Why is software so bad? (E)

0:00Forty years ago, I drove a brown Fiat Strada, perhaps one of the 10 worst cars ever made. It went 0 to 60 in about 19 seconds. It got about 20 miles to the gallon. The backseat was remarkably uncomfortable. It handled poorly, and it was really ugly. Today, for about the same number of inflation-adjusted dollars, I drive a Toyota Prius plug-in hybrid. It gets 199 miles to the gallon. It goes zero to 60 just as fast as I need it to. It's super safe, and the backseat is quite comfortable. Hey, this is Emily in the Bronx, and you're listening to a special archived episode of Akimbo.

0:57I don't really want to talk about cars today. I want to talk about software, and I want to talk about why software is so bad. I developed my first computer game as a hobby in 1976 when I was in high school, and I was super fortunate in the early 80s to do it professionally. In those days, making games for the Commodore 64, our major limitation was hardware. This is what the music sounded like.

1:31The text was so clunky it was almost unreadable. The graphics were nothing to write home about. We were busy pioneering how software might work. At the time, I was beta testing the original Mac, and the Mac was a revelation. On that Mac, I had a word processor. Soon after that, a spreadsheet. Today, 40 years later, I have a word processor that does almost exactly what that word processor did 40 years ago. I have a spreadsheet, Google Sheets, that does less than the spreadsheets I used to pay for. Sure, it's free. Sure, it's connected and can have multiple users, which makes it even more useful.

2:18But in terms of software development, if we leave aside the network effects, most of the things that I and you do with software are not dramatically better than they were soon after the Mac came out. Why is that? What happened to this industry that is no longer driven by hardware? The hardware is now as fast as we need it to be. The screens are as sharp as we are able to discern. And the colors, unless you're a sea slug, are infinite. So what all this means is that software, the architecture of software, isn't what it could be. I want to share a few reasons why I think this is. The first one is the buying cycle.

3:12Cars have made a lot of forward motion, even though they are largely hardware dependent. Well, why is that? First, there's a buying cycle. Every three, four, five, six years, we take a deep breath and we start over. The operating system of the car hasn't changed in almost 100 years. If you know how to drive one car, you know how to drive almost every car. But when we start over, we can switch brands. We can go from company A to company B. and we're about to spend$50 ,000,$30 ,000,$70 ,000 for the privilege of doing that, which means that the car companies are under a lot of pressure to create something that users think is better.

4:00That's not true with word processors or spreadsheets or video editing software. We are stuck with the operating system, with the method, with the UI, with the file format that we are used to. Number two, car companies have dealers, and dealers talk to consumers face-to-face. Dealers live and die every day. Dealers see that people are walking off the lot, and they are happy to scream directly at the people at the car companies who have no choice but to listen to them. That's not true for software. Software isn't sold for nearly the same price to a consumer, and it's usually sold directly without a middleman.

4:45Number two, as we discovered about 20 years ago, the network effect is actually the killer app. Software exists primarily today to connect us to other people. That the way Microsoft ended up destroying WordPerfect was by showing up with file formats that could be shared one person to the other. It wasn't worth it once someone in your office was using Word for you to insist that it had to work with WordPerfect. Shareable file formats meant that it went in one direction. Google gives its software away, but we'd probably pay for it because the magic of the network effect is so overwhelmingly powerful.

5:34We would forgive the fact that you can't do nice typography and other features in Google Docs. We would get over the fact that Google Sheets isn't as fast or as reliable as Excel at its best because the network effect, multiple users using the thing, overwhelms everything else. As a result, the biggest brains, the smartest people, the hardest driving focus at every software company tends to be about how do we get people to share this, not how do we architect it with care so that the experience of using it on your own is delightful. Number three, which might be the biggest one, is that software is complicated.

6:25It's complicated for the user. When I think of how many instructions I need to give somebody to be able to learn to drive a new car, it's probably three minutes. Here's one button. Here's the lighter. Here's how you turn on the cruise control. And don't forget about this. Off you go. You can walk into a rental car agency, get a car you've never driven, and drive it off the lot. Software isn't anything like that. that the number of options, because options are super easy to add, keeps increasing. One person wants to, I don't know, composite reverse type in these colors with an alpha channel, and suddenly that's one of the features.

7:08As a result, the architecture of software is significantly more complex because the architect doesn't know what the user wants to do. And there is no convention that has been accepted for turning off features so that you can have the version that you want and have it work the way you want to use it. The next idea is that culturally, we stopped giving prizes for craftsmanship. That when the early versions of Keynote came out, people moved to it away from PowerPoint because it was well crafted. But that was more than 10 versions ago. Since then, there hasn't been a lot of discussion about what it means to be good at crafting how to create presentation software.

