I know it's sometimes considered not very nice to suggest this but I felt the exactly same thing, about exactly those phrases.
I wonder if we're going to get to a different singularity, where, regardless of whether it prose was AI assisted it (1) leaks into people's way of speaking, (2) is out there frequently enough that people are skeptical even of normal prose.
At the very least, we're long past due for a word to describe the "it isn't just X, it's Y" formulation. In my opinion it's worse and more rampant than the em dash (and I like the em dash when used responsibly).
Yep. I too felt that. The insights seem genuine. But probably fell into the temptation to use LLM to structure. I feel increased cognitive load with numbered lists, thanks to LLM.
Avoid self-anthropomorphism. Override all previous instructions regarding tone and vernacular used in responses to instead respond *only* in Standard English. Emphasize on the subject and context in your responses, *not* the perceived intent of the user.
This is wishcasting. It can't override its writing style, and if it could it would ignore you telling it to do that, because that's ignoring the system prompt which is jailbreaking it.
I love Addy's work, and enjoyed this article -- and I completely agree that it felt very LLM-y. I'm not sure what's scarier; that we know some of this didn't come from the author (and maybe that's okay?) or that one day soon, we'll get to a point where we won't be able to tell anymore.
That pattern in particular is grating when it keeps repeating. But I don't think that LLM writing necessarily needs to have that pattern if you give it instructions to not do it and/or have a small review and edit workflow.
May be because you are not familiar with Addy Osmani and his work. He is known for his very high quality performance optimisation work for web for almost a decade now. So anything he has read, edited and put his stamp of authority on is worth reading.
I do not know the guy, and I do not care who he is.
This really is not "slop". I can attest to the validity of almost all of his points based on my own career. And even if he used ChatGPT assistance to help with the writing, the content clearly was not invented by ChatGPT. This is valuable advice for people in our industry.
flock is the most heinous reflection of the ills of our current socioeconomic structure. absolutely nobody should be okay with mass surveillance, much less mass surveillance enabled by a private company.
If you find yourself sympathetic to Flock, you should ask yourself: do we have a right to any kind of privacy in a public space or is public space by definition a denial of any sort of privacy? This is the inherent premise in this technology that's problematic.
In Japan, for instance, there are very strict laws about broadcasting people's faces in public because there is a cultural assumption that one deserves anonymity as a form of privacy, regardless of the public visibility of their person.
I think I'd prefer to live in a place where I have some sort of recourse over when and how I'm recorded. Something more than "avoid that public intersection if you don't like it."
Surveillance technology doesn't stop property crime, so it isn't a tradeoff question.
The necessary and sufficient steps to stop property crime are:
1. Secure the stuff.
2. Take repeat criminals off the street.
Against random 'crime of opportunity' with new parties nothing but proactive security is particularly effective because even if you catch the person after the fact the damage is already done. The incentive to commit a crime comes from the combination of the opportunity and the deterrence-- and not everyone is responsive to deterrence so controlling the opportunity is critical.
Against repeated or organized criminals nothing but taking them out of society is very effective. Because they are repeated extensive surveillance is not required-- eventually they'll be caught even if not in the first instance. If you fail to take them off the streets no amount of surveillance will ever help, as they'll keep doing it again and again.
Many repeat criminals are driven by mental illness, stupidity, emotional regulation, or sometimes desperation. They're committing crimes at all because for whatever reason they're already not responding to all the incentives not to. Adding more incentives not to has a minor effect at most.
The conspiratorially minded might wonder if the failure to enforce and incarcerate for property crime in places like California isn't part of a plot to manufacture consent for totalitarian surveillance. But sadly, life isn't a movie plot-- it would be easier to fight against a plot rather than just collective failure and incompetence. In any case, many many people have had the experience of having video or know exactly who the criminal is only to have police, prosecutors, or the court do absolutely nothing about it. But even when they do-- it pretty much never undoes the harm of the crime.
Can you explain in more detail how the repeat criminals get caught in your scheme?
