I need to say as someone who also really really cares about the craft. About doing things right. LLMs have been a massive boon to me, because they're the only entity in this world that seems willing to work with me to get things right.

If something is working but it is ugly, sketchy, or I feel there may be a more apt abstraction or generalization of the problem, every other human being on the planet has persistently turned that situation into an exhausting argument. If the code works, it works, right? And all this shit I talk about is pointless because it works. And I have to fruitlessly construct an argument that would appeal to someone who cares nothing other than a complete ticket.

LLMs are at least not needlessly combative and incurious as to think that every approach but the first one we try is a pointless waste of time.

Has anyone figured out why so many corps do this? It always feels like it's one step from:

* Of course do whatever you want! * Nah I'm a big dumb idiot who doesn't know whats good for him so ask me in 3 days when I've hopefully come to my senses.

I think a simple law would fix this

If no wasn't an option, you did not establish consent.

Because every single developer that writes one of these is a shitty incel that can’t take a “no” for an answer in their dating life so they had to force themselves upon you in the OS.

Management that struggles to acknowledge their users don't like something they wanted built.

They'll definitely realize their mistake and turn on our amazing feature! (This unfortunately is the real attitude sometimes)

You answered your own questions. They do think of their users as big dum idiots.

Not only idiots. They bank on laziness. One time you enable something by mistake then you might not spend the time to find the well hidden option to disable it.

Which reminds me, anyone know the precise location where one would disable Google's Gemini on their account?

> They bank on laziness.

And they would be correct. Less than 5% of users change defaults. That's why features get shoved in, on by default because if they didn't, very few would ever enable them. Not defending the practice, I hate it, but that's why they do it.

KPI hunting? Toxic/fake positivity?

If there has anything I have learned in my life it is this: "the world is just. solutions are easy. and if people are suffering, then they deserve it. if they would just try to solve their problems then their problems would be solved." Is deep in the thoughts of most people, and it leaks out in all kinds of ways.

you DESERVE misery. If you are suffering, then you are the cause.

I feel I should hate this line of thinking... But it's too common. I can't hate all of humanity. Instead, I know to fear others. I know to never ask for help. This belief is so common that to do otherwise is self destructive.

If you had gone to the police they would have blamed you.

Your husband resolved it because it wasn't "his" problem. He didn't "cause" it so could not be blamed. Because as far as everyone else is concerned, the reason you had that problem was because "you weren't trying"

it's not true.

I know, the takeaway is so funny, because it's not actually up to the author whether or not the solution she took would've worked. To her, "Actually Trying" and "not trying" both hinged on the success of other people helping her out. To a more hard hearted individual, the only thing that would've qualified as "Actually Trying" would've been flying out to India in person and then finding the guy first when he'd least expect it. "Actually Trying" was only determined after the fact when the result was that there was a success. If the husband coordinating with a friend hadn't worked, what then? Did he "Actually Try" if the FBI, the Consulate, and the friend were all like "we don't give a shit about this random little troll sending death threats. Did you know everyone gets death threats all the time on the Internet? Log off." What then? Is that still Actually Trying, or do we only determine that after a success? Texas Sharpshooter Fallacy, friends.

The husband was the social proof that this random junkie woman who's complaining about a guy on the opposite side of the planet was worth taking seriously. She might not have wanted to go to the cops because she didn't want to go to them, have them laugh it off and then feel both powerless and humiliated, and the conclusion is that another person who then manages the difficulty isn't even trying? Maybe not, but that's a different think from deciding that they aren't doing it because they're stuck in some old mindset.

The other example she gives is this: "These are people who could successfully launch a product in a foreign country with little instruction, but who complain that there aren’t any fun people to meet on the dating apps." Like, girl, it's not up to you whether or not there are any cool people on dating apps. There's a selection bias of who gets on dating apps going on here. You can do everything within your power and Actually Try all you want to have it be another way, but you can't really force cool people to socialize in the way that you want them to. It could be the case that there really is no one cool on those dating apps, because social climate being the way that it is means that no one feels comfortable showing their whole, unvarnished, "cool" selves and the coolest people are hidden.

