The AI Hype Cycle: Navigating Growth and Skepticism

0:00 Scott: Is it possible the AI bust is coming?

0:07 Brad: We're gonna kill time protecting me and the rest until you know us every time to become it so.

0:12 It's not a true,

0:13 big names and the reasons I'm for how many CTOs will share their stories on

0:17 Scott: All

0:17 Brad: code.

0:18 Scott: right,

0:18 Brad,

0:18 welcome back for another episode of How Many CTOs Does It Take?

0:22 My name is Scott Paul Rad and this is my trusty Hugo host.

0:26 Brad: Brad Hefta-Gaub probably

0:27 Scott: I feel like I mumbled the word co-host,

0:30 Brad.

0:32 Brad: did. That's fine.

0:33 That's fine.

0:34 Scott: Yeah, Brad,

0:35 we're here this week to do another AI roundup.

0:40 It's

0:40 Brad: Yeah,

0:40 Scott: been a while since we've done an AI roundup.

0:43 And I personally feel like since we talked about it last,

0:47 there has been a change in.

0:51 The AI vibe,

0:53 and I don't mean the vibe vibe coding way, I mean what the world is thinking about AI.

0:57 What's your take?

0:58 Brad: I 100% agree with I've seen a sea change,

1:03 at least in the way that big tech is talking about AI and in particular the way that the CEOs and the and the VCs are talking about it.

1:16 And I think that's pretty that is something interesting going on there.

1:20 Scott: Tell me more.

1:21 Brad: Well, the it was it was six months ago. It was nine months ago that the CEO of of Anthropics said that, you know, in six months, all code would be 90 percent of all code would be written by AI or something like that.

1:36 And that, you know, engineering jobs were going to go away in 12 to 18 months.

1:43 He's not saying that anymore.

1:45 Scott: Right.

1:45 Brad: um that's one example um and i think that uh you're sort of seeing this this shift start to play out a little bit you know you're starting to seeing a different tone from you know the head of all the big uh you know um ai players you know open ai anthropic you know uh google everybody microsoft um and so i think that's really interesting and the vcs are starting to kind of cool their jets a little

2:11 It's a little bit too about the way they talk about stuff.

2:14 Scott: Mm hmm.

2:15 Brad: So that's huge,

2:16 I think.

2:18 Scott: In the hype cycle,

2:19 do you think we are in the trough of disillusionment?

2:22 Brad: Yeah, I think maybe.

2:24 Scott: Let's talk about the we need to tell everybody. So there's the innovation.

2:29 Trigger is the first phase.

2:31 Then there is the peak of inflated expectations.

2:34 I

2:34 Brad: Oh,

2:34 Scott: think we're

2:35 Brad: I

2:35 Scott: past

2:35 Brad: think we're past that for sure.

2:37 Scott: the peak of inflated expectations.

2:39 Then there is the trough of disillusionment. I can't pronounce that word today.

2:44 The trough of disillusionment.

2:46 Brad: There you go. You got it now.

2:48 Third time's the charm.

2:49 Scott: The slope of enlightenment and the plateau of productivity.

2:55 Brad: Yeah, I don't know. It's interesting.

2:57 I wonder if we're actually in the full trough or if we've just started.

3:02 But this particular hype cycle has gone so much faster than most other hype cycles that we could be all the way into the trough now.

3:09 But I definitely feel like we're coming off the high of the peak of inflated expectations.

3:13 Scott: Yeah,

3:13 I don't know how to put it in the hype cycle, but I feel.

3:16 feel like we're in we're off that high and we're in a spot now where people are actually just kind of finding real things to do with it i guess that is the slope of enlightenment but i think the ai slope of enlightenment okay here's what i think has changed i think because the peak of expectations was so high that people thought the impact was going to be crazy high

3:40 But I actually, so they thought it was going to be V-shaped,

3:43 right, upside down V-shaped, like up,

3:45 I don't know how to describe it. But I actually think it's more like U-shaped, like the peak was really high,

3:51 but it's going to take, I think it's going to be a decade long grind of finding the value in it.

