So...we all know my election predictions have...shall we say, greatly improved in format over the course of this election cycle. While I developed my existing methodology back in 2016, I didn't really perfect it until now. A lot of it is because those predictions took time. Take my 2016 election prediction chart. It actually took me the whole day before election day to put this together. And then I ended up editing it as new polls came in. Again, HOURS. This was HOURS of work. For ONE chart. And while I did keep reusing that one chart and filling out the data again and again in future predictions, once again, this took HOURS.
This is one of the reason I only did election predictions once every few weeks, and earlier on this election cycle, I would be spotty with them and not be able to release them as I wanted because I could only fill out the chart properly on my desktop, and with it being up to 95F in here because summer, i would be suffering in the heat for hours just to do ONE prediction. Not wanting to ahve my predictions be dictated by the weather, I developed a solution. Gee, why don't I make this chart in excel and have it fill itself out and update itself from me just inputting new data? That way I could save hours of work, do it in the air conditioning, and release a chart when I felt like it.
So, I spent the 4th of July doing exactly that. My family was grilling food, and I was working furiously on replicating my chart in excel. And it took all day to figure it out but by that night, I was able to produce my first forecast.
At this point, it was a bit rough. If I updated it, sometimes it would break. If I added new lines for additional states, it would break. But it worked. And I was able to produce the same charts I always did exceedingly easily. Now, labor saving inventions are intended to save labor, but such is the curse of growth. Sometimes we dont just work less on the same tasks, sometimes we want MORE! We pursue more growth instead of more leisure. And because election predictions are a passion project of mine, and I'm an autist obsessed with my special interests, I ended up being so proud of my new chart that I started doing predictions more often. And talking more about the race in light of the predictions. And I kept improving my chart.
Eventually, simply having the chart wasnt good enough. Remember how I used random number generators to create hypothetical outcomes to the election? What if I could do that in excel? So by the end of july, I worked on that, eventually creating a rudimentary simulator that spat out a randomly generated outcome with the push of a button.
Now I could generate outcomes whenever I wanted.
I even experimented with, with the help of a friend who knows more about excel than me, a version of the simulator that generates hundreds or even a thousand outcomes at once. It worked...kinda, and did allow me to understand the flaws in my methodology that caused these particular simulations to have limitations, but it was cumbersome, it broke constantly simply by updating the data. And it would crash the spreadsheet or slow things to a crawl when it was used. I gave up on that because there was way too much maintenance associated with it, and returned to my simple 1 at a time simulator, but I did try to grow in that regard too. I still might bring that back just for my final election prediction maybe.
As such in a sense, a lot of time I'd be spending making simple charts once every few weeks has been spent on improving my model, with me often investing a lot of hours id otherwise dedicate to election predictions to improving my chart to give me better election predictions. This is kind of like how, when we automate a job, we end up creating new jobs associated with fine tuning the automation of the previous jobs, and we end up going from low skill grunt work to high skilled brain work dedicated to improving efficiency and making things better. As such, in a sense, more efficiency doesnt always lead to reduce time spent on certain tasks, just reallocating that time toward more complex tasks.
And this is the problem with the modern economy. We're reaching a point with automation where most low skilled tasks are at risk of automation, leaving mostly white collar workers automating the work away and using their high skills to continue making things more efficient. This will, inevitably, lead to a reduction in work over time, if we allow it, but because we have a system toward more and more endless growth, we keep subjecting people to a never ending cycle of work instead, with the work getting less pleasant and more miserable as tasks that remain are often menial and don't pay well unless youre part of the higher end of the economic spectrum that can focus on the automation and increasing efficiency of the more basic tasks.
And the growth continues. Eventually I found out how to add election maps directly into my spreadsheet, so I've done that. And to my simulator too. So now I got election maps that I dont have to make on yapms any more (although admittedly the yapms ones are a bit nicer in some ways).
And yeah. My model just keeps improving.
And again, the simulator has a map too now:
Yep. Change the outcome, change the map...
It's not perfect, it actually does crash a lot. Especially if you spam the button asking for too many simulations too fast. The raw text part of the simulator is fine but the map adds a lot of complexity that causes google sheets themselves to crash. So this isn't perfect on their end, which makes it less than perfect on my end.
And yeah. This one simple desire to automate spreadsheets to save me labor has actually made me work more on this topic. But in a sense, this is okay, because at least in my case, I'm not being forced to do this by economic anxiety. This is weaponized autism being directed toward a passion project aka a special interest, and as such, I'm using the time freed from making simple charts to being able to do that in seconds, and I'm working on fine tuning and improving my predictions, even delving into simulations to create more sophisticated predictions.
And I'm not the only one who has done this. There are other posters on reddit and the like who have even more sophisticated models than me, that calculate fundamentals, and take into account multiple polling aggregators (or even aggregate their own polls), and while I don't plan on getting that sophisticated, it's amazing what people can do. And this ins't even being driven by the economic need of capitalism, this is just stats nerds who wanna predict elections making models that do exactly that. And then we all argue online about whose model is better and who is correct and we get in stupid spats like the Nate Silver vs Allen Lichtman one.
And yeah. All of this is done without economic coercion too. Just passionate people doing things they wanna do. We act like we need paid work to motivate people. No, people would find things to do to motivate themselves if they eventually recenter their lives away from work. As I said, that's what business leaders are afraid of so they invented 9-5 to keep us working like sispyphus.
THAT is the bad kind of work, because it's driven by economic need, coercion, and exploitation. It takes people away from simple passion projects and keeps them working needlessly on crap they don't even enjoy forever and ever.
In some ways, more automation doesnt have to lead to less labor. I mean, it's true in this case. If anything I'm working harder on election predictions to the point it's taking away from other projects, and next month, i plan to focus on other things after this election is done. but until then, I'm in election mode, pouring over the data, updating my charts every time a new poll drops, and trying to make my predictions more advanced and sophisticated. I likely have hit a wall this time. Outside of maybe trying my hand at another simulator to mass produce possible outcomes for purely statistical purposes (no promises, I may choose more leisure on that front), I doubt I'll be doing more than just updating data, making screenshots, posting screenshots here, and on social media, and sharing them with friends. And omg, you want another one even though I just made one yesterday? yes, I'm getting annoying with these. I overshare as we autistic people do. Did you know Harris only has a 45% chance of winning now? I know I mention it every time I post, and I posted that exact graphic above already showing off my charts. But yeah.
Anyway. I just had a thought about the nature of growth and the tradeoffs between leisure and growth and wanted to put it into practice. But yeah, our entire economy is just my election prediction model in practice. We do stuff to save time, then we instead do more with the same time, and while with me it's driven by passion and that's fine, for most it's driven by pure economic coercion which is messed up. But yeah. Growth is like this. I could use this extra efficiency to spend less time doing the same thing, or I can spend the same time doing more things. Much like with the economy, I'm spending more time doing more things and constantly trying to improve my methods. Of course, again, it's voluntary for me so that's okay. i just resent coercion in capitalism forcing us to be economically productive all of the time when I'd rather F off and do passion projects like this. Or maybe even do nothing at all. Because that should be my right as a human being. I'm a human BEING. Not a human DOING. Remember that. Always remember that. You are a person, you are not purely a factor of production. What you produce isn't you. You can wrap up your identity in it if you want and you're proud of it, but you should always make that call yourself, society shouldn't force you. And you should be able to stop and just...exist if you want.
No comments:
Post a Comment