Ben and Ryan talk about how tiny nations are making huge money from their domain names, the US government’s antitrust case against Apple, the implications of a four-day work week, Reddit’s IPO, and more.
Small nations like Anguilla (.ai) and Tuvalu (.tv) are benefiting from their coveted domain names.
The US government is suing Apple for violation of antitrust laws, which could have a huge impact on devs, end users, and the whole ecosystem.
Reddit went public last week despite not being profitable since its launch in 2005.
How can you give feedback on a poorly reviewed PR? The Software Engineering Stack Exchange has ideas.
The four day work week is probably not the solution to our work-life balance problems.
AI-powered software development tools like Devon show promise, but their impact on code quality and maintainability remains an open question.
Shoutout to Robert, who earned a Lifeboat badge by explaining Square brackets in CSS.
Chapters
00:00 Introduction
00:31 The Impact of Dot AI Domain Space
01:07 Antitrust Cases Against Apple
04:01 Vendor Lock-in and Apple's Ecosystem
05:08 Issues with Infotainment Systems and Apple Play
06:29 The Benefits and Challenges of a Four-Day Work Week
08:03 Providing Feedback on a Badly Reviewed PR
10:00 The Importance of Clear Expectations in Code Reviews
11:40 The Potential of AI Tools in Development
14:01 Reddit Going Public and the Future of Tech Companies
15:29 AI Tool Devon and the Challenges of Operationalizing AI Projects
21:22 Shoutout and Closing Remarks
[intro music plays]
Ben Popper Hello, everybody. Welcome back to the Stack Overflow Podcast: Ben is almost on vacation edition.
RD Woop! Woop!
BP I'm the Director of Content here at Stack Overflow, and I always bring my co-host with the most, Ryan Donovan, Editor of our blog and our newsletter. Ryan, I'm headed to your homeland, going to Iceland.
Ryan Donovan That’s right, the motherland.
BP Ryan Thor Donovan, going to the motherland. Excited. There's an active volcano, but it's not too active. You get to see the lava but you don't have to dodge the lava.
RD Right. You can do some volcano tourism if you want.
BP We'll see. I have a great little opener for us here. The small Caribbean territory of Anguilla– or is it pronounced Anguilla? Anguilla made $32 million last year from the .ai domain space, 10% of its GDP. So if there's one good thing to come out of this hype cycle, it is that the people of Anguilla are having a boom. For every domain registration, they get anywhere from 140 to thousands of dollars.
RD I love the small countries benefiting from their top level domains. It reminds me of Tuvalu getting hella paid because they had the .tv.
BP Ooh, that's a good one.
RD I think Iceland has a good one, too. They're .is.
BP Oh, that's good. How many people live in Anguilla? Take a guess.
RD Oh, geez. 2 million.
BP It says it's a territory of 16,000 people.
RD Wow.
BP So 32 million a year ain't bad if you're splitting it that many ways. Pretty good. All right, so that was just a little amuse-bouche. That was just a little opener for us. Moving on to more serious things, we had this week a bunch of announcements from the US government that they are bringing antitrust cases. And you and I have discussed this before with respect to the European Union, it seems as though governments feel that now is the time to force Apple to open up its walled garden. Ryan and I do not have an opinion on this one way or the other that we will state on this podcast. This is not financial advice, consult your lawyer, but obviously it's a huge deal for developers and for the entire ecosystem and so can't ignore that news. And also I think the fact that we're talking not about just an area like search where the switching cost is relatively simple, but I am deeply bought into the iOS ecosystem. It would be long, painful and expensive if I wanted to go to Android with my whole family, with my whole collection of devices, with my entire set of cloud services. So opening that up from the point of view of a monopoly is going to be quite interesting.
RD That is a business model. We talk about in other places vendor lock-in. Everybody says that we don't want to get vendor lock-in. But I think I've talked about it before, Apple has had a very protective ecosystem. Their simplicity, their uniformity of design has been a positive, but I did hear something recently that the car makers were also mad at Apple because Apple wanted full control over the sensors and infotainment system when Apple Play connected.
BP Interesting.
RD That seems a bit dangerous to me.
BP That's an interesting area. When something is on, who's collecting the data, how's the data shared? When you're talking about sensors, I would assume we're not talking about my backup camera or the thing that's keeping me in my lane or the thing that's keeping me a certain distance from the car in front of me. Apple Play, which I love and which I have in the car, can kind of take over, for example, the button on my steering wheel that then voice activation. And when they don't play nicely together it's kind of a pain in the butt. I actually have to kind of agree with that. I'm in CarPlay, I'm doing Spotify, I'm looking at maps and then I hit the button on the car and suddenly I'm in Toyota's land and they're not talking to each other.
RD Right. And honestly it would be a testing nightmare for Apple to have to test their system on every car's hardware.
BP Oh, yeah.
