The Stack Overflow Podcast

Developers use AI tools, they just don’t trust them

Episode Summary

The home team shares what our Developer Survey respondents said about AI, spicy opinions about recent Apple unveilings, and an update on crypto regulation.

Episode Notes

Our 2023 Developer Survey explored AI’s benefits for developers. Read about the results here.

For more WWDC talk, listen to our episode from last month: Chatting with Apple at WWDC: Macros in Swift and the new visionOS (Ep. 578).

Squarespace is acquiring Google Domains.

Congratulations to Bruno Brant, who won a Lifeboat badge for answering Where can I view LINQ source code?.

Episode Transcription

[intro music plays]

Ben Popper Hello, everybody. Welcome back to the Stack Overflow Podcast. It's a home team edition. It's been so long. Very happy to have my collaborators, Cassidy Williams and Ceora Ford back on the show. How's it been going, y'all? 

Ceora Ford Pretty good. 

Cassidy Williams It's been great. I had a baby, so that's been a big deal. 

BP Congratulations! 

CW Thank you!

BP Yes, part of the hiatus was a maternity leave and much congratulations are in order. 

CW Thank you. She slept six hours the other night, so I feel like a champ. 

BP For those of you who have not had kids yet, a six hour stretch is everything. I would've taken two hours, then four hours. That's glorious, six hours. 

CW Yeah. I bragged to everybody, every single person. 

BP Amazing. So while y'all were out of town, there was a few big events. I got the chance to go to WWDC out in Cupertino which was pretty fun. If you haven't heard it yet, you can check out the podcast with some of the developers at Apple talking about what they released. So we've already done the deep dive on the software side. Let's talk about Apple Vision. Are you ready for a $3,500 spatial computer on your face? 

CW No. Okay, first of all, a lot of this whole Apple Vision stuff looked a whole lot like it's just putting your phone in the air. It's a lot of 2D screens that just happen to be in front of you, and so I was hoping to be a bit more impressed on that end of things. But then the demos that they did of an organ and a racecar and stuff being just like, “Wow, it's in the room with us.” 

CF Yeah, I'm on Cassidy’s side. I would've expected something a little bit more. I do think it's significant that Apple is taking this step though because I feel like, of the big tech companies, they're the ones that have lagged behind with the AR and the VR and all that kind of stuff. So I'm more curious about what's in store for the future, because Apple seems like they stick more with the things that they create, so I'm wondering if this is a precursor to more things they'll do in this space, which to me is interesting. 

BP I mean, I'll steelman the argument as they say on the pods, I'll try to make the case. So when the iPhone came out people were like, “Why do I need all of this? I can get a phone for a hundred bucks that does everything I want, all the bells and whistles.” And then once they got their hands on it, they were like, “Yes, I needed a computer and email and a camera on my phone. This is the jam.” But to be fair, like you said, AR and VR, lots of folks have tried it at different price points and in different form factors and so far I think they sold something like 10 million headsets last year. So it's not nothing, but it's a small sub-cohort of folks. And then to the point about HoloLens or Magic Leap and, Cassidy, what you mentioned about everything being 2D, so what they said was that they can take any Mac or iOS app and move it over. First it just shows up flat, and then if you as a developer want to, you get to add volume or you get to add motion or you get to add all these other things. They're hoping, and they released different tools for developers to do this, that when folks port their apps over or build them natively, that they'll take more advantage, like you said, of “Hey, this is in space,” but also they want to pre-populate it with stuff. So it's just like any app, you can just kind of flatscreen it on there and you can interact with it, and then they're hoping that people will actually sort of lean into the spatial side. 

CW I have to say, I do like that it can handle gestures and stuff like the pinching and everything. That being said, as I saw it I was like, “Oh, this is straight out of Google Glass 10 years ago,” which was kind of funny to see because I have a dusty Google Glass in my desk drawer right over here because I was one of those. 