3:59 It's

3:59 Brad: Yeah.

3:59 Scott: kind of like the internet,

4:00 really.

4:01 I mean, like the internet started,

4:02 there was 2000,

4:03 you and I were there, it was crazy peak,

4:05 there was a big bust.

4:06 But then really over the next 15,

4:09 20 years is when it became the thing, right? It's 2002 to 2022,

4:15 right?

4:17 It was really those 2010s, it was the 2010s when it really took off.

4:23 Brad: It started getting its flywheel going,

4:25 yeah.

4:26 Scott: It just was everywhere,

4:27 you

4:27 Brad: Yeah,

4:27 Scott: know, like, you know how I think about it?

4:31 There was a time, I don't know when it was, when you could live in America, at least without the Internet.

4:41 Brad: right.

4:42 Scott: You

4:42 Brad: What did

4:42 Scott: could still

4:42 Brad: that end?

4:43 Scott: go to a government office,

4:44 you could still send a letter somewhere, get a letter from something,

4:47 right?

4:48 And at some point, you cannot function as a.

4:52 I mean, I'm sure you can,

4:54 I'm sure there are people who do, but in general,

4:55 you need the internet.

4:57 Brad: Yeah.

4:57 Scott: You need to go renew your driver's license,

4:59 you basically need the internet,

5:01 whatever it is, right?

5:01 Brad: Yeah.

5:02 Hell,

5:03 I just got a traffic ticket. I got stuck in the box because I couldn't get all the way out of the intersection because traffic jamming.

5:15 The whole thing came to me email.

5:17 It came to me an email.

5:18 I could only pay it on the internet.

5:20 I was like, man,

5:22 like, wow.

5:24 Scott: So yes,

5:25 there was actually there's this piece that happened where there was an assumption society somehow shifted and there's an assumption that everybody will have access to the Internet.

5:36 And the smartphone was a key part of that story.

5:38 Brad: Yes, that's true.

5:40 So how does this relate to the hype cycle of AI?

5:44 Scott: It took us a long time to get to the assumption that everyone had Internet.

5:49 Like we had the boom in late 90s, 2000,

5:53 then we had the bust and then there was the grind and it took like 10 or 15 years and then all of a sudden the assumption of ubiquity was there.

6:00 Brad: Yeah.

6:00 Scott: And so now I feel like we're on the slope of enlightenment. The slope of enlightenment in my framing could be called

6:07 the slow uphill grind until the assumption of ubiquity. And we're going to get to ubiquity at some point, but we're not there and it's going to be a long time.

6:15 Brad: OK,

6:15 so this is this is going to I have some things to share that are going to be a little bit less tech and a little bit more sort of like economics and investing and things like that.

6:25 So, what was the number?

6:27 50% of GDP in the first half of 2025 was AI data centers for our whole country.

6:37 50% of GDP growth.

6:40 Scott: Whoa.

6:40 Brad: GDP growth.

6:42 Scott: Growth.

6:42 Brad: GDP

6:43 Scott: Well,

6:43 Brad: growth.

6:43 Scott: now what are you, hang on a second,

6:45 what does that mean?

6:45 So GDP is gross domestic product.

6:47 That's on

6:48 Brad: So

6:48 Scott: how

6:48 Brad: basically

6:49 Scott: much stuff,

6:49 Brad: like,

6:49 Scott: that's like our country, our country's income,

6:51 Brad: yeah,

6:51 Scott: right?

6:51 How much

6:51 Brad: so.

6:51 Scott: we made.

6:52 Brad: Yeah. So we, you know, like basically like, you know, typically the U.S. is got about it has about, you know, three to four percent annual growth is basically what we've been at for the last 10 years.

7:04 Fifty percent of that growth was in AI data centers.

7:08 And we

7:08 Scott: And

7:08 Brad: actually had

7:08 Scott: so when

7:08 Brad: shrinking,

7:09 Scott: you say it was,

7:09 Brad: we actually had shrinking, you know, we had shrinking GDP because of other economic decisions that have been made recently about tariffs, et cetera.

