No, because the issue is navigating legal agreements with your vendors, not the ease of deploying. With AWS Marketplace you are making a new legal agreement with the vendor of the product, not with Amazon.
> But I won't follow you as far as "expecting developers to have expertise in how and where their software runs is unreasonable".
In my professional experience, management tends to treat this as a dichotomy rather than a range. And its lead to situations where we're choosing to route every decision through a small number of 'experts' and expecting that to scale, or have people make very bad decisions about things like indices, data layout, etc.
I align with you on this. If you are using some sort of system, there is a sort-of minimum bar of expertise one should acquire. When using analytical platforms, having a basic understanding of how data is stored and the levers it gives you to optimize data processing should be standard.
Some examples:
MySQL - I would expect developers to be able to come up with a list of columns to index and why, understand why UUIDv4 is a bad choice for primary keys, know how to write performant SQL by hand and understand when their ORM is going to do the wrong thing, etc.
Redshift - Understand dist and sort keys and have some mental models on how data shuffles and is pruned based on each choice.
Iceberg + Spark/Trino - Understand that data tends to be stored/loaded as batch files. Understand how deletes work. Understand partitions and sort as options. Understand generally how bloom filters work and when you should decide to place a bloom filter. Understand why UUIDv7 is better than UUIDv4 when its part of a partition key.
This is where I’m at. I find AI to be useful but there’s a tendency for it to suck the oxygen out of the room. Other things still matter.
Edit: I'll also note that while its nice to see real novel AI news I find a lot of it is extremely marginal value. I mentally associate most discussion with debating what color of marker to use when marking the edge of your CDs to improve the sound.
> React can render at more than 60fps if necessary, and if the code isn't shit.
Sorry, that doesn't really come across as a ringing endorsement. React apps are typically not doing anything complex so rendering at less than 60fps should only happen if you're doing very computationally intensive things or writing bottom-quartile quality code.
Yeah, obviously your average React app has no need to render anywhere near that frequently. The ones I've worked on will generally only render as a response to user action, server event or polling.
I'm just saying they can render that fast. So if your app takes a second (or several) to render it's obviously not because react is slow, it's because the code is ass.
In most cases if a React app takes a long time to render a page it isn't really rendering that's taking time, it's a slow network call or multiple. So the app being slow has nothing to do with React at all, it's the backend code that's slow or it's the frontend code doing multiple consecutive requests or something like that.
All I'm saying is react is not the reason it's slow.
> A badly made native app is worse than a well built react-native app. There are plenty of anti-patterns that will ruin your native app.
Sure, but thats not an interesting observation unless there is something about react-native that makes those apps consistently higher quality than native apps.
On the other hand, like-for-like C/C++/Rust performs better than javascript in terms of CPU and memory use, so the null hypothesis is that re-implementing would , in fact, improve things. There's little reason to think a priori that the end result would somehow be worse.
> I don't understand why a police cruiser can sit in a public space (or even a private one) and write down licence plates and descriptions of passers-by with pen and paper, or record everything around them with dashcams and bodycams for later use, but when it comes to cameras on a pole this would require a warrant.
Scale actually matters. Things that are generally OK at a small scale become problematic at larger scales. A single police cruiser writing down license plates isn't able to track you in the same way a huge surveillance network is, and the opportunities for abuse are much lower.
> Scale actually matters. Things that are generally OK at a small scale become problematic at larger scales. A single police cruiser writing down license plates isn't able to track you in the same way a huge surveillance network is, and the opportunities for abuse are much lower.
Note that they already don't write down license plates. They use LPRs attached to their cars, and tend to be automatically running every plate immediately through the database, which tends to catch people without insurance or expired registration.
Is the line automation? Or 24/7 operation? (genuinely)
Been thinking about this today and I agree. If the actionability is reliant on a human in the loop making a decision to create some incident (e.g. traffic stop) then there is a limit to the surveillance while still scaling the law to apply to our growing population.
In that vein, I wouldn't be 100% against Flock cameras if their purpose was limited.
The problem in the concept is that they store all license plate hits for later retrieval regardless of if anything is actively tying that license plate to some form of illegal activity. Even with a short retention window, this is mass surveillance because it allows later viewing of hits based on any kind of search given. And you can never limit that search query in a 100% effective manner, but you could very well tell police departments "tough luck, we only have hits if the plate was involved in something at the time of scan".
But then scale doesn't actually matter. If you have a problem with something that "everyone does" then you should have a problem with something only a few people do.
