Part 2: Our Collective Order-Keeping Fictions
more stories and lessons from my first grown-up job
Advisory About Comments
I need all my time, energy, and attention during the week to focus on my new job. Comments were only open over the long weekend and are now closed. This will be the usual pattern—whatever I publish during the week will be comments-closed until after work on Friday until the evening before I go back to work.
I apologize for the annoyance, but this is really helping me function better in my new job, which has to be my priority. Thanks for understanding.
I recently published a post with some thoughts about the business world, occasioned by having started my second job out of college.
Unusually for me, I said that I would rely on reader response to determine if I published any more on the topic. As I said in that post:
I have enough material for at least five essays, and if this one does pretty well, I will eventually write them all. It is rare for me to make decisions about what to write and publish based on readership metrics, as I write primarily to clarify my own thoughts, but this case is a little different. I’ve been both talking about this stuff in therapy and writing about it in my journal—as well as using some of it as material for fiction—for years. So my thoughts are quite clear to me at this point. But if the statistics indicate that people enjoy this, I’ll definitely write a few more installments.
That one did in fact do very well by the statistics, and I got a few emails encouraging me to keep it up, as well. Thank you all for the feedback. It was not hard for me to guess how political topics will do, as my readers tend to be broadly anti-Woke, disagreeing primarily on the details of how non-Woke principles should be applied, not the principles themselves.
But now that I’ve pulled away from politics, it is often very difficult for me to guess what people will be interested in.
Stephen King’s book, On Writing, encourages would-be novelists to write about what they know from their own jobs and work experiences. “People loooooooove to read about work. God knows why.” (Emphasis mine, based on how he narrated it in the audiobook version.)
I suppose he got that right. I estimate I have five posts worth of material, possibly a little more, and I’ll keep writing them as long as y’all continue to enjoy reading them. 😊
So here’s part 2.
In part one, I explored five conclusions: one, my hypothesis on the real reason why huge US companies employ foreigners; two, that most managers are entirely useless; three, that stupidity is terrifyingly rampant; four, that bureaucracy exists to employ and thus collect taxes from cogs, and five, the revelatory role of remote work.
Here in part two, I will explore some specific details of “stupidity is terrifyingly rampant,” from part one. In particular, I will go into how terribly insecure your data really is, regardless of what the company’s privacy policies promise, and why; explore a bit more of the consequences of offshoring customer service, and share how egregiously overblown the managerial class plans for AI are, and what it means for you.
Stupidity Is Terrifyingly Rampant: Moar Stories
Managerial Plans: To Help AI Kill You If They Can
There came a day when I realized that the entire company, and probably all of society, was totally, utterly, and thoroughly fucked. I came to this conclusion not because of how few people could do any coding—more on that later—but from the rampant stupidity of one of the very few people in my company who could.
Not that he was good at coding. He wasn’t. He was so bad that he had no idea how bad he was, which is the real danger: not knowing what you don’t know. Very little is more important, in coding, than knowing what you don’t know.
Let me explain what I mean by that.
One way to look at coding is to put it into two boxes: developing and scripting.
Developers do things like create apps and design new software programs to use to get things done. Think of it like building a house, from the ground up.
Scripting is more about using code to do things within an existing app or software program’s capabilities, to gain new insight. It’s like designing an automated system for a house to have the lights go on and off at predetermined times, cause the robot vac to clean on a schedule, and using data to figure out the exact patterns of how much each room of the house is used, and when, so that the lights and floor-cleaning schedules are as accurate and personalized as possible.
I am a beginning developer, in the early stages of learning how to create something new from scratch. There are frameworks for building apps, that are sort of like systems that help people take a set of building plans and figure out how to build the house from them. How to calculate the amount of wood, concrete, etc., that they will need, and how to plan the building process. I’m learning from those and well on my way, but I’m not (yet) who you would hire to build you an app from scratch.
I am a strong scripter in the data science realm. I can take a massive, unwieldy, error-filled dataset and derive real insights. If you have a huge, nasty file (even an Excel file) that holds ten years of records for your online store, you could give it to me. I would be able to organize the data correctly, rigorously analyze it, and return actionable insights: things like figuring out when is your ideal time to launch new products, how the holiday season affects your refund rate, that Tuesday is consistently your slowest day (so if you need to cut down your customer service hours, that’s the safest day to close), whether coupon codes are worth the trouble, how much you’re spending on replacing lost packages from FedEx vs UPS vs the Post Office, etc. And I could produce really flashy, cool-looking graphics to help everyone on your team understand the insights, even the ones who swear they’re “not a math person.”
The guy who made me realize how totally fucked we all are is a very weak scripter who thinks he’s an expert developer.
He programmed a very simple web-scraper to go through Facebook Marketplace in his neighborhood and grab anything related to a word that would appear in ads related to one of his hobbies, then count how many relevant ads there were in each day.
It didn’t deliver him the ads.
It didn’t even link him to the ads.
It did nothing but return a number.
In his mind, his achievement, of having a script that notified him there were four ads he should look at, was the equivalent of developing a killer app that was ready for an IPO.
He once spent 30 minutes at an All Hands meeting walking everyone through how he had done this magical thing, made it so that every night when he sat down to go to Facebook Marketplace and read the ads, he already knew how many he would find related to his hobby.
I wish I were kidding.
That guy is responsible for the fact that the company no longer even tries to hire people with coding skills and is perfectly happy to get Excel users.
Why?
He convinced the higher-ups that ChatGPT means nobody needs to learn how to code.
This is beyond insanity, and means that the other two problems—how insecure your data really is in many cases, and how bad the offshoring problem is—will only get worse.