --> 00:10:30,448 [Eric] Their latest- ... greatest frontier model. 00:10:31,488 --> 00:10:36,128 [Eric] And there's a very popular video editing software called DaVinci. 00:10:36,168 --> 00:10:36,208 [John] Mm-hmm. 00:10:36,208 --> 00:10:39,808 [Eric] There's a free version that's, like, fully capable. You could edit a movie on it, and then the- 00:10:39,848 --> 00:10:40,248 [John] Yep 00:10:40,248 --> 00:10:41,868 [Eric] ... the paid version has more stuff. 00:10:41,908 --> 00:10:42,408 [John] Yeah. 00:10:42,468 --> 00:10:43,808 [Eric] I don't know enough to know exactly- 00:10:43,868 --> 00:10:47,528 [John] Yeah, 'cause I, I sent you a screenshot, and there was this guy that works for Red Bull- 00:10:47,608 --> 00:10:47,788 [Eric] Yes 00:10:47,788 --> 00:10:48,968 [John] ... that was using it to- 00:10:49,028 --> 00:10:49,328 [Eric] Yes 00:10:49,328 --> 00:10:51,288 [John] ... that was using the new, the newest AI from- 00:10:51,308 --> 00:10:51,448 [Eric] Mm-hmm 00:10:51,448 --> 00:10:53,728 [John] ... from, um, ChatGPT to edit- 00:10:53,948 --> 00:10:54,128 [Eric] Yep 00:10:54,128 --> 00:10:54,888 [John] ... using the software. 00:10:54,888 --> 00:10:55,228 [Eric] Yeah, yeah. 00:10:55,288 --> 00:10:55,408 [John] Um... 00:10:56,488 --> 00:10:56,808 [Eric] So 00:10:58,348 --> 00:11:02,848 [Eric] long story short, because I wanna talk about why epiphanies are important, but long story short, 00:11:04,128 --> 00:11:17,968 [Eric] we bought DaVinci Resolve Studio, which is the upgraded version. You can use the, um... It ha- it unlocks a number of things. Uh, you can also use an MCP server with it. So, uh, people have written different adapters- 00:11:18,028 --> 00:11:18,048 [John] Right 00:11:18,048 --> 00:11:21,148 [Eric] ... but you can basically control... You can have 00:11:22,428 --> 00:11:26,308 [Eric] ChatGPT control the editing software and edit- 00:11:26,388 --> 00:11:26,388 [John] Yeah 00:11:26,388 --> 00:11:27,108 [Eric] ... on your behalf. 00:11:28,208 --> 00:11:28,888 [Eric] And- 00:11:28,928 --> 00:11:31,168 [John] Where, where you're using words to do all the editing. That's where it comes into- 00:11:31,168 --> 00:11:32,368 [Eric] You're using words to do all the editing. 00:11:32,488 --> 00:11:32,548 [John] Yeah. 00:11:32,608 --> 00:11:37,608 [Eric] Exactly. And it was, it was absolutely astounding. I mean, in 00:11:39,028 --> 00:12:02,284 [Eric] one day, it completely redid the intro, created graphics, created a reusable template with a drop-ins, timestamps, all of that. And then we actually used it to, um, create a format for shorts, which we're gonna publish some of those. And then you used it- ... uh, in computer use mode to do audio mastering for the recording that's happening- 00:12:02,344 --> 00:12:02,564 [John] Yeah 00:12:02,564 --> 00:12:03,704 [Eric] ... right now to do a different mix. 00:12:03,924 --> 00:12:16,624 [John] Yeah. Yeah. So, ac- last week, I think last week or the week before, we had some audio issues and we, we used it and we would sit down and record a sample and it would go back and you could see the mouse moving on the screen- 00:12:16,824 --> 00:12:17,084 [Eric] Yep 00:12:17,084 --> 00:12:20,164 [John] ... dialing in our virtual audio tool- 00:12:20,424 --> 00:12:20,564 [Eric] Right 00:12:20,564 --> 00:12:22,204 [John] ... to turn the mics up and down and things like that. 00:12:22,204 --> 00:12:25,384 [Eric] Which is sort of the audio interface. It's kind of like a recording studio on a screen. 00:12:25,424 --> 00:12:26,624 [John] Yeah. On a screen, yeah. And- 00:12:27,544 --> 00:12:27,544 [Eric] Um- 00:12:27,584 --> 00:12:30,804 [John] Yeah. And, and we've checked it a couple times and seems to sound great. 00:12:30,884 --> 00:12:30,984 [Eric] See- 00:12:31,184 --> 00:12:32,224 [John] We'll, we'll see what people say. [laughs] 00:12:32,224 --> 00:12:32,784 [Eric] It seems to sound great, yeah. 00:12:32,804 --> 00:12:34,224 [John] Seems to sound great. [coughs] 00:12:34,284 --> 00:12:37,804 [Eric] Okay. Why is an epiphany important? I was thinking about this because 00:12:39,124 --> 00:12:39,704 [Eric] I think that 00:12:41,304 --> 00:12:43,284 [Eric] it can be easy to, 00:12:44,724 --> 00:12:47,264 [Eric] for lack of a better term, get stuck in a rut with AI- 00:12:47,324 --> 00:12:47,344 [John] Mm-hmm 00:12:47,344 --> 00:12:48,544 [Eric] ... and not try different things. 00:12:48,564 --> 00:12:48,744 [John] Mm-hmm. 00:12:48,764 --> 00:12:52,064 [Eric] Especially if you're not using it day in and day out for your job, right? 