7:17 And it's the AI data centers that have basically kept us from having negative GDP.

7:24 Scott: and is that, so GDP is actual like.

7:29 our income, our national income.

7:32 So that means people are paying like tons of money to data centers.

7:34 Brad: Well,

7:35 that's the thing, right?

7:36 It's all the big guys are all spending money on building new data centers for the AI.

7:41 And it's like, and it started to get into this very strange like friends cycling on top of each other type thing where.

7:51 nvidia is selling uh chips to companies and then renting them back from the companies there's like all kinds of crazy like churn and and sort of like it's not any there's nothing nefarious going on but there's all these like

8:06 partnership deals where it kind of looks like the money goes in and out of two different people's pockets and therefore that gets counted twice and that kind of creates a bigger gdp number when in fact it's really just i'm handing you ten dollars you're handing me two fives and then i'm handing you a 20 and then you hand me back you know you know two more tens or something like that it's like some you know goofy like no real money changed hands but a lot of money changed hands multiple times

8:34 And that's that's the last basically 2025 has been this massive churn of the big guys spending money mostly like all that investment that's gone into the big powerhouses.

8:47 It's all building data centers.

8:51 Scott: So this is interesting,

8:54 a minute, you know, we lived through the first dot com bust.

8:57 Brad: Yeah.

8:59 Scott: Is it possible that AI

9:00 Brad: Yes.

9:01 Scott: bust is coming?

9:02 Brad: Yes.

9:03 That's what I was about to say. That's why I don't think we're in the trough of disillusionment because we haven't had some kind of a bubble burst.

9:13 now so many people are intertwined on this so many of the big players are intertwined on this you know that they're it's in their vested interest and and honestly like i think that they're serious i mean again i don't necessarily want to get into the politics of this but there could be some very serious uh like uh

9:31 too big to fail type situations going on here um because

9:35 Scott: Mm-hmm.

9:35 Brad: what because guess what we haven't replaced software engineers but now we have like you know open ai released the you know the latest version of sora which basically creates 10 second you know viral video clips for you and they're like you know like that's where all that's the latest announcements it's just like it's it's

9:55 slop or bubble gum or it's not we're not really seeing you know like one of the things everybody's talking about oh you know this this ai will maybe cure diseases we're not that's not happening that's not i know we've had a year of people predicting that ai was going to you know come with these amazing breakthroughs of you know curing rare diseases or curing cancer or something like that and they're not any closer to that

10:21 Scott: Okay,

10:22 I saw something on the internet recently,

10:25 which is a totally unreliable source,

10:27 by the way,

10:27 Brad: Yeah,

10:27 Scott: but

10:28 Brad: the internet.

10:28 Scott: I just tell you what I saw.

10:30 Brad: Yes.

10:31 Scott: I saw it on Instagram.

10:33 Brad: Oh, the

10:34 Scott: So 100

10:34 Brad: more reliable of the unreliable sources.

10:38 Scott: % unreliable, but it caught my attention and I thought to myself,

10:42 I wonder if that's true.

10:43 This was talking about radiologists and there were a few things that caught my attention.

10:48 The first thing it said was the FDA has approved.

10:51 I can't remember if it said 70 or 700.

10:54 It must have said 70, 700 is too many.

10:56 The FDA has approved 70.

10:59 AI type models to do radiology,

11:03 computers doing the radiology as opposed to a radiologist looking for broken bones and stuff like that.

11:08 Brad: Apparently in January they had approved over 1,000 clinical AI applications.

11:15 Most aimed at radiology.

11:16 Scott: Maybe 700 was a real number.

11:18 Brad: It

11:18 Scott: Maybe it might was 700.

11:19 Brad: looks like I found an article on radiologybusiness.com that says 758 approved applications.

11:30 Scott: There you go.

