[Saadia] Pretty Good

Blogging, creativity, and AI. Perfect is nice, but pretty good will do.

An async podcast by Saadia Carbis and Joshua Wold.

Chapters

• 0:00 Intro
• 0:26 Honouring art
• 3:20 AI advancement will not be at the same rate as Moore's law
• 5:27 Jetlag (sorry, folks!)
• 6:23 Simple vs Complex apps
• 9:07 Refilling your creativity cup
• 10:30 Perfect

Transcript

0:00

Hi. Welcome to Async. This is Saadia. Async is a back and forth conversation podcast between me and Joshua. In this episode, you'll only hear my voice and in the next episode, Joshua will respond and add new ideas and topics. And we just keep bouncing forward between me in Australia and Joshua in the United States. So let's start with some follow up from the previous episode. You've been telling me about some, how are you going with your blogging really? And you're really hitting it out of the park. Your blogging has been sensational recently. Well done. I also recently wrote a piece about Pebble OS and the European Commission imposing like roadmap laws on Apple. I was pretty proud of that one. How do you write more? Along those lines. It's hard because I really am just writing as inspiration strikes. Maybe it's okay to have a couple of standout blog posts amid a bunch of different types of blog posts. Actually, I've just been writing about poetry, not writing about poetry, writing poetry itself. Um, which is kind of strange writing poetry because it could actually be terrible. How would I know? You know, it's like, you can't really tell whether you're a good singer or not. You might think you are, but actually you suck. I think it's kind of the same with poetry, but when I'm read it, I'm kind of happy with it. So I just hit publish and put it out there. There's something about honoring the art, if that makes sense, like taking art seriously. I think that's what I'm starting to realize is my answer to artificial intelligence. You know, like that channel, channeling the divine creative spark. You know, a lot of writers talk about how the, and musicians talk about how words and songs kind of flow through them from some external source, not from their own creative brain, like the muse or the universe or God or whatever. AI can't do that. AI is just mimicking humans who have done that. And, you know, maybe this creative work that we do is bad. Maybe my poetry is terrible, but at least it's human, you know, and human, like in a more important way, because economics, that's human ad based monetization strategies are human war is human. But so is like going to the beach for a swim at sunrise and enjoying that fresh cold water feeling. That's something AI can't do. And poetry, you know, can be mimicked, but channeling the creative muse that that's a human thing. But yeah, AI seems overhyped to me right now, I think, especially after this big AI ish. Well, it was a supposed to be a conference about the cloud and computing on the internet, but really, it was everything AI. And, uh, I think the hype will die down in the next year or so. And maybe the pace of innovation too. It's hard because AI doesn't really behave like regular tech. like, it's not a matter of programming. John Syracuse on accidental tech podcast was talking about how you can't just pull an all-nighter and make AI better. There are some things that it can do a possible job at, you know, not great, but possible like programming or live chat, things like that. AI is pretty good. Let's get in there. But we just, we look at it and we have this tendency to go, that's where it is right now. But think about how much better it's going to be. It's only a matter of time and that's where the hype starts. But I don't think, well, you know, like with the semiconductor based technology, you know, basically it obeyed Moore's law for many, many, many years, but training AI isn't the same thing and it won't work in the same way. Maybe because of semiconductors, we can run, we might be able to run models cheaper and faster but that doesn't mean the models will improve because they don't base, they don't improve based on chip design. They don't improve based on how fast they run. They improve based on the amount of information they can process. So it's just, it doesn't work the same way because we're running out of information and there's no guarantee that the current pace of change will be the same into the future. this might be peak AI. Anyway, back to blogging. You recently had one of your pieces quoted in what seemed to me to be a stranger's blog, someone I'd never heard of. So that's huge. That's exciting. That's like you've made it. Well done. as we say in Hebrew. Well, like, well done. That's super impressive and I'm proud for you and nice work. Maybe one day I'll get there too. Man, I got back from Germany at 1am and I don't know, it's midday now and the jet lag is setting in. I'm fading. Lots of travel, lots of like sitting on planes and watching movies. Refilled my creativity cup a bit. Iterating on Tamale and Peter, the couple of apps that I'm working on, making them better. There was some excitement about them, by the way, with some people I've been talking to. All right, back to follow up. What are some examples of simple apps and complex apps? That's something we've been talking about. Well, I said I really would prefer to work on a simple app and you said that you've been working on something that's more complex and I was thinking more about it because a simple app, that doesn't mean that the implementation itself is simple. What it means is it solves one very, very specific problem it's not so much a simple app as it is a hyper-targeted app, you know, like Flighty or Play or Call Sheet or Lumi. Like, these are apps that Lumi, like, is a great example really because all Lumi does is tell you when is magic hour. That's all it does. But it does such a good job of it. When the premise itself is like simple, you can execute on that premise in a bunch of different and interesting ways. Like, platform-specific ways, you know, live activities, Dynamic Island, interaction stuff, Apple Watch apps. So, like, these apps, they solve one problem and solve it really, really well. Really thoroughly. and then you get complex apps that require a complex approach, have a complex information architecture, rely on difficult technology that maybe doesn't always work or have, like, high reporting requirements and things like that. so, this is apps like Eralo, which is like a eSIM app that lets you add SIM cards from around the world. I use it, it's great, but it's not a great app because it's, it deals with a really complicated technology. Home Assistant, that's another, even just GarageBand and Numbers. See, these apps, like, Slack, that's another one. they've got, they do a good job. You can do a good job with these apps. That's not what I'm saying. I'm not saying they're bad apps. It's just that they have so many different problems to solve and it's just hard. You need a team. You need a lot of testing. You need a lot of time and then it's hard to get all of the features that you want in there because you've got all huge amount of requirements. So, that's what I mean by a complex app and a simple app. I definitely prefer to just focus on simple apps. Alright, so, how do you renew your creative inspiration? That's a question you asked me. Well, you know, when you're running out of things to talk about or blog about or whatever, how do you renew yourself? Because you might have a big, long reading list, sorry, writing list, but not feeling it and so, you said one thing that you do is you read or I guess I would phrase that for me as like input. I have input and I have output and when the output is starting, the flow is starting to slow down, then it's time for some input and that was your suggestion and it's good. I do the same thing. another thing that works for me is just to talk to friends, just hang out, catch up for coffee, fire night, you know, or even strangers, doesn't have to be a friend, anyone really, just be curious about other people's lives, ask them what they're doing, deep dive into, you know, a specific, interesting aspect of their lives or problem they're trying to solve, argue about it with them if you need to and I find that I get a lot of creative inspiration, a lot of refilling from conversations like that. I really love what you said also about something being good and I'll finish on this, that it doesn't have to be perfect, it's just good and then you ship it and then you learn from it, you ignite those feedback loops, it doesn't have to be to be great, just pretty good.