At what point is someone "Actually Trying"? Is it once they've succeeded? This feels like the self help, "The Secret", "Girl, Wash Your Face" of previous eras but dressed up in the language of people who use terms like "non-zero probability", "priors", and "local optimum". The takeaway should not be that someone shouldn't try, but that the serenity prayer's most difficult part is the "wisdom to know the difference". She had the power to change this, but the wisdom to know the difference here was not guaranteed. It would be a kinder message to everyone who is in a tough spot to at least acknowledge that having a husband to vouch for her, with connections, free time, and a motivated reason to help out kind of changes to what extent she personally could be responsible for Actually Trying.

Because for most people, someone reacting with disinterest for the thing they care about is a rare and upsetting event, not their entire life's experience. That's what it means to be "normal", you align better with your peers. Most people don't need what you need. Most people can work with what you can not. You are choosing to be the exception. You chose to be like this, so unchoose it and stop being a problem.

Of course... That's the quiet part. The out loud part is just dismissing everything you say and passing you over for promotion.

The objections you have raised, the things you have said. I really understand what you mean. There's evidence all around that the aspects of our experience isn't alien at all. Why can't others see that? At this point I think that not seeing it is necessary mental infrastructure for some people. It's a bridge over an abyss that for us broke.

I think the solace I get is that this line of work tends to funnel people of our disposition into it. So we find ourselves less alone than we normally would.

Just some food for thought: I was recently brainstorming ideas for building a more decentralized moderation system, and one of the ideas I arrived at was using the rules themselves as part of the flagging system.

It would work like this: When you flag a post for breaking the rules, the community's guidelines will pop up. You are then asked in this window to highlight the relevant section or sections of those rules that this post has violated. And I don't mean just "select which rule was violated", I mean "use your cursor and highlight the text of the rules that were violated." (with support for highlighting multiple sections if so desired).

This serves the following functions:

1. Communicates why something was flagged (obviously).

2. Forces the person who's flagging the submission to actually read the rules.

3. The subjectivity of the highlighting system is used to make Sybil attacks more obvious. I'll explain why after this list.

4. It differentiates flagging from downvoting. Downvoting is for saying "I don't like this". Flagging is for saying "This violates our community's rules".

As to why this helps reveal Sybil attacks: There are several subjective points on what, where, and how people will highlight rules. Should punctuation be included or not? Should the key word in the rule be highlighted? The key sentence? The whole section? What about examples? Should we include them? Or only highlight them? Users operating in good faith will cluster around common points in common areas, but will have different ways of doing so. So, if a block of users all have: the same input, in the same way, clustered around the same time, then it was likely a Sybil attack.

This system doesn't require that it de-anonymize the people who submit flags, but it does provide a form of publicly visible transparency as to why something was flagged.

Edit: I forgot to make clear, you would be able to see a heat map of the rules that were highlighted for a flagged post.

I'd be interested to hear any thoughts on this idea.

Except for the highlight the text (think people on mobile, people with disabilities etc) it sounds like a good plan.

Than other random "judges" would be asked if the reason given by the "accuser" are correct. There would have to be some "cost" in karma to flag a post (or limit of X flags / day for X karma status or smth) and some reward in karma for being chosen as a judge/jury.

Also the need to have a minimum flagging weight and a minimum of judging weight and to reconcile conflicting votes.

Anyway would love to talk about it more but tbh it's probably not gonna happen also because most people don't like jury duty... Maybe when ai gets over the "hallucinations" but well at that point we can also get our individual ai's to read everything and judge for us

I don't find highlighting text on mobile to be too difficult, so I don't see that as a barrier imo.

for disabilities well... That one I dunno. I don't have a good concept of what kinds of UI are most convenient for each type of accessibility case.

And it's a little tempting to get lost in the weeds of who watches the watchers, but to be honest even if implemented in hacker news case, the mods themselves could vet flags for anomalies. Just this on its own would serve as a force multiplication for HN mods.

For more decentralized forms of moderation. One method might just be a simple flag appeal. Circles back to the community, they can discuss if the rule that is cited is fair, and if it wasn't possibly remove or limit flagging abilities of those who cited the rule incorrectly. And possibly some increased punishment if the appeal fails? There are lots of options there. Big wide design space.