11:31 And now if you think about AI or machine learning even and radiology, it makes a ton of sense because

11:38 Brad: Yeah,

11:39 Scott: it's just a

11:39 Brad: I'm

11:39 Scott: pattern

11:39 Brad: pretty

11:39 Scott: match.

11:39 Brad: sure I'm pretty sure those are those are we're calling them AI nowadays,

11:44 which is fine,

11:45 but those are probably classic machine learning hugging face,

11:48 you know, various models are not LLMs doing that.

11:52 Scott: That's

11:52 Brad: So

11:53 Scott: right.

11:53 Brad: let's be clear when we say AI,

11:55 we're not.

11:55 Not always talking about LLMs.

11:56 Scott: 100% everything's AI now literally you know I got new tires the other day and they're called they're AI tires I mean everything it's

12:04 Brad: They tell you when they need to spin.

12:05 Scott: I mean it's amazing there's a there's a there's a I watch for our listeners out there I'm a fan of European soccer and there's a European phone company called auto and they're a sponsor of one of these teams or something or one of these tournaments I don't know

12:21 I don't know. They have their AI phone.

12:22 It was like AI hadn't even been out 15 seconds and they were like, boom,

12:26 it's an AI phone now.

12:27 OK,

12:27 but that's not what I wanted to say about the radiologists.

12:29 They got all these models and they thought radiologists were going to all get fired because the computers were going to be doing it.

12:37 Instead, but what they what this thing said, and maybe your radiology today or whatever you're reading says the same thing. They said, actually,

12:44 the cost of reading an x-ray has gone way, way down.

12:47 And so there's just more happening.

12:50 There's more need for radiologists to do this stuff because they can do it faster,

12:55 which I thought was pretty fascinating.

12:56 So there are.

12:57 some good things happening out there but you're right there's a lot of make me another viral clip so when does the bubble so now i'm thinking as an investor not political i'm thinking just like how do we get rich off this and i'm thinking

13:09 So how does the bubble actually burst? Is it burst when one of these companies like Microsoft or Oracle or whatever starts to say,

13:17 oh, yeah, we've way over invested in data centers and we've got to write some of this stuff off?

13:21 Is that when it bursts?

13:22 We need to get an economist on here.

13:24 Brad: Mostly, a lot of this stuff is happening in the private markets,

13:29 right?

13:29 Like,

13:29 you know, there's very few public companies that you can invest in that.

13:35 are 100 ai plays right so you can obviously invest in nvidia you can obviously invest in microsoft apple google but you can't invest in open ai you can't invest in anthropic right that's it's it's it's all private and unless you're unless you're you know a billionaire hedge fund you're not you're not investing in them um so yeah i mean that's one of the other things that you know it's kind of interesting is like

14:02 The dot-com bubble that we all know about,

14:06 that also was the hey, hey, you know, rowdy, you know, open market, you know, public going public and everybody IPOing and then, you know, a lot of those companies blew up.

14:15 We're talking about, you know, $100 billion valuation seed rounds these days. That's not,

14:21 those companies don't make sense on the public market ever,

14:24 really.

14:25 So I don't know.

14:26 But but you but but let's let's pivot back to let's let's let's get let's get to the technology of AI for a second.

14:34 Scott: Yeah, yeah.

14:34 Brad: You you suggested that we should look at Hacker News.

14:38 Scott: Oh, yes.

14:39 Top 10 articles in Hacker News.

14:40 How many of them are about AI?

14:43 Brad: Well,

14:43 I already looked.

14:45 Scott: Whoa, rats.

14:46 One, only one on my view of Hacker News.

14:49 Brad: Well,

14:49 that's funny.

14:50 I got two.

14:51 Scott: What are your two?

14:52 Maybe

14:52 Brad: I

14:52 Scott: I should be

14:53 Brad: got

14:53 Scott: you.

14:53 Brad: number three,

14:54 Claude skills are awesome.

14:56 Scott: Yeah.

14:56 Brad: Maybe a bigger deal than MCP.