RD Are they going to force all the cars to have some sort of standardization? I've known friends who've developed on mobile and one friend would go to conferences just so he could get a mobile phone to test his games on. Hardware variability is no joke.
BP I think that Stack Overflow is the only place I've ever worked that has a fair number of what we call ‘learn, share, grow days,’ where we get a Friday, not off of work, but to invest in personal growth. They offer classes and lectures, and we have summer Fridays. And so we have sometimes what amounts to a four-day workweek with a little bit of one day reserved for other stuff. A good friend of mine, the Chief Strategy Officer at Kickstarter, was in front of a Senate hearing, the Committee on Labor something something last week, to preach to the virtues of the 40 workweek. Apparently it's done nothing but increase productivity and employee engagement and reduced employee turnover at Kickstarter, where he works. And they had a professor there who said that this seems like a kind of remote digital knowledge-worker only thing, but we've actually tested this across manufacturing and food services and police departments and it works. They had some naysayers there from the largest HR organization in the United States or something like that. It was definitely cool to see people bringing hard data to the table after now about three or four years of pilot programs where people have switched. And in all of these cases, not seeing a loss in productivity but have seen a big boost in employee engagement, happiness, and employee retention.
RD I have also heard a lot of these statistics that four-day workweeks boost productivity per day, and I'm for it, but I think a lot of folks will have a hard time adjusting to it. I think a lot of people who have made their brand workaholism or made themselves a brand think of themselves in terms of what their work is will have a hard time working less.
BP There was an interesting thing said that if you mandate it, it's going to be extremely difficult because a lot of organizations would have to rethink all their economics. And what about teachers and people who have five-day workweeks and now their kids. So I think it would be interesting if the mandate was that every employee has the option to choose. If I want a four-day workweek, I will. If I don't, I'll work five days. If I work a four-day workweek, my salary will be X versus Y. Not a full one-fifth less, but less. And you would then have people, instead of having the same teacher five days a week, you have one teacher two days and one teacher three days, and they could work corresponding Fridays or something like that. And the same on a manufacturing line. You just hire a few more people. You pay the same amount overall and they show up four days a week instead of five.
RD And I think having more people doing the same job would definitely be the solution. But I think not mandating it could fall into the trap of the flex time where you're given all this free PTO, take all the PTO you want, and people don't go on vacation unless you mandate it, unless you say you have to take at least two weeks.
BP No, that's true.
RD So not mandating, people will be working five days a week and are you going to do four days a week and people kind of sneer at you like, “Oh, are you being lazy trying to cut out?”
BP Right. Ryan, you put a good question in our Stack Overflow nose this week– “How to get feedback on a badly reviewed PR.” Asked three days ago, already viewed 8,000 times. “In my current workplace, I act as lead developer and architect on a team of software developers. The general policy for merge pull requests is that one person has to review, this person must not be the author. Sometimes code is merged which I personally would not have approved for one or multiple of these reasons: edge case is not considered, changes/additions not designed in a maintainable way, there are inconsistent changes, not all code received proposed changes leading to inconsistent program behavior and bugs, the changes didn't achieve the goal of the overarching business requirements. How can I most gracefully give feedback to both the author and the reviewer? In the past I have been faced with discontent or resentfulness when pointing out that the review did not meet the quality standard I expect for the team.”
RD Anytime you give feedback, it's hard, but also I think one of the things I took from this question is that it didn't meet the quality standard they expected, but they didn't also make those expectations clear. There was no sense of what ‘good’ was in this question. So a lot of the answers were saying to put out a document that says best practices. Here's what we expect, here's what you should do in a review. Otherwise everybody's just winging it.
BP I liked that when they said that for a while, sit down and go over the edge cases with people if they’re new or experienced and then when they know what you expect and hopefully you trust them, then maybe it won't be such a big issue. So I thought that was good advice. Also you're a manager, and when you give people constructive criticism, they're not going to like it and you have to deal with that negative emotion.
RD I always feel like I'm a bit of a weirdo in this case because I almost prefer the negative feedback. I almost prefer, “How can I do better?” If you're just like, “You're doing great,” I’m like, “Oh. How so?”
BP I'll take that to heart. I'll be sure to be more critical of you in the future. All right, once I get back from vacation recharged and refreshed.
RD That's right.
BP Shout out to our sort of peer, our web forum peer. Reddit went public, is now a public company. They sold a bunch of shares to the people who were the most prolific contributors of Reddit. So if you're a big time poster, you got in.
RD Reddit Gold finally means something, huh?
BP The Reddit Gold finally means something. All those shite posts are now worth shares.
RD Share posts.
BP Share posts, exactly. And it was a success. We'll see how it goes. You never know with stocks over the next six months, year, two years, but folks are saying, “Okay, now shows that the window is open,” meaning technology companies can float the idea of going public, even if they're not profitable. And as we all know, the tech industry has been roiled by a lot of layoffs and a lot of downsizing within the last year, so maybe this is a healthy sign for our industry for the world of software developers.
