I am also in favor of catching violent criminals. And as such I join you in the downvote fest. Of course, no one puts forth a cogent argument about why catching violent criminals with this technology is bad. Can it be abused? Yes. Was it abused here? No. Luckily, we live in a democracy where we make the rules. If we want a policy that law enforcement can't buy these systems, make that the law. If we want a policy where law enforcement can't use these systems in any capacity, then make that the law.
Funny how when one of us finds a workaround it's regarded as a hack and clever and a sign of how talented and 'relentlessly resourceful' we are. When a cop does it, oh god no the sky is falling. Asking a neighboring dept for an assist is within the law and from this own reporting not abused at all. It was used "at least 5 times" (ok, that is lazy, what is the upper bound?) and "no matches were returned". It's not some crazy surveillance state. It's cops trying to catch criminals. Which we need
> Funny how when one of us finds a workaround it's regarded as a hack and clever and a sign of how talented and 'relentlessly resourceful' we are.
Indeed. Which is why when somebody gets pulled over for speeding and points out that because of the theory of relativity speed limits are meaningless without a frame of reference, but the law doesn't explicitly specify one, and laws are required to be interpreted most favorably to the accused in a criminal case, the police always let them go instead of giving them a ticket.
> It's cops trying to catch criminals.
Corporations dumping mercury in the river are "businesses trying to create jobs". Muggers are "disadvantaged youths trying to put food on the table". They're supposed to do the second thing but not in that way.
Your first bit isn't an example of a legal workaround, it's an example of attempted sophistry for illegal behavior. Are you implying that the cops in this case broke the law? I don't think that is clear.
Cops should be able to do what is within the law, and if we don't like that then we should change the laws.
> Your first bit isn't an example of a legal workaround, it's an example of attempted sophistry for illegal behavior.
This is exactly what the police are doing.
> Are you implying that the cops in this case broke the law?
Are you implying that it's illegal to drive with the flow of traffic? That seems like an unreasonable interpretation, it would imply that everybody is constantly breaking the law! The interpretation where the speed limit is relative to other traffic makes much more sense, because the main issue is speed differences. If everybody is driving 70 that is much safer than some people driving 25 and some people driving 115, so the law sets a maximum speed difference, e.g. if the speed limit is 55 and some people are driving 40 then you could drive up to 95 but 100 is too fast. This is clearly much more in line with observed behavior.
Anybody can come up with reasoning to justify whatever. You can say that we should follow the letter of the law and let people get away with it if they can find a loophole, or that we should have judges reinterpret what the law actually says to try to get the result the legislature intended, but don't try to say we should do different ones for different people.
I'll also add that I think it's a bad law if everyone breaks it, yet police look the other way. The beauty of a democratic republic is our elected leaders can change laws and pass new ones! So, just like this case where the people have a choice in what cops can do with surveillance tech, the people also have a choice in what our highways are limited to. The ideas are all really quite simple, and you don't have to go reaching for arguments about Einstein's theories to engage with them.
> The beauty of a democratic republic is our elected leaders can change laws and pass new ones! So, just like this case where the people have a choice in what cops can do with surveillance tech, the people also have a choice in what our highways are limited to.
That isn't actually how it works.
There are in practice multiple constituencies. Some people want higher speed limits, because they prefer to waste less of their time in a car. Other people want lower speed limits, because they believe them to be safer. Politicians have learned how to lie to people in the most adaptive way: You lower the speed limit but don't vigorously enforce it. Then you can tell the people who like lower speed limits that they've won, while the people who want higher speed limits don't actually slow down.
This deceitful compromise is also supported by the police, because when everybody is constantly violating the law they always have a pretext to pull over anybody they want. So the status quo is sticky. But it's also unreasonable. It's just not something that democracy is well-suited to fixing.