14:59 And then number seven, Claude code versus codecs. I built a sentiment dashboard from Reddit comments.

15:05 Scott: Yes, I just refreshed my Hacker News and now I see that.

15:12 Brad: And then just below the fold is number 11 asking AI to build scrapers should be easy,

15:18 right?

15:18 Scott: Now on my reload,

15:20 that's number 10.

15:20 So three of the top 10.

15:22 Brad: Oh, Claude skills are awesome, isn't a bullet. It's jumping up to the top.

15:27 So yeah.

15:29 Scott: Number 13 on mine is something I'm very interested in, but slightly afraid to click.

15:34 Brad: Meow camera.

15:35 Scott: meow.camera what do you think is what do you think is behind that link it's

15:40 Brad: A little afraid to click.

15:43 Scott: like oh it's exactly what you would expect it is a cat it's a cat cam

15:49 Brad: Oh, wait.

15:51 Oh, it's multiple cat cams. Look down the side. You can look at Mr.

15:55 Happy.

15:56 Scott: I wonder if this is a product that people sell a little.

16:01 Now, okay,

16:02 we need to pause for a second here.

16:04 Everybody needs to know something that I spent five years of my life working on a website about cats.

16:11 Okay.

16:12 And doing, redoing ridiculous things on the internet.

16:15 And I, this is,

16:16 I wonder if this is like a cat feeder that you can, because there's hundreds of them you can look,

16:20 well, not hundreds of them. There's many of

16:22 Brad: they

16:22 Scott: them all you

16:22 Brad: look

16:22 Scott: can look

16:22 Brad: the same

16:22 Scott: at here.

16:22 Brad: yeah

16:23 Scott: They all look the same. I wonder if this is a product that you can buy and then you can plug yourself in and these people who are running this are genius because they know people love to look at pictures of cats.

16:35 Brad: i i think you're right though i think they all look like they have the same

16:38 They have the same product.

16:38 They Yes. all look like it's a bowl that's got a camera.

16:40 Scott: Very interesting.

16:42 So it looks like Meow.com is an alternative viewer for the feeders from the HelloStreetCat app.

16:50 Brad: Ah,

16:50 Scott: we're down we are now down the cat rabbit hole

16:52 Brad: we're down the cat rabbit hole. And see,

16:54 Scott: i'm this taking us

16:55 Brad: is

16:55 Scott: how out

16:55 Brad: it

16:55 Scott: of

16:55 Brad: works.

16:55 Scott: the cat rabbit hole this is how i made a living for five years was the cat rabbit hole it turns out that people really like to look at pictures of cats on the internet and you can actually run a business doing that yeah

17:08 Brad: Just showing cats on the internet.

17:10 So let me ask this question back to

17:12 Scott: yeah

17:12 Brad: the AI topic.

17:13 What is the coolest thing you've done with AI?

17:15 in the last month.

17:18 Scott: I haven't done it. I've seen it.

17:20 Brad: Okay, tell me.

17:21 Scott: So, because you know what? I haven't done anything with AI in the last month. I haven't even done any.

17:26 computering in the last month we do the other things going on in my work and life but I'll tell you something I saw see people in this podcast know we I work at a company called TaskRabbit right now TaskRabbit has an API for partners who would like to integrate with TaskRabbit it's like a it's a public API in the sense of like

17:48 other people use it but it's private in the sense of we don't just let anybody use it you got to like call us up and say hey we want to partner with you but one of our engineers in like less than an hour with three prompts essentially built an API client

18:03 He said, here's our documentation.

18:05 Here's the API.

18:07 I want you to make a thing called Henry's discount car delivery service.

18:13 And here's the requirements.

18:14 I think what impressed me most is it was three prompts.

18:18 Brad: Yeah.

18:19 Scott: And

18:19 Brad: Yeah.

18:20 Scott: it did it. I mean, look,

18:21 it was just a little demonstration site for users of the API to see how it worked.

18:26 But the fact that.

18:28 he was able to do that and was really impressive to me.