There hasn't been an explicit ruling from the supreme Court that applies here, but there are indications that they would have ruled that the fourth amendment applied. They've explicitly avoided closing the door on that possibility.
> Also left unanswered was the broader question surrounding the privacy implications of a warrantless use of GPS data without a physical intrusion – as might occur, for example, with the electronic collection of GPS data from wireless service providers or factory-installed vehicle tracking and navigation services.[27] The Court left these matters to be decided in some future case, saying, "It may be that achieving the same result through electronic means, without an accompanying trespass, is an unconstitutional invasion of privacy, but the present case does not require us to answer that question."
> Ultimately, in Carpenter the court determined that the third-party doctrine could not be extended to historical cell site location information (CSLI). Instead, the Court compared "detailed, encyclopedic, and effortlessly compiled" CSLI records to the GPS information at issue in United States v. Jones, recognizing that both forms of data accord the government the ability to track individuals' past movements.[24] Furthermore, the Court noted that CSLI could pose even greater privacy risks than GPS data, as the prevalence of cellphones could accord the government "near perfect surveillance" of an individual's movements. Accordingly, the Court ruled that, under the Fourth Amendment, the government must obtain a search warrant in order to access historical CSLI records.[1]
Of course there are no guarantees on how they'd rule today, but in the past they've ruled that scale and ease of access and compilation are significant enough to mean that fourth amendment protections should apply.
It's not about many people vs. few people. It's about a single person doing little with a lot of effort vs. a single person doing an enormous amount with minimal effort.
Because the outcome is meaningfully different when every citizen can be passively tracked at all times versus a few cops sitting around writing down license plates.
The former is massively damaging to our rights, the latter is almost useless and thus barely a problem worth talking about.
I didn't have a problem because the damage that can be caused by doing a little is very limited, and if someone puts too much effort it becomes very visible and they're put under extra scrutiny, preventing them from doing more harm. Also because outlawing standing on a street and taking notes is a very slippery slope.
that is not the case. It is not about effort: The court has ruled that single point of time monitoring is fine, but continuous monitoring (a covertly placed gps tracker, e.g.) is a violation of privacy.
Needing more people directly translates to needing more energy and money and to having less complete picture. So reducing it to the simple "how much a single person can do" is very useful I'd say.
By my understanding, this would be similar to the situation with marketing calls. In the US is it legal for a human to call anyone and market to them. But due to issues from scale, it is illegal for either a program to make the call, or a prerecorded message to be played.
To me, that is precedent in US law that computerized methods of executing human tasks do not fall under the same rules.
I don't believe so. The whole purpose of a license plate is to deanonymize vehicles and their drivers. Vehicles must display license plates while driving or parked on public roadways.
This is not the same as faces. Faces are not like license plates. People are not given faces by a state licensing authority and then required to display them.
The scale of facial surveillance is a problem only because the surveillance of a single face is already a problem.
People who don't like license plate tracking should take a bicycle or public transit.
When you are given the privilege of riding a two ton murder weapon on a public road, you trade privacy for that.
It's obviously different in that Chatrie was about google location data, and there actually was a warrant. SCOTUS ruled that warrant was too broad since it captured everyone in that location, not a specific person(s).
Constitutional scholars have pointed out the ALPR are next in line to face challenges based on the same principles as Chatrie - LEO can get a warrant for a specific person(s) to search but cannot search the entire universe.
That would not prevent ALPRs from existing, but it would require a warrant to search them. Which seems reasonable and completely in line with the 4th amendment.
OK sure, but do they now also need a warrant to look at their own footage from their body cam, or their dashcam? What about the camera on the side of the municipal building?
Compelling parties to give up data is different than looking at your own data. And it is firmly established that recording and reviewing footage in public space is totally fine.
So we are back to square one...How do we define scale such that a police camera on a police car is OK, but a police camera on every corner is not. Or that individuals can record in public but police cannot.
SCOTUS gets into it in their opinions on Chartrie, you should check it out.
They wrestle with the balance there of 4th amendment protections and mass surveillance. It's evolving and generally courts are moving to not allowing mass surveillance -or less so than they have allowed in. Courts have been split on it, and someone already pointed it out that SCOTUS will eventually come up with tests for it.
There were some really interesting parts of those Chatrie opinions that really will test the idea that someone in public gives up their right to any privacy - the fundamental underpinning of why governments can currently point cameras at public places and dont need a warrant to search that video.
