Automating a broken 15-step process just makes it fail faster. Fix the workflow first, then automate it. Otherwise you're speeding up something that shouldn't exist.
I think with AI, the number one mistake I see people doing is just trying to move faster with AI. And if there's a 15-step workflow process that you're automating, it's, and it's still 15 steps, but now it goes very quick, quickly. You're just going to run into the wall that much faster if that's the direction you're going to begin with. You can drive around town with your tire very low on air. You get on the highway, you're going to crash and burn. Same thing with systems. And so number one is making sure whatever you're working on, you take the time to analyze that process to ensure that it's good enough, number one, to be automated and that it's where you want to go, not where you are, right? So build where you're going.