18:33 And I think the other thing I want to say about that, you and I have talked about this a lot.

18:40 It seems to me that the more talented engineers seem to be getting more out of AI.

18:49 And this is one of the most talented engineers on the team, one of the most senior engineers on the team who has invested time in figuring out how to make it work.

18:59 And someone said we need a demonstration app for our customers.

19:03 And he was like,

19:03 I can whip that up quickly.

19:06 And maybe it would have taken him a week to do it by hand.

19:10 Maybe it would have only taken two and a half days.

19:12 Who knows? He did it in one hour with three prompts.

19:15 Brad: Yeah.

19:16 Scott: Which I thought was really impressive.

19:18 What is the most interesting thing you've done or seen?

19:23 Brad: It's funny as you were saying that, like, I don't have I shouldn't have even asked you this question because like I don't have a good answer myself. But what you know what I've seen that I am I'm I'm really pleased with is that.

19:38 I am just seeing the engineers that I work with on a really regular basis,

19:43 they are more and more they're up leveling their game by pushing the way the AI tools that they're using to be better written AI code.

19:58 So I'm just.

19:59 Scott: them do harder and harder things and getting good quality higher and higher quality results and

20:05 Brad: Mm-hmm.

20:06 Scott: you can see this because here's how i here's how i noticed it so so i noticed that the commit messages and the pr descriptions by all the developers are all now clearly being written by by ai

20:24 And I've done this myself where I've got like a, you know,

20:27 I was using the.

20:31 The approach that Michael Natkin gave us about the idea of like, you know, have it build an arc,

20:37 you know, write yourself an architecture.md document,

20:40 then give another AI a chance to like, you know, break that down into tasks and then assign those tasks to AI agents.

20:48 And the AI agents do a really good job of their commit messages are very specifically written in a way that's like obvious.

20:57 like the stepping the thought process they went through and

21:00 Brad: Mm-hmm.

21:01 Scott: so you can read a commit message or a a pull request and it was you can definitely see that it was written by ai not a develop not a human developer and so as i track that i'm seeing more and more and more of the commits and the prs that are clearly the messages were written by ai and and

21:21 The things that the developers were doing were really quite advanced,

21:27 quite advanced things,

21:28 and they clearly used AI to do the bulk of it.

21:31 And code quality is quite good.

21:35 One thing that's really funny is when you when you said because,

21:38 you know, we talked about this a couple months ago,

21:40 how the AI code review agents were pretty good.

21:44 Right.

21:44 Hmm.

21:44 Yes,

21:45 Get

21:45 that's right.

21:45 an AI code review agent to do a code review of an AI agent's code.

21:52 And you see in really

21:53 Okay.

21:54 interesting dynamic.

21:55 What do you find out?

21:57 Well, six months ago, the AI code review agents were complaining about a lot more things.

22:06 Oh,

22:06 really?

22:06 Now the AI code review agents.

22:10 tend to complain less about the code written by AI.

22:12 And it's not because they're getting laxer. It's because the code is getting better.

22:18 And I've just been watching this. As I watch

22:20 Brad: Hmm.

22:20 Scott: the commits fly by,

22:21 you can see that it is

22:23 Brad: It's learning.

22:23 Scott: getting better.

22:24 It's getting better.

22:26 Brad: Do you have any naysayers on your team?

22:28 Scott: I don't have any naysayers on my team.

22:30 The people that I thought were the biggest naysayers.

22:33 Um,

22:34 I've now, you know, we at least on the at least on,

22:38 you know, my main team, you know, we joking, we have a term in our language we call.

22:43 my robot overlord suggested this that's what we say we say

22:46 Brad: Yeah.

22:46 Scott: that so everyone's now you know integrated that into their language they're like you know me and my robot overlord worked on this you know this spec document me and my robot overlord worked on this um uh um dr and everyone now is fully embracing that as the as kind of like it's kind of a joke but

23:07 they're using it they're really using it as a tool and some of the folks that were the most naysayers i see them using it more and more now so i i'm pretty i'm pretty impressed with the way that and people are and i feel like we're going a lot faster than we were so oh

23:20 Brad: I got one naysayer on my team. I got to figure out how to handle that situation because I feel like it's a little corrosive.

