Hacker Newsnew | past | comments | ask | show | jobs | submit | firefoxd's commentslogin

Without even looking at the AI part, I have a single question: Did anybody investigate? That's it.

Whether it's AI that flagged her, or a witness who saw her, or her IP address appeared on the logs. Did anybody bothered to ask her "where were you the morning of july 10th between 3 and 4pm. But that's not what happened, they saw the data and said "we got her".

But this is the worst part of the story:

> And after her ordeal, she never plans to return to the state: “I’m just glad it’s over,” she told WDAY. “I’ll never go back to North Dakota.”

That's the lesson? Never go back to North Dakota. No, challenge the entire system. A few years back it was a kid accused of shoplifting [0]. Then a man dragged while his family was crying [1]. Unless we fight back, we are all guilty until cleared.

[0]: https://www.theregister.com/2021/05/29/apple_sis_lawsuit/

[1]: https://news.ycombinator.com/item?id=23628394


The thing about the legal system is there's no incentive to investigate to find the truth.

The incentive is to prosecte and prove the charges.

Speaking from the experience of being falsely accused after calling 911 to stop a drunk woman from driving.

The narrative they "investigated" was so obviously false, bodycam evidence directly contradicted multiple key facts. Officials are interested only seeking to prove the case. Thankfully the jury came to the right verdict.


There needs to be consequences for shitty, procedure-ignoring police work. Period.

Minimum 1 year of jail time for grossly wrongful arrests that could be avoided with standard procedure or investigation tactics that were not applied.


I agree with this sentiment but when you start punishing this sort of thing you create more incentive to cover it up. It's a tricky problem and I'm not sure there's a perfect solution.

What we really need is a change in police culture.


Medicine has a culture that adapts to this quite well. If you make an honest mistake and communicate it, you are often persecuted by your peers but not hung out to dry legally by your hospital and generally your actions are always defensible.

Similar practices are used in law enforcement, but the legal implications are seemingly more severe


Then the system should be redesigned such that transparency is a priority and cover ups are not feasible. And when cover ups eventually get found out, the punishments even more severe.

We already have administrative punishments for the police when they incorrectly assign blame and cause a public relations mess.

Is the termination of your career and/or potential retraining and social embarrassment not already an incentive to cover up?


> change in police culture

until then, there's a simple rule which works well: never talk to a cop. Or at least say the minimum number of words possible, give them nothing to use against you. Present ID if they ask for it, but never admit anything. If they persist, "lawyer". That has worked for me.


These dialogs always prompt me to chime in with my solution: make the police be self-insured, backed by their pension fund.

The police today have zero incentive to serve the public, they have zero skin in the game and can literally get away with murder.

Any time you hear the call for "law and order", that is the audience that supports the current system, because they like it like this.


Police in some states are actually self-insured, though not backed by a pension fund.

> These dialogs always prompt me to chime in with my solution: make the police be self-insured, backed by their pension fund.

I'm curious, what exactly do you mean by "self-insured"?

(Is the idea to combine literal insurance underwriting for retirement planning with a monetary incentive system for ongoing work performance)?


Great idea, Except that this will never happen because public sector unions are important voting blocks. Public sector unions should be abolished (don’t have a problem with unions) but the conflict of interest is just too great.

Great point. Obviously can't expect them to vote against their own interests, because higher standards, higher accountability, and higher transparency will always be against those interests.

> The thing about the legal system is there's no incentive to investigate to find the truth.

The truth is much more complicated and involves politics. For example Seattle (and possibly other cities?) enacted a law that involves paying damages for being wrong in the event of bringing certain types of charges. But that has resulted in some widely publicized examples where the prosecutor erred by being overly cautious.


And then you have Florida who will bill you about $100 a day for finding yourself in a Florida jail, regardless of whether charges were dismissed, you were found not guilty or any such thing.

And to nobody’s surprise, failure to pay this bill is in itself a Class B felony…


That sounds like a recipe for domestic terrorism - the systemic disenfranchisement of people who have done nothing wrong for no apparent reason other than sheer greed. How long has this been in effect there?

The system doesn't push the issue on people who can't afford it. Blood from a stone and all that.

I'm confused. Are you suggesting such a ridiculous system is letting class B felonies slide here? That would certainly be the pragmatic approach to being evil but in that case simply treating it as regular debt and going through civil channels would be more than sufficient.

Are you letting stuff in your backlog that you'll never get to before the product is gone or irrelevant "slide"?

