Posted by surprisetalk 6 hours ago
I have never, ever, bought anything on TEMU, because anytime I open the website it gives me a gambling roulette for some supposed discount thing. I hate gambling, so everytime I see the roulette I just close the webpage and go do something else.
Well, the whole cheap-generated content craze feels basically like gambling to me to this point, too. Folks are just randomly generating content: books, videos, articles, even applications, and gambling that something will pay for itself.
So now we have Enshitification, Eemufication and Gamblification of the Human Society. Damn I might write something on Substack about that.
And the thing is, Temu - I was going to buy something from you. I saw an ad or whatever and thought "sure, I'll pay $13 for that keyboard." But then you get pulled into the dopamine labyrinth hell-realm and by the time you're out they've wasted 15 minutes of your time and you don't even remember what you came in for.
All that said, there have been a decent number of things that I bought from them which I've found incredibly useful. Just like, phone stands, ODBII reader, sticky notepads that look like Windows 95 Notepad windows, teeny tiny wireless earbuds I use while sleeping, etc.
Isn't that just true about most things over the past century or so? They left out "staggeringly uglier" though.
Here's an example:
Based on what I've seen in industry, I think the answer could be very likely worse. They'll generate bad code at an even faster rate, and very quickly nobody in the company will have any idea how anything works.
What this proves is that "human made" is not automatically good. Better is a complex question. LLMs have a magnifying affect on the quality of human decisions.
Codebases degenerate when the authors do not understand the system in which they operate and make changes that degrade the cohesion of the system as a whole. Humans do this all the time; at my company we had a 10k line lua function (within a 250k loc lua service) that grew through ten thousand point modifications. It's bad and humans caused it. Now, it's 15k lines of code and 10k lines of misleading slop comments.
LLMs do not go through the struggle of understanding and don't experience the distress of "wtf am I reading?" and don't experience the drive to make things less horrible for the next person. That suffering loop is how many codebases halt degradation and in the absence of suffering LLMs will pile bugs and rot endlessly.
> It's bad and humans caused it.
LLMs do the same thing. I review the code and interrogate the model about the bandaid fix it applied until the proper architectural solution becomes clear. Then the model implements the correct solution which usually involves some major refactoring.
> LLMs do not go through the struggle of understanding and don't experience the distress of "wtf am I reading?" and don't experience the drive to make things less horrible for the next person.
Of course not. That's the human's job.
And it's not every human either. Way too many humans are actually quite content to ship whatever works then go the hell home. Just a job after all.
- 30+ committers on <5K LOC
- software is robust/rock solid but the code is essentially unreadable
- changes take a long time b/c there are no tests
Whenever folks mention that LLMs write buggy code I think of the above.
> all held together with thousands of flags
> ridden with mysterious macros that one cannot decipher
> needs to understand the values and the effects of 20 different flag
> Sometimes 100s
> only reason why this product is still surviving and still works is due to literally millions of tests
> The fact that this product even works is nothing short of a miracle!
Sorry, but I can't think of any adjective for this mess other than just "bad". That's some truly industrial grade slop, and I hope whoever worked on that got paid handsomely for it.
Simply FINDING good media becomes the problem. I have no interest in AI slop, but it's everywhere.
I don't choose to consume bad media, but finding good media is getting harder and harder.
I still feel bad for what is happening though, it's wrong and anyone with our best interests in mind would agree generated content needs to be firewalled from the human internet.
Log in, pay for premium, and use the like and subscribe buttons liberally. Use the downvote button too. It learns.
My feed is really fantastic. Once a month or so I notice some AI slop (usually with a handful of views - clearly the algo testing something out). Thumbs down, problem solved.