You can get AI to generate the best code you have ever seen. It just takes time and direction. I can write "poetic" code, but it takes orders of magnitude more time. I can also write beautiful code with AI, but it is also time and brain intensive.

It generates terrible code when used in a nearly open loop manner, which all coding agents are currently doing.

You could literally get knifed by mechahitler in your own kitchen.

I wonder how much is/was paid out for The Excursion.

The US military is the world's best killing machine, but the US as a country cannot win a war. These are different things. Hell, I'd say it didn't even win the first civil war. And with the bullshit that has been perpetrated on the world in the last year, it might not be possible for the US to ever win another one.

I don't know what format your spec is in, but it should be graph for this very reason.

Exactly, and it is a DAG (specs and tasks in the toml plan). Check the QOIzig example and its task graph if you’re curious!

How about you link to it so people can judge.

The article says nothing about all ideas being good. It doesn't talk about quality at all. It talks about how people in a team approach ideation.

Discussing and solving problems is different than firing off Thought Terminating Clichés to kill someone's project.

You do have to get to a working system, even if it is on a naive happy path before you can start preparing for the things that can go wrong.

Agreed, but it’s been my experience that anything that isn’t enthusiastic agreement, is considered “negativity.”

When I encounter someone that is so fragile, that they literally fall apart, if we aren’t cheerleading, then I can’t work with them. I’m a creative, myself[0]. I’m familiar with the process.

It’s my experience, that we often can’t see things that will kill the project, until we start drilling into the details; sometimes, not too far. Finding these things is not a death sentence. It’s the first step to success. They exist; whether or not we choose to see them.

If your feet are wet, and you see pyramids, you’re in de Nile (or a fountain in Vegas).

I have learned that this kind of introspection needs to happen, as soon as possible. It’s the way that I have been trained to work, and I’ve been shipping stuff, for my entire adult life. Shipping is not for the faint of heart. If your idea is meant to be displayed on a refrigerator, then you have the luxury of chasing off disagreement. If it's supposed to be something that people stake their careers on, and pay for, then we need to have a thicker skin.

[EDITED TO ADD] A practical reason for finding the issues early, is who will be solving the problem. The earlier you find the problem, the more likely the person addressing it, will be creative. They will have more room to work, and they will be more invested in the vision, as opposed to the implementation. Late fixes are ugly.

[0] https://news.ycombinator.com/item?id=40917886

Lol, Mercurial is written in Python (and now some Rust).

Most of the time, when someone raises a hypothetical performance criticism, it is either to further their pet language or as a cheap shot to shoot something down.