Chatrie protected the aggregated viewpoint versus any single public moment - And SCOTUS said the aggregated viewpoint (mass search of location data with no specific person(s)) is not allowed under the 4th amendment. A very interesting precedent.
Gorsuch had an interesting take in that he argued that in Chatrie the location data was Chatrie's personal property and therefore protected from illegal searches from government.
There are a lot of similarities to how ALPRs work, and key differences. Mostly Chatrie deals with a private firms data collection of locations - google vs ALPRs are generally on publicly owned property and recoding things in public.
Still, many justices opened the door to the idea that the government can't mass surveil citizens. We'll have to wait and see how it all plays out.
I am very much of the opinion the government should need a warrant to search through ALPR data (and video for that matter). So, I am hopeful that the tenets of Chatrie extend to ALPRs, video, etc.
Edit: Gorsuch argues that the data is personal and not the government's data - this is what prevents the government from searching for it without a warrant. He did not get into why a body cam or something like that is different.
Do yourself a favor and submit a FOIA request for some body camera of something you're trying to understand. If the footage is too long, they'll deny your request for being unduly burdensome because of the need to review/redact those records, even in public.
It took 8mo for Chicago Police to review and redact bodycamera footage from a protest in Chicago. It was 95% blurred. Almost hilariously, one of the only parts that wasn't blurred was a moment where one cop commented on another cop's mustache.
So from the FOIA perspective, the public isn't able to get gobs and gobs of footage just because it's public.
Sometimes the law is vibes, SCOTUS will probably create a set of tests where lower courts will have to look at how many points it hits and decide whether it's over the line on a case by case basis.
You don’t need to rigidly define scale to demonstrate at a common sense level that there is a major difference between a couple cops writing down your license plate and a nationwide network of them entering your plate into a searchable ledger.
Given that in the majority of places police are typically busy doing things like writing speeding tickets and going on calls that leaves a lot less time for them to sit still for an 8 hour shift and record every single plate they can see 24/7/365. Those 5 cameras would represent 15 to 20 officers which is a huge taskforce no matter how large the department.
The number of cameras matters much less than what those cameras are being used for. If police body cam footage was being indiscriminately scanned for any and all license plate data and all of that data was being fed into a national database which was being routinely accessed and abused that would also be problem. As far as I know, that isn't the case.
The point at which we decide it's dangerous to the safety, freedom, and happiness of citizens. Every law we have is a balance between "provides good" and "imposes bad" - and we have to adjust where that balance lies on an ongoing basis as things (technology) change.
But nobody wants their minds connected and all of them to immediately be able to see, comprehend, and act on anything any of them sees. This is what a surveillance dragnet backed by AI is.
The only people who want that are the ones that want to subjugate large portions of the population, or those that want those large portions subjugated even if they don't get to do it.
If we simply mandated that everyone install an app on their phone so that the government can track them scales very well. The people without the app would probably be criminals,
I don't follow that... posting a police officer in everyone's house is pretty much a Third and Fourth Amendment violation and not primarily constrained by finances.
Youre the one that made the intellectually dishonest argument that scale is about cost and efficiency while the discussion was about scale of the surveillance.
from your point, nothing is cheaper and more efficient then mandating an app on all phones through which the governments can track everyone, and consequently whoever doesnt have it would be a criminal. it completely tracks from your argument. You may need to make the app stop tracking while theyre in the home, but thats a technicality. voila. no longer an issue according to your point
No it doesn’t track. In his example, observing license plates is perfectly legal without a warrant. Looking in your phone is not. These are not at all alike.
Here's my rebuttal: I don't want police sitting down and writing every license plate down 24/7 on every street corner tracking my every movement. That's bad. And they don't do that now, so saying "it happens today" is false.
It isn't possible for them to do that without sensors and AI everywhere. That's why using this technology is a red line.
Of course? We already have private entities doing this today. Private toll road operators such as United Bridge Partners already administers camera systems that collect license plates for toll payments. I would assume parking lot operators also do this, and private citizens could create a nonprofit network too if they wanted.
Are you suggesting that people couldn’t do that for some reason?
The challenge here is to ensure that any controls that apply to law enforcement operating their own hardware for this purpose would also apply to them obtaining this information from a third party. We don't need the mayor setting up "Totally Not The Cops, LLC" that collects this data and "sells" it to his police department to bypass warrant requirements.
People either have privacy in public or they do not, but I don't know anyone who seriously claims that there's an expectation of privacy on public roads. What freedom are you talking about?