Thanks for all the downvotes with no cogent argument against my points. I will explain it very simply why you are engaging in nothing more than sophistry:
- My original post puts forth an interesting double standard, where when a non-cop finds a workaround to achieve their aim it's considered a 'hack' and a sign of being 'relentlessly resourceful' (a fundable, valued attribute!). When a cop uses a workaround to achieve their aim, which is to find criminals, it's considered grounds for outrage.
- Rather than engage on this point, you put forth analogies that don't address the hypocrisy, but rather use fine sounding language (relativity!) to make a completely different point that all laws should be treated in the spirit of how they were written, and not by what they actually say. And this takes some thinking on the reader's part to get there, as it's diverged significantly from the thrust of the first post, which, again, was to point out an interesting double standard.
- I try to get clarification on your now changed point
- You double down on the speed analogy, even further removed, with whataboutism and the logic that 'everyone is breaking the technical law so we need wide latitude in how laws are interpreted'. It was so far from my original observation that I don't even know why I responded. Perhaps you were trolling, IDK.
It's also interesting to me that my initial post got a couple quick upvotes and then got buried 5 min later. I don't claim foul play or anything. Just interesting. Kinda reminds me of the post against applying to YC that rocketed to the front page and then got flagged for removal. It's quite amazing how hiveminded this place is. And you know, I really have to question the value that I've gotten from this site in 15 years of reading it. I am positive that my life would be richer without it. So so long. This last part has nothing to do with you, just doing some reflecting. I'll see myself out.
> My original post puts forth an interesting double standard, where when a non-cop finds a workaround to achieve their aim it's considered a 'hack' and a sign of being 'relentlessly resourceful' (a fundable, valued attribute!). When a cop uses a workaround to achieve their aim, which is to find criminals, it's considered grounds for outrage.
That isn't a double standard. People are constantly complaining about "loopholes" when someone other than the police have found one, and law enforcement are constantly claiming that the one you've found isn't allowed.
The real double standard is not police and everybody else, it's normal people and people with power. Apple doesn't have to pay taxes because they can use some complicated shell games, but you can't do the same thing because it's reserved for international megacorps with fancy lawyers. It would be much better if we were consistent and always allowed "loopholes" to be used by everyone, so that the law had to be fixed instead of ignored or selectively enforced. But the police wouldn't be the beneficiaries of that, because they're in the group with power.
Notice how the police who did this are being berated rather than incarcerated.
The article mentions a guy "charged with aggravated assault for allegedly charging toward someone with a knife". I'd say getting that guy into trial is good for NOT harming people
I would like to see one on all Apple's latest advancements. The hype: "SwiftUI, async/await and actors, and declarative animations make building a robust app easier than ever". The shower: "Imperative programming is really easy for our little brains to handle. UIKit, Core Animation, and GCD are easier to work with and reason about"
> My guess is that none of this can easily be fixed. Systems like DALL-E and ChatGPT are essentially black boxes. GenAI systems don’t give attribution to source materials because at least as constituted now, they can’t.
Is it necessary to fix in the model itself? It seems a gate in the post processing pipeline that checks for copyright infringement could work, provided they can create another model that identifies copyrighted work (solving the problems of AI with more AI :/)
I should maybe preface this by saying that I probably agree that this is the way this will shake out ultimately.
But I also would say multiple odd post processing stuff (obviously completely obscured for security reasons) bolted onto a giant black box model will erode the trust in the results. If a robot was unveiled and the question of "what prevents this robot from using it's superhuman strength from smashing my head in" the answer of "don't worry there is a post processing step in the robots brain whereby if it detects a desire to kill we just cancel that" would be a little disconcerting.
The more satisfying solution is: the model / robot is designed to not be able to produce specific images / to smash human heads in. It just might not really be possible.
Exactly; there is no need to do this in the model, you just need well understood token retrieval methods for identifying copyright infringement that ChatGPT's competitors already have.
You will get into some murky definitions of what is exactly required for copyright infringement vs fair use, etc, but we already do this for ContentId for YouTube and text is far simpler.
