Welcome to DU! The truly grassroots left-of-center political community where regular people, not algorithms, drive the discussions and set the standards. Join the community: Create a free account Support DU (and get rid of ads!): Become a Star Member Latest Breaking News Editorials & Other Articles General Discussion The DU Lounge All Forums Issue Forums Culture Forums Alliance Forums Region Forums Support Forums Help & Search

highplainsdem

(61,205 posts)
27. No, the problem using genAI is with genAI and its inherent flaws. Anyone who's ever used genAI should
Mon Feb 16, 2026, 09:16 AM
Feb 16

be aware that the exact same prompt can get different answers, often wildly different answers. Every image generator that offers multiple options as results given just one prompt proves that.

And text generators can do the same thing, but are rarely set to do so because multiple varying text responses make it painfully obvious (painful for AI peddlers) that genAI has no real awareness of what is correct. And what can be excused as "creative" when there are very different images offered at once can't be excused away when it's text, no matter how hard the AI companies have tried to market their flawed tech's hallucinations as "creative" responses.

I would argue the biggest problem is the fuckhead AI companies paying off the pols so they can rush to market before there's ANY protective standards for humanity.


The first gigantic problem was the worldwide theft of intellectual property.

The second gigantic problem was release of tech they knew could and would be widely used for cheating and criminal fraud.

The third was releasing AI that hallucinated and still sounded convincing.

The fourth was releasing chatbots designed to become addictive.

I don't consider badly flawed tech that can be used for fraud amazing.

And there was a paragraph in your previous message that I should have responded to:

AI doing coding is 100% here to stay, so discussing the non-perfection in using AI to do it is a bit pointless at this juncture. There's probably nothing it's better at doing, and it relies on publicly available/non-copyrighted documents to do that work, so it's also among the more morally acceptable things to use it for.


"Publicly available" is not the same thing as "non-copyrighted" even though AI robber barons would love people to believe they are the same.

AI models that code also have legal problems.

https://www.bloomberglaw.com/external/document/X4H9CFB4000000/copyrights-professional-perspective-ip-issues-with-ai-code-gener

Recommendations

1 members have recommended this reply (displayed in chronological order):

Thx to you & Goonch SheltieLover Feb 13 #1
Lol for fun I asked an AI the reason for the discrepancy and gave it the passage from your post AZJonnie Feb 13 #2
Not sure what you mean. It quoted one of my replies in that Science Fiction thread, or it quoted highplainsdem Feb 13 #4
I mean it literally sourced from this very thread AZJonnie Feb 13 #7
Claude is clueless. There is an actual story. I posted links about it in the earlier thread I linked to. There highplainsdem Feb 13 #9
It's not surprising that it got confused, this entire discussion is extremely circular AZJonnie Feb 13 #13
I would've expected any bot to at least follow the links in both threads, which would have shown that highplainsdem Feb 13 #15
Claude would have seen this bit, in Goonch's follow-up, which said this: AZJonnie Feb 13 #17
Interesting seeing you try to defend Claude's inane answer, when this thread links to the older thread highplainsdem Feb 15 #19
I guess I am, given you don't know what the actual prompt was, yet are arbitrarily coming up with a strawman AZJonnie Feb 15 #20
If there was anything close to intelligence in Claude, the highplainsdem Feb 15 #21
Why must you insist it's a delusional tangent AZJonnie Feb 15 #22
No, the problem using genAI is with genAI and its inherent flaws. Anyone who's ever used genAI should highplainsdem Feb 16 #27
And, The Chatbot Is Still Wrong ProfessorGAC Feb 13 #3
+1. It's patent nonsense dalton99a Feb 13 #5
I don't know about that one, Professor :) AZJonnie Feb 13 #8
I'm Going To Say No ProfessorGAC Feb 13 #11
Obviously I know I don't know nearly as much on this topic as you do, so I generally defer, Sir :) AZJonnie Feb 13 #14
Pretty Much ProfessorGAC Feb 13 #16
Can you name a movie in which one person occupied two places at the same time? Orrex Feb 16 #29
BTTF Is A Prime Example, Yes ProfessorGAC Feb 16 #33
Right, but that's not the same Marty in two places at once Orrex Feb 16 #34
Not Getting You ProfessorGAC Feb 16 #35
I think we're differing on what qualifies as the "same" person Orrex Feb 16 #37
Oh wait. I just caught your bit about "lack of matter available" Orrex Feb 16 #38
These tools don't just fabricate fiction. They fabricate citations in law and science pieces. RockRaven Feb 13 #6
Yes. I mentioned that in the earlier thread I linked to. I've posted lots of warnings here over the last few years highplainsdem Feb 13 #10
+1. AI is essentially a smooth-talking buzzword-spewing bullshitter with an unlimited capacity for plagiarism dalton99a Feb 13 #12
Exactly. highplainsdem Feb 15 #18
When ChatGPT became popular, people said AI systems really need to provide sources. Renew Deal Feb 15 #24
Good catch Renew Deal Feb 15 #23
Thanks - but I wouldn't have caught it if I hadn't already looked at a number of websites about the story so highplainsdem Feb 16 #25
Yikes. So how do we combat this? It's only going to get worse. Scrivener7 Feb 16 #26
Yeah, it's getting scary. I posted something I didn't know was AI. I took it down as soon as mucifer Feb 16 #28
It's really bad in the political commentary video space Renew Deal Feb 16 #31
I suspect AI systems will be like fish farms. cachukis Feb 16 #30
I miss the old Google. Hell, I miss the old Alta Vista. haele Feb 16 #32
Anyone can make up a quote. GoCubsGo Feb 16 #36
Latest Discussions»General Discussion»You've probably heard tha...»Reply #27