> People either have privacy in public or they do not
Thankfully, the law isn't as overly simplistic and black and white as that. It'd be very stupid if it were. We have the ability maintain a general principal like "there's no expectation to privacy in public" and still prohibit certain things that are deemed harmful, like taking upskirt photos of people while they are in public spaces.
> Nobody is talking about cops or flock cameras taking upskirt photos.
Just as we have the ability to pass laws prohibiting that specific harmful use of cameras while in public spaces, we have the ability to do something about the harmful uses of police and flock cameras while in public. The conversation is about the harms of mass surveillance and the exceptions necessary to protect the American public.
>I don't know anyone who seriously claims that there's an expectation of privacy on public roads
Hello, I'm here.
Practically, we do have privacy in public pre-AI. We weren't entitled to complete anonymity, but the ability to track every single person at all times was impossible.
Now that AI and sensors are making it possible, people are saying "No, I don't want that to be the norm".
You should go read the United States v Jones concurrence and try to figure out why five Justices disagree with you. Have some curiosity about the world!
You can also read Carpenter v United States for some background thoughts on why it's not as simple as Justice Fluidcruft condescendingly imagines it to be.
In fact if you're really curious, you could even look at the state cases that are bubbling their way up like Commonwealth v McCarthy.
If only everyone were as confident and simplistic as you, we could avoid all these darn cases :(
idk how installing a GPS tracker is even remotely comparable to taking photographs of publicly visible license plates in plain view in public but please do go on refusing to engage facts.
People don't have any expectation that the license plate on their car cannot be seen by law enforcement.
People do not expect their phones to be feeding their locations to cops and they do not expect things hidden on them unknowingly to exist. But it really stretches believability that people have no idea that cops check license plate. Or that cops communicate with each other about license plates to coordinate searches or that cops have databases of license plates. Heck there's a whole trope in media that you need to swap license plates to hide from cops.
Next you're going to be telling us we need a warrant for an Amber alert.
I’d like it if you provided relevant quotes from these opinions that support your assertions. It’s not very persuasive otherwise. These cases are highly nuanced.
> I’d like it if you provided relevant quotes from these opinions that support your assertions... These cases are highly nuanced.
That nuance is why it's worth reading the entire opinion.
Anyway here you go:
> GPS monitoring generates a precise, comprehensive record of a person’s public movements that reflects a wealth of detail about her familial, political, professional, religious, and sexual associations. See, e.g., People v. Weaver, 12 N.Y.3d 433, 441–442, 909 N.E.2d 1195, 1199 (2009) (“Disclosed in [GPS] data . . . will be trips the indisputably private nature of which takes little imagination to conjure: trips to the psychiatrist, the plastic surgeon, the abortion clinic, the AIDS treatment center, the strip club, the criminal defense attorney, the by-the-hour motel, the union meeting, the mosque, synagogue or church, the gay bar and on and on”). The Government can store such records and efficiently mine them for information years into the future. Pineda-Moreno, 617 F. 3d, at 1124 (opinion of Kozinski, C. J.). And because GPS monitoring is cheap in comparison to conventional surveillance techniques and, by design, proceeds surreptitiously, it evades the ordinary checks that constrain abusive law enforcement practices: “limited police resources and community hostility.” Illinois v. Lidster, 540 U.S. 419, 426 (2004).
> Awareness that the Government may be watching chills associational and expressive freedoms. And the Government’s unrestrained power to assemble data that reveal private aspects of identity is susceptible to abuse. The net result is that GPS monitoring—by making available at a relatively low cost such a substantial quantum of intimate information about any person whom the Government, in its unfettered discretion, chooses to track—may “alter the relationship between citizen and government in a way that is inimical to democratic society.” United States v. Cuevas-Perez, 640 F.3d 272, 285 (CA7 2011) (Flaum, J., concurring).
> I would take these attributes of GPS monitoring into account when considering the existence of a reasonable societal expectation of privacy in the sum of one’s public movements. I would ask whether people reasonably expect that their movements will be recorded and aggregated in a manner that enables the Government to ascertain, more or less at will, their political and religious beliefs, sexual habits, and so on. I do not regard as dispositive the fact that the Government might obtain the fruits of GPS monitoring through lawful conventional surveillance techniques. See Kyllo, 533 U. S., at 35, n. 2; ante, at 11 (leaving open the possibility that duplicating traditional surveillance “through electronic means, without an accompanying trespass, is an unconstitutional invasion of privacy”). I would also consider the appropriateness of entrusting to the Executive, in the absence of any oversight from a coordinate branch, a tool so amenable to misuse, especially in light of the Fourth Amendment’s goal to curb arbitrary exercises of police power to and prevent “a too permeating police surveillance,” United States v. Di Re, 332 U.S. 581, 595 (1948).[1]*
=======
Now the question is: is this characterization true only of GPS monitoring? Or is it also true of a sufficiently broad network for ALPRs? The answer of course is the latter.