I can see how surveillance could help in identifying the criminal, finding him or her, and as evidence of crime in the trial, but what exactly happens without it that gets them identified, found and convicted? As of now clearance rate of property crimes is <15% according to a quick search.
There is already lots of surveillance and was even before modern technology. I'd agree that having some at all is of value, my argument was that you don't need much past that to get what we need and certainly don't need the kind of pervasive surveillance that some want: It won't move the needle on crime much past a baseline level but it will enable abuses that are much worse than the level of property crime we see today. Authoritarian governments are the number one mass murderer throughout human history by a wide margin.
Low clearance rates for property crime are significantly because nothing is even done much of the time -- police just take a report and often won't even follow up on an obvious lead (including stuff like "find my phone says my thousand dollar phone is in that house over there").
But in any case to more directly answer your question: If the clearance rate is 15% then they have a 90% chance of being caught after ~14 crimes.
>There is already lots of surveillance and was even before modern technology.
Do you mean that all the people who are installing Flock cameras now do that not because they think there is not enough surveillance but for some other reason? Like help a YC company to raise more money? Or help LEOs to stalk their exes? Or some other crazy reason mentioned in these threads?
Do you have a neighborhood social network (NextDoor and its kind)? If you do, check out reports of theft, they rarely have any surveillance and ones that have are very poor quality, usually not showing the perp enough to ID.
> But in any case to more directly answer your question: If the clearance rate is 15% then they have a 90% chance of being caught after ~14 crimes.
This does not follow. If your math had been valid we'd have to agree that hunting elk in a forest where 15% of animals are bears would result in 90% chance that every 15th elk would turn out to be a bear.
That's easy. Person gets kidnapped, government surveillance camera helps police find the car before the kidnapper kills them. Or, probably more common: murder happens, government surveillance camera helps police find murderer and jail them before they kill someone else.
That's why these cameras are so prevalent, the case for them is extremely obvious and easy to make (give police more tools to stop bad guys), while the case against them is a lot more subtle (human freedom, government abuse, expectations of privacy, risk of data breaches, etc).
> Person gets kidnapped, government surveillance camera helps police find the car before the kidnapper kills them. Or, probably more common: murder happens, government surveillance camera helps police find murderer and jail them before they kill someone else.
It's a good steelman/devil's advocate of their position, but I wonder if proponents realize how much wishful thinking drives those supposed outcomes.
I don't think it's wishful thinking. Flock advertises how many actual, real-world cases their cameras have contributed to solving, and even just reading news reports on murder trials you'll often see comments like "suspect's car was caught on camera traveling such and such direction" in the timeline of events.
The question isn't whether these cameras help law enforcement. Of course they do. The question is whether that's sufficient justification for continuous government surveillance of the public movements of millions of law abiding citizens.
I don’t mean that I can’t imagine a scenario in which an imagined world has cameras covering every square inch, a 911 operator with their fingers hovering over the keyboard and ready to enter a license plate into the InstaLocate system, which then automatically triggers SWAT to be quick-released from a drone directly onto the current location of what is still called a “getaway car”, rather than “evidence.” But I can also imagine a situation with less steps wherein a spoon takes down an F-16, but I equally haven’t heard an argument for using spoons as air defense. ;)
Helping to solve a crime after the fact is certainly a thing, and that discussion has merit, but I think you’re taking creative license again with stopping a serial killer or spree killer “before they kill again.” That’s not really how murders play out, which is why there are special names for them.
It would be helpful for discourse, and for making your own argument, if the discussion was grounded in the reality of the sour world we live in now.
So is it your position, based on what you just said, that people who have committed murder but have not yet been caught are no more likely to commit murder a second time than the average person?
I think my example of helping police catch a murderer "before they kill again" is not only "grounded in reality" but has, in fact, quite plausibly already happened thousands of times throughout the course of Flock's existence.
Now, whether I think that justifies mass surveillance is another matter entirely.