8:06I think you and I could sit down and come up with 20 ways to make it significantly better. Then when a new company like Prezi comes along, what we notice are a couple gimmicky moves. not that it is perfectly crafted the way perhaps a Porsche is a better crafted car from the user car experience point of view. Not that you asked, but as someone who's given more than a thousand presentations using Keynote, here are some of the things great presentation software would do to make my job easier and to make the experience of consuming the presentation better. Number one, how come there isn't a timer built in that shows me, compared to what I expect, where I am through my presentation?

8:54Show me with colored lights that I'm behind or ahead. How come there isn't a way to group up subsections of a presentation so I could sub them in, turn them on, turn them off easily, not with some sort of clunky hierarchy? Number three, why can't I instantly glance through all of the material I've used in the past to find which subsets are available, which ones I want to add? It goes on and on. None of these features have been added. Instead, what I've got is now the ability to change the outline on a font, which no one has any business doing. That's not what it's for. but nobody seems to be in charge of making Keynote more elegant and useful.

9:44Instead, it's sort of a random collection of ways to get more people to share it and use it together. It's not becoming more powerful or more beautiful. It's simply becoming more clunky. The same thing is true with almost all the software I use. Apple, which used to lead the way in figuring out how to give us power and leverage, now dumbs things down because they're a luxury brand, not a group of people trying to craft tools for folks who are trying to change the culture. And then back to this idea that software is complicated. We've been trained to put up with it. So a specific example. About four years ago, some versions of the Mac laptop began to lose connection with Wi-Fi networks, sort of randomly.

10:37No one's exactly sure why. And you can find posts online from 2015, 2016, 2017, 2018, with people complaining about the same bug, with solutions that verge on witchcraft. delete this, do that, stand on one foot, over and over again for years and years. Try to imagine a car that didn't work at many gas stations on a regular basis and no one knew why the filler cap wouldn't open at some places under some circumstances. Try to imagine that this went on for year after year after year of the car being sold. I think it's pretty easy to believe that that car would be recalled, that that brand would plummet in value, that our standards for what we're looking for from our$40 ,000 car are really different from what we're looking for from the software that we make a living using every single day.

11:41Now, users are complicit because as software companies stopped caring about the quality of experience, users started stealing the software because the network effect is powerful. Because incrementally, a second copy of a piece of software in the world doesn't cost the company money the way it would cost Ford Motor money if you went to a dealer and drove a car off the lot without paying for it. Conceptually, digital goods have always had this marginal cost problem. It doesn't feel as wrong to steal software as it does to steal a car. None of this would matter except that software drives the culture.

12:29When they came up with new ways to do editing, new ways to do special effects, the movies we watched began to change. The Terminator, that guy with the silver skin, happened because someone made software that would enable it to happen. That changed our understanding of how the future might look. That changed the kind of cars that got made. That changed our expectations of what tomorrow would look like. Software changes not just the way our movies look. It changes the way that we tell a story. It changes the way we consume a story. It changes our attention span. That the network effect at all costs means that we've got clickbait.

13:16It means that people are hunting around for hours every day on their smartphones, a device that didn't even exist when I sat down with that first Mac that was on my desk. Our attention span has been shifted because software engineers are in a hurry. And this lack of care about architecture and user delight combined with the overwhelming effect that networks have means that we are victims of a cycle that has been driven by a hundred or a thousand cutting-edge software engineers who are making decisions. Making decisions about whether to make something more beautiful or more profitable. making decisions about what tools we're going to have or not have, about how long we should spend on something before we get to the next thing.

14:09We began with software being our tool, a tool to help us do our work. But over time, we have become software's tool, that we exist to enable software to reach its goals. And its goals are to turn us into subscribers, networked paying subscribers who keep making more stuff that enables the network to thrive. It is entirely possible that hardware isn't going to get much better. That back when I was making games for the Commodore 64, I saw a hard drive burst into flames because our software was making it work too hard. And yet, just last week, Google updated Chrome, which caused the editing stations of hundreds of companies in California that were making cutting-edge commercials and movies to stop working on their Macs.

15:09Totally stopped working, unable to reboot, to do anything. That's 40 years later. Software is a mess. Software is complicated. Software is driven by the network effect. And we are the victims of it. If we stand up, speak up, and argue that this thing that we are spending our entire day using ought to be better, if we establish standards, awards, heroes, maybe we'll find another Andy Hertzfeld. Maybe we'll find another Susan Cain. Maybe we will find another series of architects and designers who will insist that software can be beautiful, that it can be powerful, that it can help people express the ideas they want to express, as opposed to being victims of a commercial system that doesn't have our best interests at heart.

16:06That's my rant. Thanks for listening. We'll see you next time. Hey Seth, it's Maria. Hey Seth, my name's Kyle. Greetings, Seth. This is Stephen out in Madison, Wisconsin. Hi Seth, Alicia from Charleston here. Hi Seth, this is Anupam. Hi, this is Caitlin. Hi Seth. Warm greetings from Curacao. Hey Seth, my name is Nick Ryan from Pittsburgh, Pennsylvania. Hey Seth, this is Rex. Hey Seth. Hi, this is Vasilis from Greece. Hi, this is Roberta Perry. My question is... And that completes my question. I really do love hearing from you, and it's okay if you want to ask a question that isn't about this week's episode.