I do think the direct text highlighting has a few important features. The Sybil attack resistance is one. That was one of the OP's primary concerns. Also, clarity on what rule was broken and why is very important, and a given rule can be verbose. It might not be obvious what specifically in a given rule was the reason for the violation. Direct highlighting lets flaggers more directly communicate what the issue is, without opening the communication channel up for a flame war.

The most effective argument I have for getting other developers to comment their code is "The agent will read it and it will give better suggestions".

Truly perverse, but it works.

I agree with you... but the reality is that there's a wide contingent of people that are not capable of understanding "people don't know the same things as me". So they need some other reason.

It's made my project documentation so much better. If I write out really good acceptance criteria, 9 times out of 10 I can point Claude at the ticket and get a workable (if unpolished) solution with little to no supervision.

They understand it just fine; they are acting selfishly, because it does not benefit them. Helping the coding agent does.

They really might not understand it fully. That's very much in line with my understanding of how autism works.

several ironies here:

1) an AI agent is less likely to notice than even a junior is when the docs are out of date from the code

2) AI boosters are always talking about using language models to understand code, but apparently they need the code explained inline? are we AGI yet?

3) I frequently hear how great AI is at writing comments! But it needs comments to better understand the code? So I guess to enable agentic coding you also have to review all the agents' comments in addition to the code in order to prevent drift

HOW IS ANY OF THIS SAVING ME TIME

Well... Yah. For the record I'm saying this to trick humans into making better comments for humans. It is very difficult to convince people to do this otherwise, in my experience.

buuut...

I will also mention that these agent files are typically generated by agents. And they're pretty good at it. I've previously used agents to dissect unfamiliar code bases in unfamiliar languages and it has worked spectacularly well. Far far FAR better than I could have done on my own.

I have also been shocked at how dumb they can be. They are uselessly stupid at their worst, but brilliant at their best.

Think of it like it's saving future you time if you just let the AI centipede feed off of you. Surely it'll eventually regurgitate perfect code.

I assume you used these against a relational database? Did you commit those ids with the prefix still attached? or did you `.split()[1]` or something?

I think it's a pretty good idea. I'm just wondering how this translated to other systems.

We were using Mongo and stored the ids with the prefix in the DB as the primary key. Pretty much everywhere we were passing them around as strings, never as 128 bit int, so there was no integrity checking outside of the app layer.

The only drawback was marshalling the types when they come out of the db layer. Since the db library types were string we had to hard cast them to the correct types, really my only pain. That isn't such a big deal, but it means some object creation and memory waste, like:

    // pseudo code:
    const results = dbclient.getObjectsByFilter( ... );
    return results.map(result => ({
        id: result.id as ObjectId,
        ...
    }));

We normally didn't do it, but it would be at that time you could have some `function isObjectId(id:string) : id is ObjectId { id.beginsWith("object:"); }` wrapper for formal verification (and maybe throw exceptions on bad keys). And we were probably doing some type conversions anyway (e.g. `new Date(result.createdAt)`).

If we were reading stuff from the client or network, we would often do the verification step with proper error handling.

I've been looking for a tool like this that can publish. I was thinking of some way to create a help doc system for end users, but interleaved with technical information and discussion for devs. IE to make the help documentation a single source of truth for application behaviour.

All this needs to work is the ability to mark blocks of the document as "public" so only that gets published properly. Any possibility of doing this currently or interest in supporting in the future?

Keep chrome installed and fall back iff forced to. That way the majority of usage statistics show up as other browsers so when developers are making guesses at which browser to support, those statistics will push them away from chrome.

Additionally: you would be surprised how infrequently you have to switch to chrome

Can't remember the last time I actually had to open a website on chrome for compatibility reasons. Is that still a thing?

The F1TV site didn't work on Firefox earlier this year but send to be fixed now, other than that I haven't had any issues.

I only have to switch to chrome for e-transfers. Everything else seems to work

Btw, the 'website requires chrome browser' problem is often solved if you just make Firefox user agent say it is Chrome.

The problem is this needs to be a standard Firefox feature.

There's one site I have to switch to Firefox for. And it's a big one that handles a lot of money, so that's kind of surprising. Can't log into their site in chrome, no matter how hard I try. Nor edge.

So, hypothetically here, say Apple was still extracting extortionate rents: Why wouldn't they just choose to make more money with ads?