In computer programm single threaded programs are pretty predictable (apart from human errors). As soon as you have multi threading that goes out the window. Modern CPUs in most devices you use have what’s called a scheduler that schedules when to let different things actually use the CPU so you can actually do multiple things at once. It’s a super important concept for what we want to do with devices. But because of that you have no guarantee about when (or if) other threads of your own code will execute. Apart from truly insane edge cases, single threaded programs act pretty deterministically. Multi threaded ones do not. It’s very similar to the “it’s alive and dead until you check” idea because you just don’t know. So much so that there are data types we use called things like Maybe where the result is either a success or a failure and you write code for both.

Also much like the cat in a box thing, programmers don’t really view it as magic, it’s just sort of a side effect of the uncertainty.

Well, I work as a bartender, and here in Finland it’s strictly against the law to serve alcohol to, or even allow a “visibly intoxicated person” to enter the premises (a law which almost every bar breaks at some point, intentionally or no), and I think I’ve witnessed multiple times myself how a customer’s level of intoxication reveals itself only after you have served a drink to them and they’ve payed for it. Could it be called a Schrödrinker’s cat?

The Heisenbug. Once you try to observe this kind of software bug with your technical means, it simply goes away.

Not quite Schrödinger’s cat, but in programming we have Heisenbugs named after Schrödinger’s peer.

It’s when you have a bug/crash that is not reproducible when debugging it. Might be that you’re reading some memory that you’re not supposed to, and the debugger just sets it up differently. Maybe you have a race condition that just never happens with the debugger attached.

Employee salaries in HR; they are both correctly paid(employer perspective often), underpaid (employee perspective often), and overpaid (company and co-worker perspective). Depending on how and how often you open the box, any of these views can be accurate.

The drill bit is good, until you get it to surface.

As a bicyclist, I see that we have Schrödinger’s Cyclist: Too poor to be able to afford a car like “normal” people, but also a rich elitist who can afford to commute by bike.

Also, Schrödinger’s Bike Lanes: A conspiracy by car-hating politicians to punish drivers, but also an amenity that only rich elitists get in their neighborhoods.

For work I use a database written in COBOL. Reports are simultaneously running and frozen until I either get the report results or sufficient time has passed that I’m certain the system has crashed.

I guess the best one for me may be elite university students are “just smarter” than others until I have to read their term papers.

For some reason it’s always the non-native English speakers who write well.

Autonomous vehicles are at times both amazingly advanced and bedshittingly idiotic.

I’ve ridden ~25k miles in them for work, and I trust them more than 95% of the drivers on the road. But I’ve also experienced them acting in ways that are still quite far from the way humans would.

Code is both great and terrible until it compiles.

Print jobs are both completed successfully and failed until someone checks the queue.

A person that has a lot of certs or a high title is both extremely smart or extremely unintelligent. You don’t know until you start talking with them about things more than surface level.

Projects will either be done next month or take at least a year to complete. Also, if you ask my team to calculate how long a project will take, and then ignore the estimate, the project will take infinite time because you are an insufferable moron.

“The Computer never makes a mistake” is true and also probably responsible for people believing LLM-hallucinations uncritically