CF I think pushing it as something that just your everyday person can use and enjoy and have fun with is not smart because the everyday person is not going to spend $3,000 for something like this that's essentially a plaything. But I think that this could have a big impact especially in medicine and education, and I would be much more excited if I saw more of a push on that front. Imagine being a medical student and instead of having to be in the operation room, you could do something with a VR headset or something like that instead. Those are the kind of things that I could imagine would be like, “Okay, yeah, I could see some university paying 3k to have a bunch of these in a classroom and have people use them for these kind of cases,” opposed to someone who just likes expensive toys paying for this kind of thing just for something like that.

BP Yeah, I think that's totally right. To both of your points, I mean, the folks over at Apple, they're no dummies. They're the most valuable company in the world and they've sold a lot of hardware. They know at a $3,500 price point, it's not going to be mainstream. I mean when the iPhone initially came out, people were like, “Oh, it's $6-700. Do I need this?” And then it was like, “Well, if you sign up for a new plan with your carrier, now it's $200.” And then it became, “Hey, if you change carriers you get a free phone.” So there are ways for people to get an iPhone that didn't end up costing a lot, and like you said, almost everybody in this day and age is like, “I need a mobile phone of some kind.” That had already been decided. Everyone had a phone, whereas everyone does not have VR, AR anything. It's still pretty fringe. I think that they'd been working on it for so long that they had to release something and it was kind of known. Why after all that effort and all that time, and I saw so many people who worked on the project celebrating on Twitter after like, “I have been working on this for like eight years and now I can finally share it with the world.” And it's like, just get it out there. See how it does or doesn't do, let the power users adopt it, and maybe in medicine or maybe in gaming or maybe in architecture they kind of love it and then they have to hope that 1, 2, 3, 4 years down the line, we can cut the price in half, or the size in half, and then cut the price in half again, or the size in half again. That's what they have to hope. I mean, people don't spend $3,500 on a computer. That's a high-end pro computer. That's not even what a laptop costs. So I'm sure they understand where it sits in the market at this moment. They hope that people will try it, feel like it's magic, and that this is the tip of the iceberg for whatever they're doing in R&D. 

CW Yeah. And I also noticed how they didn't use the term ‘metaverse’ or anything like that. They barely used AR, VR, XR, any of those mixed reality type terms. They used branded spatial computing and stuff like that, which I think is very intentional on their part, both for marketing, but also just for that space so that way, as they release it and then figure out who wants to use it the most, whether it be the architecture angle, the medical angle, what have you, they can kind of have the terms owned in those spaces. 

BP Interesting.

CF Yeah. I will say, I think of all the bigger tech companies, looking at Facebook and Google in comparison to Apple, I do think Apple is very smart with the strategy they use for marketing their products and things like that. So I don't think that this is going to be a flop. I don't think this is going to be something that is 1000% not successful and was a waste of everyone's time. I just think we have to wait and see how this evolves. And I might sound like I'm being pessimistic or a hater, but I'm really not. I do think that this has potential and if anyone's going to take it somewhere then it's going to be Apple in my opinion. 

BP It's hard because the bar is so high. I remember when Apple Watch came out and I was just like, “Do I need all this?” And now they're extremely common and extremely affordable and people have them for different reasons. I know older folks who have them just for heart rate and sleep and that's all they do with them and otherwise it's just a watch. And some people use them for notifications or hiking or biking or a million different things. But anyway, my point being that it has to find its purpose and they probably don't expect to sell tens of millions of these like they do with other devices right away. 

CW Right. I do think that the Microsoft HoloLens is the closest thing to it, but they're being branded as such different devices and stuff, so I think we'll see in the next few years how things change. 