Sure they could round those people up pretty easily just by following up on any contact with the system that they have, but why, for what, to cost the state more money that will likely never be repaid? Especially when sticking a body on DUI detail is hugely in the black. They'll just let that debt, it's accruing interests and the threat of further incarceration linger on the books indefinitely. If the person ever gets their life together they'll have to pay it or face incarceration.

I'm sure someone somewhere has written a DB query to select from outstanding balance where <exists in some other DB that is a proxy for people who have money to pay> and prioritize those cases.


Are you suggesting that Florida it’s to go ‘soft on “crime”’? That would fly in the face of almost all available evidence.

I have extended family in Florida. The system absolutely can and does and will push the issue. There’s a reason that it’s a crime not to pay for your incarceration even if you have a finding of factual innocence against you.


Your family isn't sleeping under a bridge or whatever. Of course the system wants your money or the money of people on comparable economic footing you associate with. If you can work as a debt slave to the system it wants you to do that even if it means a never ending cycle of robbing peter to pay paul, sleeping on other people's couches, etc. The man sleeping under a bridge cannot, so the cops and the DA and everyone else just go fry bigger fish. Maybe they push the issue 1/100th of the time and incarcerate someone every now and they but they absolutely do not prioritize it the way they do someone who could pay even if only by moving heaven and earth. The system doesn't want to manufacture yet another felony and then incarcerate someone for it out of thin air, that just costs the system more money.

Source: my tiny keyhole view into the system. The parties involve always have have discretion to downgrade stuff to something else, or not pursue it at all and are incentivizes.


That sounds absolutely terrifying.

> The narrative they "investigated" was so obviously false, bodycam evidence directly contradicted multiple key facts. Officials are interested only seeking to prove the case. Thankfully the jury came to the right verdict.

I don't get it, if they only care about prosecuting and proving the case, wouldn't they go by the bodycam evidence? They didn't prove the case. Maybe if their incentive was to prosecute and prove the charges, they'd go by the obvious evidence. Or am I missing something here?


There’s a judge down in Texas, Dallas area I believe, who is in social media a lot because he will excoriate prosecutors who bring bs in to his court room. He’s not soft on crime but hard on rights and process. If a defendant did the wrong thing, he will have the appropriate amount of sympathy, down to zero. At times he will tell them, we all know you got lucky here, do better. But he won’t let prosecutors slate by on garbage charges or statements or investigations by police. Which leads to my primary point at least for this discussion in particular:

To me the scariest part of this as a process is how many times (I’d casually estimate at least 75%) it is blindingly obvious that the prosecutor has not read the statement of charges or officer statements until everyone is in front of the judge. I get on one hand this judge seems to often be handling probable cause hearings but so many of these should never have resulted in any paperwork being turned in to the prosecution, let alone anyone having to show up in court.



It's fascinating to me that judges are elected in Texas, and what's more, run as members of a political party.

There is an incentive . It’s called fraud by negligence. I’m hoping she sues everyone here.

That’s seems to be in the realm of poissibility here if I am understanding things correctly (imo)


I would absolutely never call the police on a woman. Simply walk far away and let her be someone else's problem.

Unless it’s a Karen chasing you and yelling and threatening to call the police on you for some asinine reason?

Imo they're right, if you're faced with the option of running away from some crazy person or interacting with the police in the USA, the safer option is to run.

A police interaction can escalate to ruinous heights within seconds due to no fault of your own. Remember that cop that got scared by an acorn falling and started shooting at random? I don't care how many "good cops" there are, I'm not rolling the dice on encountering an acorn cop.


Yes, of course someone should have investigated, but the larger point here is that people don’t because they are being sold a false narrative that AI is infallible and can do anything.

We could sit here all day arguing “you should always validate the results”, but even on HN there are people loudly advocating that you don’t need to.


I don't think people on HN think "AI is infallible", I think people on HN believe HN is sufficient enough for "most tasks". In the context of HN "most tasks" refers to programming tasks, not arresting and jailing people tasks.

You should always validate the results, but there is an inherint difference between an AI generated tool for personal use and a tool which could be used to destroy someones life.


The problem is that the people who will put this in place rate capability on a linear scale: in their view the ability to write software is sufficiently magic, so such an ability is obviously good enough to recognize criminals. From their perspective, there are hurdles to be crossed (like probable cause) and an AI flagging a suspect feels like a magical intelligence crossing those hurdles and allowing them to continue in the process.

They don't validate the results of their fellow officers, or the validity of warrants, or anything else that predicates an arrest. Why would they start with this?


