AI code generation is like a 3D printer: it makes individual parts fast, but humans have to assemble them into something that works. AI struggles with the dynamic moving parts of system design and orchestration. The real value is in what you do with the pieces it gives you.
Use AI like a 3D printer. 3D printers are super innovative, very clever stuff. But you'll find that people who build complex things using a 3D printer actually build parts and then they assemble those parts together. The assembly is really what requires the human touch. And the, you know, the creation of the individual pieces is part that you can outsource to the 3D printer. This, I think, is a really important concept when it comes to like developing, you know, using AI to generate code, that what it can do super well is put together the pieces otherwise can take quite a long time to put together. Like imagine 3D printer trying to make that piece of plastic, molding it yourself, you know, taking a little knife to, you know, whittle away at it. Instead, it creates it for you. What it does not do well though is going to be the assembly. Similarly, like a 3D printer, you want to think moving parts won't work very well at all. Consider that metaphor. Consider your job of composing in the middle as being very important to gain the maximum from these tools.