23:27 Scott: that's too bad well what what is that what's the objection

23:30 Brad: Well,

23:31 it's all the objections.

23:32 It's in, you know, it's incorrect sometimes.

23:34 uh it'd be faster to do it myself uh it uh you know can't do very complex things etc i mean they're they're not that's not that things aren't untrue but i feel like it's a mindset it's

23:51 Scott: A hundred

23:52 Brad: kind of to me if they feel like excuses of like i don't want to do this and so i'm gonna as opposed to saying

24:01 interesting somebody's created a new tool to help me do my job better I'm going to figure out how to use it they're just saying all the reasons why they shouldn't use it

24:08 Scott: You know, it's funny.

24:09 I wonder if I think some of the people that were the this the early naysayers in my teams have been the people that were.

24:19 are the type of people that are paying really close attention to the latest and greatest tech gadgets.

24:25 And so they were exposed to them really early.

24:29 Brad: Mm-hmm.

24:30 Scott: And I think the people that are embracing it more robustly are folks that either just have growth mindsets anyway or came to it later when it was already better.

24:43 And so a part of the sort of like if the naysayer is a naysayer because they tried it early,

24:49 right,

24:50 then I would say focus on trying to get them to try it again because the new stuff is just so much better.

24:58 Brad: Here's something that they wrote the other day.

25:04 Scott: Okay.

25:04 Brad: So we use a tool called CodeRabbit for automated code reviews.

25:08 Scott: Yep.

25:09 Yep.

25:09 Brad: No relation to TaskRabbit.

25:12 Scott: Yep.

25:12 Brad: I suggested to our lawyer, by the way, that we should contact them with like a copyright thing just to see if we could get free licenses.

25:20 but he he didn't first of all he didn't find that funny and second of all uh he was like that'll be so much work blah blah i don't think we'll have a good case i'm like i don't care about the case i just want to see if they'll get you we'll threaten them and they'll give us some free software i

25:33 Scott: Have you learned anything from the current political climate? It's not about whether you actually have a case or not.

25:40 Brad: wasn't even serious i was just joking and he didn't find it funny anyway uh code grab it's great at catching stupid typos that inevitably make it into

25:46 get into poll requests.

25:47 However,

25:48 its accuracy at providing relevant comments to even slightly complex issues drops dramatically in my experience.

25:54 Scott: That feels like I feel like there would be several easily retrievable examples of legitimate bugs and problems that CodeRabbit caught.

26:08 Brad: Mm

26:09 Scott: happens

26:09 Brad: -hmm.

26:09 Scott: all the time to my team some of my best engineers like literally like we we're now click we click that we click we asked for the code review on all the things and like you know it's like before i even look at a pull request i'm i'm looking at what the what the you know the code ai code had to say about it and sure enough like i had one of my top engineers wrote this cool thing it seemed like it was testing and i i ran it and it's like it said oh you missed a you basically

26:34 accidentally had a knot in a place that you didn't want to have a knot right it was like pot like an exact opposite error that was a it was a it was an uh hard to get to you know branch in the code but it was it was exact opposite wrong and the developer's like oh shit yeah it caught like a legit straight bug there like yeah that's a problem

26:57 Brad: Well, now I'm bummed out right now.

26:59 We talked about such cool stuff a minute ago, and now I'm talking about naysayers.

27:03 Scott: Yeah, well,

27:04 I mean, there's always going to be what is a hater's going to hate, right?

27:07 Brad: Haters are going to hate.

27:09 Hey, can we finish on a more positive note?

27:11 Scott: Yes, finish on a positive note.

27:13 Brad: Also on Hacker News,

27:14 there's another important article.

27:16 Scott: I closed it already.

27:17 Tell me.

27:18 Brad: No, don't look.

