I think @WeirdWriter@caneandable.social made that first point very beautifully in https://sightlessscribbles.com/the-colonization-of-confidence/ and I liked it so much I want to share it to no end
CC: @screwturn@mastodon.social @tante@tldr.nettime.org
Post
Replies:
22
@tante I'm also really frustrated. It's constantly in my face. I console myself with anti ai content, but it's all so bleak. Can I work in the tech industry without having to deal with it? Seems unlikely. I'm also waiting for it to fall to bits. I don't even care how bad the fallout is anymore. Just make it stop
@tante It seems to be this time of the month. This morning I've had quite intense thoughts of quitting my IT-related job, starting to study something like medicine to wind up as a forensic medical expert, thusly having waaaaay fewer people tell me about AI and fancy and stuff...
These hype cycles are exhausting in general, this one is actively destroying everything I strive for (in short: enlightenment as explained by Kant - in short: use your brains).
It. is. so. tiresome. 🫂
@tante skills atrophy is what frightens me most.
As I am getting older, there are many things I have learned and quite a few I forgot by not using them.
French.
Windsurfing.
Poems I used to know by heart.
Lately my English started to deteriorate. I blame my use of DeepL.
I use LLMs at work - carefully- and I am convinced I could not manage the workload without but there are days I can feel my brain getting mushy.
@tante feeling every word of this despondency.
every time i'm asked to rate some stupid company's AI assistant, i take the opportunity to tell them to stop using them & i give them all the reasons. do human eyes see (let alone, read) my words? i highly fucking doubt it.
@tante I am tired. This is very tiring. Good luck..
@tante I understand the sentiment, it is so very, very tiring.
Yet, I hold out hope, because I have seen the future.
Our twins are second year in elementary, their teacher used to use ChatGPT to give them homework. The assignments were full of misspellings and other errors (ChatGPT sucks at Hungarian). We helped the kids correct those mistakes, and when they showed their homework, the teacher saw the red lines under the mistakes, and the corrections - also in red. Kids didn't say anything. Didn't call her out. They made her uncomfortable.
By the next week, the entire class was doing this, and the teacher stopped using ChatGPT. By the end of the month, the entire school was having great fun correcting AI mistakes.
Today, "AI", for the kids, is synonymous with "liar", "idiot", "wrong", "bad". And they teach it to their parents. I've seen some of them turn, and it's becoming harder and harder to be pro-AI in our little town. Being one is considered ridiculous.
It's a small thing. But it gives me hope.
@algernon @tante Agree with you both! There's hope here, too, with most the high schoolers and many middle schoolers openly mocking AI generated content. They can sniff it out in a heartbeat at school, too. It has caused them to lose respect for some teachers that use AI in class, and at least some of those teachers are coming to the realization that using AI is not the way to teach kids the intended lessons about content, critical thinking, discussion, and working together.
@tante I’m with you, thanks for capturing my feelings and thoughts so well.
Feeling discouraged is understandable but here is a little perspective.
About two weeks ago I was talking to my mom on the phone. She's a mathematician and has been programming since there were punch cards but she never really got into the internet. A month ago we had a bit of a falling out since she was using chatGPT to help her edit this book on Civil War history that's she's doing as a hobby project. I was worried it'd destroy everything that might make the project good.
God, I hate to spoil this but here I am, to tell you that using an LLM to help with editing a manuscript can be really useful.
Every defect found in a manuscript is a defect found and corrected, and getting an LLM to do a line by line technical edit catches a lot of defects I may not have noticed anymore, and which my beta-readers might also miss because they are focused on plot and character.
The LLM also picks up when I called "Agatha", "Agnes" etc.
@screwturn Is it worth destroying our planet for? We are entering a climate emergency. We need the water, etc, that data centres supporting this need. @futurebird @tante
@CStamp
Nothing is worth destroying the ecosystem, but does my use of the LLM to edit a manuscript do more harm than doing it all on paper? I don't know, but it doesn't seem so.
If I get a steel hip joint, like so many of my friends, does that do less ecological harm than time chatting with an AI? I doubt it.
What I do know is that there is a trade, and I am unsure which trades are worse, but I do know that my use seems light
@screwturn @CStamp @futurebird @tante
> If I get a steel hip joint, like so many of my friends, does that do less ecological harm than time chatting with an AI? I doubt it.
One of those is a necessity. The other isn't. Like, those aren't even in the same league, and this argument feels almost intentionally obtuse.
It makes most writing worse IMO. Sometimes it finds errors but it also just drains the life out of everything.
There are contexts where people what their writing to be lifeless. So it's good for that.
I think @WeirdWriter@caneandable.social made that first point very beautifully in https://sightlessscribbles.com/the-colonization-of-confidence/ and I liked it so much I want to share it to no end
CC: @screwturn@mastodon.social @tante@tldr.nettime.org
My mom thought I was being paranoid. But then she calls me last week and asks "What's a data center can you explain those to me?"
So I did.
"I didn't know it was using THAT much electricity. Is that what they are trying to build?"
(she lives in rural PA and there is a big fight about a data center going up)
It was only at this point that she really "got" what the LLM was or how it worked.
"So they need to process millions of documents of text just to make it work?"
"Billions, mom"
She's stopped using chatGPT. In part because of knowing more about it and the "magic" being destroyed... but also because "it seemed like it was saving us time editing but I need to check everything so many times... so it's not saving us time."
Some of us have gotten deep in with this new tech right away and we've thought about it and made decisions and a lot of people are still just finding out "what is a data center?"
Don't get discouraged.
@futurebird @tante There's also the problem of semantic ablation: https://www.theregister.com/2026/02/16/semantic_ablation_ai_writing/
Honestly, if you are going to use AI, only use it for something you are no good at and have no interest in getting better at. Then, because it's basically average at everything, it works ok if you have no other option. [This paragraph ignores the other environmental, ethical, and societal reasons not to use it].
@futurebird @tante Thank you. Been feeling the same tiredness. This helps.
@tante I'm in a similar place. It's strange how selective people can be in how they gather and assess data. Slop AI is one area in a series where this has been screaming in my face how stupid collective human culture can be.
Maybe we need to move on. Look ahead, and build based on what we expect going forward rather than try to change the direction. Everything changes and almost most everything is a cycle. The period can be too long for individuals to span, but we can still contribute goodness.
@tante feeling this so much.
like there have been a lot of crazes in tech. going to agile. automatisation. putting everthing into containers.
and my tech friends believe it is just another one of those. but AI is destroying so much more then any of those other things could.
I will continue running against that wall with you.