What about cops and legislators? They thing AI is infallible and thats very convenient for them since they can thus not mandate cops having to double check tmwhat the AI suggests

We can barely convince powers thar be that eye-witness testimony is unreliable, after all.

Where are you seeing people being told that AI is infallible? AI is being hyped to the moon, but "infallible" is not one of the claims.

To the extent people trust AI to be infallible, it's just laziness and rapport (AI is rarely if ever rude without prompting, nor does it criticize extensive question-asking as many humans would, it's the quintessential enabler[1]) that causes people to assume that because it's useful and helpful for so many things, it'll be right about everything.

The models all have disclaimers that state the inverse. People just gradually lose sight of that.

[1] This might be the nature of LLMs, or it might be by design, similar to social media slop driving engagement. It's in AI companies' interest to have people buying subscriptions to talk with AIs more. If AI goes meta and critiques the user (except in more serious cases like harm to self or others, or specific kinds of cultural wrongthink), that's bad for business.


> Where are you seeing people being told that AI is infallible? AI is being hyped to the moon, but "infallible" is not one of the claims.

I see all kinds of people being told that AI-based AI detection software used for detecting AI in writing is infallible!

You want to make sure people aren't using fallible AI? Use our AI to detect AI? What could possibly go wrong.


Where did you see this claim about AI-based AI detection?

> To the extent people trust AI to be infallible, it's just laziness and rapport (…) that causes people to assume that because it's useful and helpful for so many things, it'll be right about everything.

Why it happens is secondary to the fact that it does.

> The models all have disclaimers that state the inverse. People just gradually lose sight of that.

Those disclaimers are barely effective (if at all), and everyone knows that. Including the ones putting them there.

https://www.youtube.com/watch?v=Xj4aRhHJOWU


Society went through the necessary lessons with DNA and fingerprints. Putting people in jail because the computer produce a match is a terrible idea, especially when its done by an proprietary dark box that no one really understand why it claims there is a match. It can be used as a tool of investigations to give the investigators an hint to find real more substantial clues, but using it like in fiction where the computer can act as the single truth is terrible for society and justice.

A month ago or so people on HN discussed facial recognition when looking victims and perpetrators in child exploitation material, and people were complaining that meta did not allow this fast enough. Neither the article or the people in that discussion draw any connection that the issues in this article could happen. People seemingly want to think that the lesson is "Never go back to North Dakota", as that is a much easier lesson than considering false positives in detection algorithms and their impact on a legal system that is constrained in budget, time, training and incentives.


I think you missed many important points.

"The trauma, loss of liberty, and reputational damage cannot be easily fixed,” Lipps' lawyers told CNN in an email.

That sounds a LOT like a statement you make for before suing for damages, not to mention they literally say "Her lawyers are exploring civil rights claims but have yet to file a lawsuit, they said."

This lady probably just wants to go back to normal life and get some money for the hell they put her in. She has never been on a airplane before, I doubt she is going to take on the entire system like you suggest. Easier said than done to "challenge the entire system", what does that even mean exactly?


It was worse than that, the reporting from an earlier story[0]

  ...Unable to pay her bills from jail, she lost her home, her car and even her dog.
There is not a jury in the country that will side against the woman. I am not even sure who will make the best pop culture mashup - John Wick or a country song writer?

(Also, what happened to journalism - no Oxford comma?)

[0] https://news.ycombinator.com/item?id=47356968


As an aside AP Style is not use an Oxford comma, and that's been the rule for 50+ years https://www.prnewsonline.com/explainer-how-to-use-oxford-com...

This is upsetting.

Yes, finding out how badly wrong you were is never fun. Of course the lack of ubiquitous Oxford comma use is itself and separately displeasing.

AP Style is simply wrong on this, then.

You have more faith in the country than I do.

Normally, I would be a bit more grim, but people love their animals. I pray even the staunchest authoritarian would see the injustice of losing a dog.

You're not aware of Noem killing her dog by shooting it in the face, lining up three horses and shooting them, while being proud of it all?

Indeed let out on Christmas Eve with no money 1000 miles from your homeland.

Where your home was lost to foreclosure because one JUDGE did not look at the paperwork.

There should be a way to personally sue somebody when they don't do their job. Protecting the innocent. The JUDGE failed badly here.

Flimsy evidence would mean no warrant. Do your basic investigation please... Rubberstamping JUDGE caused this.

Why are they not named? Like they are a spectator. Infact they are the cause.


