Transcript: The Reality of AI in Software Engineering and the Future of Work
0:00 Brad And this is the next revolution.
0:06 Scott We're gonna kill them, protect me,
0:08 Scott and throw a message, and tell us every tale in the comments,
0:11 Scott so we'll start a few big names and a few times for how many CTOs will share the stories untold.
0:18 Scott Brett,
0:19 Scott Scott.
0:19 Brad An article came across my desk this week about AI coding that I thought was very fascinating.
0:26 Scott Yes.
0:27 Scott Tell me about it.
0:28 Brad Okay,
0:29 Brad so OpenAI,
0:30 Brad the company behind ChatGPT, wanted to find out how effective their models were.
0:37 Brad So they went out to Upwork and they got...
0:42 Brad 1,488 jobs, freelance software engineering jobs off of Upwork.
0:49 Brad These were worth a collective amount,
0:51 Brad a million dollars.
0:52 Brad They ranged from $50 bug fixes to $32,000 like entire projects.
0:58 Brad And they fed these jobs into two of their models and into Claude.
1:03 Brad to see how effective AI could be doing this.
1:07 Brad And they did this because a lot of the tests are versus sort of contrived greenfield type of projects and they wanted more real world examples,
1:16 Brad more diverse examples to really test their model.
1:19 Brad Would you care to venture a guess at what percentage of the jobs were successfully completed?
1:28 Brad By the models,
1:30 Brad the AI coding agents.
1:32 Scott Let me think about this.
1:33 Scott I'm going to try to use the hype scale and the new trend towards vibe coding to predict the success of this.
1:45 Scott Seeing how OpenAI just raised.
1:51 Scott more billions of dollars on an even larger market cap and they keep saying that uh you know we're going to replace all engineers you know with this i gotta imagine it was 100 right scott like they were completely successful in all of these things
2:09 Brad It was not 100%,
2:10 Scott wait
2:11 Brad but
2:11 Scott what i
2:12 Brad I think you're joking around.
2:13 Brad So what do you think it really was?
2:15 Scott don't know in my experience if i ask it to do small well we did a test on this right we asked we asked ai to update our website in three sentences it did a very simple thing very quickly and well and then we gave it a slightly more complicated thing and it went off the rails so at least 50 50
2:34 Brad Okay,
2:35 Brad so credit to OpenAI.
2:37 Brad I'm just going to read you what it says.
2:39 Brad Results indicate that the real world freelance work in our benchmark remains challenging for frontier language models.
2:46 Brad The best performing model,
2:48 Brad Claude 3.5 Sonnet, so not their model,
2:52 Brad resolves 26.2.
2:56 Brad percent of the issues yeah
2:57 Scott 26% of the issues,
2:59 Scott a
3:00 Brad a
3:00 Scott quarter of the time it succeeded.
3:03 Brad quarter of the time it succeeded higher reliability is needed for trustworthy deployment they
3:09 Scott Yep.
3:09 Brad said that right here in the paper that's pretty interesting if i had an engineer who only completed 26 of their tickets successfully
3:18 Scott And the others, other 74%.
3:21 Scott What caused
3:22 Brad unsuccessfully.
3:22 Scott the bugs?
3:23 Scott Unsuccessfully? Couldn't do it?
3:25 Brad Yeah, I don't think they would last very long.
3:27 Scott They'd be on a pip.
3:28 Brad They would be on a pip.
3:30 Brad So here's a little bit further data just digging into it.
3:34 Brad It was most successful at server-side logic.
3:40 Brad That's 42%. It was least successful at system-wide quality and reliability tasks.
3:49 Scott Hmm.
3:50 Brad Now, that's a small sample set.
3:52 Brad There was only,
3:52 Brad I can't tell if it was either three or five in that sample set,
3:56 Brad so that's not very many.
3:59 Brad Then next up at 24% was client-side application logic.
4:04 Brad At 32% was UI UX.
4:08 Brad Those were on bigger samples.
4:11 Scott Interesting.
4:11 Scott Well,
4:12 Scott what I was going to say.
4:13 Scott This maybe speaks to this.
4:17 Scott Um,
4:17 Scott so I've been using a ton of cursor lately,
4:20 Scott um,
4:20 Brad Mm
4:20 Scott in agentic
4:20 Brad -hmm.
4:21 Scott mode.
