I recommend taking a look at this video to get an idea behind the through process (or lack thereof) law enforcement might display when provided with a number of "AI" tools, and even if this one example is closer to traditional face recognition than LLMs, the behavior seems the same. Spoiler: complete submission and deference, and in this specific case to a system that was not even their own.
I can read that "submission and deference" at the casino as conflict avoidance, the arresting officer says to his peers at the station that he "kind of believes" the suspect. He also states at some point that he can't cite (and I infer then release) the suspect because he is not certain who he is, and therefore has to arrest him as a "John Doe" so that his identity can be established. The fact (?) that the suspect now has a police record for this possible farce won't be settled until after the facts are determined in a court of law.
This video demonstrates that when it comes down to it the blunt end of law enforcement is oftentimes a shit show of "seems to work for me" and that goes for facial recognition, shot spotter, contraband dogs, drug & DNA tests, you name it.
An absurd decision with dangerous second order effects, many of which lead to VPNs and other privacy tools being next, just look at UK hyping and building that up right now. I hope they will vote accordingly when they're of age, not forgetting what liberties were taken away from them in the name of very dubious benefits, easily circumvented, and prone to exposing them to greater danger going through unofficial channels. Trying to really address the issues younger generations are facing is clearly too difficult for the geriatric, decrepit ruling class that just won't let go, and this helps them further every government's ambitions of increasingly regulating the means of communication between people. Actually, it's not that it's difficult, they simply don't care.
This is increasingly many cars, starting minimum in the past decade and an half, and not limited to EVs. It's definitely something you need to research before purchasing one so you can dodge the worst offenders. Automotive engineering has been a clown show for years, and greatly suffered from becoming too reliant on digital technology without being willing to invest and spend for robust systems, going for low-cost, low-quality, proprietary parts made in small numbers and unique to each production run. The traditional expectations that you could have options in regards to your vehicle being serviced are on their way out without consumers doing something about it. A future where only the manufacturer and its authorized shops can perform maintenance means they can set any price for it, a price that's already been skyrocketing, and that would effectively allow them to collect far more revenue than previously possible.. and if you can't extract value from customers through heated seats and high-beam subscriptions, maybe you can just have their cars full of black boxes break down more often?
There's no way to determine whether a contributor used LLMs in part or full, not without them being honest about it. With that in mind, this seems like a reasonable position. Been using KeePassXC since forever and will continue to do so. It might feel wrong to some, but these changes are inevitable and it's best to be prepared and become acquainted with that now rather than later.
This was a greatly unpleasant post to read, likewise for all the others from this substack until I could not anymore. Its unrestricted, excessive usage of obvious LLM patterns was so unbearable I wonder how much of it had any human input at all.
As for the topic: software exists on a spectrum, where the importance of safety and stability is not equal from one point to another, does it not? Safe software is probably the safest (and most accessible) it's ever been before, meanwhile the capacity to produce low-effort has increased massively and its results are most obvious outside of a browser's safe space. And CrowdStrike is a terrible example because nobody ever had any love for them and their parasitic existence, even before that accident their track record of disastrous bugs and awful handling of disclosures.
And your operating system's Calculator apps have always been buggy pieces of crap in some way or another. You can find dozens of popular stories on this website talking about <platform>'s being garbage over all of its existence.
Imgur is a joke. They block VPN users with an intentionally obtuse "Imgur is temporarily over capacity. Please try again later.". Most importantly, its value for the average person has plummeted ever since its 2021 acquisition, and when they started deleting inactive content. UK's regulations have no place on a free internet, but the company running it is anything but worthy of praise.
You don't "get back to normal". 77 million people voted for this, and 90 million more did not care enough to stop it. This is the "normal" now, what they voted for, and you don't just forget about it.
I doubt they voted for _this_. They voted for a vague promise of a better life through some cool sounding measures, at the time. Now that people are starting to realize the demon they have given full power to, I doubt he would win another (unrigged) election.