This is bogus. Now you require that every piece of copywriter be registered and indexed in a central authority?
What if I write a story and publish it on my blog. Should I be required to submit this to openAI's copywrite model to ensure the story is never used in openAIs other models? What about the other 100 AI model companies that are going to spring up in the next year?
It should be on the curators of the training set to ensure all material inside is fair for them to use.
I'm not sure that's actually true, even though we hear it often. The artists I know (about a dozen) are all trying to figure out how to make _more_ money from their art so that they can continue making their art
I'm grateful for the last several days spent w my family, including my 18 month old nephew who made us all absurdly happy. He is the first of the next generation in my family. Just a wonderful time
Learning the contract is great, thank you for the work! How about the infra stack used by imessages? Does anyone have intel on that? The scale is incredible, which always makes me wonder how it can be so good while other apple web services (forums, dev portals, etc) can be so buggy and half baked
The actual mind-blowing scale is that Apple's push notification service isn't just carrying iMessages. It's also carrying push notifications for every third-party messaging app.
And the non-messaging apps with notifications too.
And the silent internal notifications. You added a meeting to your calendar on your Mac? Push notification to your iPhone to tell it that the iCloud data changed and it needs to update. Changed a file on iCloud Drive? Push notification to sync your other devices. Got a phone call, and it starts ringing on your Mac too via Continuity? Push notification (encrypted like an iMessage).
Just how many messages are going through that service every second?!
Centralized notifications is not a "genius solution", it is the "only possible solution" for power-constrained devices, if you think about it. Same thing applies to Android: in ideal world it keeps a single connection open to GCM servers to listen for notifications for ALL apps on the device, and then routes messages to the appropriate applications they are intended for.
I never said centralized notifications is conceptually a genius solution. I said actually building a system that can handle so many billion notifications is impressive. I'm curious how it works internally...
I'll bite on "just a fad in webdev circles", in case you are serious and not trolling. Railsworld is celebrating the 20th anniversary of Rails this year. As I'm sure you know, Rails is built on Ruby. 20 years does not fit my definition of fad.
I see what you mean, but I counter with Plone. It's been around for 24 years, and the annual Plone Conference is coming up in a couple months. Still, you don't see a whole lot of new Plone sites rolling out these days, and there's no Automattic-scale company with click-here-to-deploy convenience.
Nothing against Plone (although I was very happy to put its foundation layer, Zope, in my rear view mirror). It's a fine program that's very good at what it does. But just because something's been around a while doesn't mean that it's still vibrant and growing.
Just because there are people who still use it, doesn't make the statement that RoR was a fad less true. RoR was "the thing that everyone was using" at one point. It was that for several years. Anyone who's anyone was building stuff with Rails.
Now it's probably the third option, when building websites... and a fourth option, when building APIs.
> It's still one of the most popular languages for web development.
But is it? Stats on major job boards such as Indeed show a steep decline in recent years. There are consistently twice as many Django/Flask roles listed compared with Rails. Node is the most popular but Spring, ASP.Net and PHP occupy the next level below Node.
My point was more that Ruby exploded because of a period when Rails was growing super fast. I don't know what the growth in RoR space is but I'm certain it's nowhere near what it was in late 2000s/early 2010s. And outside of that there's really not much in Ruby space.
Python was all over the place when RoR came out, meanwhile I only remember hearing about Ruby on random forum posts or by some enthusiast before RoR.
I don't think moonchrome was disparaging Rails or minimizing its popularity, but, rather pointing out that Ruby is only as popular as it is because of webdev, because of Rails (and the Rails-likes that came after)—outside of that specific niche, it's not very commonly used just about anywhere else, especially when compared to Python.
Mh. Honestly, I've encountered ruby in about 3 primary ways. RoR and Sinatra, which are kinda the same webdev thing, and Chef as a config management. There are additional things, but more supporting things - Rake and Capistrano as deployment systems, Berkshelf and such as supporting things.
