« GRITZ Home | Email msg. | Reply to msg. | Post new | Board info. Previous | Home | Next

Re: AI Model Collapse Spreading. What is AI Model Collapse?

By: monkeytrots in GRITZ | Recommend this post (0)
Wed, 28 May 25 10:59 PM | 14 view(s)
Boardmark this board | Grits Breakfast of Champeens!
Msg. 08877 of 09138
(This msg. is a reply to 08853 by Fiz)

Jump:
Jump to board:
Jump to msg. #

The Internet already suffers from TMI - when human beings have a difficult time separating TRUTH from garbage who really believes that AI can do any better ?

The woke left already suffers from 'model collapse'.

A 'consensus' of opinion does not a reality make.




Avatar

Finally, brethren, whatsoever things are true, whatsoever things are honest, whatsoever things are just, whatsoever things are pure, whatsoever things are lovely, whatsoever things are of good ...


- - - - -
View Replies (1) »



» You can also:
- - - - -
The above is a reply to the following message:
AI Model Collapse Spreading. What is AI Model Collapse?
By: Fiz
in GRITZ
Wed, 28 May 25 12:16 PM
Msg. 08853 of 09138

http://slashdot.org/story/25/05/28/0242240/some-signs-of-ai-model-collapse-begin-to-reveal-themselves
excerpt:

"Welcome to Garbage In/Garbage Out (GIGO). Formally, in AI circles, this is known as AI model collapse. In an AI model collapse, AI systems, which are trained on their own outputs, gradually lose accuracy, diversity, and reliability. This occurs because errors compound across successive model generations, leading to distorted data distributions and "irreversible defects" in performance. The final result? A Nature 2024 paper stated, "The model becomes poisoned with its own projection of reality." [...]

We're going to invest more and more in AI, right up to the point that model collapse hits hard and AI answers are so bad even a brain-dead CEO can't ignore it. How long will it take? I think it's already happening"

http://opentools.ai/news/ai-model-collapse-alarm-bells-ring-as-self-training-ai-faces-degrading-accuracy


Fiz: There are many, many articles out there about this. Here is one from just a few days ago. It sounds a lot like psychosis...or a conclave of the Demon-Rat/WOKE Party....

Now, here's what I want you to think about: because of

(1) training AI systems on original data sets involved and involves MASSIVE copyright violation. (Which, is an open secret in the AI community, and dismissed as 'well, they have to catch us first, and prove it in court, before they can stop us from violating original copy-right...and by then we will all be billionaires...)

(2) "Everyone" is already using AI to write "original" material, including to polish "original research" for publication, and I doubt this trend can be caught, much less stopped, anytime soon...if ever!

(3) The legal response to AI stealing copyright materials is just leaving the station and is likely to get more and more onerous, SO THE USE OF SYNTHETIC AND PSYCHOTIC DATA SETS IS LIKELY TO BECOME MORE AND MORE INEVITABLE.

Ergo -- and this is what I think is worth thinking about -- eventually all the 'data' in the world will become suspect as stolen, or psychotic, or both ...! But AI systems write SO WELL, and are so good at mimicking things like writing style, so how is anyone going to really know the difference?????


« GRITZ Home | Email msg. | Reply to msg. | Post new | Board info. Previous | Home | Next