What To Automate?

So with AI technologies progressing the way they are, a lot of you folks are really excited about building cool stuff with it and automating different tasks from your life. The problem is you're automating the wrong stuff. And there's two major reasons why you're picking the wrong tasks to automate in the first place. One of them is in this XKCD comic, which is what you usually end up doing is you pick a task that's not even worth automating because you're, you know, you're spending way too much time trying to automate a certain task than you were spending doing the task to begin with. And as a programmer, you know, we're builders, we're programmers, we're engineers, we love to build, right? And so we are very perceptible to falling into this trap. And we love to build shit, right? And so if you tell a programmer he's going to save, you know, he's going to automate a task that takes him an hour and automating that task, like building the actual automation is going to take him, let's say, 10 hours, he's not even going to think about the one hour he's going to save, right? He's going to think, oh yeah, 10 hours of building stuff, 10 hours of what I enjoy, right? And so as programmers, we're very susceptible to falling into this trap. And if you don't believe me on that, just look at all of the dead side project folders on your desktop right now Stuff that you built and it has never seen the light of day It has never had any users besides just yourself right You probably have like a large fucking graveyard of software projects that never even see the light of day because you enjoy this building process so much. So that's the first reason, right? You're spending time, you're spending more time trying to automate a task than it took to do the task in the first place. And secondly, you're automating a task that doesn't produce any kind of economic value, right? You're trying to automate a task that you probably didn't even do in the first place, or you did it very rarely, or it's just not a very important task to begin with. And so you're just wasting your time trying to build an automation to do it for you because it doesn't really make sense for you in your life, in your work, right? And so in light of both of those issues, I'm going to share with you a little, I guess you could call it a mental workflow or something that helps you decide what exactly you should be automating in your work, in your life, whatever, right? And this is going to take you like five or 10 minutes to go through, probably like five minutes to go through, but it's going to save you weeks of your time because you won't be wasting all that time spending, you know, trying to automate random tasks that don't necessarily need to be automated. It's going to save you weeks of your time and more importantly, all of your LLM tokens that you were going to waste on building these useless automations in the first place So let get into it The very first thing I want you to do is you should not try to brainstorm tasks to automate This is a huge mistake A lot of you are making this mistake. You think about trying to build an automation and the first thing you do is just sit there and think, hmm, what should I automate? That is not the right approach because your mind is just going to give you useless ideas and it's just going to waste your time, right? don't operate on this gut instinct kind of a method. Instead, collect some data first. And to do that, literally just track everything you do over a week, right? A week is a long enough time frame that you'll, you know, figure out exactly what you're doing, where your time is being spent the most, and what kind of tasks would be most optimal for you to automate in the first place. So get some sort of a digital planner, some spreadsheet or whatever you want to use. figure out what tasks are most repetitive figure out what tasks are taking up most of your time and figure out what tasks are simple repetitive stuff and you know stuff that can be automated very easily the more manual the more repetitive the more boring a task is the easier it's going to be to automate with ai and again to figure out all three of these things all you have to do is just track your activities over a week and so this is an example of what that weekly calendar layout might look like right This is not a real timetable by any means It just an example to show you what i mean all you have to do is just have like time blocks over here like this and you're just going to track every single thing you're doing within those time blocks it's going to tell you exactly where all of your time is being spent and then you can figure out well what tasks am i spending the most time on that's probably something you should automate so you can free up the most amount of time and you can also see over here what tasks are most repetitive, right? What tasks am I doing constantly? What tasks are, you know, prime candidates for automation? What tasks are going to be the easiest to automate, right? Repetitive, boring, manual kind of tasks that, you know, kind of follow a deterministic path or a deterministic workflow. They're going to be extremely easy for you to automate. And so that's the low-hanging fruit with the highest ROI that you should be focusing on the first. Not, you know, sitting there trying to brainstorm some dumb tasks for you to build an automation for that you're never going to use because it's not a useful task in the first place. Actually tracking what you do over the course of an entire week and then figuring out which tasks are the most important for you to automate and then building the automations for that. So hope that helps. Do that for a week and then you'll have a clear idea of what exactly you should be building as opposed to just sitting there trying to come up with useless stuff to build. Hope that helps. See ya.