BP Okay, cool. All right, let me steer us towards a new topic. So we had the annual developer survey come out and there's lots of great stuff in there. 90,000 people took it, some interesting changes in what languages folks are using, all that kind of stuff. But we did a deep dive on AI, because that's the topic de jure, and interestingly, we found that a lot of developers feel pretty positive about it– that it increases productivity, speeds up learning, improves accuracy, those kinds of things, and that a lot of them are trying them out at home or at work but that they don't trust the output. The output is something that is a starter or whatever, and then you have to go in and fiddle with it, which is fine. Same for when you copy and paste from Stack Overflow. But just curious to get both of your takes from folks you're talking to in the industry, from personal experience, or from what you're sort of observing, how are these AI tools being adopted and to what degree is there hype or kind of substance to it?

CF Yeah, I'll say I think that a lot of people are looking into them. And I've said this before, but I definitely think this has more promise than some of the other trends we've seen in the industry like with crypto and NFTs and stuff. I do think there's a lot of overhype, I do, because you don't know how many articles I've come across that have been like, “Is AI going to replace everybody?” which to me is a bit of fear mongering and things like that. I also do think that we need to be a bit more cautious with some of the ways that we're trying to apply AI. I've also seen a lot of weird investments in AI companies that haven't really done anything but because they have that AI spicy word.

BP There was this French company. It's three people. They were formally from DeepMind or Meta, all of them, and they got like $150 million. It's just like, “We're going to build the best AI. We're going to do it better than everyone else.”

CF And that's essentially it and they've raised all that money. And that to me is like, “Okay.” 

BP They didn't have a new original idea. They were just like, “Trust us. We'll do it better than everybody else.” And it's like “The three of you? Really? Why? Why the three of you?”

CF Yeah, and that's the kind of thing that's kind like, “Uhh… I don't know if this makes so much sense.” But I think that what we're seeing is AI being used more as a tool to enhance the work that we do instead of necessarily replacing a whole development team. I think it's really silly how in some, not even necessarily in the tech industry, just in our world today, I've heard of a lot of people being fired or departments being downsized because they're like, “AI is going to do this for us now.” I don't think that's necessarily smart. I think we still very much so need actual people to do some of these things. I think that more so than anything, I see AI being more of a tool like Google or a Stack Overflow that's going to help us do our jobs better. But no one would ever say, “Oh, Stack Overflow means that we don't need developers anymore. We don't need as many developers anymore.” No, it's just going to help us to move faster and work more efficiently. So that's kind of how it is in my mind, but I'm happy to hear what everyone else thinks. 

CW What you just said reminds me of a headline I saw over the past couple weeks where there was some crisis hotline that laid off all of its workers because they had AI to respond, and then one week later they were just like, “We have turned off the bot because it ended up being a racist,” or something like that where, unfortunately, you can't trust it yet because there's edge cases. And so I do think that there are elements of prompt engineering. There's elements of human-assisted AI stuff that I do think is very helpful. And I think using AI as a tool, like something like GitHub Copilot for example, is really useful because it can get you part of the way there as long as you are guiding it. But just letting the AI do an entire job, that's something that I don't trust yet and I think companies have to earn that level of trust. I was talking to this founder who was accepted into Y Combinator for this next batch, and he was saying that his team is making something that generates entire code environments based on what you want. And you can say, “Okay, I want to add a database to this,” or, “I want to add tests to that,” and as I was talking with him I was saying, “This sounds cool, but I don't trust it in any way, shape, or form.” And so I would need to see a whole lot of outputs and see how much it actually helps me before I leaned into the product and said, “This is a part of my monthly workflow and cost centers.”

BP That makes a lot of sense. All right, so I think we can move on to our next topic. I personally have been the crypto bull in the past and other folks have not been as bullish. And the SEC, which is the Securities and Exchange Commission here in the United States, recently announced some pretty big actions against some of the biggest exchanges to the point where they're basically saying, “We're going to have to fight this out in court.” Some people are being charged with serious crimes and other people are saying, “If we can't sort of come to a resolution with this, we're just going to have to leave the United States. The United States will no longer be a place that has crypto companies, at least not at scale.” So, curious to get your take if you've been hearing this or if you have friends who went into crypto or are still in there. It feels like this is sort of kicking somebody when they're down. The crypto industry is already kind of in a winter and now the government, which allowed certain companies to have IPOs and allowed them to do business for years, is now like, “You're basically an illegal business.” This opens the discussion generally on what you're hearing in crypto or what you're feeling. 