4:21 Scott And, um,
4:22 Scott so what that means of course, is that when I get it, when I ask it to do something,
4:27 Scott it kind of goes off and starts doing it and try and do things like run its own tests or run its own linting or run its own compile. And then whatever it sees the problem is, it says,
4:39 Scott Oh, let me fix that. And it kind of runs in these loops where it tries to.
4:42 Scott to kind of keep going until it completely solves the problem.
4:45 Scott And one of the things I've started doing now when I use it is I explicitly tell it, give me your proposal,
4:52 Scott do not implement it until I agree with your proposal.
4:57 Scott because and and because it often will get stuck in a loop like i've seen it where like it like for example for whatever reason in our code base we have pretty strict linting rules and it's you know a lot of typescript and um the agent will like write some code it'll lint it
5:16 Scott the lint will fail with a bunch of errors about you know some of our strict rules like you know unsafe any's or whatever and it'll go and try to fix the linting error and it'll create more lint errors so then it'll go try to fix those lint errors by reverting back to the original code which then has the original lint errors and then this gets stuck in this infinite loop of like beating itself up trying to
5:37 Scott undo what it already did and so much so that I've like I said I've started this practice of basically saying do not proceed do not pass go do not collect $200 until I tell you that you can proceed
5:54 Brad Well, yeah,
5:56 Brad that's smart because it needs that.
5:59 Brad And we would do that with our junior engineers also.
6:02 Brad Go make me a plan. Let me look at it before you go.
6:06 Brad You know, work on it.
6:07 Scott Yep.
6:08 Brad So I saw this,
6:10 Brad it came across my desk. And when I say it came across my desk, I mean, I saw it in my LinkedIn feed.
6:15 Brad And I thought it was interesting in regard to the context of the announcement made by Spotify last, not Spotify,
6:23 Brad Shopify last
6:24 Scott Yeah.
6:25 Brad week, where the guy who was the CTO says, or maybe it was the CEO,
6:29 Brad I can't remember.
6:30 Scott It was CEO,
6:31 Scott yeah.
6:31 Brad Prove before you hire someone you have to prove AI can't do it.
6:36 Scott Oh, I didn't see this one.
6:38 Brad Yeah,
6:38 Brad that's what they said before you hire someone you have to prove that AI can't do it
6:42 Scott Oh,
6:43 Scott interesting.
6:43 Scott When I saw another memo that was something from the Shopify CEO that was like, I use AI and I've become using it for everything I do every day, and I expect it.
6:57 Scott You know, you all of you should be doing that.
6:59 Scott You should all be using AI for everything because learning how to use AI is a skill. It doesn't do it on its own. You need
7:06 Brad Yeah
7:06 Scott to know how you need to have the skill to know how to use it.
7:10 Scott And it's, you know, that memo was kind of like I had previously encouraged people to research this.
7:17 Scott Now I am mandating you will be using it in all of your.
7:21 Scott in
7:21 Brad Yeah,
7:21 Scott all your tasks.
7:22 Brad it is a skill. And that's absolutely right.
7:25 Brad I don't know that it's going to be I'm not yet convinced it's going to replace people in the same way that others are.
7:31 Brad I've definitely found it saved me tons of time on some stuff.
7:36 Brad I still have to edit and own what it's doing.
7:41 Brad I can't go blind.
7:43 Brad I've
7:43 Scott Yeah, 100%.
7:44 Brad used the chat GBT deep research for some interesting things.
7:49 Brad It's surprised me, but not for coding.
7:52 Brad You know, when you do it to get yourself some information,
7:55 Brad there's it doesn't have to be perfectly right.
7:57 Brad Code is exacting. It has to be right.
8:01 Brad There's no or it's a bug. Right.
8:03 Brad So but this this shows that it's not there yet to replace people,
8:07 Brad I don't think.
8:09 Scott I think it's a long ways away from.
8:11 Scott replacing um engineers and i think that let's talk about vibe coding for a second right so uh vibe coding is all the rage now where um a combination of people that are not technical at all are using things like lovable or other vibe coding tools and they're just kind of just chatting with the ai telling it to write stuff and they're getting code and then they run it in their sample you know the ai will run it you know like
8:37 Brad Mm
8:37 Scott literally
8:37 Brad -hmm.