00:12:52,144 --> 00:12:52,164 [John] Yeah. 00:12:52,164 --> 00:13:02,924 [Eric] And so a good example of this is, I have a great friend, he's involved in a, a medical startup and he's like, "I just love GPT for helping me write emails," which is great. 00:13:03,024 --> 00:13:03,044 [John] Yeah. 00:13:03,064 --> 00:13:04,044 [Eric] That is an awesome thing. 00:13:04,104 --> 00:13:04,784 [John] Mm-hmm. 00:13:04,884 --> 00:13:05,204 [Eric] Um, 00:13:06,244 --> 00:13:09,084 [Eric] but I don't think he's had the epiphany moment yet, right? 00:13:09,164 --> 00:13:09,344 [John] Okay. 00:13:09,344 --> 00:13:41,464 [Eric] Where you start to realize, oh, well, this could actually do so many more things, um, could automate all sorts of interesting processes, right? And so I think it sort of gives you... it opens up the world of possibility and I think it can give you, like, a healthy respect and even concern for, like, the power of, the power of this technology. Um, but what do you think? Do you think it's, do you think it's good for people to have an, uh, an epiphany moment? 00:13:42,724 --> 00:13:57,644 [John] Yeah. I, I think so. I, I think it's, I think it's important to, to the, to your understanding. If, if you don't have that and you only have academic or secondhand knowledge, I think there's a lot of people that have second and third-hand knowledge, maybe just through the news. 00:13:57,884 --> 00:13:58,164 [Eric] Yep. 00:13:58,164 --> 00:14:04,124 [John] Um, that, yeah, I, I think it's really important not just for you to be able to use it, but for you to understand it better. 00:14:04,464 --> 00:14:08,804 [Eric] Yeah, totally. My dad recently had an epiphany moment. He actually took a class on AI. 00:14:09,084 --> 00:14:10,224 [John] Yeah, we were talking about this. 00:14:10,504 --> 00:14:11,604 [Eric] Mm-hmm. And 00:14:12,664 --> 00:14:26,924 [Eric] he... I saw the light go off and as the light went off and he... they came over for dinner and he was telling me about what he's seen these things do and what he's actually been able to do with them. And my dad's 70. He runs an automotive shop. 00:14:26,964 --> 00:14:27,084 [John] Mm-hmm. 00:14:27,104 --> 00:14:27,584 [Eric] Right? So, 00:14:28,684 --> 00:14:38,904 [Eric] uh, but he just, he couldn't stop saying, "This is crazy." Like, you know, "Do you realize how crazy this is?" You know, which I was like, "Dad, I definitely do. It is crazy." 00:14:39,584 --> 00:14:39,744 [John] Right. 00:14:39,744 --> 00:14:48,624 [Eric] But it raises questions about... It, it raises questions that we should ask about AI, because I think the 00:14:50,024 --> 00:14:51,104 [Eric] dangerous path 00:14:52,124 --> 00:14:57,084 [Eric] to go down is, uh, w- we mentioned one of them, which is- 00:14:57,684 --> 00:14:57,684 [John] Mm-hmm 00:14:57,684 --> 00:15:02,004 [Eric] ... sort of a scarcity mindset. This is gonna displace a bunch of workers. 00:15:02,064 --> 00:15:02,604 [John] Mm-hmm. Mm-hmm. 00:15:02,604 --> 00:15:04,584 [Eric] You know, or, um, 00:15:05,604 --> 00:15:14,364 [Eric] extreme, extreme techno-optimism to the point where, you know, you think that, um, you know, people aren't gonna have to work anymore. 00:15:14,744 --> 00:15:16,684 [John] I have a saying for this that I'm trying to adopt. 00:15:16,984 --> 00:15:17,124 [Eric] Okay. 00:15:17,124 --> 00:15:17,564 [John] You ready for it? 00:15:17,664 --> 00:15:18,004 [Eric] I'm ready. 00:15:18,224 --> 00:15:23,624 [John] Okay. Just because you don't know what the work of the future is going to be doesn't mean there's not going to be any work. 00:15:24,684 --> 00:15:25,464 [Eric] Wow. 00:15:25,484 --> 00:15:26,004 [John] Isn't that good? 00:15:26,044 --> 00:15:26,604 [Eric] That's a great- 00:15:26,704 --> 00:15:27,524 [John] I don't know where I heard that- 00:15:27,604 --> 00:15:27,604 [Eric] That's a great- 00:15:27,604 --> 00:15:27,784 [John] ... but- 00:15:27,844 --> 00:15:28,244 [Eric] Yeah 00:15:28,244 --> 00:15:29,184 [John] ... um, you know. 00:15:29,184 --> 00:15:29,684 [Eric] Claim it. 00:15:29,744 --> 00:15:30,044 [John] Yeah. 00:15:30,124 --> 00:15:31,244 [Eric] As an original. [laughs] 00:15:31,244 --> 00:15:31,284 [John] [laughs] 00:15:31,284 --> 00:15:36,964 [Eric] That's a John Wessel original. Yeah, totally. But that's... but you can... when you have an epiphany, 00:15:38,164 --> 00:15:38,484 [Eric] you, 00:15:39,904 --> 00:15:43,844 [Eric] I think, rightly start to extrapolate that experience. 00:15:43,884 --> 00:15:44,744 [John] Yes. Right, right. 