That said, it has definitely pulled some real humans out of the woodwork to give their real opinion that "yes, I'm influenced/duped by context and that's a good thing". And that's an accomplishment.

[1] By which I mean predicting the behavior of error-prone code that requires good knowledge of all the quirks of the language to correctly answer.

To illustrate with an example, your point is like saying that if we had a prediction market for "Will the United States cede Texas to Mexico in 2026?", then the US government would give up Texas just to get that sweet sweet prediction market payoff.

I would agree with a smaller point, that an org would accept minor tweaks it doesn't care about in order to game a market, but this just means it can tolerate being unpredictable about lower-order bits of its decisions. You see that in cases like Trevor Noah making a minor change to a speech to influence a particular bet.

I'm always like "JFC, can't you cache the notifications, so I can see it there while waiting for the app to gets its act together?" But no, that's never an option.

So I'm getting a laugh out of how notifications last long enough to be extracted by someone just not the person that they're for. (Though to be fair, it could be a case of a notification that was never tapped, and therefore hadn't been purged yet. I couldn't tell from the story.)

Not really, even in America. Like, take alcohol regulation. Your model would be "drunken bar fights are already illegal, so just prosecute that, problem solved."

Except that, historically, there's so much of that that it overwhelms the ability of law enforcement to keep up. So we try to remove the driving factors: "Okay, you can drink in public, but only[1] at these licensed places that are heavily incentivized to prevent fights before they start."

I'm not advocating any particular position, I'm just saying that if there's a persistent situation that heavily incentivizes violence, then it's not unreasonable to push back on that mechanism rather than just try to mop up the violence after the fact. Which specific situations merit that is up for debate, but it shouldn't be controversial that some situations should be handled this way.

[1] Yes, I'm simplifying, just focus on the general point here.

Oh crud I just opened a can of worms with that, didn't I?

https://en.wikipedia.org/wiki/Pirates_and_Emperors

https://www.youtube.com/watch?v=UQBWGo7pef8

>“Once the email arrives and Iran completes its assessment, vessels are given a few seconds to pay in Bitcoin, ensuring they can’t be traced or confiscated due to sanctions,” FT reported, citing Hosseini.

https://beincrypto.com/iran-bitcoin-toll-hormuz-strait-tanke...

Putting quotes around a quote isn't hard.

It's not fair to expect everyone to be intimately aware of every political gaffe, and instantly make the connection when you repeat it so they know not to reply to the comment seriously -- as the original comment was doing.

FFS, just put it in quotes so people know you're quoting someone. (Or if it's not a direct quote, mention that it's from the mentality of the person you're mocking.) Is that so hard? Is it so important that you feel special as someone who knows about that incident that you just have to provoke a confused flamewar?

At the very least, they could have put the comment in quotes to indicate they're quoting someone and it shouldn't be read at the object level.

Edit: Seriously? Am I wrong here? Are you all really okay with Poe's Law-ifying HN, where people post comments that are easy to read as serious when they were intended as ridicule of the person who said them, and the comments erupt in confusion and flames? That's not what we should be expecting out of HN.

None of the management-level desiderata he appealed to require that the user experience be broken this bad. There is very little bot deterrence from prevention of typing at that stage, while it heavily impacts user experience, especially on mobile.

I elaborate here: https://news.ycombinator.com/item?id=47575982

There have been times when, across about ten minutes of usage, most of which is me typing on iOS Safari, it drained 15% of my battery. There is no functional justification for this beyond poor code quality. (It was on a long conversation FWIW.)

This when I'm logged in, with a paid (Plus) account, connected to a very old email address with a real user profile. That can't be the result of super-clever bot defense measures, because it's merely an inconvenience on desktop. And if you genuinely believe that email has been compromised, why aren't you reaching out the to the account owner, as the account isn't otherwise connected to fraud by your heuristics?

However brilliant the LLM agent it is, I'm seeing a lot of unforced errors regarding how you implement a web interface to it. If it makes you feel any better, it doesn't really register compared to all the bloat I see on other sites.