TBF isn't it rather unreasonable that our system permits your home to be foreclosed while you're detained prior to a hearing?

Also rather unreasonable to arrest someone who is clearly neither violent nor a flight risk. You could literally hold the trial via video conference at that point and there would be no downside.


At the risk of sounding like more of an anarchist (irony, autocorrect went with absurdist which isn’t entirely wrong either) than I might usually feel, that all depends on who you believe the system is for and works for? If you believe it’s “capitalism” as been so often proven, then it could be said that it’s entirely “reasonable”.

> depends on who you believe the system is for and works for

We are still enough of a democracy to blame ourselves for this. We could choose that the system is of the people, by the people, for the people. I think too many of us simply don't agree with that, except in the narrow situation where we are talking about ourself.


We could just overcome the tens of billions shoved into our faces aimed at undermining it and brainwashing us, and choose that the system is of the people?

The deck is so unbelievably stacked against it.

Another thing: many people hav e been permitted to vote in let's say 40 elections (at different levels), out of which maybe 1 had a candidate that indeed supported a "system that is of the people", and 39 didn't. Gets tough then doesn't it.


anyone in the chain of responsibility should be punished so severely that they will be still crying about it in 2030

The real problem here is she'll get money, who knows how much, but that ultimately does nothing to actually address the problems in the system.

Effectively it just raises taxes to cover the cost of these failed prosecutions.

Everytime one of these cases happens, a cop and a prosecutor should be out of a job permanently. Possibly even jailed. The false arrest should lose the cop their job and get them blacklisted, the prosecution should lose the prosecutor's right to practice law.

And if the police union doesn't like that and decides to strike, every one of those cops should simply be fired. Much like we did to the ATC. We'd be better off hiring untrained civilians as cops than to keep propping up this system of warrior cops abusing the citizens.


> The false arrest should lose the cop their job and get them blacklisted

There is actually a federal register for LEOs that have been terminated for cause or resigned to avoid termination.

The police unions that operate in the jurisdictions that employ 70% of US police have negotiated into their CBAs that the register “cannot be used for hiring or promotional decisions”. Read into that what you will.


I'm generally pretty for unions, but the police union is one that's a complete cancer on society. It pretty much solely exists to make sure cops are free to harm the public without any sort of accountability.

Agreed. And I think we really, really need to put more effort into a "police the police" organization. Someone who has power only over the police, who the police do not have power over, to act as a check.

We might call this the administration of the executive. Maybe we can vote for that or something.

> police unions

... test my support for the idea of unionization. I have even said in the past that I think public sector unions are especially important because their boss (the people) are the most capricious and malicious of all.

Maybe we could find a way to put guardrails on what they could and could not negotiate into a contract. Wages, benefits, basic job environmental conditions, stuff like that -- okay. But administrative policies which exist to prevent bad behavior should be non-negotiable.


>No, challenge the entire system.

Agree in principle. But people like her does not have the resources, financially and emotionally to go through the legal system again. Unless there are charitable lawyers who are willing to do it on her behalf for free.


IANAL but AFAIK custodial interrogation triggers Miranda, lawyers, and those awful awful civil liberties we’re trying to get rid of.

Better just to apply Musk or Altman software to the problem and avoid it entirely.


> Whether it's AI that flagged her

It absolutely was. There's no question of this. Now we need to ask how was the system marketed, what did the police pay for it, how were they trained to use it?

> anybody bothered to ask her "where were you the morning of july 10th between 3 and 4pm.

Legally that amounts "hearsay" and cannot have any value. Those statements probably won't even be admissible in court without other supporting facts entered in first.

> we are all guilty until cleared.

This is not at a phenomenon that started with AI. If you scratch the surface, even slightly, you'll find that this is a common strategy used against defendants who are perceived as not being financially or logistically capable of defending themselves.

We have a private prison industry. The line between these two outcomes is very short.


>Legally that amounts "hearsay" and cannot have any value.

How is that hearsay if she's directly testifying to her own whereabouts?

Hearsay would be if someone else was testifying "she was in X location on july 10th between 3 and 4pm", without the accused being available for cross


No!

"I was at the library" is firsthand testimony.

"I saw her at the library" is firsthand testimony.

"I saw her library card in her pocket" is firsthand testimony.

"She was at the library - Bob told me so" is hearsay. Just look at the word - "hear say". Hearsay is testifying about events where your knowledge does not come from your own firsthand observations of the event itself.


You don't know what hearsay means

> Legally that amounts "hearsay" and cannot have any value. Those statements probably won't even be admissible in court without other supporting facts entered in first.

