It's incredibly interesting content, but I can't but help but notice the "AI-isms". I noticed it when he said "And that's the first mental shift", so I extracted the transcript and..
> "But print f is not really just a function... It's a formatting engine."
> "A format string is not just a string. It's a program."
> "A good log file is not just data. It's an emergency instrument panel."
> "...how print f has evolved into a diagnostic language and not just a formatting function."
> "...remember that you're not just printing an integer. You're invoking a tiny machine..."
> "print f of user input is not just bad style. It's an engraved invitation to chaos..."
And then there's the "whimsy" that AI liked to throw into this type of content.
> "It's a bit like giving a blindfold machinist a box of random parts and instructions and saying the third item is a carburetor."
> "...at which point it becomes a swamp full of math crocodiles."
> "The number 0.1 looks pretty innocent in decimal like a little cherub sitting on a cloud. But in binary floating point, it is a repeating fraction wearing a fake mustache..."
> "...letting the user hand your tiny formatting machine a bag of burglary tools."
Apologies to the author if this is incorrect, but this very much feels like videotaped AI spam. Even if I really, really do enjoy the subject matter.
If it's interesting then it's not spam. And why should you care if AI was utilized or not if you found it interesting?
The worst part of GenAI seems not to be AI slop (I can easily close the tab if the content isn't interesting). It's the fact that every...single...submission(!!!) on HN now has someone questioning and dissecting the content to dismiss it as AI generated.
I'd much rather people gave submissions the benefit of the doubt, or just clicked `Flag` if it is obviously worthless slop.
Just like those spam 'articles' that may at their core be interesting or have some value - but force you to click past 4 ads and scroll over/filter out another 17 just to extract the promised value - noticing that content you're consuming is obviously AI generated results in two things:
1. resentment that your time and attention was wasted by machine generated word-padding, and
2. a loss of confidence in the accuracy of the information presented
> Just like those spam 'articles' that may at their core be interesting or have some value - but force you to click past 4 ads and scroll over/filter out another 17 just to extract the promised value
That's a different problem entirely and predates the recent GenAI craze.
> a loss of confidence in the accuracy of the information presented
Then it isn't interesting any longer ;)
---
You're missing my general point though. People like yourself moaning about AI is far less interesting and useful. If you don't like the content, then flag it. If you do, then don't flood the comments with analysis about whether-or-not this was AI.
> It's the fact that every...single...submission(!!!) on HN now has someone questioning and dissecting the content to dismiss it as AI generated.
Then, when finally neither the topic nor the content has anything to do with AI, "It's so nice to read something on HN not mentioning AI" in the comments.
HN has made a clear decision on when AI content is acceptable on the site itself, it'd be nice if there was a clear decision on the linked content as well. Regardless whether it's the policy I'd personally prefer or not, it'd do a lot in regards to avoiding the same discussion appearing everywhere.
It's moreso the "AI-isms" that irk me. It's interesting, but I'm not finishing the video because once I notice it -- I can't help but focus on it. Instead, I tl;dr'd the transcript.
People question my use of AI when I double `-` with an iPhone on the internet constantly.[0] I get it, it's annoying.
However, if our barrier for quality is "at it's core, the content of this is interesting", then the quality of this place will fall off a cliff. This is factoid-level interesting. It's not a hacker writing something profound or presenting a breakthrough in garbled grade 8 English. It's a fun fact being presented in an acceptably, inoffensive, reasonably produced format.. Is that the bar?
Dave (the guy in that video) is an older timer Microsoft engineer. I don't know if he used AI to help compose his script but you can guarantee the subject matter is from his own experience.
I suspect the AI-isms you identified were really just more his own personal presentation style. I've watched a few of his videos over the years and from what I recall, they were similarly written.
> "But print f is not really just a function... It's a formatting engine."
> "A format string is not just a string. It's a program."
> "A good log file is not just data. It's an emergency instrument panel."
> "...how print f has evolved into a diagnostic language and not just a formatting function."
> "...remember that you're not just printing an integer. You're invoking a tiny machine..."
> "print f of user input is not just bad style. It's an engraved invitation to chaos..."
And then there's the "whimsy" that AI liked to throw into this type of content.
> "It's a bit like giving a blindfold machinist a box of random parts and instructions and saying the third item is a carburetor."
> "...at which point it becomes a swamp full of math crocodiles."
> "The number 0.1 looks pretty innocent in decimal like a little cherub sitting on a cloud. But in binary floating point, it is a repeating fraction wearing a fake mustache..."
> "...letting the user hand your tiny formatting machine a bag of burglary tools."
Apologies to the author if this is incorrect, but this very much feels like videotaped AI spam. Even if I really, really do enjoy the subject matter.
The worst part of GenAI seems not to be AI slop (I can easily close the tab if the content isn't interesting). It's the fact that every...single...submission(!!!) on HN now has someone questioning and dissecting the content to dismiss it as AI generated.
I'd much rather people gave submissions the benefit of the doubt, or just clicked `Flag` if it is obviously worthless slop.
Disagree.
Just like those spam 'articles' that may at their core be interesting or have some value - but force you to click past 4 ads and scroll over/filter out another 17 just to extract the promised value - noticing that content you're consuming is obviously AI generated results in two things:
1. resentment that your time and attention was wasted by machine generated word-padding, and
2. a loss of confidence in the accuracy of the information presented
That's a different problem entirely and predates the recent GenAI craze.
> a loss of confidence in the accuracy of the information presented
Then it isn't interesting any longer ;)
---
You're missing my general point though. People like yourself moaning about AI is far less interesting and useful. If you don't like the content, then flag it. If you do, then don't flood the comments with analysis about whether-or-not this was AI.
Then, when finally neither the topic nor the content has anything to do with AI, "It's so nice to read something on HN not mentioning AI" in the comments.
HN has made a clear decision on when AI content is acceptable on the site itself, it'd be nice if there was a clear decision on the linked content as well. Regardless whether it's the policy I'd personally prefer or not, it'd do a lot in regards to avoiding the same discussion appearing everywhere.
People question my use of AI when I double `-` with an iPhone on the internet constantly.[0] I get it, it's annoying.
However, if our barrier for quality is "at it's core, the content of this is interesting", then the quality of this place will fall off a cliff. This is factoid-level interesting. It's not a hacker writing something profound or presenting a breakthrough in garbled grade 8 English. It's a fun fact being presented in an acceptably, inoffensive, reasonably produced format.. Is that the bar?
[0] https://news.ycombinator.com/item?id=48151641
I suspect the AI-isms you identified were really just more his own personal presentation style. I've watched a few of his videos over the years and from what I recall, they were similarly written.
https://www.lispworks.com/documentation/HyperSpec/Body/22_c....