CF Yeah, this definitely feels like kicking someone while they're already down. As much as I kind of hated on crypto –I'll call it hating this time– I would never wish this on any of them, especially to this degree. I think the thing that I've seen the most with the people I used to keep up with or follow, or some colleagues who were into crypto, if they haven't left that behind completely, they've probably focused more on blockchain as a technology instead of the application of it in crypto. The thing that I'm thinking of when I hear this is I wonder if something similar may happen to certain AI companies in the future. And I say this because, not even necessarily in tech itself, but more so when we look at AI's application to things like music and art and writing scripts and books and things like that, I wonder if it's going to come to a point where some entity is like, “We would like to sue this company because they've used my art, my music, my whatever to create their product,” and the government will have to step in and be like, “No, we're not allowing this anymore. You're no longer legally able to function as a company.” That's the only thing that I really can think of, but otherwise, this just sounds terrible and I do think that some of the crypto stuff was very scammy and maybe kind of targeting vulnerable people, but it makes me kind of sad to see something like this happening.

CW I think it’s definitely the result of a whole lot of regulators who don't fully understand the technologies in general. Whether we're talking about the AI side or the crypto side, for a long time people were kind of getting away with it in terms of the good and the bad and how fast they were able to build, innovate, change, scam, not scam, that sort of thing. And now that the SEC has kind of caught up on that knowledge end, they're just like, “Oh wait. We’ve got to slap some regulations on this.” And I have a feeling that's the same thing with AI, that's the same thing with all kinds of different technologies, where it certainly helps to have someone who's knowledgeable in the room be able to explain the technology so that it can be regulated in a safe way. 

BP Yeah. I mean, the legislation is always going to lag behind the technology. That's kind of inevitable. And in this case, it was a technology and industry where a lot of people basically started gambling, and so that I think is kind of part of it too. There was a big upswing and everyone was like, “Crypto seems great!” and companies went public and the SEC allowed that, and now everyone has lost a lot of money and they're like, “Crypto is wrong, somebody must be punished for this.” And it's kind of like, “Well, all right. You should have done that two years ago when everybody was hyping the coins.”

CW And putting all their money in it. 

BP Yeah, exactly. Before they mortgaged the house. 

CF Yeah, that ship has sailed so it really feels like overkill at this point. I wish that instead of using it to still be like, “Let's just punish these people, extremely. Punish them really badly,” I wish they would take this as a learning experience and then be like, “Let's try to apply this to some other emerging technologies to see if we can prevent more harm being done in AI, for example.” It's just fascinating how little the people who make the regulations know about the technologies, like you mentioned, Cassidy. I don't know if you guys kept up with the whole trial with TikTok. 

CW It’s embarrassing. 

CF It is. It really, really is. 

BP Will TikTok connect to my home Wi-Fi and send data over it? 

CW Oh, gosh. I watched a clip of a stream that Alexandria Ocasio-Cortez –not trying to make this political– but she made a statement that kind of stuck with me where she said, “You don't know how difficult it is to explain what livestreaming and Twitch is to someone who had a brand new TV when TVs were coming out.” And I was like, “Oh, that's a good point.” 

CF Yeah, it's very interesting. I've talked about this before about how just generally in our world as a whole there's a huge lack of understanding of how technology works, and as much as we use it, the knowledge people have is not proportional to how much it plays a part in our daily lives, which I feel like maybe we should try doing something about that eventually. I don't know, how many people had computer classes in school that would teach you how to use Microsoft Word and stuff like that. We need an update of that that actually talks about how the web works, because you don't know how many times even people who are in their early twenties will say stuff that I'm like, “No, this doesn't work like that at all.” And the stakes are even higher when it comes to these things like the government questioning, political officials questioning the CEO of TikTok about stuff and asking him questions that just make no sense, which probably has a lot to do with the fact that a lot of those people in those hearings are super, super old.

