Re:The Collapse of GPT
by ComplexSystems on Saturday May 17, 2025 @09:57AM
(#44014376)
So does human content. Much of the original data that GPT was trained on is Reddit posts and web pages on the internet - stuff that isn't exactly known for being a source of high quality facts.
[ Parent
| on HN ]
Re:The Collapse of GPT
by lelanthran on Saturday May 17, 2025 @12:48PM
(#44015513)
That's still very different to "drift error compounding".
If you output a mere 5% drift error and then use that as input you only need a few cycles (single digits) before your output is more erroneous than correct.
We are already partly into the second cycle. By the fifth the LLM would be mostly useless.