RD Perhaps. This is the way that things were in the past. There was a lot of IPO fever. I hope it is a balancing towards companies that actually have a future instead of another Pets.com.
BP Right. Reddit's future is bright. If my time on Reddit is any indication, its future is bright.
RD No, I love the niche posts on there.
BP Yeah, I love them too.
RD I ran across another AI tool called Devin, which actually does the AI agent thing. It's a developer in a box.
BP Cognition Labs. Yes, I've heard of it.
RD We joke about AI not going to take your jobs but this one seems like it's actually looking to take developer jobs.
BP Right. Did you get to play with it or you just watched demos and stuff?
RD Just watched demos. It's not publicly available yet, but they have demos of it doing Upwork jobs or fixing issues in open source repos or getting a blog post and then implementing the code.
BP My conspiracy theory at the moment, if I were to float one, is that a lot of the big AI labs are holding back the next generation of tools because they fear the repercussions and response, and that if you were to just bring out the next version of ChatGPT-5 today, ChatGPT-6 today, that the improvement over 4 would be so striking that it would really be unsettling. And so I think in a responsible way, they're trying to– and Sam Altman did an interview this week and he said, “We're trying to do releases in a way that doesn't catch people by surprise or make them feel uncomfortable and that we're going to be more iterative, that you'll see 4.6, 4.75, 4.8 before you would see us just say 5.” So I don't know, we'll have to see. I think we need to get our hands on Devin. It brings together, from a team of medal winning coders and mathematicians, what folks have been toying around with ever since ChatGPT-4 came out, which is creating an autonomous agent and adding layers of planning, critique, and retries on top of what Gen AI is good at, which is offering up potential solutions.
RD To tap the brakes a little bit, it is just demos. Everybody kind of knows at this point that the demos look good, but let's see it, let's get somebody else to play with it. So I am interested to see where this goes, but it's also a sort of societal change if these sort of things happen.
BP We had a great discussion with Bill from GitClear. He made a lot of really interesting points. He is a software developer, has been, still remains an engineer. End of the day when he's feeling lazy, he's accepting more of the suggestions from the AI, but he finds that that code churns more quickly, that it gets reversed, that it's not as high quality and that it doesn't use existing methods and so it doesn't maintain at the sort of coherence that you get in a code base over a while. And so I thought that was really interesting because two things can be true at once. Yes, it makes developers faster and more productive. Yes, sometimes it makes them feel happier. But B, the quality may decrease and the tech debt may increase. And so we don't know yet how those two sides of the seesaw balance out. That was an interesting one. I hope people listen to that episode because he had a lot of good stuff to say and they looked at a pretty huge amount of work from developers across a pretty sprawling set of code bases to make that happen. And then the other one, speaking to your Devin example, and this is a striking demo, but we haven't actually seen it out in the real world to get good feedback. We have a talk coming up with some folks from Google Cloud and MongoDB. We haven't done the podcast yet, but one of the things that they were sort of saying which I thought was striking was that industry surveys and data that they've been looking at says 10% of the Gen AI projects that people are working on have actually been operationalized or put into production. The other 90% are like, “Is this going to work for us? It's just a toy demo. We tried it and it didn't work. It's not worth our time.” Not to say that 10% is not a big number in one year. If we're one year in and 10% that’s big.
RD And I think operationalized at scale, too. So it's something that is able to run with a number of customers. And I think that either speaks to the vast number of people that are playing around with generative AI or just the difficulty of getting it to be something that is usable for customers.
[music plays]
BP All right, everybody. Thanks as always for listening. If you hear my voice the week after next it means I survived my trip to the volcano. And by the time I return, we'll probably have made a few interesting announcements and posted a few interesting blog posts, so I look forward to getting back on the pod with Ryan. Let's shout out somebody who came on Stack Overflow and helped us out. A Lifeboat Badge was awarded to Robert two days ago. In classic Stack Overflow form, the question is, “Square brackets in CSS.” I wouldn't say that's written in question form, but that's okay. “What does it mean when something between square brackets in CSS?”
RD Square brackets in CSS?
BP Right. “What does it mean when something between square brackets in CSS? E.g. input[type="radio"]” Well, Robert, we appreciate you coming in, helping out with this question, even if it wasn't phrased as a question. Gave a great answer and helped over 44,000 people, so appreciate you spreading some knowledge. As always, I am Ben Popper. Find me on X @BenPopper. Email us with questions or suggestions. If you've played around with Devin, have thoughts on Gen AI, got an interesting coding project that you want to discuss, podcast@stackoverflow.com is the email. And if you like the show, you can leave us a rating and a review.
RD I'm Ryan Donovan. I edit the blog here at Stack Overflow. You can read it at stackoverflow.blog. And if you want to reach out to me, I am on X @RThorDonovan.
BP Thanks for listening, and we will talk to you soon.
[outro music plays]