Too Long; Didn't Read
AI projects, especially automation, can easily slide into one of two, or both, bad outcomes. The first bad outcome is the project that never ends. The second is when an automation is built that doesn’t do what’s intended. This is when you’re left with a bad process that is automated, so you must create a new process to fix the mistakes. This means that you, the end user, and client, never have to touch it again. In an AI world, we want to frame the problem so that the problem is solved once-and-for-all.