Here, Sotomayor + 4 other Justices are saying it's actually not as simple as Fluidcruft's "you're in public so plain view doctrine answers it"
Remember, though, that a concurrence is not controlling precedent.
The actual opinion in U.S. v. Jones hinges on the fact that "The Government physically occupied private property for the purpose of obtaining information," and also stated "[t]his Court has to date not deviated from the understanding that mere visual observation does not constitute a search."
There were two concurring opinions in that case: One was signed by Justices Alito, Ginsburg, Breyer, and Kagan. The other--the one whose quotes you cite--was Justice Sotomayor's alone.
There can't be a "concurring opinion" signed by five Justices. An opinion signed by five Justices is a majority, and would constitute the Opinion of the Court.
I think that the most we can say right now is that although the "plain view" doctrine has been weakened somewhat over the last couple decades, it's not completely dead yet.
The fact is that there hasn't been a case on point to answer the questions as to whether 1/ALPRs constitute a "search" under the Fourth Amendment (a threshold question that must be answered "yes" to proceed further), 2/whether use of the technology can be "reasonable" under any circumstances, and 3/whether there are any exceptions to the warrant requirement. We have a long way to go before we know what the law is on this technology. Remember, too, that the Fourth Amendment has only a "reasonableness" and a warrant requirement. It does not ban technologies for surveillance outright. (See, e.g., Kyllo v. U.S., where the use of thermal surveillance technology wasn't the issue in and of itself, but rather constituted a search necessitating a warrant.)
So how about you tone down the attitude a bit? You're arguing from shaky ground, and there are more constructive ways to contribute to this debate.
Yes you're right. The "signed by five" was a misphrasing on my part.
What I meant is that both the concurring opinions (signed by 5 collectively) agreed on this same point:
> society’s expectation has been that law enforcement agents and others would not—and indeed, in the main, simply could not—secretly monitor and catalogue every single movement of an individual’s car for a very long period. In this case, for four weeks, law enforcement agents tracked every movement that respondent made in the vehicle he was driving. We need not identify with precision the point at which the tracking of this vehicle became a search, for the line was surely crossed before the 4-week mark.
So no, I'm not arguing from shaky ground. And no, ample attitude is warranted against people not just advocating for unbounded state surveillance, but insisting that it is already a foregone conclusion.
> What I meant is that both the concurring opinions (signed by 5 collectively) agreed on this same point: [...]
I agree with you that the two concurrences are aligned about this. Interestingly, it did not make it into the majority opinion, even though there was a majority on this point. It probably wasn't included because it wasn't needed in order to reverse the lower court's decision; the physical violation was enough to merit a unanimous decision.
> ample attitude is warranted against people not just advocating for unbounded state surveillance, but insisting that it is already a foregone conclusion.
No, it isn't. It is possible, and better, to disagree with someone while remaining courteous--even on subjects you have strong feelings about.
From our Guidelines:
> Be kind. Don't be snarky. Converse curiously; don't cross-examine. Edit out swipes.
It's actually far more dangerous to presume we are already protected (your position) particularly based on reading tea leaves in flimsy ancillary opinions when the Court has shown itself fully happy to blow precedent out of the water, particularly with regards to privacy rights which are grounded in interpretation of the Constitution and not actual text (see Dobbs).
Obviously not true given that a wannabe dictator would literally say what you are saying ("it's just plain view doctrine! Always been this way! 4th Amendment doesn't prevent me from flying a drone over you to and from your home every day!")
A wannabe dictator, inversely, would not say nor be happy with what I am saying: You have a right to exist as a private individual, even if going out in public would leak vast amounts of information to a drone hovering overhead.
Sure, I'd love to have more policemen around, but I don't want large automated blanket surveillance network tracking the positions of everyone everywhere. There is a huge difference.
I don't, 95% of what US cops do is harrass and extort people despite most of these people causing nobody else any concern or even a second glance or thought.
Yes, this is why a warrant is required to access carrier location data for a phone but one is not required to have a police officer physically following you around town despite the outcome (the police get a trace of your location) being the same. Cell phone location data is too easy and cheap for the government to scale and so courts have found that it demands tighter protection.