BP Not that we're against old people. Old people are great. 

CW No, we're not against old people. But if you're going to make laws, you better educate yourself on how something works. 

BP Right, but it's hard to know everything about everything. 

CF Yeah. I think this exposes a huge lack in education that we maybe have to try a little bit harder to fill the gaps. And a lot of us, like Cassidy and I do a lot of education work for developers, but we’ve got to figure out something for non-developers to kind of at least be able to have semi-coherent conversations about this kind of stuff because it doesn't seem like that big of a deal until you realize that these are the people who could possibly be making laws about AI and TikTok and what apps we can and can't use. 

BP Totally. I will say one thing that's interesting about AI, which is that it came out of the gate from a lot of these companies with an ethics team and sort of an impact team built in, which was not true of social media when that came out or mobile devices. They were just like, “We got new technology. I hope you like it,” and now they're like, “Technology really can impact the world at a global scale so we better think about this before we release it,” which is interesting. Not to say that they're going to solve the problems, but for where we are in terms of AI adoption, it feels like the conversation about regulation, self-regulation, and government regulation actually proceeded pretty quickly. We're already talking about it. I would say, if a hundred million people have used Chat-GPT, that's not a big slice of the world. 

CF Yeah. I will say that I definitely feel like more people have been raising red flags with AI compared to social media when it first became a thing. Are people listening? Eh, I don't know. I've heard some kind of icky stories about some of these ethics teams, ethics and AI teams that work for these big companies that got completely fired and things like that. But we are, I would say, very aware of a fraction of the damage that AI could potentially cause if we don't use it correctly. Are we taking the right preventative measures? To be determined. I don't know about that yet. 

BP We've been trained by pop culture movies and comic books and sci-fi novels to be afraid of AI, so we come in intuitively thinking, “Better watch out for this thing that seems smarter than me sometimes.”

CW Yeah. I had to kind of explain to my parents that it’s not like the AI is going to come alive and take over the world like in the sci-fi movies, it's how will people use the AI in ways that we don't want them to and do not great things if we're not careful? 

BP And they're like, “What do I do with the AI?”

CF Yeah. I'm kind of mad about the Avengers, the second movie, Age of Ultron. I think it broke a billion dollars or something like that. Because Ultron is basically an AI that turns sentient and ruins the whole world. I'm probably pretty sure that's a huge reason why a lot of people look at AI the way they do.

BP Oh, completely. How many billions of dollars have been invested in AR glasses because people want to be like Tony Stark. They just want that scene where they're like, “And I move the computer around with my hands.” Yeah, exactly. Well, optics and glass. No programming can solve that.

[music plays]

BP All right, everybody. It is that time of the show. Let's shout out somebody who came on Stack Overflow, shared a little knowledge, and helped to save a question from the dustbin of obscurity. Awarded June 12th to Bruno Brant, “Where can I view LINQ source code?” Well, if you've been curious about that question, Bruno has an answer for you and won a Lifeboat Badge. Congrats, Bruno. You've helped over 14,000 people. We really appreciate it. As always, I am Ben Popper, Director of Content here at Stack Overflow. You can find me on Twitter @BenPopper. Email us with questions or suggestions, And if you like the show, you could leave us a rating or a review, because it really helps.

CF My name is Ceora Ford, and I'm a Developer Advocate over at Okta. You can find me on Twitter. My username there is @Ceeoreo_. And I also have a blog– you can find that at 

CW And I'm Cassidy Williams. You can find me @Cassidoo on most things, and I'm CTO at Contenda.

BP All right, everybody. Thanks for listening, and we will talk to you soon. 

CW Bye! 

CF Bye!

[outro music plays]