8:37 Scott with curse
8:38 Scott cursor in agentic mode it's like you could start with a blank directory and tell it to build a you know a full stack web app and it'll set up docker containers and set everything up and go boom go go to your local host you know 5147 and and here's your web app right and
8:55 Brad Yeah
8:55 Scott it it definitely if you're not a developer who's familiar with things like where do you put your api keys don't put your api keys in your source
9:04 Scott source code you know in your javascript right uh and you don't know that it that's what it does for its first implementation it just does that right and so there was a i just saw this post on um
9:16 Scott I don't know if it was on X or LinkedIn,
9:18 Scott but somebody who basically in 47 minutes, who's not necessarily a, you know, a red team or a hacker type person in 47 minutes basically found exploits in the top 20 lovable apps that are out there on the website. And it was things like API keys were exposed,
9:39 Scott you know, prompts were exposed,
9:42 Scott people's personal.
9:43 Scott personal information was exposed including bank account information and it's it's these types of things that I think are the other side of this coin which is you can go really fast to get a prototype up and running
10:02 Scott Because the agents will do that for you. But if you don't have the wherewithal and the technical acumen to say,
10:10 Scott wait a second,
10:12 Scott don't put my API keys in the source code.
10:17 Scott You don't know what you don't know.
10:19 Brad Yeah,
10:20 Brad that's exactly right.
10:21 Brad Or I've been finding this. I was doing a little thing yesterday. I was trying to get Gemini to write some code for me. I wasn't trying. I was having it do it.
10:28 Brad And it just made a mistake.
10:29 Brad And this thing I was trying to run it wasn't working and I went back to the Gemini chat I was like, are you sure that's right?
10:34 Brad And they were like, oh, yeah,
10:35 Brad that good catch.
10:37 Brad That's that's not right.
10:40 Brad And not I knew a little bit of what I was doing.
10:43 Brad It saved me a ton of time because I got me the scaffolding. But when I tried to run it and then it wasn't it was giving me a strange error.
10:51 Brad I knew how to sort of debug it a little bit. So.
10:55 Brad We'll see.
10:56 Brad I do agree with whatever that was saying. It is a tool that people are going to have to learn.
11:00 Brad I've been talking recently to people,
11:02 Brad you maybe remember this, but I'm going to have to dig up the article from the olden times about digital natives versus digital immigrants.
11:10 Scott Right.
11:12 Brad And that was when people were talking about the internet.
11:15 Brad And for anybody listening who maybe doesn't know,
11:18 Brad a digital native is somebody who grew up with the internet.
11:23 Brad and a digital immigrant is somebody who sort of came of age before the internet and then immigrated to the internet.
11:30 Brad Brad,
11:30 Brad you and I are both sort of digital immigrants.
11:33 Brad Well,
11:33 Brad we're bread on the cusp actually.
11:35 Brad It's like we were brought to the country at a young age.
11:37 Brad But like my dad,
11:39 Brad so immigrants have accents.
11:41 Brad My dad, of course, he's 80 and he is a digital immigrant for sure and he has an accent. You know what his accent is?
11:49 Brad He types in all caps too often,
11:51 Brad you
11:51 Scott Right.
11:51 Brad know.
11:53 Brad um because he writes in all caps that's how he writes he writes in all caps why wouldn't i write why wouldn't i type that way so um well we're doing the same thing again with ai right
12:02 Scott Right.
12:02 Brad you know uh i am going to be in you and i are going to be ai immigrants right we spent a whole life whole career before ai came along there are people who are in the eighth grade today who are going to be ai native and um it's just gonna be part of their lives and they're going to know how it works
12:19 Brad And so it's an interesting thing, I think.
12:21 Brad So the point, the reason I think it's interesting is coming back to what you were saying about,
12:26 Brad I don't remember what you're reading saying,
12:27 Brad I use this every day,
12:28 Brad I'm learning how to use it.
12:31 Brad Just like an immigrant has to learn the language and learn the society and the norms and all these things, it's kind of a similar kind of thing.
12:40 Scott Yeah, I mean, it's interesting.
12:41 Scott There's a part of me that wonders.
12:45 Scott will the you know there's no doubt that there's the huge hype machine right the hype machine talks
12:51 Brad I've
12:51 Scott about how
12:52 Brad never seen so much hype.
12:54 Scott yeah it's it's the hype machine um and you know i mean i i've told you this before too like when folks come to me and say oh what are we going to do about the national debt you know eventually everything the economy is going to collapse because of the national debt and i'm like
13:10 Scott I, you know, I think I think a major revolution in work that comes from AI is more likely to arrive sooner than national debt.
13:22 Scott Right.