I just want to understand your argument: you believe that any alibi provided is hearsay, and has no legal value, and that they can't even take the statement in order to validate it? That's your position?


The condition here being she was already arrested. You don't arrest someone first and then try to establish their alibi second. That would be an investigation which would be prior to getting a warrant which would allow you to arrest someone. You will never talk yourself out of an arrest, you might talk yourself out of an investigation.

You can offer your story to the police but the fact that you did or what you said to them will not come into evidence in court. You cannot call the officer to the stand and then ask them to repeat in court what you said. That would be "hearsay." So, for a lot of reasons, if you're already arrested, you probably don't even want to tell them any of that. It can only be used against you and never for you. Get your lawyer and have them ready the case to prove that alibi for you.


> but the fact that you did or what you said to them will not come into evidence in court.

What?? Isn't it that everything you say can be used in court? Aren't interrogations and arrests recorded?


The experience is a lot like when you are talking with a friend, then they decide to ask siri or google a question using voice. The result is always imprecise. Meaning they either have to repeat their query, or end up typing it anyway.

If you want to buy a Walmart product, the easiest way is to go to Walmart. Why add an imprecise middle man in between?


I think the cult of personality always backfires, pun intended. Our company biggest product was a celebrity making fun commercials for the actual product. Works wonders. Personally I don't have a problem with him, I enjoyed his movies in the past. But not everybody does. Internally, the company tried to push this cult so deeply that it was part of the hiring process, part of the onboarding, even obscured some of the CEOs messaging. And you wonder, what happens when you hire someone who doesn't like this celebrity?

Many of us are mature enough to follow the principle of, "if you don't have something nice to say, don't say anything." But not so when you have young developers flowing in and out of the company. In one of the town halls, a 24 year old dev, was put on a mic, and simply said, "I don't like X, he is super annoying, why do we keep plastering his face everywhere."

I've never seen an entire company freeze before. There was no way forward, no way backwards. The script had been broken. The dev, thinking he wasn't heard properly, sent the same message in our townhall slack channel. I did what I believe 90% of other people did. I screenshoted it.

The kid got another job a few months after. For once we saw the emperor wore no clothes.

Edit: million typos

Edit 2: in case it wasn't clear, no was not fired, he just found another job.


I've been that 20-something myself.

I was working as a programmer at some high flying merchant bank in London in the 90's and at the pub with my workmates one night I started tearing strips off of the IT director because he was comically incompetent. Everyone was kicking me under the table because unbeknownst to me his close friend was at the table taking in my rant. Everyone agreed that I was toast and bought me drinks.

In the morning, at about 10am, security went into his office and marched him out of the building, right past my desk. I turned around and said to my team and said "See! Don't fuck with me!"

It was hilarious.


Was his close friend the MD or something? I’m curious how this wound up playing out exactly backward from how everyone expected it would.

It was a coincidence. A few weeks earlier there was a fault at the bank's data center. The very expensive backup data center failed to go online. Management was not amused.

Another memory from that time: a stressed sounding trading desk assistant rang me asking after a trade confirmation that went missing and the client was demanding. I determined that the system I worked on didn't handle those kinds of trades. Out of curiosity I looked up the trade. It was for 2 billion GBP of UK Gilts (government bonds), thats about $5 billion USD in today's money.


I don’t often literally LOL here.

Bravo.


The interview for the next job must have been interesting.

Anyway, good for him. Too many agree to too much because they fear they'll lose their job.


> Many of us are mature enough to follow the principle of, "if you don't have something nice to say, don't say anything."

This isn't maturity, this is selfishness. A group often benefits from someone challenging the status quo, but the individual doing that gets punished. In your view, Germans during WW2 were "mature" by not saying anything that wasn't nice about nazism, and nowadays Russians are "mature" when they don't want to discuss a war that left a million people either dead or wounded - both cases are individuals acting out of self-preservation, not "maturity".

If you're American, then maybe a good example is Martin Luther King Jr. - do you really think that he should've had the maturity to shut up and not say anything that wasn't nice about racism? Well, he got killed, just like your junior employee got fired, so I guess he was indeed a loser in a sense.

In general this is a very common pattern in corporations where everyone is "just doing their job" and "being mature" but the end result is atrocities - for example Nestle literally killing babies.


I might generalize this as “push against actions, not personalities”. Don’t like someone because you just don’t mesh? Keep your mouth shut. You have a different outlook than they do, and yours is as wrong from their POV as theirs is from yours. Someone you think is interesting and fun to be around does something bad? Resist it.