00:15:44,784 --> 00:15:50,544 [Eric] And so I think asking good questions about AI help create good guardrails for- 00:15:50,564 --> 00:15:50,644 [John] Right 00:15:50,644 --> 00:15:57,784 [Eric] ... that extrapolation. So h- I'll start with a question that I think is a good one to ask, which I'm just gonna ask it to you 'cause I'm really- 00:15:57,824 --> 00:15:57,824 [John] Okay 00:15:57,824 --> 00:16:04,224 [Eric] ... interested in the answer. Do you think AI is the most transformative technology in the history of the world? 00:16:06,244 --> 00:16:06,804 [Eric] Just a light- 00:16:07,024 --> 00:16:09,604 [John] G- are, are we starting, like, at the wheel? [laughs] 00:16:09,604 --> 00:16:11,804 [Eric] [laughs] Yeah. Well, that, that's actually g- 00:16:11,944 --> 00:16:12,264 [John] Right. 00:16:12,644 --> 00:16:14,504 [Eric] I, I, I appreciate where you're going with that- 00:16:14,524 --> 00:16:14,544 [John] Yeah 00:16:14,544 --> 00:16:16,204 [Eric] ... because I think that's a good way to think about it. 00:16:16,504 --> 00:16:16,684 [John] Yeah. 00:16:18,304 --> 00:16:31,544 [John] I, I was thinking about this recently, uh, bec- of our lifetimes, yeah, of recent generations, I think so. But it's really hard to put yourself in the shoes of, of somebody 00:16:32,564 --> 00:16:33,384 [John] that was 00:16:34,404 --> 00:16:37,224 [John] prior to maybe electricity. 00:16:37,284 --> 00:16:37,644 [Eric] Mm-hmm. 00:16:37,664 --> 00:16:38,064 [John] You know? 00:16:38,124 --> 00:16:38,304 [Eric] Mm-hmm. 00:16:38,304 --> 00:16:46,484 [John] I think especially that age, that turn of the century type time where you've got guys like Edison, Nikola Tesla- 00:16:46,644 --> 00:16:46,824 [Eric] Mm-hmm 00:16:46,824 --> 00:16:47,584 [John] ... like, people like that, 00:16:49,384 --> 00:16:54,904 [John] uh, that, that to me is an era that maybe something like electricity around that era- 00:16:55,004 --> 00:16:55,004 [Eric] Mm-hmm 00:16:55,004 --> 00:17:02,144 [John] ... might, might be... because part of it... so think about flight. We all have the concept because of birds about flying. 00:17:02,204 --> 00:17:02,564 [Eric] Mm-hmm. 00:17:02,564 --> 00:17:07,444 [John] Like, it seems amazing that a person can be in the sky flying, but it... but the concept was there. 00:17:07,484 --> 00:17:07,644 [Eric] Yep. 00:17:08,824 --> 00:17:13,104 [John] Electricity, like, what's the concept? It's like, well, it's like the sun, but- 00:17:13,144 --> 00:17:14,264 [Eric] Or like fire or- 00:17:14,304 --> 00:17:15,784 [John] Yeah, or like fire, but it's not fire. 00:17:16,024 --> 00:17:16,364 [Eric] Yeah. 00:17:16,364 --> 00:17:29,284 [John] So that's the closest one I can think of that's extremely transformative, but not in our lifetime or in recent memory that, that maybe would be close, but, but it's still not a good analogy, I don't think. 00:17:29,424 --> 00:17:29,464 [Eric] Yeah. Yeah, I agree. 00:17:29,464 --> 00:17:35,384 [John] It's not close enough to explain with how electricity had impact to go back and say how AI is having impact. 00:17:35,404 --> 00:17:35,484 [Eric] Yeah. 00:17:35,544 --> 00:17:36,344 [John] I don't think it's close enough. 00:17:36,964 --> 00:17:48,764 [Eric] I, I don't know if you can answer that question. I agree in recent times, absolutely it is. I, I mean, at a minimum, there's no question that the amount of investment that has gone into this technology- 00:17:48,924 --> 00:17:49,024 [John] Right 00:17:49,024 --> 00:17:50,264 [Eric] ... has outpaced, you know- 00:17:50,584 --> 00:17:50,684 [John] Yeah 00:17:50,684 --> 00:17:52,604 [Eric] ... has, is, is eye-watering. 00:17:52,644 --> 00:17:53,184 [John] Yeah. Mm-hmm. 00:17:53,984 --> 00:17:55,084 [Eric] Um, but I think the point... 00:17:56,324 --> 00:18:00,508 [Eric] as I think about that question, I think the point is- 00:18:00,508 --> 00:18:02,548 [Eric] Are we going back to the wheel? I don't know. 00:18:02,568 --> 00:18:03,068 [John] Mm-hmm. Mm-hmm. 00:18:03,088 --> 00:18:08,008 [Eric] Right? Uh, is, is electricity a good analogy? Eh, it's imperfect. I don't know. 00:18:08,068 --> 00:18:08,468 [John] Yeah. 00:18:08,468 --> 00:18:11,368 [Eric] But I think the point is that it is a hard question to answer. 00:18:11,448 --> 00:18:11,828 [John] Mm-hmm. 00:18:11,928 --> 00:18:14,868 [Eric] And that forces you to think critically about it- 00:18:14,988 --> 00:18:14,988 [John] Right 00:18:14,988 --> 00:18:17,708 [Eric] ... as opposed to just extrapolating, you know- 00:18:17,748 --> 00:18:17,788 [John] Right 00:18:17,788 --> 00:18:20,228 [Eric] ... to a, to a conclusion. Um, 00:18:21,768 --> 00:18:32,888 [Eric] what, what questions come to mind for you as we think about having an AI epiphany? What is another good question to ask to set up those guardrail- guardrails? 00:18:34,148 --> 00:18:36,988 [John] I think it's along the same lines. I've been pretty obsessed with, 00:18:38,428 --> 00:18:40,668 [John] um, f- how do you frame it in analogy? 