I include a line in AGENTS.md that says "We *always* add docstrings to methods, classes, structs, and namespaces - there should be 100% coverage of docstrings."

It will make better choices about what functions to make or remove if you force it to justify why the function exists.

And when you have to go back and understand, it becomes easier.

I used to check once a quarter. With AI, I devote 1-2 days a week to the meta-task of improving my tooling, but I'm sure that will fade once AI tooling becomes more standardized.

It's not worth being one of those guys who gets too into the tools, because it can keep you from the actual valuable work. Plus, you are then judged not as a regular engineer, but as a person who is supposed to have some secret productivity sauce, and when people see that it doesn't make more of a difference they can start to discount your opinion.

Maintaining a code path with a snippet for each possible viewing experience, back when REST was all the rage, was a bit obnoxious though.

[1]: https://qz.com/average-car-payment-by-state#average-car-paym... [2]: https://www.lendingtree.com/auto/debt-statistics/

If you could avoid a $700 car payment and still get to school/job/grocery store on time, a lot of Americans would.

But if you can't, and you are paying $700 anyways, why not at least use the fact that you're paying $700/month to travel anywhere and get a bigger parcel of land all to yourself? There's a big incentive to have single-family homes for rent as long as that math is true.

And if you can live in a dense area and get around OK, but paying $700/month saves you from $800/month in private school payments, then the math might still make renting SFH better even if the transit is there.

Rural homeowners and people with single family homes in streetcar suburbs or cities don't, but most new developments already have an organization with dues for shared maintenance.

If we modeled Spain, we could have more condos. That would require good, safe transit, condos with green space and multiple rooms at a SFH price point, and a cultural change among the older generation to view cities as having less crime.

Overly strict tests enforce specific implementation details not specified in the prompt, invalidating many functionally correct submissions.

Underspecified prompts omit requirements that hidden tests enforce and that are not reasonably inferable.

Low-coverage tests under check the requested feature, so incomplete fixes can pass.

A misleading prompt points models toward the wrong behavior or contradicts what tests require.

If the goal is, "how does my model compare to real SWEs", these are pretty reasonable situations that your model will have to encounter. It's a little like making a nursing exam and then flagging that some of the tests required you to ask the attending doctor for additional information that's not in the chart, or that the patient's family didn't fully explain their aging grandma's medical history.

I can understand why they might want a tighter benchmark, but if you're OpenAI and you promised your model as a replacement for real workers, this isn't the best look. It seems like you would want to test these things.

You can be correct that your method makes code more DRY, and miss the point that the other person believes that things are going to diverge significantly over time and doesn’t value DRY.

You can be correct that your method is more resilient to failure, and miss that the other person believes that some level of failure is OK and wants an option that is less technically complex.

I’ve seen people get upset that they were correct and yet the room shifted against them. Most times, it seems like they are correct. But they are correct on a narrow axis, that misses the motivations of the other people in the room.

This is part of the reason high level account reps focus on the mix and viewpoints of people in the room over technical specs. Get the lay of the land first, and then you can tailor your pitch to be correct in the way that the audience will be receptive to.

The most successful strategy is to make a virus that spreads fast, with few visible symptoms until the late stages of the disease. A deadly virus, early will just cause borders to be locked and the international research community to swarm on a cure.

Scholarly article for reference if you want to learn more: https://www.jstor.org/stable/827888