It happens very rarely that there's someone I don't like as a person but I know they're competent in their domain. Sure, it does happen, but it's very rare.

For real? I’ve known all sorts of brilliant bastards. Or even people who weren’t actually jerks, just incompatible with me in an oil and water sort of way, like they were exceedingly uptight, or so happy-go-lucky that they thought I was the uptight one. We just didn’t get along at all, but they weren’t bad people, and certainly not incompetent, just not a good pairing with me.

There's a difference between criticism and personal insults, and honestly if you can't tell the difference you might be a liability for your employer.

In a perfect world yes. In reality, people often take offense in being criticized. Or ignore the feedback until the situation becomes disastrously bad.

I am guessing this is Salesforce and the celebrity is Matthew McConaughey who’s real chummy with Benioff.

Brave of the developer to bring it up. This cult of personality is pervasive throughout the tech industry.


I was thinking of Mint Mobile. I guess there are many examples, not even counting the cases where the CEO becomes a celebrity (like Steve Jobs).

That's why Kagi is winning! I can't wait for LinkedIn "influencers" to start using this seriously.

Last year, we had layoffs. I spent a whole lot of time on linkedin writing recommendations for my ex-colleagues. I couldn't help but read so many LinkedIn stories, and I was surprised how supportive the comments where. I couldn't believe it. So naturally I've come up with my own [0]. I don't think I have the courage to post it there with a straight face though.

[0]: https://idiallo.com/blog/the-show-on-the-other-foot-linkedin...


Not sure what you mean by 'seriously'. People don't dish out this level of advice for fun--they do mean business; they are serious.

They are also mindless corporate drones. Better to invest time in actual skills.

That's very funny actually. The one shoe ceo.

Not the author, but last year I wrote about my experience being on #1 on HN [0]. I created a visualization of the requests hitting my server.

[0]: https://idiallo.com/blog/surviving-the-hug-of-death


Very cool post! I enjoyed it a lot! Do you have any more backend insights like this on your blog? I'd happily read those.

Btw. was the writing assisted by AI or is it all you?


In Brave browser on Android device, the visualisation media(?) is blacked out. just wanted to give you a heads up.

Thanks for letting me know. It seems like when Brave is set to aggressively block tracking, it also blocks the canvas since it could be used for fingerprinting a device.

I'll see if an avg version is viable in the future.



If you are serious about sharing written ideas, I suggest you avoid using this type of prompts at all cost. I've worked with LLMs to write on my blog and they are pretty good at first glance [0]. But do it a few time and you'll notice that those tropes are the least of your problems. Not only all your articles will sound the same, but you'll see that same voice on other blogs, news articles, white paper, etc. It's as if they were all written by Mo Samuels. Readers are often here for the author's voice, not just the content of the text.

I often hear this here: "if you don't bother writing, why should I bother reading?" In fact, save us some time and just share the prompt.

[0]: https://idiallo.com/blog/why-we-hate-llm-articles


I have seen people suggest that the problem is that LLMs let you express any of your ideas, but the number of people with ideas worth expressing is limited.

In a sense I think this is accurate, but not inevitable. I think there is a lack of creative thinking, but it has come from a world that doesn't value it and suppresses difference.

There is a brilliant line in Treehouse of Horrors IV where Principle Skinner says "Now I've gotten word that a child is using his imagination, and I've come to put a stop to it." Which is just the perfect comment on the modern education system.

Models trained on the lack of diversity will push one way, but I think it will also avenues for expression that didnb't exist before. The balance will come from how we react and support what we would like to have happen


I think it has more to do with LLM's being statistical models than human creativity lacking in the input. The creativity and millions of voices and tones may be there, but since these models tend to go for the most likely next words, polishing this away becomes a feature.

A text by a human mind may be seen as a jagged crystal with rough edges and character. Maybe not perfectly written but it's special.

An LLM takes a million of crystals and trims the most likely tokens to be chosen into what would rather appear as a smooth pebble; the common core of all crystals. And everyone using the LLM will get very similar pebbles because to the LLM, regardless who is speaking to it, it will provide the same most likely next tokens. It's not that creativity is lacking in the input, but the LLM picks the most commonly chosen words by all humans in given contexts.

For that to sound imaginative and great as you go, it would have to not only exist in the data, but be a common dominating voice among humans. But if it was, it wouldn't be seen as creative because it would be the new normal.