00:18:40,828 --> 00:18:41,428 [Eric] Mm-hmm. 00:18:41,528 --> 00:18:47,028 [John] Uh, one of the popular ones right now is in, in coding developer land is, is factory. 00:18:47,328 --> 00:18:47,568 [Eric] Mm-hmm. 00:18:47,908 --> 00:18:52,388 [John] So everybody in, in my world is talking about development factories or coding factories- 00:18:52,428 --> 00:18:52,628 [Eric] Mm-hmm 00:18:52,628 --> 00:18:58,088 [John] ... eight, whatever you wanna call it. So I've been thinking about that bigger than just for coding 00:18:59,228 --> 00:19:06,988 [John] and what people would have thought in a factory when, when machines, when you're, you know, you have this, all this industrialization happening- 00:19:07,228 --> 00:19:07,288 [Eric] Mm-hmm. Mm-hmm 00:19:07,288 --> 00:19:23,528 [John] ... and what they thought about machines. I think they for sure thought by now that a factory, there'd be all these machines doing all these things autonomously, and then something, finished part would come out the other side. I think they would have laughed if they said, "No, there's people standing around, walking around factories- 00:19:23,528 --> 00:19:24,188 [Eric] [laughs] Yeah 00:19:24,188 --> 00:19:25,808 [John] ... doing stuff," to this day. 00:19:25,908 --> 00:19:26,608 [Eric] Yeah, yeah. 00:19:26,628 --> 00:19:30,368 [John] And then I thought about early internet days, like the moment the internet 00:19:31,628 --> 00:19:33,828 [John] is a thing and, and people start connecting their computers. 00:19:33,828 --> 00:19:34,688 [Eric] Mm-hmm. 00:19:34,708 --> 00:19:39,488 [John] Well, of cour- all, all the computers are gonna be able to connect together. Everything's gonna be seamlessly integrated in a couple years. 00:19:39,648 --> 00:19:40,028 [Eric] Mm-hmm. 00:19:40,068 --> 00:19:40,928 [John] It won't take long. 00:19:40,968 --> 00:19:41,688 [Eric] Mm-hmm. 00:19:41,728 --> 00:19:44,328 [John] And you have the mess we have right now where nothing talks to each other. 00:19:44,408 --> 00:19:44,688 [Eric] Right, right. 00:19:44,688 --> 00:19:46,448 [John] It's like the number one problem for businesses. 00:19:46,488 --> 00:19:46,608 [Eric] Yeah. 00:19:47,168 --> 00:19:50,788 [John] So I think with those two analogies, there's gotta be something there- 00:19:50,868 --> 00:19:51,248 [Eric] Yeah 00:19:51,248 --> 00:19:53,168 [John] ... with just maybe something about human nature- 00:19:53,268 --> 00:19:53,608 [Eric] Hmm 00:19:53,608 --> 00:19:58,248 [John] ... where we start with the f- the, a machine in a factory or early internet- 00:19:58,408 --> 00:19:58,408 [Eric] Mm-hmm 00:19:58,408 --> 00:20:01,968 [John] ... and then we just extrapolate out, "Oh, we'll, we'll, we'll have these." 00:20:01,968 --> 00:20:02,088 [Eric] Right. 00:20:02,108 --> 00:20:07,548 [John] You know, we have... 'Cause think about it, we'll have a machine that can do the whole thing that a human used to do the task. 00:20:07,668 --> 00:20:08,028 [Eric] Mm-hmm. 00:20:08,088 --> 00:20:11,308 [John] And then same thing with the internet. Oh, well, you know, of course they're all gonna be connected. 00:20:11,828 --> 00:20:12,468 [Eric] Mm-hmm. 00:20:12,508 --> 00:20:16,148 [John] And, and, and there's no... On the internet, there's really no theoretical barrier. 00:20:16,968 --> 00:20:18,108 [Eric] Hmm. Yeah. 00:20:18,128 --> 00:20:20,288 [John] It's all company lines and people- 00:20:20,388 --> 00:20:20,588 [Eric] Mm-hmm 00:20:20,588 --> 00:20:21,928 [John] ... lines is why things don't connect. 00:20:21,948 --> 00:20:22,008 [Eric] Yeah. 00:20:22,008 --> 00:20:23,728 [John] There's, there's no, there's no theory problem. 00:20:23,808 --> 00:20:24,488 [Eric] Yeah. 00:20:24,488 --> 00:20:34,968 [John] And that's the question, where are the people lines going to cause problems with AI where it's not gonna work how people want it to, and it's because of how companies are drawn up and how people wanna operate? 00:20:35,028 --> 00:20:36,048 [Eric] Yep. It's such a good- 00:20:36,088 --> 00:20:36,908 [John] And we don't know the answer. 00:20:36,908 --> 00:20:40,648 [Eric] Those are great examples of assumptions. I think it's the same extrapolation, right? 00:20:40,668 --> 00:20:40,708 [John] Right. 00:20:40,708 --> 00:20:42,048 [Eric] Epiphany to extrapolation- 00:20:42,148 --> 00:20:42,268 [John] Right 00:20:42,268 --> 00:20:50,628 [Eric] ... uh, on the factory side and the internet side. I think that's great. Another one that comes to mind for me is, 00:20:51,788 --> 00:20:55,948 [Eric] what are the really good impact- impacts that AI could have? 00:20:56,008 --> 00:20:56,268 [John] Hmm. 00:20:56,288 --> 00:20:57,048 [Eric] Right? If we think about, okay- 00:20:57,188 --> 00:20:57,428 [John] Hmm 00:20:57,428 --> 00:21:06,348 [Eric] ... AI is going to have an impact. It already is having an impact. What are the really good things? And so I think about things like unlocking creativity, right? 