Asked why Things built its own sync engine instead of leaning on iCloud, Jainek argued that the foundation has to be provably correct, and then turned to what current AI coding tools cannot do. He runs a company that had just shipped an AI-powered Siri integration, which makes the dismissal less predictable than it looks. His co-host's reply treated it as the clearest statement in the episode of what separates a human developer from a machine.
that's right. Yeah, use your
own. Why did you choose to do that as opposed to use iCloud? Why did you choose to build your own app? Why on earth?
…that it works. And that's what we've done and what we love, because then you can start building on top of it. And to this day, users write in and they just rave about the sync being excellent. And that they sadly don't have quite the same experience with other apps. So we love this. And because you asked about AI specifically, I think this is an area where current AIs are just not good at. Like the code they produce works, but there's no guarantee you have how well it works. It's almost certainly not thought through well because there's nobody thinking. There's no brain, actually. Yeah. Like there's no formulation of a concept first that you then manifest in code and then you adhere to that concept strictly and stringently to make sure that it actually works Survives and so forth. None of that you can get with current AIs. So I
could be wrong, but I feel like you just articulated kind of base level what the difference is between a human developer and an AI machine. And while the AI can do certain things, but like you're saying, there is a code, there's like a core part of the human developer speaking into the app that is…
Werner, one question I had to ask you is: you make a productivity app, but what are the other productivity apps that you use?
Okay, that's a good question. So I like MindNote, for example. I love that. I'm sure you guys know it. Yeah, that's great. I use a launcher. Currently, I'm using, what is it called?