It comes up fairly often here since the HN audience tends to be mostly programmers/computer people, but it's really important to remember that the law is not a series of rules a computer can directly evaluate to determine whether something is or is not legal; judges frequently use their judgement to balance the assumed intent of laws and the competing interests of various parties. I don't say this to you specifically but really just the HN audience as a whole :)
"Yes, this is why" is in reference to scale here? I don't see how it's "why". Cell phone tracking follows you where a police officer physically cannot follow you (on private property, for example). This is why, not because of scale.
Scale does not matter. Here’s a quote from Chatrie:
> Still more fundamentally, we have never understood Fourth Amendment protections as kicking in only once an intrusion “goes too far.” Pennsylvania Coal Co. v. Mahon, 260 U. S. 393, 415 (1922) (adopting that approach for regulatory takings). Where the Fourth Amendment applies, it applies—regardless of “the quality or quantity of information” the government obtains. Kyllo, 533 U. S., at 37.
It's not about scale, it's about oversight. A warrant puts a lock on the data until the request is approved. Leaving it wide open is the reason we have constant news stories about police being arrested for stalking people. Nobody wants to limit the police's ability to investigate crimes, but we do want them to use their power to investigate crimes, not to stalk their baby mama, crush, rival, etc. I fully understand and accept your point that there is no real difference between a cop watching and the cop setting (or hiring) a camera to watch, and I even believe that the more data available the more humanity benefits, and that a total absence of privacy as we know it is inevitable, I think we are not there yet culturally, so strictly requiring warrants for police to access these systems is desirable.
> Nobody wants to limit the police's ability to investigate crimes.
This is not true of privacy advocates, they do want to limit the ability to fish for circumstantial evidence that targets innocent people and causes suspicion.
I mean it's just not true at all. We don't want police to torture people for information. We don't want police to detain people for questioning, even briefly, without reasonable suspicion. We don't want police to search people or their property without a warrant. And so on.
This is non-sequitur. "Investigating crimes" entails knowledge that a crime has actually occurred, and deliberately seeking information specifically related to that crime; "fishing for circumstantial evidence" does not.
Being located in an area of a crime is circumstantial evidence. Any info on what drivers license plates drove by a camera is absolutely circumstantial evidence to a crime in the area that is known to have occurred, almost definitionally.
Well you've insulted me like 50 times, so I think you're the one who should chill, my good sir. All I did was ask a couple of sincere questions and I'm met with a barrage of insults, yet no answers. Interesting. Also, I honestly have not heard or read anything Donald Trump has said in years.
Yes but the key here is doing proper risk assessment. "What is the consequence if the LLM gets this wrong?" "How do we verify the output?" "What are the legal ramifications for using the LLM in this way?" "Who is responsible when the LLM fails?" "Whats the expected accuracy here?" etc.
In the current AI mania, there's a lot of due diligence simply being ignored. Plenty of "Well humans make mistakes too!" going on here on HN too.
The due diligence not being done is people putting cats in microwaves. It‘s not the dangerous part. The real danger is risk assessments coming to wrong conclusions, because it is still terra incognita. Talented engineers were in this situation before, doing mistakes with cars, airplanes, buildings etc.
No, I'm sorry but I think thats a cop out. The fact that LLM are stochastic and can give incorrect answers is not particularly difficult to comprehend, and the risks that fall out of that are reasonably understandable. The issue is entirely down to bad choices by the people driving LLMs, because they are engaging with what they wish LLMs do instead of what they actually do.
So you say that it is easy to understand what LLMs are, yet the issue is that people don‘t understand it. Either you put too many people in the bucket labelled „idiots“ or you are blowing that problem out of proportion, at the same time seeing only the tip of the iceberg.
Risks are not binary. Most people are not idiots. They do understand that LLMs aren‘t thinking humanly or 100% logically. You are talking about some sort of faith, I think this is more about trust, which is built on observations. Statistically, the outcomes of LLM work may hit the goal quite a lot. More of that in the beginning of the journey than in the end. For many reasonable people that becomes a trap, where minor, acceptable deviations accumulate into a fireball. Still, saying that LLMs are failure with inevitable fireball at the end and giving up is a trivial and stupid solution. Staying within razor-thin distance from that fireball and managing to deliver a working solution is what everyone tries at the moment. This is how it always happened and always worked. It‘s the art of engineering.
reply