daily log 03-25-21

Hi computer!

It’s been a few days. I miss this. I’m sorry I’m so bad at keeping things going.

Why am I so reliant on streaks?

Because I like the momentum.

Why do I like the momentum? Why do I need the momentum?

Because it’s harder to break promises to myself when that many past Kendra’s have kept their promises…

Why do you break promises to yourself?

I break promises to myself because in the moment I want to do other things…

Why do you want to do other things in those moments?

Because the thing I’ve promised myself I WILL do is much less attractive than [insert specific distraction here]. And it’s not like those distractions are video games (yes, sometimes they are) or TV (also, sometimes they are) – frequently the distractions are actually things I really do want to do I just don’t NEED to do them.

For example –

It all started when I was filling out my daily GTD doc with my time. I couldn’t remember what I was doing at what time so I consulted our nest cam footage (something Colin turned me on to). This footage could tell me when we started and ended our evening entertainment. It could also tell me how long I was sitting at my desk. How long was I actually “Working” vs. how long was I actually on the phone. Theoretically, I could also use this information to determine who uses “their office” more – Kendra or Colin. I could track “If Colin in living room” (than not in office) – “If Colin outside – either backyard or front door area” (than not in office) etc etc.

But most importantly, I wanted to track my own time. How often was I on the phone (could track volume etc.) how often was I actually sitting at my desk typing (at desk without zoom on and sitting in the specific way etc.

It would take a very long time to do this manually. Going through each day and marking when I started and stopped something.

It would take considerably less time to offload this (delegate?) to a computer. Especially when I love image recognition so much.

So, first things first, how do I get that nest cam footage? I request it from Google Takeout. Cool. I do that. And then I wait for FOUR DAYS while google assembles the 191 “packages” of 2GB data that I can now download onto an external hard drive. For those of us counting, that is 382 GB of video data.

So now all I have to do is comb through that data and timestamp it based on specific conditions!! Right?! Easy!!

Wait.

Definitely not “easy.” First of all, the way it’s “packaged” from google is… interesting. It doesn’t seem to have a pattern. Days aren’t contiguous. This shouldn’t be a problem as I’ve already written an aggregation script for my larger “External Hard Drive Consolidation” projects (basically (1) dating every file (2) flattening all nested folders (3) pulling out files of a similar file type).

Now we start to run into the problems that I’m not as well equipped to solve – where does all of this video data live while I process it? Does it live on the hard drive? How can I label specific video frames? How CAN I label it?? Should I find specific frames and say THIS is “kendra-at-desk” so we label it as “kendra-at-desk” and record that timestamp and then… magic? No, we’d record the timestamp and then also record the timestamp when “kendra-at-desk” changed… by a certain percentage? What would that percentage be? What would that metric be? This reminds me, I should call Xan and see if he wants an assistant lol.

Once again, here is where I’d love to have a rolodex (sp?) of people who I could ask about these next steps. I feel OK doing basic image classification. I technically haven’t done video classification yet, so my first step should obviously be a video classification tutorial. Then I should take a small segment (one day) and see how well I can navigate that one day.

Also, feeling “fine” about image classification might not be enough to do a project of this size.

In summation, I’m lazy and I don’t like doing a bunch of rote tasks. I’d much rather train a computer to do it for me. This is an exciting project and might just be the impetus I need to get into video classification (as I’ve always said, it’s super important to find your own selfish motivator).

But before we sign off, let’s revisit a very important question – Why. Why are we doing this? We are doing this so I can record on my GTD spreadsheet how I’m using my time.

Why am I doing that?

So I can see how much time I spend “working” vs “playing”

Why do you want to see that data?

Because I’m hoping that if I’m confronted with that data, I will be startled into using my time more efficiently.

Why do you think THIS is what it will take to get you to use your time more efficiently and effectively?

Because I will be confronted with my own data instead of relying on my own (poor) self reporting. Remember, as we learned in Predictably Irrational, Humans are bad at self-reporting. I don’t want to be one of the humans caught in that trap and making decisions based on false (self-reported) data.

Why do you think you want to do this?

So that I can say “look how hard I’ve worked even while unemployed” to everyone who might be questioning.

Why do you feel like you need to do that?

Because I feel the need to prove that I’m smart.

Why do you feel the need to prove that you’re smart?

Because that’s where I’ve placed my self worth…

Why did you place your self worth there?

Because it’s easier to put my self worth into something I can “control” than self-actualizing. Because I respect people who are smart. Because this is how I value other people and I want them to value me like that, too.

Why do you value people based on their intellect? Why is this what you admire? Why not “kindness?”

At the end of the day, I definitely respect kind and loving people more than I respect intelligent people. However, at the middle of the day, I respect intelligent people because they gift me knowledge and knowledge is energizing.