13:22 Scott Like and so now,
13:23 Brad Yeah.
13:24 Scott is that part of the hype machine?
13:26 Scott Is that me falling for the hype or is that me just saying, wow,
13:29 Scott man, I I can see what's happened in the last two years of using this stuff.
13:37 Scott That trend line is moving very quickly and it's reshaping the way people work.
13:44 Scott It's reshaping the way that kind of white collar work is done.
13:48 Brad So a commenter on a post of mine on LinkedIn described it as the industrial revolution.
13:58 Brad So we had the Industrial Revolution, of course, prior to that we had the Stone Age, the Iron Age and all these things, right?
14:04 Brad And then we had the Information Revolution. That's what we were just talking about a few minutes ago in terms of the Internet.
14:09 Brad And this is the next revolution. It's not the industrial revolution.
14:13 Brad Maybe you and I can become the guys that coin the name.
14:17 Brad A
14:17 Scott That's
14:18 Brad revolution.
14:18 Scott what it will be.
14:19 Brad Because AI revolution isn't sexy enough. We need a better name. Remember there was the information superhighway. We need to come up with something like that for this.
14:26 Scott It's a bunch of tubes.
14:29 Brad It's a tube revolution.
14:31 Brad I want to ask you something about cursor.
14:33 Brad Do they really call it agentic mode?
14:36 Scott Yes.
14:38 Brad Okay, I've got some grammar issues with that. It should really be agent mode.
14:44 Scott Uh,
14:45 Scott let me just double check.
14:47 Scott I'm pretty sure they call it Agenic Mode,
14:49 Scott but um...
14:50 Scott But
14:50 Brad Because
14:51 Scott you
14:51 Brad in
14:51 Scott might be right.
14:52 Brad our language that I see that tick at the end means like, right?
14:55 Brad It's like an agent.
14:56 Brad But this
14:57 Scott Oh,
14:57 Brad is an agent.
14:57 Scott maybe they call
14:58 Brad It should
14:58 Scott it
14:58 Brad just be,
14:58 Scott agent.
14:59 Brad they call it agent.
15:00 Scott They call it agent.
15:00 Brad Good for the grammar officials at Cursor.
15:05 Brad I was worried about that business.
15:09 Brad Well,
15:09 Brad the problem is we've heard Cursor has been made like by three people,
15:12 Brad right?
15:12 Brad And a bunch of agents.
15:13 Brad You hear about them. They got a billion dollars of revenue with three people.
15:15 Brad people the problem is they didn't actually go hire a grammar person and that's Yeah, the problem
15:19 Scott maybe.
15:21 Scott That could be.
15:22 Brad so i want to since we're on this ai topic in our little mini bonus episode here i want to tell you the thing that's actually been really bugging me lately
15:30 Brad it's i mean i worry about the jobs and careers and what this is going to mean for our industry but the thing that actually really is making me uncomfortable is is the climate change piece of it the energy consumption i mean every one of these queries is you know like running a light bulb for three hours and you know i can't only imagine how much one of these deep research queries costs that's
15:51 Scott Yeah.
15:52 Brad cost a fortune and now we're just sitting here typing it all day long just burning energy
15:58 Scott Yeah. Well,
15:59 Scott I agree that's concerning.
16:04 Scott I don't have an answer to what to do about that.
16:08 Scott It's a little bit like the genie's out of the bottle.
16:11 Brad She knows
16:12 Scott I
16:12 Brad the bottle.
16:12 Scott think that, you know, maybe one of the things to consider is that this does, maybe this is part of the answer of.
16:23 Scott How do we really commit to these reusable forms of energy,
16:29 Scott you know,
16:29 Brad For
16:29 Scott or,
16:30 Brad sure.
16:30 Scott you know, some of the less polluting forms of energy, you know, without not trying to get political on that, but I do think that that's one of the one of the things we know is that technology.
16:49 Scott almost always drives other technology,
16:54 Scott right?
16:54 Brad Yeah, that is true.
16:56 Brad Well, look, if the cost of carbon emission were in the price of energy,
16:59 Brad it would sell itself in an instant.
17:01 Scott That's true.
17:02 Brad And I think we're all getting away on a free ride where we're not having to pay the full price.
17:09 Brad I think that stepping away from energy for a second, I think we're.