I have some MAGA-leaning family and professional acquaintances. Here is what I hear from them:
> I doubt they voted for _this_.
They most certainly did.
First, they love the retention of the tax cuts and the OBBB.
Second, they were repulsed by most/all things that liberals embraced: DEI, language police, gender topics, loose immigration policy, an obviously over the hill Biden, perceived weakness abroad (Afghanistan, Ukraine, Israel/Gaza), etc.
> They voted for a vague promise of a better life through some cool sounding measures, at the time.
Ehhh… I don’t think so. They wanted tax cuts and anything that was the opposite of liberal social policy. They got it.
As for the people who didn’t vote, I’m not sure anything has changed that would get them out to vote for an opponent to Trump (Biden, Kamala, or whoever other milquetoast candidate lined up).
> Now that people are starting to realize the demon they have given full power to, I doubt he would win another (unrigged) election.
I think he wins by more if the election is tomorrow.
Large swathes of the monied classes in the US are largely ok with the direction things are going.
Tariffs, harassment of immigrants, neutering the federal government, and chaotic diplomacy just seem like unfortunate collateral damage to them.
Imho, the US is in the midst of a constitutional crisis. If the Dems or some other group don’t approach 2028 accordingly, they may be watching from the sidelines again for who knows how many more years.
That's really sad to hear, in a way. I don't judge the choices people make, but it seems many have gone sour - for lack of a better word. This is something that was in the makings for many years then, people generally not happy with US politics and the way the country is doing things. But T is not the answer long term.
Imho, “T” only has enough gas in the tank to enrich himself and his family by any means necessary while protecting the same group from legal consequences.
The real question is whether p2025 is the answer. The deafening indifference to p2025 makes me think that this is the direction we will be moving for a while — likely past 2028 if an opposition party does not create a compelling and unified response.
This. Reinventing the wheel at every opportunity, forgetting about or ignoring the expected way to do something, mixing patterns, you name it. The author may call it "vibe coding", that's fine but it has little to do with LLMs. The tool has the same amount of care anyone rushing to get something done, or that hasn't build the project themselves, or maybe doesn't have enough experience would. I can only assume it's a not-very-subtle complaint about a specific person in their team, "written in a way no developer on the team would" is telling.
I'd be extremely careful about applying this thinking anywhere else. There's enough baseless finger-pointing in academia and arts already.
> The author may call it "vibe coding", that's fine but it has little to do with LLMs.
Humm.
Maybe if we say that this is not an issue from vibe coding it wont be?
Maybe if we pretend that maybe a naive junior would make these mistakes (true) we should be happy to accept them from senior developers (false)?
LLMs are extraordinarily bad at doing these things.
I’ve seen it.
You've seen it.
The OP has seen it.
You’re in a rush so you wrote some classes in a code base in a language which supports classes but has no classes in it?
Really? Did that get past code review before? Did you deliberately put up a code review that you knew would be rejected and take longer to merge as a result because you were in a hurry?
Of course not.
You did the bare minimum that still met the basic quality standards expected of you.
I get it. We all get it. When youre in a rush you cut corners to move faster.
…but thats not what the OP is talking about, and its not what I see either:
Its people putting up AI slop and not caring at all what the content was.
Just a quick check it compiled and the tests pass if youre lucky.
Too lazy even put a “dont use classes” in their cursor rules file.
Come on. The OP isnt saying dont use AI.
Theyre saying care, just a little bit about your craft ffs.
It's well known some things will trip those flags, probably not what all of them are or why, but most of them inappropriately (e.g. rating IP trustworthiness, but also simple HTTP requests that look "odd"). It's also well understood you have little to no options available as contacting support, live human or not as it may be, is made intentionally opaque and difficult or completely impossible. They just don't care, there's no reason to when you're one of the many hundreds of millions using their service most likely at no cost, and it's not unique to Microsoft. It's not that they became incompetent (they are, objectively), they simply never cared about you.