So I'm not sure how there's a good way out of this. You could push LLM temperature high so that it becomes more "creative" by picking less popular tokens as it writes, but this instead tend to make it unpredictable and picking words it shouldn't have. I mean, we are still dealing with statistical models here rather than brains and it's a rough tool for that job.


>I think it has more to do with LLM's being statistical models than human creativity lacking in the input. The creativity and millions of voices and tones may be there, but since these models tend to go for the most likely next words, polishing this away becomes a feature.

I have always thought this is a rather misguided view as to what LLMs do and indeed what statistical models are. When people describe something as 'just statistics' I feel like they have a rather high-school-ish view of what statistics represents and are transferring this simplistic view to what is going on inside a LLM. Notably they do not find the most probable next word. They find the probability of every word that could come next. That is a far richer signal than most imagine.

And ultimately it's like saying that human brains are just chemical bonds changing and sometimes triggering electrical pulses that causes some more chemicals to change. Complex arrangements of simple mechanisms can produce human thought. Pointing at any simple internal mechanism of an entity without taking into account the structural complexity would force you to assume that both AI and Humans are incapable of creativity.

Transformers are essentially multi-layer perceptron with a mechanism attached to transfer information to where it is needed.


> They find the probability of every word that could come next.

If we're being pedantic, they find a* probability for every token (which are sometimes words) that could come next.

What actually ends up being chosen depends on what the rest of the system does, but generally it will just choose the most probable token before continuing.

* Saying the probability would be giving a bit too much credit. And really calling it a probability at all when most systems would be choosing the same word every time is a bit of a misnomer as well. During inference the number generally is priority, not probability.


I was using the term word to be consistent with the previous comment. It need not be a word, or even text at all.

Most systems choosing the high probability thing is what probability is.

They're just relative scores. If you assume they add to one and select one based on that it's a probability.


> I have seen people suggest that the problem is that LLMs let you express any of your ideas, but the number of people with ideas worth expressing is limited.

It doesn't just have to be one problem.

1. Laundering your "ideas" through an LLM makes them less of your ideas, at best you get the classic two sentences of content embedded in two pages of padding.

2. LLMs removed a filter that help cut down on the amount of useless writing we'd have to wade through. The difficulty of expressing an idea acts as a filter to weed out many (but not all) ideas not worth expressing. That applies to both to people with ideas worth expressing and those without.

On the former, I've had the experience of having an idea, then witnessing it fall apart as I try to express it, as I think about it more deeply. LLMs let you avoid that.


> I often hear this here: "if you don't bother writing, why should I bother reading?"

That is an opinion somebody shared on X which has been mindlessly repeated over and over again in other places such as this site.

Why do you value those comments when all they are doing is parroting something they didn’t think themselves? It seems to undermine your point entirely. There is zero originality or effort in those comments. Why are you bothering to read them?

Copying and pasting somebody else’s opinion from one social media site to another is no more virtuous than what you are complaining about.


I value that opinion because it resonates with me. When I use an LLM to write an article, it's usually because I don't have the time or energy to go through my normal process of writing.

Sure, I still end up with a polished article, but a lot of it is not entirely my idea or something I would have written through the filter of my own experience. So in order to share my true take on a subject, I have to go through the struggle of writing and bouncing of ideas in my head, which almost always results in a better output.


"Sharing the prompt" is a category error. It assumes the value of a piece is in the instructions given to the model, rather than the proprietary input or the iterative editing that follows. There is a hard line between using an LLM to generate content from a void and using it to synthesize specific ideas.

If someone asks a model to "write a post about X," they are outsourcing the thinking, which results in the homogenized voice everyone is tired of.


As someone working for a telco, not Vodafone, this would be my assumptions: A developer mistakenly grabbed a real MSISDN, instead of a QA one, while testing a promo still in development.

I only say this because there's no identifier to differenciate a real phone number from a test one. Subscribers often called to report those gibberish text messages they received. It's always a dev entering an incorrect number while testing.


When I worked for an Australian telco (not Vodafone), some developers on another team had used a very conspicuous mobile phone number in their integration tests, which actually connected to a real SMS service somewhere else in the company. No idea why they would do this. It turned out that this number belonged to a real person, who got absolutely buried in test SMS messages, when the integration tests ran as part of a CI/CD pipeline. The owner raised a complaint to the ombudsman, which led to all kinds of trouble for the developers.

In case anyone else here is curious, the ACMA maintains a list of reserved numbers for use in creative works, which you can use for dummy data: https://www.acma.gov.au/phone-numbers-use-tv-shows-films-and...


