It's Work that taught me how to think

On the first day of my college CS class, the professor walked in holding a Texas Instruments calculator above his head like Steve Jobs unveiling the first iPhone. The students sighed. They had expected computer science to involve little math. The professor told us he had helped build that calculator in the eighties, then spent a few minutes talking about his career and the process behind it.

Then he plugged the device into his computer, opened a terminal on the projector, and pushed some code onto it. A couple of minutes later, he unplugged the cable, powered on the calculator, and sure enough, Snake was running on it. A student raised his hand. The professor leaned forward, eager for the first question of the semester.

"Um... is this going to be on the test?"

While the professor was showing us what it actually means to build something, to push code onto hardware and watch it come alive, his students were already thinking about the grade. About the exit. The experience meant nothing unless it converted into points.

That was college for me. Everyone was chasing a passing grade to get to the next class. Learning was mostly incidental. The professors tried, but our incentives were completely misaligned. Talk of higher education becoming obsolete was already in the air, especially in CS. As enthusiastic as I had been when I started, that enthusiasm got chipped away one class at a time until the whole thing felt mechanical. Something I just had to get through.

I dropped out shortly after the C++ class, which had taught me almost nothing about programming anyway. I was broke and could only pay for so many courses out of pocket. So I took my skills, such as they were, to a furniture store warehouse. My day job.

When customers bought furniture, we pulled their merchandise from the back and loaded it into their trucks. They signed a receipt, we kept a copy, and those copies went into boxes labeled by month and date. At the end of the year, the boxes went onto a pallet, the pallet got shrink-wrapped, and a forklift tucked it away in a high storage compartment.

Whenever an accountant called requesting a signed copy, usually because a customer was disputing a charge, the whole process ran in reverse. Someone licensed on the forklift had to retrieve the pallet, we cut the shrink-wrap, found the right box, and sifted through hundreds of receipts until we found the one we needed. The process took hours.

One day I decided enough was enough. After my shift, I grabbed the day's signed receipts and fed them into a scanner. For each one, I created two images: a full copy and a cropped version showing just the top of the receipt where the order number was printed. I found a pirated OCR application, then used VBScript and a lot of Googling to write a script that read the order number and renamed each image file to match it.

I also wrote my first Excel macros, in Visual Basic this time, which wasn't so different. When everything was wired together, I had a working system. Each evening, I would enter the day's order numbers, scan the receipts, and let the script match them up with a preview attached. When the OCR failed to read a number, the file was renamed "unknown" with an incrementing number so I could verify those manually.

From then on, when an accountant called, I could find and email them the receipt in under a minute, without ever leaving my desk.

When I left that warehouse, I was ready to call myself a programmer. That one month building that system taught me more than two years of school ever had. But the education didn't stop there.

Years later, now considering myself an experienced developer, a manager handed me what looked like a giant power strip. It had a dozen outlets, and was built for stress-testing set-top boxes in a datacenter. "Can you set this up?" he asked.

A few years earlier, I would have panicked. I would have gone looking for someone who already knew the answer, or waited until the problem solved itself. But something had changed in me since the warehouse. Unfamiliar problems no longer felt like a barrier. They felt like the first receipt I ever fed into a scanner. It was just something to pull apart until it made sense.

I had never worked with hardware. I had no idea where to start. But I didn't need to know where to start. I just needed to start. I brought the device to my desk and inspected every inch of it. I wasn't looking for the answer exactly. Instead, I was looking for the first question. And I found one: an RJ45 port on one end. Not exactly the programming interface you'd expect, but it was there for a reason.

I looked up the model number of the device, downloaded the manual, and before long I was connected via Telnet, sending commands and reading output in the terminal. Problem solved. Not because I knew anything about hardware going in, but because I had learned to spend time with unfamiliar problems.

None of this was in the syllabus. Nobody graded me on it. There was no partial credit for getting halfway there.

That's the difference between school and work. School optimizes for the test, like that student who couldn't look past the grade to see what was actually being shown to him. School teaches you the shape of a problem and gives you a method to solve it. Work, on the other hand, doesn't care about the test. Work hands you something broken, or inefficient, or completely unfamiliar, and simply waits.

Often, there are no right answers at work. You just have to build your own solution that satisfies the requirement. You figure things out, not because you memorized the right answer, but because you thought your way through it. Then something changes in how you approach every problem after that. You don't flinch at the next problem. You understand that facing unfamiliar problems is the job.

---