00:21:06,368 --> 00:21:06,828 [John] Yeah. 00:21:06,828 --> 00:21:10,728 [Eric] More people bringing, being able to bring their ideas, um, 00:21:11,848 --> 00:21:31,708 [Eric] you know, being able to create the ideas that maybe they've had for a really long time. And just a couple quick examples of that. Um, there are people who... I was actually talking to someone the other day. They're an artist, and they said, you know, idea exploration, it's been so helpful, like, just for- 00:21:31,728 --> 00:21:31,788 [John] Hmm 00:21:31,788 --> 00:21:38,248 [Eric] ... idea exploration. It can go so much faster now. And so they're creating more art than they were- 00:21:38,288 --> 00:21:38,328 [John] Mm-hmm 00:21:38,328 --> 00:21:40,068 [Eric] ... before because they're able to iterate through- 00:21:40,148 --> 00:21:40,188 [John] Mm-hmm 00:21:40,188 --> 00:21:44,628 [Eric] ... a bunch of different ideas, like, way, way, way more quickly visually, right? 00:21:45,708 --> 00:21:57,048 [Eric] Um, and test how different things go together or don't go together. That was super interesting. On the technical side, and there's people who have said, "You know, I've been, I've had this side project in mind for 10 years, and I just haven't had the time to do it." 00:21:57,508 --> 00:21:57,528 [John] Yeah. 00:21:57,548 --> 00:21:58,568 [Eric] They're doing those things [laughs]. 00:21:58,628 --> 00:21:59,228 [John] I hear that a lot. 00:21:59,228 --> 00:21:59,848 [Eric] You know? I mean- 00:21:59,948 --> 00:22:00,308 [John] Yeah 00:22:00,308 --> 00:22:07,888 [Eric] ... which is really, really cool, right? And on the negative side, we've talked before on the show about AI psychosis and- 00:22:07,988 --> 00:22:08,168 [John] Right 00:22:08,168 --> 00:22:13,788 [Eric] ... things like, you know, uh, an AI girlfriend or boyfriend, um, you know, 00:22:14,828 --> 00:22:28,208 [Eric] uh, AI leading you down really negative or, um, you know, sort of, uh, errant lines of thought, right? Or, you know, um, you following a path- 00:22:28,368 --> 00:22:28,548 [John] Yeah 00:22:28,548 --> 00:22:33,528 [Eric] ... that's inaccurate because input that you gave at the beginning, you know, isn't accurate, right? Or- 00:22:33,548 --> 00:22:33,888 [John] Right 00:22:33,888 --> 00:22:41,528 [Eric] ... or negative, you know, really negative things that could happen, right? And that's just a microcosm. But I think those are... When you have an epiphany, 00:22:42,628 --> 00:22:44,768 [Eric] those are really good things to step back and think about. 00:22:44,848 --> 00:22:45,088 [John] Right. 00:22:45,128 --> 00:22:54,448 [Eric] Right? What is the im- what is, uh, uh, how transformative is this? Even if you can't answer that, you're thinking about it critically. And then what are the good things and the bad things which sort of help you- 00:22:54,768 --> 00:22:55,468 [John] Right 00:22:55,468 --> 00:22:58,928 [Eric] ... um, place AI rightly in your thinking- 00:22:59,128 --> 00:22:59,268 [John] Right 00:22:59,268 --> 00:23:01,988 [Eric] ... uh, and not just, not just follow a pathway of extrapolation. 00:23:03,228 --> 00:23:03,468 [Eric] Okay, 00:23:04,588 --> 00:23:14,148 [Eric] let's end by asking ourselves some questions, uh, which we love to do. Uh, so for all of us humans who are using AI- 00:23:14,148 --> 00:23:15,868 [John] [laughs] All of the humans that are listening. 00:23:16,028 --> 00:23:16,408 [Eric] All, for all- 00:23:16,428 --> 00:23:18,388 [John] Do we have any AI followers or listeners yet? 00:23:18,408 --> 00:23:20,048 [Eric] I don't know. They're all AI. It's all bots. 00:23:20,048 --> 00:23:20,768 [John] [laughs] Oh, no. 00:23:20,768 --> 00:23:35,188 [Eric] [laughs] The... What is, what's a question that comes to mind for you that's really healthy to ask of ourselves? And I wanna think about this in the context, again, of having had an epiphany. So we asked some questions about AI. What should we ask about ourselves? 