We should perhaps be comforted by the knowledge that everything will eventually break. Despite our best efforts, things will go wrong. Here are several such ways that I've found them going wrong.

...This feels oddly like I'm reenacting Bart Simpson writing on the chalkboard for atonement.

1. JavaScript bundlers and the runtime dilemma with browsers.
2. Not preparing enough for technical interviews.
3. Using the canary/release candidate versions of stuff.
4. Focusing all efforts only on a narrow subset of "work."
5. Expecting social media to increase in quality over time.
6. Expecting social media to stay at the same level of quality over time.
7. Taking out a hot stainless steel pan from the oven without sticking a silicone thing over it, or at least marking it as being hot.
8. Expecting an LLM to improve over the course of any one chat instance.
9. Trying to add stateful stuff to everything.