I worked for an Australian insurance company and we physically DDOSed a poor man's real mailbox with printed policy documents as we used their address during e2e testing and we mistakenly didn't put a testing flag somewhere.

Our CTO had to personally apologise to him


I lost a Nationwide Building Society account I've had for forty years last year because the bank bought some extremely poor online-ID-verification system.

The bank forgot it had customers in a Crown Dependency. It forgot those countries issues their own ID, their own passports, their own drivers' licences. It forgot it closed its branches in those countries: it told me I had to go into my branch. My nearest branch is a £200 airfare away. It was not paying, naturally.

The crappy online-verification tool only recognises UK documents. It can't handle Isle of Man ones. They did not think.


This sounds hilarious, how many physical items are we talking? Like his whole front porch full up of contract boxes?


The documents pack is like an A4 folder 1cm thick. He received close to 100 in one day. Enough for his mailbox to get full and for the postie to dump most of it on the lawn


In the 1970s, a German rock group had a one-hit wonder with a protest song against Munich's sex trade licensing (Skandal im Sperrbezirk). In the lyrics, they had a made-up (so they thought) phone number 32-16-8 that fit the meter of their lyrics in German.

Unfortunately, that was a real phone number in many cities, you could dial the short/local number directly without a 0 and the area code back then. Cue prank calls across the country and quite a few scandals since the topic of the song was, after all, the sex trade.


Same thing happened in the USA with 867-5309/Jenny.

https://en.wikipedia.org/wiki/867-5309/Jenny#Popularity_and_...


Online, I believe that one bit of Beverley Hills has the highest number of online users in the USA.

I am one.

I have never visited California or the West Coast of the USA in my life.

But I have used dozens of websites which need a Zip code.

I have never had a Zip code. I have postcodes, like my old one IM2 3EW. That has letters. It won't fit a Zip code field.

So I and millions of others use the only Zip code we know:

Beverley Hills 90210.


I worked at a grocery retailer, and we had the same exact thing. The CI/CD pipeline was firing out order related SMS messages to a contractor's number during test runs for years.

I wonder how common something like this is.


0412 345 678

That was one number we were told to stop using at Internode. I heard similar stories from Optus and Telstra employees.


Someone I know who works at a telco (no idea if Vodafone is a thing in Belgium, but whatever: not Vodafone) was talking about a number someone has: 0411 11 11 11, and they got over a hundred operator messages every day.



I was slightly more inclined to think it might be some bored employee somewhere acting in a sort of Robin Hood capacity just because it's unusually accurate and thorough for a test message. I'd expect more like TEST TEST test DFOIUHDFUOHDFOIUHDFROIHDSFOIHDSF LOREM IPSUM 999999.

Sometimes enthusiastic or particularly bored developers do put in the effort to write things out like a real message though.


Testing vs prod bugs are always FUN.

In my first job we had warehouse management system, and for testing new versions we allowed users to log-in to test environment.

Some employees didn't knew they were supposed to only log in to prod and happily worked in their warehouse accepting deliveries, stocktaking, moving stuff in real world using test db instead of the prod one. We only realized when they moved so much stuff that the inconsistencies db vs reality triggered alarms.


I have a phone number that is +1(gapless non-decreasing sequence). I’m entertained every week by developers testing.

(Also, people using it as a fake number for some appointment/reservation - which I sometimes update to change their name or add a special request :))


There was a tool shared here that could show which accounts belong to the same person based on the writing patterns. Can't remember the name, but it found my old accounts on HN pretty accurately.


https://news.ycombinator.com/item?id=33755016

Way simpler than hnprofile from the sibling comment. This one used cosine similarity between user vocs - https://web.archive.org/web/20221126225241/https://stylometr...


Oh yes, this is the one I was referring to. Too bad it's shutdown.


Hnprofile.com which has since closed down - lettergram was the author - https://news.ycombinator.com/item?id=17942981


I used to take Uber to work daily in 2016. It cost around 3 - 4 dollars per 5 miles ride. Now the same ride cost $24 [0]. There's no indication that AI coding tools won't follow the same path given they are funded by VC.

But I think what matters is that the new generation of coders will adopt it as the norm. Gone are the days where you download a free text editor and just trial and error with the documentation one tab away. Every bootcamp is teaching react with clause and cursor. You have to pay to for a subscription to build your BMI calculator.

[0]: https://idiallo.com/blog/paying-for-my-8-years-old-ride


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: