Minor things that bother you

[kphoger]

A fun thing to do is to ask it a question about the details of minor highway, especially one containing a plausible misconception a non-roadgeek might have, and then watch it FritzOwl an answer to the question instead of telling you it doesn't know or that your question is based on a faulty premise.

The only thing more frustrating than when A.I. makes up some bullshit and passes it off as fact is when A.I. tells you it's impossible to give a definitive answer about something when there's obviously a definitive answer out there.

[kphoger]

If I ignore a popup long enough, then it should just go the heck away on its own.  I shouldn't have to click "No thanks" or "Not now" or whatever.

[Scott5114]

If I ignore a popup long enough, then it should just go the heck away on its own.  I shouldn't have to click "No thanks" or "Not now" or whatever.

If they were content with you ignoring the message, they wouldn't have popped it up to begin with. And automatically dismissing a popup requires more code than the user clicking a button, so, yeah...

[formulanone]

Remember when the interwebs kind of rejected pop-up ads around 2004 or so? They seemed to appear around 1998 and nobody liked it, but inline ads were seen as less intrusive. Then they made a much more annoying comeback a few years later, probably has the ability to deny them in a simple browser setting now has no effect.

Then there's the ones that seem to appear three seconds and 4.6 microswipes after the page loaded. (I'm two sentences into your article, how can I make a decision to join or get this annoyance right in my Inbox? Is the rest of the world this impulsive?)

[Molandfreak]

Remember when the interwebs kind of rejected pop-up ads around 2004 or so? They seemed to appear around 1998 and nobody liked it, but inline ads were seen as less intrusive. Then they made a much more annoying comeback a few years later, probably has the ability to deny them in a simple browser setting now has no effect.

Then there's the ones that seem to appear three seconds and 4.6 microswipes after the page loaded. (I'm two sentences into your article, how can I make a decision to join or get this annoyance right in my Inbox? Is the rest of the world this impulsive?)

Nowadays you have to deal with clicking out of a pop up on the bottom left corner of the screen when watching YouTube. And that's after you've skipped or watched the ad which preceded it. This decade is just the worst.

[bugo]

It was that way with 60s and 70s music too - Japanese album covers and artwork for Western artists were always cooler than the original Western ones.

The cover is the same except for the album title, but Judas Priest released an album called Killing Machine in 1978. The record label thought the name was too edgy for the puritanical US so they released it as Hell Bent for Leather in that country. Judas Priest also released an album in 1981 called Point of Entry which had a different cover in North America, Brazil, Australia, Japan and Hong Kong. Why they changed the design is unknown, but the normal cover wasn't offensive in the least. Then again, it is the US, where folks get offended over everything and nothing, so it's possible that some bible thumpers got triggered by an abstract image. Judas Priest doesn't get enough credit for influencing thousands of metal bands over the years. They practically invented thrash, and bands like Metallica wouldn't exist without them. They're undoubtedly one of the greatest bands of all time.

Normal cover:


US cover:

[bugo]

Not everything about A.I. is bad...

I can't agree. The negative aspects of AI far outweigh any positive traits it might have. I can't think of anything good about it.

[Molandfreak]

Not everything about A.I. is bad...

I can't agree. The negative aspects of AI far outweigh any positive traits it might have. I can't think of anything good about it.

The medical field has tailored AI to its advantage and I'm sure analytical AI is a net positive to the researchers who make use of it, but generative AI really offers nothing beyond what has already been available for years since it relies on crawling through information that already exists on a website.

The advice on car seat installation ultimately came from a human anyway, whether it is coming directly from the user manual of the vehicle or some random Reddit thread discussing the topic, assuming somebody else had the same question.

Nobody really needs the AI overview from Google since the source it pulls information from is almost always just the top result from the search query anyway. Maybe the concept of directly jumping to a highlighted passage based on your search is a good idea, but an entire cliff notes version based on the text written by a human? Nah.

[kphoger]

Not everything about A.I. is bad...

I can't agree. The negative aspects of AI far outweigh any positive traits it might have. I can't think of anything good about it.

Maybe the bad outweighs the good.  But that doesn't negate the existence of the good.  If A.I. can make menial tasks easier or faster, then I'd say that's a good thing;  maybe the bad aspects outweigh that, but that aspect is still a good thing.

In my wife's case, how I mentioned she uses it for work sometimes, I can't think of any bad aspects about that use case.

[JayhawkCO]

In my wife's case, how I mentioned she uses it for work sometimes, I can't think of any bad aspects about that use case.

Except the boiling of oceans.

[wxfree]

Not everything about A.I. is bad...

I can't agree. The negative aspects of AI far outweigh any positive traits it might have. I can't think of anything good about it.

Here's some perspective from someone with no actual knowledge of the topic.  I think it's a mistake to focus on the "I" part of the label, and it's better to just call it a computer.  There is no "I," there's only pattern recognition, something computers are incredibly good at, and re-arranging information, which can be a home run or a strike out, and you never know which to expect.

I expect that the medical field can benefit a lot from the computers.  They have the ability to hold in their memory millions of images, X-rays and MRI scans, along with diagnoses, treatments, and outcomes, and compare and contrast them in great detail, something no human can do.  It isn't intelligent, it's just analysis of large blocks of data, which is what computers do.  I'm sure there are other fields, such as weather forecasting, that can benefit from this ability.

On the other foot (something a computer might say), I think it's a mistake to think of the computers as intelligent or as a person or as a way to avoid thinking or tasks.  I don't know why people have the obsession with assuming that anything we call intelligence, even if in error, must be human like and conversant.  If we could let it be a beep beep, boop boop robot like C-3PO instead of an android like Data I think we'd be better prepared to make good use of them.  Using them as friends, or to displace the human capacity to generate ideas, is self-destructive.  Even if they had the ability to produce beautiful paintings and videos and music, what's the point of it?  Is the purpose of life to look at and listen to beautiful things, or to produce beautiful things?  Are we here to think, face down challenges both menial and monumental, and accomplish, or are we here to pump our minds full of endless streams of distraction?

[vdeane]

The modern trend where people who make videos end them by putting their hand over the camera instead of just clicking the button to stop recording.  What's the point of that?

[JayhawkCO]

I was buying some train tickets today for my trip to Uzbekistan. I bought a ticket from Khiva to Bukhara for May 27th. When I looked at the ticket once it got emailed to me, the ticket was for the 26th. I figured I must have screwed up, so I paid the $3 cancellation fee. Then I went to try again, and the exact same thing happened, and I assuredly had the date correct this time. So, I had to eat the $3 fee again and then book the ticket through a 3rd party site. Now I won't have both tickets as e-tickets on the Uzbekistan Railways app and will have to print them out at home and bring them over there.

[Rothman]

I was buying some train tickets today for my trip to Uzbekistan. I bought a ticket from Khiva to Bukhara for May 27th. When I looked at the ticket once it got emailed to me, the ticket was for the 26th. I figured I must have screwed up, so I paid the $3 cancellation fee. Then I went to try again, and the exact same thing happened, and I assuredly had the date correct this time. So, I had to eat the $3 fee again and then book the ticket through a 3rd party site. Now I won't have both tickets as e-tickets on the Uzbekistan Railways app and will have to print them out at home and bring them over there.

Should have made it for the 28th.

In Russian, there's some weird things they do with date formats and the like (which was fun to navigate when I lived over there).  Wonder if something like that was lost in translation in their website design.

[TheHighwayMan3561]

The modern trend where people who make videos end them by putting their hand over the camera instead of just clicking the button to stop recording.  What's the point of that?

Since this is common as an editing technique to hide jump cuts, maybe they think that makes it look like a "smoother" end. I agree it sounds weird.

[Molandfreak]

The modern trend where people who make videos end them by putting their hand over the camera instead of just clicking the button to stop recording.  What's the point of that?

Since this is common as an editing technique to hide jump cuts, maybe they think that makes it look like a "smoother" end. I agree it sounds weird.

I remember this being a trope in Rocket Power and similar shows with plot points involving camcorders or found footage. And it was more annoying than a jump cut then, too.

[JayhawkCO]

I was buying some train tickets today for my trip to Uzbekistan. I bought a ticket from Khiva to Bukhara for May 27th. When I looked at the ticket once it got emailed to me, the ticket was for the 26th. I figured I must have screwed up, so I paid the $3 cancellation fee. Then I went to try again, and the exact same thing happened, and I assuredly had the date correct this time. So, I had to eat the $3 fee again and then book the ticket through a 3rd party site. Now I won't have both tickets as e-tickets on the Uzbekistan Railways app and will have to print them out at home and bring them over there.

Should have made it for the 28th.

In Russian, there's some weird things they do with date formats and the like (which was fun to navigate when I lived over there).  Wonder if something like that was lost in translation in their website design.

You'd think. But I also bought a ticket from Bukhara to Samarkand on their super fast train and it went just fine.

[Rothman]

I was buying some train tickets today for my trip to Uzbekistan. I bought a ticket from Khiva to Bukhara for May 27th. When I looked at the ticket once it got emailed to me, the ticket was for the 26th. I figured I must have screwed up, so I paid the $3 cancellation fee. Then I went to try again, and the exact same thing happened, and I assuredly had the date correct this time. So, I had to eat the $3 fee again and then book the ticket through a 3rd party site. Now I won't have both tickets as e-tickets on the Uzbekistan Railways app and will have to print them out at home and bring them over there.

Should have made it for the 28th.

In Russian, there's some weird things they do with date formats and the like (which was fun to navigate when I lived over there).  Wonder if something like that was lost in translation in their website design.

You'd think. But I also bought a ticket from Bukhara to Samarkand on their super fast train and it went just fine.

Eta zhizn

[JayhawkCO]

Eta zhizn

Haqiqatan ham shunday. (Uzbek)

[bugo]

Maybe the bad outweighs the good.  But that doesn't negate the existence of the good.  If A.I. can make menial tasks easier or faster, then I'd say that's a good thing;  maybe the bad aspects outweigh that, but that aspect is still a good thing.

Would you take a pill that gave you cancer if it got rid of your headache?

[bugo]

I personally don't want AI being involved in my medical care in any way. I don't trust it any further than I can throw a car. Unfortunately, the capitalists are shoving it down our throats, and we can't avoid it. Using AI in medicine will lead to many deaths and unnecessary injury and misdiagnosis. We did without AI for thousands of years, why do we need it all of a sudden?

[Scott5114]

Generative AI like LLM chatbots and image generators is of the devil.

Analytical AI is an entirely different thing that has nothing to do with generative AI, and in fact has been around for a long time, has a proven track record, and had its good name ripped off by the generative AI people to make a buck.

[Rothman]

I personally don't want AI being involved in my medical care in any way. I don't trust it any further than I can throw a car. Unfortunately, the capitalists are shoving it down our throats, and we can't avoid it. Using AI in medicine will lead to many deaths and unnecessary injury and misdiagnosis. We did without AI for thousands of years, why do we need it all of a sudden?

I'm not so sure.  Studies are starting to show that computer programs may be better at diagnoses than human doctors.  If that keeps panning out, I'd rather be properly treated...

[formulanone]

It was that way with 60s and 70s music too - Japanese album covers and artwork for Western artists were always cooler than the original Western ones.

The cover is the same except for the album title, but Judas Priest released an album called Killing Machine in 1978. The record label thought the name was too edgy for the puritanical US so they released it as Hell Bent for Leather in that country. Judas Priest also released an album in 1981 called Point of Entry which had a different cover in North America, Brazil, Australia, Japan and Hong Kong. Why they changed the design is unknown, but the normal cover wasn't offensive in the least. Then again, it is the US, where folks get offended over everything and nothing, so it's possible that some bible thumpers got triggered by an abstract image. Judas Priest doesn't get enough credit for influencing thousands of metal bands over the years. They practically invented thrash, and bands like Metallica wouldn't exist without them. They're undoubtedly one of the greatest bands of all time.

Normal cover:


US cover:

There's been many reasons why one release of an album doesn't look like the next; sometimes it's because of obscenity, but I wouldn't be surprised if artist copyright is the primary reason. In most cases, the visual album art has little to do with the musical artists, while usually they'll choose an idea of concept, in some cases they don't really like it and but just roll with it due to distribution deadlines. Or the artwork was contracted for one type of release, but not another (for example, international versions of the album). So it's not always because someone clutched at pearls over obscenity, there were probably legitimate legal reasons...though it is funny how how brains can perceive "album art" as something tied specifically to the music contained within the packaging, yet it can often be envisioned by another artist years before a single note was played nor word sung.

In the case of photography, sometimes the artist(s) really like an image and let the copyright and redistribution permissions get sorted out later...supposedly.

[Beltway]

Generative AI like LLM chatbots and image generators is of the devil.
Analytical AI is an entirely different thing that has nothing to do with generative AI, and in fact has been around for a long time, has a proven track record, and had its good name ripped off by the generative AI people to make a buck.

Decision support systems (DSS) are exactly the kind of analytical AI you're describing -- rule‑bound, domain‑specific, and designed to augment human judgment rather than imitate human language. They've been around for decades in aviation, logistics, medicine, and infrastructure management. They don't "generate" anything; they evaluate inputs against known models and return structured outputs.

One that I used around 2002-08 --

ESTIMACS was an early structured IT project estimation tool developed by Howard Rubin in the 1980s. It used a detailed questionnaire to score factors such as application complexity, data structures, user interaction, performance requirements, and team experience. Those weighted scores fed a deterministic model that produced effort estimates, staffing curves, cost ranges, and risk indicators. ESTIMACS wasn't predictive in a modern sense, but it provided consistent early‑stage sizing and comparative analysis across alternatives. It functioned as a classic Decision Support System, helping organizations standardize planning long before agile metrics and parametric models became common.

It was reasonably accurate for its era and for early‑stage estimates. But it was never intended to be precise, and it had known limitations. Think of it as a structured sanity-check tool, not a forecasting engine.