00:23:35,388 --> 00:23:35,688 [John] Yeah. 00:23:37,368 --> 00:23:54,728 [John] I'll back into a question a little bit. I've also been thinking... We, we both have young kids, and I've been thinking about there's something about being a child where it's really positive for them to be in environments where they can learn and grow and they can test boundaries, fail in small ways- 00:23:54,788 --> 00:23:54,848 [Eric] Mm-hmm 00:23:54,848 --> 00:23:57,508 [John] ... learn from mistakes, you know, et cetera. I think everybody agrees- 00:23:57,668 --> 00:23:57,828 [Eric] Mm-hmm 00:23:57,828 --> 00:23:58,228 [John] ... that 00:23:59,648 --> 00:24:07,968 [John] with AI, and especially at work, especially with coding and writing, um, which is really what we do for most of our days- 00:24:08,008 --> 00:24:08,048 [Eric] Mm-hmm 00:24:09,284 --> 00:24:19,004 [John] It's weird 'cause we're adults, so it's su- it's super weird. And this goes a little bit to the good or bad, and it also goes to the personal impact piece of 00:24:20,344 --> 00:24:22,764 [John] work is just not set up like that. 00:24:22,964 --> 00:24:23,684 [Eric] Mm-hmm. 00:24:23,684 --> 00:24:23,984 [John] In that 00:24:25,324 --> 00:24:28,484 [John] we are-- we wanna be optimally efficient, right? 00:24:28,504 --> 00:24:29,244 [Eric] Mm-hmm. 00:24:29,244 --> 00:24:34,064 [John] We want to do things the best, you know, the best way we can. And- 00:24:34,084 --> 00:24:37,824 [Eric] Doing it right the first time generally makes most things go really well- 00:24:37,984 --> 00:24:38,284 [John] Yeah 00:24:38,284 --> 00:24:39,144 [Eric] ... in your job. 00:24:39,204 --> 00:24:45,564 [John] Right. Yeah. There you go. Yeah. And there's all these incentives and structures around that. And when everything changes in the digital world- 00:24:45,644 --> 00:24:45,924 [Eric] Mm-hmm 00:24:45,924 --> 00:24:48,784 [John] ... it, it just, it just messes a lot of stuff up. 00:24:48,824 --> 00:24:49,124 [Eric] Mm-hmm. 00:24:49,124 --> 00:24:51,084 [John] 'Cause it's not really-- Most companies are-- 00:24:52,224 --> 00:24:56,824 [John] Vast majority of companies are not really set up for learning, trying, failing. 00:24:56,864 --> 00:24:57,984 [Eric] Mm-hmm. 00:24:58,104 --> 00:24:58,764 [John] Um, and- 00:24:58,784 --> 00:25:00,084 [Eric] Yeah, 'cause failure is costly, right? 00:25:00,124 --> 00:25:00,144 [John] Yeah. 00:25:00,144 --> 00:25:01,304 [Eric] And so you try to optimize- 00:25:01,304 --> 00:25:01,324 [John] Yeah 00:25:01,324 --> 00:25:02,084 [Eric] ... your way against that. 00:25:02,144 --> 00:25:02,664 [John] Mm-hmm. 00:25:02,784 --> 00:25:09,824 [Eric] And it is, it is the exception rather than the rule that companies build margin into the system for experimentation and failure. 00:25:09,904 --> 00:25:10,144 [John] Right. 00:25:10,164 --> 00:25:10,644 [Eric] Right? 00:25:10,704 --> 00:25:29,524 [John] Right. Yeah. Yeah, I think that's true. And especially in coding, it's really weird to have the whole process changed, um, you know, for me, like mid-career, to have the whole process need to be reworked about how you, how you do the work and, and certain steps getting really fast and other steps getting way slower. 00:25:29,564 --> 00:25:30,264 [Eric] Mm-hmm. 00:25:30,364 --> 00:25:31,524 [John] Um, that's weird. 00:25:31,984 --> 00:25:32,104 [Eric] Yeah. 00:25:32,124 --> 00:25:40,924 [John] Um, so I think personally, 'cause one of the things we were thinking about is like, are you underestimating AI or are you overestimating AI? 00:25:41,064 --> 00:25:41,584 [Eric] Hmm. 00:25:41,624 --> 00:25:42,944 [John] And the answer is yes. 00:25:42,944 --> 00:25:43,964 [Eric] [laughing] 00:25:44,064 --> 00:25:50,364 [John] Uh, like, like what-- Or even, even personally, what can I do in a day? Answer that question. Or plan a day- 00:25:50,664 --> 00:25:50,964 [Eric] Hmm 00:25:50,964 --> 00:25:59,584 [John] ... for task. And your capa- your personal capability changing on a quarterly basis is weird. 00:25:59,764 --> 00:26:00,864 [Eric] Yep. 00:26:00,864 --> 00:26:05,544 [John] Because you're used to, like, I can do about this-- You have these mental, um, 00:26:06,724 --> 00:26:07,684 [John] I don't know, markers- 00:26:08,104 --> 00:26:08,164 [Eric] Yep 00:26:08,164 --> 00:26:11,704 [John] ... of about how much I can do in a period of time, in a day or a week or whatever. 00:26:11,744 --> 00:26:12,404 [Eric] Yep. 00:26:12,404 --> 00:26:13,864 [John] And it's all broken. 00:26:14,284 --> 00:26:14,504 [Eric] Yep. 00:26:14,524 --> 00:26:20,804 [John] Becau- again, some things really fast, some things slow. Some things are just big question marks 'cause I haven't done it in, in a number of months. 00:26:20,884 --> 00:26:27,144 [Eric] Yeah. Yeah, yeah. Totally. No, no, no. Overestimating and underestimating I think is great, and I think it's good to know whether you tend to do 00:26:28,184 --> 00:26:28,804 [Eric] one or the other. 00:26:29,064 --> 00:26:29,284 [John] Right. 00:26:29,784 --> 00:26:38,264 [Eric] Right? Um, underestimating more like my friend who uses AI to write emails, and so probably not thinking about all the possibilities, right? 