something something. DAY 73 [OC, brainmade]
WARNING!!! this post is FAKE!!! i fabricated it.
yeah yeah i will make an actual post which isn’t just silly cheap stuff like this. i could have made this into a fancier thing but i don’t think it would have helped much. sadly, tweetgen was down so i had to use a really bad other one which uses the old twitter style and doesn’t support pfps. ugh. whatever whatever. at least it’s not a repost, amiright? hahaaaa-----
this one may come across a bit more theatrical than usual. i didn’t want it to, but that’s how the text turned out. hmm…
alt image (catbox, high quality)
yea i will make an actual post today, you will get that, don’t worry. i have the idea already but got stuffs to to outside of posts rn… at least a bit.
i wish you a wonderful day in which someone listens to your concerns <3
EDIT: added the warning
thats the version I specifically wanted to point away from.
the goal for users obviously isnt misinformation, so presenting them with something else that can also misinform feels like we are making fun of the user, even tho the joke is meant to be more sweet-leaning.
to me it reads like this bad mixture of insulting the target with a sweet afterburn which doesn’t sit right with me.
I guess for me there are plenty of people out there that take what an LLM tells them at face value and run with it, regardless of the warnings, both from the LLM itself, “Gemini is an AI and can make mistakes”, the wider community at large. We see it happening in USian politics (a recent article I scrolled pasted mentioned something out AI telling the USian defense group that a target was a target when it turned out to be a school).
So, jokes aside, even though the goal isn’t misinformation, it’s what the output could potentially be, so if the output has to be checked anyway, why not do the work upfront, rather than waste resources, time, and environmental damage only to arrive at the same place as having just done the thing yourself anyway?
Also, 196 so I had to post something in your comments before I left.
to me at least, the benefit of LMs is the ability to process lots of data, and then use its output as pointers to the actual source. such pointers like “the site mentions your problem in section 3.5” are much more likely to be correct than false.
it is unfortunate how many peeps do treat the output as correct by default. but to me, that says more about peoples expectations and behaviours toward the software rather than the software itself. Which… yea, is a problem.
I think it is fair to say that LLMs do actually have purposes they can serve, and functions they can do, so I’m sorry if my response seems to be outright dismissal. I didn’t mean it like that.
I’m just kinda burned out in my every day life of “I asked chatgpt, I asked Gemini”… Etc. that my snark levels are off the charts.
❤️
It’s absurd that people ask it about anything and everything. I suck at coding and sometimes I suck at muddling through error message logs and it’s been immensely helpful for those things but that’s about it. I wouldn’t ever think to use them for anything else and it’s not worth burning down the planet for just that.
Fair, and that would be useful IF it didn’t waste obscenely huge amounts of power and water while doing so.
We used to have a pretty decent information indexing system that people seem to have forgotten, it was called Google and it used to be pretty awesome at that task, before it became evil and greedy and started caring more about ad revenue than actual search results.
Can I get please some recommendation for a non-romantic manga? Maybe a female MC being bad ass?