17:13 Brad we'll start to see a correction is at some point these companies are going to have to make money i mean google and microsoft they can fund this stuff forever but open ai can't raise 80 billion dollars a year i can only imagine how much they're burning every month they must be burning four billion four billion a month
17:29 Scott Yeah, I don't know. I don't I haven't looked at their financials, obviously,
17:33 Scott but it's it is interesting.
17:36 Scott And that's kind of the point about the hype machine. Right. So like, I think that the
17:42 Scott This is a classic case where the middle always loses,
17:45 Scott right?
17:45 Scott Like these types of technology revolutions, the big established players are the ones that have the advantage to actually be the providers of the platform or the technology.
17:55 Scott And then there are real benefits for the bottom of the curve where it's like those of us that just get to use it,
18:01 Brad right? Mm
18:02 Scott So the
18:02 Brad -hmm.
18:02 Scott two guys in the garage or, you know, whatever,
18:05 Scott the one artist basically gets to use it.
18:07 Scott use it and the technology makes them become more effective and potentially you know have success that way so those are the two ends like the huge corporations of the Googles and the Microsofts and the bottom of the curve with the individuals who get the benefit but anybody that's in the middle is going to get squashed because they're you just can't compete with the big guys right
18:32 Brad Yeah, or they're going to find some way to make it more energy efficient.
18:36 Brad Right,
18:36 Brad and they're gonna the big guys in the middle.
18:39 Scott Yeah, I don't know.
18:41 Scott I think the second that the second use anybody that solves any problem in the middle that's in any way interesting in scale.
18:49 Scott immediately usurped by the big guys right
18:51 Brad That's true.
18:52 Brad That's true.
18:53 Scott that's the challenge and that's part of the reason i would argue that's one of the reasons for the hype cycle right the one reason for the hype cycle is to further protect the big guys and make it that much you know create these you know essentially these barriers to entry to becoming bigger because if they can hype it up then
19:14 Brad I see the most hype coming from people in the middle who have something to sell.
19:19 Scott You're kind of considering OpenAI to be in the middle in this case because they're too small.
19:23 Brad Yeah.
19:23 Scott They're not the Microsoft.
19:24 Brad No,
19:24 Brad I mean, they're huge.
19:25 Brad No,
19:25 Brad but like if I just look at my LinkedIn feed,
19:27 Brad the people who are posting the most about AI is going to replace everything are,
19:33 Brad if you look, what company does this person work for?
19:36 Brad They typically work for some company that sells some AI.
19:39 Brad thing so
19:41 Scott Yeah.
19:42 Brad they
19:43 Scott Yeah.
19:43 Brad got a vested interest there it's
19:45 Scott Yeah,
19:45 Scott for sure. Well,
19:46 Scott that's called capitalism.
19:48 Brad called capitalism all right should we call it a day
19:51 Scott Let's call it a day on the mini episode,
19:53 Scott the mini bonus episode on this week in AI.
19:57 Brad And this week in AI,
19:59 Brad I almost feel like we could do this every week, do a little this week in AI.
20:03 Brad I love it. A bonus episode
20:04 Scott this All right.
20:04 Brad week in AI only for the Patreon supporters.
20:07 Brad Gang,
20:07 Brad I want everyone who's listening to know that you still have the opportunity to be our first Patreon supporter.
20:13 Brad And it costs nothing.
20:15 Brad We are free on Patreon right now.
20:17 Brad So you could be number one. And the first three people.
20:21 Brad where our Patreon supporters,
20:22 Brad if I recall,
20:23 Brad get a boat ride this summer.
20:25 Scott They get a boat ride on your boat.
20:26 Brad That's right.
20:27 Scott All right. Great talking to you,
20:28 Brad See
20:28 Scott Scott.
20:29 Brad you, Bye. Brad.
20:29 Brad Bye.
20:31 Scott This episode of How Many CTOs Does It Take podcast was edited by Safran Epigal.
20:36 Scott Our theme song was AI Generated by Brad.
20:39 Scott Please follow us on all the social media platforms at at HowManyCTOsPod.
20:44 Scott You can find us on YouTube,
20:46 Scott LinkedIn,
20:47 Scott Instagram,
20:47 Scott Facebook,
20:48 Scott Threads, X,
20:49 Scott TikTok,
20:50 Scott and Patreon,
20:50 Scott or at our website at HowManyCTOsPod.com.
20:55 Scott And remember to rate, review,
20:57 Scott and subscribe on the podcast platform of your choice.
21:01 Scott Thanks for listening.