27:19 I'm just going to tell it to you.

27:20 It'll be a quiz.

27:22 Scott: okay yeah

27:23 Brad: The Rapper 50 Cent. Do you know 50 Cent?

27:26 Scott: get

27:27 Brad: Are you familiar with 50 Cent's music?

27:29 Scott: rich or die trying get rich or die trying i

27:31 Brad: There you go.

27:31 Scott: got a couple of 50 50 cent albums okay

27:34 Brad: Adjusted for Inflation.

27:36 Scott: like when the first album came out you mean get rich or die trying what is it adjusted for inflation when did that come out 98

27:46 94.

27:46 Brad: According to the 50-cent Adjusted for Inflation Tracker,

27:51 which

27:55 cites Wikipedia,

27:58 the

27:59 Scott: Here we

27:59 Brad: name go. 50 Cent was coined in June 1994.

28:04 Scott: Okay,

28:04 so 50 cent is now $2.73.

28:11 Brad: And 109 cents.

28:14 Scott: Uh,

28:14 well,

28:14 I mean, I'm not that far off. I mean, you know,

28:17 Brad: You

28:17 Scott: for

28:17 Brad: were over

28:17 Scott: a magazine.

28:18 Brad: 100% off.

28:19 You said 273 cents and it's 109 cents.

28:26 Scott: I mean,

28:26 inflation's been, look,

28:28 the rent's too damn high,

28:30 Scott.

28:30 The rent's too damn high.

28:32 Brad: Apparently the rent's cheaper than you think it is.

28:37 Scott: I love

28:37 Brad: Well,

28:38 Scott: it.

28:38 Brad: this episode

28:38 Scott: All right.

28:38 Brad: has

28:38 Scott: 109

28:39 Brad: been bonkers.

28:39 Scott: cent. 109 cent. Get rich or die trying.

28:43 Brad: We should see if we can get 109 cent on this podcast.

28:46 That would be the day.

28:48 Scott: that would

28:48 Brad: When

28:49 Scott: be

28:49 Brad: we're big enough that 109 cent is coming on our podcast,

28:52 I

28:52 Scott: that'll be the day all right scott uh

28:54 Brad: feel like

28:55 Scott: this

28:55 Brad: we got to stop

28:55 Scott: was

28:56 Brad: it

28:56 Scott: an

28:56 Brad: now.

28:56 Scott: oddball i you gotta stop this episode this was out of this was all out of left field but uh the ai hype cycle something's going on it's it's a interesting tectonic plates are moving and uh yeah but i'm still coding like crazy with it because it's it's makes me faster

29:13 Brad: Yeah.

29:14 All right.

29:14 Scott: All

29:15 Brad: Well,

29:15 I'll

29:15 Scott: right.

29:15 Brad: talk to you next week, Brad.

29:16 What

29:16 Scott: All

29:17 Brad: are we

29:17 Scott: right.

29:17 Brad: supposed

29:17 Scott: Talk to to you next

29:17 Brad: do?

29:17 Scott: week, Scott.

29:19 We're rate,

29:20 review,

29:21 subscribe,

29:21 smash that subscribe button.

29:23 Give us five stars and yeah,

29:26 to the top,

29:27 Brad: Well,

29:27 Scott: baby.

29:28 Brad: I mean, inflation,

29:29 give us a 10.9 stars.

29:31 Scott: 10.9 stars. Exactly.

29:33 Brad: All right. See you, Brad.

29:36 This episode of How Many CTOs Does It Take podcast was edited by Saffron Eftegout. Our theme song was AI Generated by Brad.

29:44 Please follow us on all the social media platforms at at HowManyCTOsPod.

29:49 You can find us on YouTube,

29:52 LinkedIn,

29:52 Instagram,

29:53 Facebook,

29:53 Threads, X,

29:54 TikTok,

29:55 and Patreon,

29:56 or at our website at HowManyCTOsPod.com.

30:00 And remember to rate, review,

30:02 and subscribe on the podcast platform of your choice.