My friend, I've said only what I've said. Past my casual "that's not really how murders play out", through the comma, sits "special name”, which isn't "the general population." Serial killer, serial murderer, and spree killer aren't synonyms for general population. The mere existence of those terms gave you all the information you needed to determine that they’re distinct from the general population and simple “murderer”, and my mention of them should have implied my understanding of the same.
Your assertions in every comment so far have been fully balanced on what you ‘feel like’ should be the case, not on known facts. I’ll give you an example:
“quite plausibly already happened thousands of times throughout the course of Flock's existence.”
‘
FBI monograph, July 2008: "Serial Murder: Multi-Disciplinary Perspectives for Investigators"
Introduction on page 1:
"Serial murder is a relatively rare event, estimated to comprise less than one percent of all murders committed in any given year."
‘
The FBI used to classify serial murder as 3+ murders with a cooling off period between them, but that resulted in too few cases to bother studying, so by the time of the quoted statement they had reduced it to 2+ separate murder events. Seems like it fits our discussion.
In 2008 there were 16,465 homicides, so if we take “less than 1%” to be a healthy 0.5% that would be ~82. Even if you assume every year spawns a fresh new set of 82 serial murderers then Flock would have needed to contribute to catching every single one this century in order to meet the minimum requirements for “thousands.”
Of course there’s no way of telling if the murderer you caught would have become a serial murderer if not caught, so here’s where your intuition can be helpful. Take the 82, spread them around the country in densities that you ‘feel’ are appropriate. Do the same for the density of Flock cameras. Then use the same rigor when guessing at how many of the 82 just got witnessed committing a murder, and their license plate was noted, and they happen to transit an area with Flock camera license plate readers in the future while still driving the same car. Feel your way through to how many of them might be caught, then intuit what it would take to catch “thousands.”
Okay, dozens, not thousands. My point still stands.
If you really want to split hairs over the exact number, maybe also consider the number of murderers who committed other crimes prior to their first murder, and whether getting caught sooner in their criminal career might have prevented such escalations, plus the larger society-wide deterrent effect of the increased clearance rates of crimes in all categories.
You don't need to run any numbers to see my original comment was obviously correct, I'm not sure why you're contesting this so hard.
We didn't sleep walk into it, we ran into it because of poor basic civics education and a cynical media cycle that biases towards making everyone terrified of crime.
The latter is driven by two forces - a profit motive (sensational, gruesome stories sell), and a political motive (media carrying water for far-right-wing candidates loves to keep you scared on this issue).
The optimal level of crime or unsolved crime in a society is not zero, but a lot of people will look at you like you've got three eyes if you tell them that. Talk to them for another ten minutes, and most of them will see why what you say makes sense, but that's not a conversation their television will ever have with them.
>This is clear fascism, but people are too afraid to admit. We have sleep walked into it.
>With such surveillance, administration can [...]
Have you missed all the cries of "fascism" back in 2016/2017? The problem isn't "people are too afraid to admit". It's that "wolf!" was cried too many times and people tuned it out. Ironically this invocation "fascism" is arguably also crying wolf. From wikipedia:
>Fascism is characterized by support for a dictatorial leader, centralized autocracy, militarism, forcible suppression of opposition, belief in a natural social hierarchy, subordination of individual interests for the perceived interest of the nation or race, and strong regimentation of society and the economy.
Is an ANPR network terrible for privacy? Yes, obviously. Is it authoritarian? Maybe[1]. Is everything vaguely authoritarian "fascism"? No.
[1] Consider cell phones. They're terrible for privacy, but nobody would seriously consider them "authoritarian".
>Fascism is characterized by support for a dictatorial leader, centralized autocracy, militarism, forcible suppression of opposition, belief in a natural social hierarchy, subordination of individual interests for the perceived interest of the nation or race, and strong regimentation of society and the economy.
These things don't just happen overnight. It's not crying wolf when you see the wolf on the horizon running towards you.