16:42A question that isn't related to any specific episode, just something that I was going through. My question is more of a general question rather than the topic of this podcast. To ask your question, just visit akimbo.link, that's A-K-I-M-B-O dot L-I-N-K, and press the appropriate button. Hello, hello, Seth. Ederson Oliveira here from Toronto, Canada. You keep mentioning here and there throughout your podcast that we are in an open marketplace moment in time, but maybe this will change in the future. When we have an open marketplace, as we do now, but maybe not for long. It seems that you are painting a future that we are inviting the gatekeepers back in the building.

17:30This is worrying. And what are you referring to? Is that about net neutrality? What is it? Are we inviting the gatekeepers back in the building? Something dramatic has happened in the last 15 or 20 years, and it's this. 100 years of gatekeepers, all right, 500 years of gatekeepers are being replaced by systems that are dramatically more open. If you wanted a business profile written about your company in 1974, well, you could use Businessweek, Forbes, or Fortune, and that was pretty much it. Now, you can write your own. Now, you can be featured on more than 500, 10 ,000, a million websites that could talk about you.

18:14In the 1980s, if you wanted to be on TV, there were three, maybe five people who could put you on TV. Now you can put yourself on TV. One medium after another. Sound, video, text, all of it. Wide open. But my instinct is that that can't last. It can't last because over time, people seek to consolidate, to lock in, to create monopolies and oligopolies. And we're seeing it, for example, in what Netflix is trying to do in the business of television. Of course, Apple and Amazon are racing after them as hard as they can. But this idea that there would be an open place where someone could make a thing and get paid to make it, Well, already we're getting boxed in.

19:09We're getting boxed in as podcasters consolidate, as people in the movie business consolidate, because that is where the stock market wants them to go, because they want to extract the maximum amount of revenue. So my argument begins with that and then proceeds to the idea that there's only a finite number of people, of voices, that a listener can follow. That permission is a self-limiting function. Once you've got your problem mostly solved, you stop giving permission to new voices, to new options. And so we have this wide open area, this land that's being carved up. And I think if you seek a thousand true fans, if you seek to change the culture, waiting is not your friend.

19:59Diving into it now makes a lot more sense. Hey Seth, this is Neil from Portland, Oregon. I am a huge Akimbo fan and anxiously await each new episode each week. I feel like I'm a living, breathing, embodiment of one of your core messages, which is if you do good work, people will follow it and they will share it with others. I find myself talking with friends, family, and colleagues about your episodes and share each podcast episode with them. Akimbo is a really interesting example, which brings up my question is how do you create content that is more shareable? I was talking with someone about Slack in their business and wanted to share an Akimbo episode with them, but found it very hard to search for that episode so that I could send it to them.

20:46Thanks again for all the work that you do. Looking forward to the next episode. Yeah, shareability is one of the factors in how ideas spread or how viruses spread. We know, for example, that the measles has an R0 of up to 18. One person with the measles can infect up to 18 others. That's why it's such an impactful virus as it spreads through our ecosystem. And the same thing happened on Twitter. Because Twitter is so bite-sized, so easy to share. people share it. And because people are sharing it, an idea can rocket through the Twittersphere in just minutes. Making your idea more shareable without giving up any other element of it is generally a good idea.

21:33But too often, to make an idea shareable, we dumb it down and make it less sticky. And it turns out that stickiness is even more important if you want your idea to persist. So back to the idea of the measles. The thing is that hepatitis has an R-naught of only two when it's peaking, which means that far fewer people are infected by someone who has hepatitis, but it can stick with someone for the rest of their life. That stickiness means that it is a notable and important disease, not because it spreads widely, it's easy to share, but because it sticks around. So part of what we need to do is not just make our idea in a package that's easy to say, hey, look at this, but also to have the guts to put ideas into the world that stick with people.

22:28Hi, Seth. This is Dan from Florida. The question essentially is, why is it so easy to help others edit their work, but then it's so hard for us to edit our own? If you have any advice, I'd love to hear it. Thanks and have a great day and thank you for all that you do. I think there are three reasons why it might be difficult to edit your own work compared to editing someone else's work. Idea number one is our context blindness that happens all the time. There will be a typo in a blog post of mine. After I've read the sentence 10 times, I still don't see it. So fresh eyes get us through that problem.

23:05But more important than that is the idea that we don't see things that are important and others might. We might not see them because we're afraid of them. And we might not see them simply because our point of view is different than theirs. So asking someone to look at something with fresh eyes, who is open to giving us that sort of generous insight, that's precious. And then the third one is this. We don't like to be wrong. We don't like to be wrong, and editing our own work requires us to admit that we could have made it better. So one of the things that it takes to be a good writer, a good creator, is willing to suspend your desire to be right and instead embrace an instinct to be better.

23:58Hope that helps. Thanks for listening. We'll see you next time.