How's that for an ambiguous title? While I'll give my two cents on the whole Mac vs PC debate in a second, this article is not primarily about Windows with a capital W. But since I was hoping you'd think it was and this post is actually about saving money, here's my two cents. Yes, you can save lots more than two cents by purchasing a PC rather than a Mac. But you should totally replace Windows with your choice of a Linux distro. Particularly if you're a software developer, you would probably save more money on all the additional software you might want than on the actual "machines" (that's in quotes because you're definitely paying for more than just the machine when you buy a Mac). Alright I'd better stop before I lose 99% of my audience. Let's move on to what this post is actually about.

As a pretty brand new adult, I have just begun to experience the joys of paying for stuff. I know it sounds awful, but for a mathy, data nerd like myself, it can be - sort of, almost, kind of - fun sometimes. There's lots of data to play with, and some of it can ultimately be used to save money. It's almost like gamifying life. For instance, I have a spreadsheet that can show the impact of an extra $X payment on a loan. A little less common probably is my log of basically all of my purchases. It's quite detailed so someday I might be able to use it to tell myself to eat fewer Oreos or... to order more Oreos. I like to think that someday it could be used to advise my financial decisions as well as my health decisions.

One thing that I don't log though, is my monthly utility bill. My budget just uses a constant over-estimate of my monthly bill. But as spring arrived, I started preparing for the increasing bill because my heat is covered in my rent but the A/C is not. So I turned off the heat and let it get into the 60s to offset the upcoming heat. Once that initial cooling effect was overcome by the increasing temperatures, I determined to be better than I had previously been at opening windows when it made sense to do so. But then I remembered the first great virtue of a programmer (as defined by Larry Wall et al.):

• Laziness - The quality that makes you go to great effort to reduce overall energy expenditure. It makes you write labor-saving programs that other people will find useful, and document what you wrote so you don't have to answer so many questions about it. Hence, the first great virtue of a programmer. Also hence, this book. See also impatience and hubris.

And now a message from xkcd.

Know where I'm going with this? Basically, I decided to automate away as much of the window opening process as possible. Unfortunately, my robot butler is not yet ready to handle actually physically opening and closing windows, but that's actually probably not the biggest pain point for me. The biggest problem is usually just not thinking to open windows unless it's insanely hot in the apartment already. In which case, it's fairly likely that it's also insanely hot outside and opening a window won't do much. But fortunately, there are some free weather APIs out there that give programmers like myself access to current weather forecasts. So using the hourly forecast data, I can determine what windows of time make sense for opening windows (of glass). I now have this program set to run automatically everytime I log in to my laptop, and it will display a notification if it wants me to take action. See it in action below:

SPOILERS: if you didn't have your volume on you probably missed that it talks to me too. Other than that though, it's fairly simple and stupid by just using thresholds for temperature and probability of precipitation, but it's a good start that over the course of my life would actually probably save me a decent amount on A/C costs. I definitely hope to get more out of it someday though. Instead of stiff rules like "only if the temperature is below X and probability of precipitation is below Y," I could generate a pretty nifty dataset someday. It could incorporate temperature, wind, precipitation, and conditions such as overcast or sunny. Furthermore, it could label the dataset with the actual changes in temperature as a result of opening the windows during certain conditions. That way it could eventually learn what conditions are good for opening windows. And obviously eventually I will need my robot butler to do all the heavy lifting for me.

Finally, some of my nerd heros like Sal Khan and Randall Munroe inspire me to throw a bunch of guesstimate numbers together to make fun little guesstimate calculations. So I set out to figure out a very approximate guess to how much I could save with this program over my lifetime.

First I looked up 'utility bill pie chart' to figure out what percentage of an average person's bill can be attributed to A/C. I averaged a few different number to get 12%. Then since I don't pay for heat right now, 12% would be a significant underestimate of my current bills. So I took the average percentage for heating out of the equation. The heating average I found was 45% of the utility bill. So to adjust the 12% cooling for my bill which shouldn't include heating, I would take 12/(100-45) to get 21. So we'll say 21% of my current bill is due to A/C. Then let's use $50 as a conservative monthly utility bill. So let's round down and say I would ordinarily spend $10/month on average on A/C.

Next let's say my program will buy me one month's worth of A/C over the course of a year. I realize I could be doing a lot more to adjust this for the seasonality of A/C, but I don't want to break out a spreadsheet for this. So if everything I've said up until now were the case up until I die at the ripe old age of 102, I'll have saved $800 ($10 for the next 80 years) from this silly little program I put together in a couple days. Let me know if I have made any mathematical blunders anywhere. And if you think my estimate is stupid, feel free to leave a better one in the comments.