There's no mention of this in the article, so be aware there's multiple posts online about QC issues. Rotring quality has been going down over the years, or their name outgrew the actual quality of the product. Current generation of 600s especially suffer from: cracking of the body (0, 1); but most importantly for pens, the joint part that screws into the bottom and upper part of the pen is extremely susceptible to wearing out the thin and fragile upper part's threads, as they are two different metals. So you should be prepared to exercise your warranty if you purchase one.
Oh, that's a very specific crack. This is an extrusion error. The extrusion temperature is dropping too low while it's still over the internal die. The thicker peaks cool more slowly than the thinner flats, remaining at a weaker temperature longer, and they're pulled apart by accumulating contraction.
These cracks usually aren't obvious until they meet a conflicting load. For example, tapping threads up the end without supporting the work correctly. It's not like this is a load bearing part, they could get around this issue with a little care. Holding the work in a hex collet during tapping is cheap, adds efficiency, and would solve the problem. Sending feedback to the extruder is free and usually effective. Or maybe the product is moving well enough on brand equity that it's not worth bothering.
> Rotring quality has been going down over the years, or their name outgrew the actual quality of the product.
This story is so common that I wish there was an established economic term for it. Something like "reputational arbitrage" or perhaps "sentiment stickiness".
The basic idea is that a business can change its quality much faster than its reputation changes. If the business rapidly cuts costs and quality, their sales will reflect their reputed quality more than their actual quality for some amount of time. That gives them a window of very high profits where they can basically sell shit like it's gold.
Eventually the reputation catches up with them, but it seems to take a very long time to do so, if ever, so it's an extremely tempting business model.
There is a related but different effect where a brand establishes some level of cachet or meaningful emotional attachment back when the product was good. The product tanks, but people keep buying it even while knowing it's garbage just because of the emotional associations they have with the historical product.
The line between these two effects can be blurry. I think Pyrex leans more towards the former where people keep buying it simply because they don't realize it kind of sucks. But Jeep is the latter where it seems like everyone knows they'll spend half the time in the shop but people just like Jeeps anyway.
Legitimately curious, could you tell me more about Pyrex? I only heard that they changed the formula years ago to trade some impact resistance for thermal shock resistance, not that there are all-around better options. I haven’t had issues with mine, though I haven’t dropped them, and the only other options I notice in stores are store brand tempered glass that seems to be competing only on price.
I'm not an expert but what I've heard is that the glass isn't as good. I know they say it's because they traded thermal resistance for impact resistance, but I don't know how much I buy that.
I do know that when I bought a Pyrex measuring cup last year, the labeling faded in months. Meanwhile the old one I inherited from my mom is still usable.
No, it should be the same. In software you can't really lower quality. Instead, stuff your product with ads and raise prices. In hardware you can lower production quality, but you can't really put ads on it. The outcome is the same.
No, enshittification is a different economic thing where a continuing service provides less and less for more and more (either attention or money) over time.
What I'm talking about is buying concrete objects at discrete moments in time.
With enshittification, you aren't just coasting on historic brand reputation. You actually have users locked into your product with pervasive network effects.
People don't still use Facebook because of any lingering nostalgia from its halcyon glory days. They use it because that's where their friends are and they can't get away. That's what enshittification is about.
There's no lock-in making boomers buy Harleys. It's just brand nostalgia despite the fact that the bikes suck now.
The original Rotring is gone; Rotring was bought by Newell Brands and is now just a label on copies made elsewhere. (They did the same with Parker and Waterman.) 20th-century Rotring is the real thing.
I had an old one and it broke, so I went looking for a new one,and it just felt cheap. Then I saw a pair at a flea market for $10, and only one had been used. I use a black rubber tape on them for comfort and it makes it less prone to cracking on the handle, and store them in a black box. ( Less UV ). I do keep going back by the store, and hoping their cheap ones get discontinued... But they are still. There.
This is correct. I have NOS Rotrings made in Germany. They sell for a pretty penny online. People who love the pens know the modern production isn’t the same.
https://www.youtube.com/watch?v=B9M4F_U1eEw
reply