Outside of this, it's been perl and python, with python replacing perl after that whole perl 6 fiasco while the language and distros started understanding packaging python.
Only one of the items in your "such as" list is unique to python's philosophy, and that is "there should be one way to do it". Ruby embraces the many ways to accomplish something, it's true. However, the three others are not unique to python. Ruby especially embraces "readability counts". And I can't think of anything in the ruby language itself that is implicit over explicit. Perhaps you are thinking of rails and comparing that to python.
For me, the Ruby community's comfort with monkey patching was a big turn off. In Python, you can hack around on class and replace its methods, but if you do, it's expected that your coworkers might stick you in a dunk tank and sell tickets. It's just not the thing that's done.
Monkeypatching is maybe more of a Rails thing than Ruby. I think the biggest problems with Ruby around the time Python began to take off were slowness (MRI was a disaster, slow, and full of memory leaks), plus patchy documentation (lots of things were in Japanese).
Still, I preferred and prefer Ruby. Python has fantastic libraries, but it is a mediocre language. Ruby feels like a simpler version of Perl + Smalltalk, and it is a joy to use. Python has intentionally crippled anonymous functions plus syntactic whitespace, which often leads to long and ugly code.
I think it is a shame Guido hated functional programming and he did not embrace an Algol-like syntax with begin/do end blocks. Those two things could have vastly improved Python. Ruby's block, procedure and lambda design is a stroke of genius that yields beautiful code and makes DSLs trivial.
For me, the Ruby community's comfort with monkey patching was a big turn off
If it helps, the Ruby community really soured on monkeypatching in general quite a while back. You don't see it much these days and every decent book or guide warns against it.
It was definitely a crazier place 10+ years ago in the early Rails days.
I certainly wouldn't blame a person for experiencing those crazier days and thinking the community didn't have the greatest technical vision.
I played around with Rails and Ruby when Rails first blew up. But I didn't start doing it fulltime professionally until 2014. By the time it seemed to me that the community was maturing. I think it's in a good place now.
An easy way to start taking the fight back to the enemy is by overriding __new__ and tinkering with things before/after Django's metaclass __new__ does its own metafoolery.
Also overriding mro() is a fun way to monkey with the inner workings.
> And I can't think of anything in the ruby language itself that is implicit over explicit
This is surprising to me. I love ruby, but it embraces implicitness more than any other language I've used. method_missing, monkey patching internals, and other types of metaprogramming make things very implicit at the interface level
Scala's implicits are a lot less unpleasant because the type system means that you (and, perhaps more importantly, your IDE) know exactly where they're happening.
Python has a default None return, Ruby returns the value of the last expression.
Neither (except maybe the None case in Python) is really implicit, Ruby is just an expression oriented language while Python is statement-oriented.
OTOH, in Ruby the keyword “return” is superfluous except for altering control flow for an early return, while in Python it is the mechanism for supplying a return value.
Doesn’t seem like that big of a deal. You’re going to have a really hard time getting your first test to pass if you make a mistake.
I don’t mean using testing as a poor man’s type system, I mean even the tests you would write in a statically typed language will quickly uncover any mistakes you could possibly make there.
> In statically typed systems, this wouldn't even compile.
Still, it's not like you're going to miss that a function doesn't return something expected. Even in statically typed languages you are going to have tests to validate that much.
> I'm pretty sure I'd see multiple cases where the function would return different types depending on a code path.
That's another problem entirely. Although even then it would be pretty hard for an errant type to not blow up your tests. Again, not tests specifically meant to check that your types or sane, just the same tests you would write in a statically typed language.
I can also vouch that what the author is saying is true. I've started publishing my notes, and I found that after three weeks none of them were indexed (and yes I submit a sitemap through google webmaster tools). I have since started using the URL inspector tool and adding them one by one. That does work.
