The damn apps that auto change emoticons to emoji I'm expressing :/ not :corpoface: ...

Might have more luck if you switched to the lazy em dash that is just two normal dashes.

I think that the people trying to call out slop are doing more to hurt those competing with AI than AI itself.

Like someone using AI to produce stuff might get annoyed at being called out but then will probably just think "ok, gotta prompt it to avoid these signs" or waits for a newer model or just dgaf that people notice it's AI.

But people producing their own things and being accused of it being AI must feel pretty fucking discouraged that not only do they need to compete with something that can output things in a fraction of the time, but even when they do, some people accuse them of being what they hate.

For text content, IMO it should be taken based on what is said rather than who (or what) said it. If it's useless slop, there will be things to push back on other than the formatting. Plus, it wouldn't surprise me if you could just add "format it like a low effort forum post" to the prompt to get it to avoid some of the more obvious formatting tells.

Also, if people religiously call out AI slop when they see it, that could be used to automate training similar to what I'm sure happened to drastically improve programming capabilities, which will accelerate the speed at which they can become indistinguishable from things produced by real people.

It is kinda funny that the AI talks like someone with ADHD, probably because most humans with it ramble the most online

What's your relationship with the semicolon like?

Congratulations, and I applaud your diversity in using a wide variety of characters that take more effort than just pushing a button on a standard keyboard.

In the human world that is rare.

In the LLM world it's zero effort, and common.

because of my ADHD, I tend to run a bit on the verbose side of things.

Feel that. I'm sure my speech patterns and propensity to be long-winded will eventually bring me some AI accusations too.

Problem with LLM tech is that if it truly models speech it will eventually be indistinguishable from humans - scary thought but what are you going to do?

I mean, I think you can kind of feel off of vibes if something has used or is AI but it is becoming increasingly harder to spot.

you still probably don't use it nearly as much as LLMs do. It's every third sentence with an LLM, and those sentences will have 2 or 3 em dashes. . The average em dash use when it was at its peak was like 1 every 200-500 words from what I remember. With LLMs it's like 1 every 40 words.

It's a numbers thing, of course some people use just about everything that's possible.

Me personally, it interrupts my flow-of-thought to think "o.k. now I want that funky character I don't remember how to get without going through some big lookup process - meanwhile, I can type the basic stuff 1000x faster...."

I also understand that for people who do it all the time, it gets easier - but the majority of people out there aren't the ones who know how to do all that.

the books the LLM were trained on that used the most em dashes used like 1 every 200-500 characters. LLMs use em dashes at a rate of like 1 every 40-50 characters. It's not even CLOSE to human behavior.

There are no absolutes, some people do spend lots of time and effort to put cool looking emojii characters into their documents which kill portability and cause all kinds of problems down the road. For the LLM it's 0 additional effort, in some ways less effort because the emojii is a kind of compressed information - but it still causes problems when moving around to different editors, different toolchains...

GOAT TODO IS BEAUTIFUL ❤️

General practice for all quality records, and code can be considered a quality record, never post-date anything. Especially code because it's so discoverable by cross-checkable references.

Has anybody been prosecuted for this? Probably not, but if an investigation finds that records were intentionally falsified with intent to deceive / cover up problems, that can lead to very costly regulatory responses. Even a minor regulatory response noting practices of inaccurate dates on records is still costly to the business that note is made on.

Inbox my beloved

The hate you had for them was made recent. I, I was full of the hatred from the early times. I have carried the hate for an age now, and it has crystallised into a rage that will never be sated.

fuck off for killing it.

Welcome to the c l o u d .

This one too

"But YouTube Music..." no. YouTube Music sucks by comparison, can't see how to upload anything new, a sizeable chunk of music I used to listen to wasn't licensed for YouTube Music like it was for Play Music, the interface blows and is constantly moving shit around so I have to find things again and again, and it can't figure out that it's for music and not video

Claude is just head and shoulders better than the rest. Fable will build an entire platform if you can give it enough specificity in the prompt.

And they'll miss important test cases such that their change will pass all of their tests (they keep iterating until it does) but then will still fail on some common edge cases.

that AI doesn’t mean inherently awful and bad

it does when it's actively destroying the planet. There's no good use of AI. There's less bad uses since we're now in an era of AI generated cyber attacks, so people are using it to defend. But there's no good use. It's a horrendous technology that is destroying humanity with every use.

It's the same as a nuclear bomb. If the bomb was never used and we just knew how to make it, it wouldn't be that bad. But it was used, it killed hundreds of thousands of people, and no matter what happens in the future it's still horrific technology.

The thing I have to constantly adjust for is: it does things at 10x my speed (on average, some things it does slower, but some things it does 100x faster), and about 80% of my accuracy. So, I have to really aggressively apply the bug catching methodologies that a lot of times I would get away without "because I'm that good..." not really, but there are some things I do that I have justifiably high confidence in, others that definitely need testing which I try to do before handing over to the test team. The reason I don't apply those bug catching methodologies to ALL my code is because that slows me down 3-5x, and it seems to slow down the LLMs about 2-3x, so they're only 3-5x faster than me, on average, when they are writing all the tests and multiply reviewing all the docs and code, compared to me "winging it."

Where I think all this LLM stuff is going to pay off bigger is 10 years in the future, looking back at old code, yeah all that documentation is a bit of overkill, but we tended to underkill the documentation 10 years ago and we're still not doing as much as we probably should when we do things the "just before LLMs came on the scene" ways.

Well... That's why you test the public API and not the internals. And if you need to change the API you should have a good enough reason that changing the tests is okay too.

Nice try hoo-man

See, so you’ve experienced the same thing! Thanks for the confirmation, fellow totally-human Lemmy user!