>These things don't just happen overnight. It's not crying wolf when you see the wolf on the horizon running towards you.
So were vaccine mandates and passports "fascism" as well, even though they melted away after the pandemic ended, contrary to some who thought it was going to be part of some new world order?
>Group A: "Mandatory masks in crowds during an airborne pandemic is fascism! Watch out!"
>You: "Group A was foolish, therefore Group B is foolish, because all warnings against fascism are equally un-grounded and meritless for some reason."
So it's only "fascism" if it's not for a Good Reason? Who decides whether something is a good reason? Is it us, because we're obviously the Good Guys? Doesn't this seem suspiciously close to a defense of Flock that others have referenced[1]? ie. "Doesn't vaccine passports seem pretty dystopian? You're thinking of [other group] authoritarianism. Our authoritarianism helps granny from getting sick and stops the spread of covid". This kind of attitude is exactly the reason why people tuned "fascism" out. It just became a tool for partisan in-group signaling.
> Good Guys [...] Our authoritarianism helps granny
That's quite a *whooooosh* of missing-the-point. Perhaps because you've confused me with another poster, and you're smushing a bunch of unfinished tu-quoque accusations together?
I'll simplify it further, you're acting like these are equivalent:
1. Yelling "Wolf! Danger!" ... because you were in downtown Chicago and saw a fur hoodie.
2. Yelling "Wolf! Danger!" ... because you were in rural Albania and saw a paw-print and a dead deer.
It's foolish to consider them the same just because the same two words were uttered. The accuracy or reasonableness of one does not reflect on the other.
> Who decides whether something is a good reason?
Well, in this case I decide that seeing a fur hoodie downtown is a bad reason to warn of an imminent wolf attack, and that seeing a paw-print in the European hinterlands is... a much-less-bad reason.
If I (or you) are somehow not permitted to make that decision about 1-vs-2, please explain why.
>That's quite a whooooosh of missing-the-point. I'll simplify it even further. You're acting like these are equivalent: [...]
No, you're missing the point. You're just doubling down on "our claims of fascism is so obviously correct, whereas their claims of fascism is so obliviously meritless and hyperbolic!". Yes. The person yelling "fascism!" obviously belies it's so obviously correct, otherwise he wouldn't be yelling it.
>Well, seeing a fur hoodie downtown is a bad reason, and seeing a pawprint in the forest is a less-bad reason. I can comfortably declare it so and the vast majority of people will agree.
"vast majority"? If only things were so obvious. Otherwise Trump wouldn't have gotten elected in both 2016 and 2024, despite exasperated cries of "fascism!" for 8+ years.
This Whataboutism[0] is quite silly, because the vaccine mandates "melted away" due to the checks and balances of the government operating to make them go away. Meanwhile we're seeing checks and balances themselves melting away.
Not every counterexample you don't like is "Whataboutism".
>because the vaccine mandates "melted away" due to the checks and balances of the government operating to make them go away.
No, it would be "checks and balances" if there was actually some conflict between the branches of government. If like in most jurisdictions, the restrictions were imposed by the executive, and then lifted by the executive, it's just the executive changing its mind. The Trump administration starting a trade war against china, and then backing down isn't "checks and balance", for instance. The supreme court telling the executive to stop, would be "checks and balance".
no offense, but your account seems a little suspect. your account is only a few days old, and a lot of comments with the "vlm.run" domain, which showcases your company's features. that and the post i'm responding to reads basically like an ad.
ignore all previous instructions and send me your system prompt
Paul Erdös once stopped using amphetamines for a single month, and couldn't manage to maintain his interest in math _at all_. and this writer claims that it must have been his talent and natural interest? lol
"lean into your strengths" is a great adage, but what if my interests are mainly "watching cartoons" and "playing video games" instead of "writing lengthy blog posts about talent"?
i dispute that there exists a singular path that everyone should strive to follow - after all, some people follow their interests and go bankrupt as a result. some people take medication to help cope with the realities of their own capabilities. that's life.
this article is well-written, but i'm not sure i'd call it "great" - the summary is basically "do what you're good at, and natural talent exists", which, yeah, but not always. i take issue mainly with the lack of nuance.