00:26:38,304 --> 00:26:39,184 [John] Yeah. Right. 00:26:39,304 --> 00:26:48,484 [Eric] Um, and then overestimating, you know, thinking that this is a panacea for every type of friction and every type of work, right? 00:26:48,684 --> 00:26:49,384 [John] Right. Right. 00:26:49,424 --> 00:26:57,024 [Eric] And then I think people in the-- I would say you and I are more in the middle where we probably vacillate between over and underestimating- 00:26:57,064 --> 00:26:57,264 [John] Yeah, right 00:26:57,264 --> 00:26:58,804 [Eric] ... multiple times in a day. [laughing] 00:26:58,804 --> 00:26:59,124 [John] Yeah. Right. 00:26:59,164 --> 00:26:59,624 [Eric] Right? 00:27:00,304 --> 00:27:00,304 [John] Right. 00:27:00,444 --> 00:27:09,404 [Eric] Um, you know, which is, which is kinda crazy. One question that I th- that I asked myself as we were talking about this episode and doing some prep for it was, 00:27:10,624 --> 00:27:16,644 [Eric] what do I want the future with AI to look like? I don't know if I'd ever asked myself that question before. 00:27:16,664 --> 00:27:17,484 [John] Hmm. Yeah. 00:27:17,484 --> 00:27:30,444 [Eric] Uh, this kinda goes into what is the [clears throat] good impact and bad impact. Very similar question. Just pointed it at myself, and that one was very thought-provoking to me. 00:27:30,484 --> 00:27:30,944 [John] Mm-hmm. 00:27:31,064 --> 00:27:36,544 [Eric] Um, and I didn't spend a huge amount of time thinking about it before we started recording, 00:27:37,784 --> 00:27:39,804 [Eric] but there-- 00:27:41,004 --> 00:27:42,144 [Eric] Humans are 00:27:43,324 --> 00:27:46,364 [Eric] creative, productive creatures- 00:27:46,684 --> 00:27:46,964 [John] Mm-hmm 00:27:46,964 --> 00:27:51,384 [Eric] ... in, like, in their nature, right? And 00:27:52,784 --> 00:28:03,164 [Eric] w- the, the first thought that came to mind, my gut reaction to what do you want the future to look, look like with AI is that there's sort of a flourishing- 00:28:03,544 --> 00:28:03,744 [John] Mm-hmm 00:28:03,744 --> 00:28:08,504 [Eric] ... right? Um, which I don't think we've seen 00:28:09,764 --> 00:28:22,304 [Eric] signs of that en masse in the way that it impacts people's personal experience. I think more what we've seen is a drive towards extreme efficiency. So this idea, th- one of the great ironies that's been discussed 00:28:23,324 --> 00:28:30,543 [Eric] a number of times and that people have pointed out is that AI was supposed to make all these things easier, but I'm burnt out because I'm working more than I ever have, right? 00:28:30,543 --> 00:28:30,543 [John] Mm-hmm. 00:28:30,543 --> 00:28:33,564 [Eric] It's like, well, that, [laughs] that didn't really deliver on the promise- 00:28:33,584 --> 00:28:33,904 [John] Right 00:28:33,904 --> 00:28:52,504 [Eric] ... of, you know, sort of removing all this friction and, and creating more margin, right? It increases capacity, but that increases demand for that capacity, and so you have this problem of AI burnout, uh, which is a very real thing. I mean, probably once a week on Twitter you see someone relatively famous in their sphere say, "I don't know if I can keep doing this. I'm exhausted." 00:28:52,544 --> 00:28:53,144 [John] Right. Mm-hmm. 00:28:53,144 --> 00:28:56,864 [Eric] You know? "And I don't like, [chuckles] you know, I don't like this way of working." 00:28:57,284 --> 00:28:57,464 [John] Right. 00:28:57,564 --> 00:29:01,784 [Eric] Um, and so I don't know. That, that to me really... 00:29:02,824 --> 00:29:04,304 [Eric] That was my gut reaction. Like, uh- 00:29:04,544 --> 00:29:04,844 [John] Mm-hmm 00:29:04,844 --> 00:29:07,044 [Eric] ... I wanna see more flourishing and not- 00:29:07,464 --> 00:29:07,584 [John] Yeah 00:29:07,584 --> 00:29:13,404 [Eric] ... just this drive. Efficiency is fine, and it's great, but there's a, there's a limit to the value that that can add. 00:29:13,484 --> 00:29:13,744 [John] Yeah. 00:29:15,044 --> 00:29:20,184 [John] I think, I think mine is more collaboration and more multiplayer 00:29:21,224 --> 00:29:22,164 [John] abilities with AI. 00:29:22,324 --> 00:29:24,184 [Eric] Hmm. Right. Becoming less individual- 00:29:24,704 --> 00:29:24,844 [John] Yeah 00:29:24,844 --> 00:29:25,944 [Eric] ... which most of it is now. 00:29:26,044 --> 00:29:28,004 [John] I think it's so individualistic- 00:29:28,124 --> 00:29:28,264 [Eric] Mm-hmm 00:29:28,264 --> 00:29:34,064 [John] ... and there, there are a ton of dangers with AI at a macro scale. 00:29:34,324 --> 00:29:34,404 [Eric] Yep. 00:29:34,404 --> 00:29:37,864 [John] Cybersecurity, you know, whatever, you know- 00:29:37,944 --> 00:29:38,084 [Eric] Mm-hmm 00:29:38,084 --> 00:29:44,364 [John] ... lots of things in the macro. But, but on micro scale, I think the number one danger is essentially, uh, promoting isolation- 00:29:44,624 --> 00:29:45,124 [Eric] Yeah 00:29:45,124 --> 00:29:46,004 [John] ... between people. 00:29:46,064 --> 00:29:46,704 [Eric] Yeah. 00:29:46,884 --> 00:29:54,424 [John] At work, personally, whatever. And creating... And I think the best place to be with AI, 00:29:55,724 --> 00:30:01,104 [John] uh, w- are interacting with other people, and we don't really have any of those- 00:30:01,424 --> 00:30:01,764 [Eric] Yeah, the form- 00:30:01,784 --> 00:30:03,724 [John] Nobody has nailed that form factor. 00:30:04,124 --> 00:30:04,204 [Eric] Yeah. 00:30:04,204 --> 00:30:05,944 [John] I think there's a cost component to it. 00:30:05,984 --> 00:30:06,504 [Eric] Mm-hmm. 00:30:06,624 --> 00:30:14,244 [John] Um, because in, in-- if everything is already very constrained with the computing resources and data centers- 00:30:14,824 --> 00:30:14,824 [Eric] Mm-hmm 00:30:14,824 --> 00:30:28,764 [John] ... that they need to run what they have now, if you start layering on... 'Cause think about it. Maybe amongst your friends, you're in the top 1% of people that are good at AI. If, if you started being able to have these multiplayer experiences- 00:30:28,804 --> 00:30:28,864 [Eric] Mm-hmm 00:30:28,864 --> 00:30:30,884 [John] ... with other people who are just not as into it- 00:30:31,124 --> 00:30:31,184 [Eric] Sure 00:30:31,184 --> 00:30:37,564 [John] ... like, you have, you could have this takeoff effect, and I think it would have... Like, I think it would melt down most of the AI companies- 00:30:37,604 --> 00:30:38,104 [Eric] Yeah, I agree with that 00:30:38,104 --> 00:30:38,444 [John] ... by capacity. 00:30:38,824 --> 00:30:39,444 [Eric] I agree with that. 00:30:39,844 --> 00:30:39,924 [John] But- 00:30:39,944 --> 00:30:45,164 [Eric] Democratization of that is a compu- like a compute constraint. 00:30:45,224 --> 00:30:46,384 [John] Yeah. It's a nightmare. 00:30:46,444 --> 00:30:46,944 [Eric] Yeah. 00:30:47,064 --> 00:30:53,724 [John] Um, and it's... But I think that is absolutely the best part for humanity- 00:30:53,924 --> 00:30:54,284 [Eric] Hmm 00:30:54,284 --> 00:30:57,264 [John] ... where you're working with other people- 00:30:57,504 --> 00:30:57,504 [Eric] Yeah 00:30:57,504 --> 00:31:05,944 [John] ... with AI versus I, I feel in a, in a lot of companies, there's this secret recipe, get my AI- 00:31:06,184 --> 00:31:06,524 [Eric] Mm-hmm 00:31:06,524 --> 00:31:18,144 [John] ... in this amazing spot. And maybe this is mostly just in developer land, but you might see this too. Get my AI or AIs doing all this amazing amount of work and attributing credit to what- 00:31:18,204 --> 00:31:18,344 [Eric] Mm-hmm 00:31:18,344 --> 00:31:19,544 [John] ... I was able to accomplish. 00:31:19,744 --> 00:31:20,644 [Eric] Yeah, yeah. Yeah. 00:31:20,744 --> 00:31:21,064 [John] And 00:31:22,244 --> 00:31:26,844 [John] part of that is just human nature. It's always gonna be true. Part of that is, what's the other option? 00:31:27,124 --> 00:31:27,364 [Eric] Right. 00:31:27,364 --> 00:31:28,644 [John] How would you even make it multiplayer? 00:31:28,724 --> 00:31:28,984 [Eric] Yeah, yeah. Yeah, yeah. 00:31:29,004 --> 00:31:30,524 [John] And that's a hard problem right now. 00:31:30,604 --> 00:31:32,204 [Eric] All right. Collaborative flourishing. 00:31:32,824 --> 00:31:33,724 [John] Wow. There we go. 00:31:33,744 --> 00:31:34,064 [Eric] That's the future we envision. 00:31:34,064 --> 00:31:34,564 [John] I like it. 00:31:34,564 --> 00:31:35,484 [Eric] [laughing] 00:31:35,484 --> 00:31:37,144 [John] [laughing] That, that sounds very, um... 00:31:39,084 --> 00:31:39,424 [John] I don't know. 00:31:39,584 --> 00:31:43,184 [Eric] Wendell Berry. [laughing] It's very Wendell Berry. We- 00:31:43,444 --> 00:31:43,444 [John] Perfect 00:31:43,444 --> 00:31:48,704 [Eric] ... we can't not talk about AI without... [laughing] Wendell- 00:31:48,904 --> 00:31:50,564 [John] Bob, Bob Ross. [laughing] 00:31:50,584 --> 00:31:53,704 [Eric] Bob Ross. [laughing] There you go. 00:31:53,744 --> 00:31:53,824 [John] Yeah. 00:31:53,844 --> 00:31:55,304 [Eric] Wendell Berry, Bob Ross- 00:31:55,404 --> 00:31:55,404 [John] Right 00:31:55,404 --> 00:31:56,844 [Eric] ... and creative flourishing. 00:31:56,924 --> 00:31:57,084 [John] Right. 00:31:57,084 --> 00:31:57,684 [Eric] You heard it here- 00:31:57,784 --> 00:31:57,864 [John] Tagline 00:31:57,864 --> 00:32:00,444 [Eric] ... on the Token Intelligence show. [laughing]