linting tools, static analysis, CPD, etc. These are all old things you can continue to use and are much more robust than anything you can prompt. These should be standard when using LLMs. In fact, you can tighten the rules even more enforcing more restrictions so you ONLY get the output you want. put this behind a pre-commit hook and a CI job that runs on a PR, and it will work wonders.
You can have all the prompts you want on top of this, but if you don't have this automated stuff running behind the scenes, you aren't serious about these issues.
Looking through some of these comments here, I see lots of people rewriting concrete rules in markdown willing to spend tokens on the hope AI won't miss it where an actual program won't.
2009 is a generation ago. Asking why a new generation why they might not have petitioned 17 years ago seems like asking where a 21 year old was on 9/11.
As for a better reception, the assumption was RFK Jr. would take it more seriously.
> If you've believed all the government data for the last 10 or 15 years I got a bridge for sale.
It's not about belief. It's about trust. And as they say: trust, but verify.
We could trust the data, but we should, of course, verify it. And that's generally how it worked. And both sides of the aisle happily trusted this data and verified it. It wasn't always correct, but not so much that you should start by not trusting it at all.
But now we are at that point. I've taken the attitude of simply assuming the opposite and starting from that position, and it has served me well most of the time. I take what the current administration is claiming, assume the opposite and start from there.
To be fair, I agree with your sentiment, don't just believe what the government is saying. But there has been a dramatic shift since January 2024.
Not in a serious way, no (hence my 'mocking' commentary). It's always playing off the joke or idea of "bribe". But if someone said "I bribed a police officer" wouldn't take it as a joke.
Edit: And to be clear, I wouldn't describe either of those are "morally neutral or even positive situations."
But please, by all means tell me how "Foo Company bribed professional Bar" is used positively and frequently enough that the gp makes sense.
> In the US we sometimes use the term “bribe” in morally neutral or even positive situations.
I live here in the US. I've NEVER heard the term bribe in a neutral or even positive way. It might be used in a mocking way, as if to mock the idea of bribes, but never seriously.
So, unless you are confusing that mocking nature as morally neutral or even positive, this is incorrect.
It's not that it's bad writing. It's that people are used to bad writing. People are used to garbage. For example, your comment is awful from a literature or writing perspective. It needs to be edited. This comment that I'm writing right now is awful. We're used to crap writing and people who don't care about editing and putting out good, thoughtful words. When all you're used to is garbage, things that actually are better are generally going to seem odd and out of place. However, for those people that are used to better, more elegant writing, it's going to seem fine.
Of course, the difference here is context. In a comment, you're not expecting well-written sentences, structure, and editing. So we jump at these things that seem out of place because of the context.
> sometimes that writing makes me really upset.
You know what makes me upset?
- No writing
- Or bad documentation
- Or just no documentation
- Or just nothing being written down about something
For me, at its core, the most important thing is accuracy. Is it accurate? If so, good. We can start from there. If your issue is style, fine, but that's a personal judgment. As long as it's accurate, I'm fine.
> Seems like Wikipedia sucks at enforcing its policy from what I am reading?
More like policies evolve and older articles are grandfathered in by the fact that they aren't edited and people aren't going back and reviewing old articles that don't meet the newer standards.
> Huh? As a parent with kids who take the bus, people ignoring the flashing lights on buses absolutely does matter.
This doesn't stop that. And the penalty is too low to stop this. Nor does this target the people running these lights.
I'd prefer to stop it from happening rather than merely fining people.
Further, to my direct point: you can literally respect the flashing lights and stop at buses and still get penalized for this.
So why do you want to punish people for obeying the law?
> Bus drivers call the cops on the cars who do this already. What evidence do they have, other than the license plate?
If they can get the license plate sure. But they rarely can. And they don't, which is why it rarely goes to trial.
I'm curious, do you want to stop this or just let people make money off this scheme? As a parent with kids, I'd think it would be the former, but you are arguing for the latter.
> how hard is it to simply not be an asshole and not pass school buses when they have their flashing lights out?
As a parent with children who take the bus... this actually doesn't matter. You can't assume that the car's owner is passing the school bus. So, this is a case of finding someone guilty with no physical evidence. And the real fear there is that suddenly you are guilty because someone else was using your Wi-Fi, and you suddenly have the burden of proof to prove your innocence.
> Where I live every intersection has cameras.
And now you are guilty of crimes. Prove you didn't do them.
You aren't answering the question. You are reframing the question so you can give the answer you want, thereby avoiding giving the real answer to the question presented.
> I would be surprised if Trump didn't complete his term by pardoning his family and all those that he works with.
As we have learned, we can ignore all of that, or they can find ways around this. The Supreme Court has made it clear the President has amazing powers.
You can have all the prompts you want on top of this, but if you don't have this automated stuff running behind the scenes, you aren't serious about these issues.
Looking through some of these comments here, I see lots of people rewriting concrete rules in markdown willing to spend tokens on the hope AI won't miss it where an actual program won't.