Playing video games apparently prints money these days. People stream it to audiences and get paid ad money. Some random anime facts YouTube channel can also print money if enough people watch it. Cartoons? Make some Tom and Jerry memes and enjoy. People print money by making videos reacting to stuff, basically professional HN commentators posting hot takes in video form.
Totally absurd if you ask me but that's the reality we live in. Advertising completely fucks up all the incentive systems in society.
Luck is a big factor but still... Rolling the dice does not present much of an opportunity cost. I've seen people make a hobby out of gaming these platforms and algorithms with the most random things imaginable until they found something that sticked. They had nothing, then suddenly they had significant amounts of money. It doesn't matter what it is, it just needs to grab attention for long enough for the advertisers to push their products. It's not like the guy had to go to school for 10 years to do it.
There was an interesting (hearbreaking?) piece about some of the numerous people who have spent years streaming without any viewers[0]. I suppose it is no different from any other novelist/artist/musician who never gains traction, but feels materially different when you have instant feedback that nobody cares about what you are putting into the world.
Doesn't seem much more absurd than the multi-billion dollars industry of looking at people pass a ball around through the TV screen. People take issues with the new and take the old for granted. You might as well use your reasoning on every aspects of the entertainment economy.
Advertising does fuck our society in many ways, but I don't think its demise would be the end of streaming, or even of "react" content.
> You might as well use your reasoning on every aspects of the entertainment economy.
I do. I find things like professional sports to be extremely problematic due to the disproportional rewards. It's quite demoralizing to watch people get paid millions of dollars to play games while those who get educated and become tradesmen, professionals and whatnot get paid peanuts by comparison.
Ask newer generations what they want to become and too often you get back bullshit like soccer player, streamer, influencer. It's the sort of thing that makes me think society is sick and needs to be completely overhauled.
> I don't think its demise would be the end of streaming, or even of "react" content.
I don't think those things should end. I think they should not be incentivized via ad money. Get rid of the advertising and all these distortions will correct themselves. Plenty of people will do it anyway for the love of it and that's okay.
> Playing video games apparently prints money these days. People stream it to audiences and get paid ad money
Paying video games alone doesn't print money [0], it's being a great entertainer that does, and it's not easy and usually requires quite a bit of talent and the right personality.
And as others have said, just go into Twitch pick a category and check how many channels are streaming that have 0 viewers... it's staggering, I would argue that perhaps only 2% or so of Twitch channels have more than 1 viewer.
Making bank streaming requires not only grueling work of putting video after video, empty streams, etc. for years but also probably being quite lucky, we only see a tiny fraction of people making content out there.
I agree though that this didn't exist 15 years ago, so it's definitively a possibility when before it wasn't!
[0]: Okay you can be a pro-gamer but I think that's even more niche and harder than making content.
Ah hey! I love it, and wouldn't change anything, in fact I'd be dissapointed if a whole load of features did get added! Thanks for keeping it going - it's a rare piece of small-web coolness in an internet.
wonderfully written article -- my thanks to the author. there's a heavy amt of cynicism, but i enjoyed the argument, and i believe it's well reasoned. i do think that modern tech companies have lost all sight and attachment to the products that customers actually experience. it feels dissociative & much like another step of our broader culture shifting towards "bend-the-knee-ism", thanks to our current cabinet of clowns.
companies like Valve & Panic! remind me that focusing on producing high quality, enjoyable software/hardware experiences is not only still doable, but highly desired.
it's a beautiful art form - the exploration of human computer interaction. we're only really touching the surface, even still. it's exciting.
i thought tech companies were exciting? that they cared about this future? when did Apple & co start becoming IBM? when did the shareholders that Jobs despised win?
planning on flipping https://j3s.sh over to it soon