The most prestigious law school admissions discussion board in the world.
|
 |
 
|
Slop tell: stacked abstractions
| The Penis | 08/17/26 |
| phenotype respecter | 08/17/26 |
| The Penis | 08/17/26 |
| phenotype respecter | 08/17/26 |
| agenticman | 08/17/26 |
| phenotype respecter | 08/17/26 |
| Natalie Harp's husband | 08/17/26 |
Poast new message in this thread
Date: August 17th, 2026 12:43 PM Author: The Penis Shit like: "This reframes the issue as a question about the structural conditions governing the relation between representational architecture and explanatory adequacy." (http://www.autoadmit.com/thread.php?thread_id=5894068&forum_id=2/en-en#50073313) |
Date: August 17th, 2026 12:49 PM Author: phenotype respecter why yes, i do agree that all academic papers are slop (http://www.autoadmit.com/thread.php?thread_id=5894068&forum_id=2/en-en#50073327) |
 |
Date: August 17th, 2026 1:01 PM Author: The Penis ai slop i mean (http://www.autoadmit.com/thread.php?thread_id=5894068&forum_id=2/en-en#50073358) |
|
Date: August 17th, 2026 1:05 PM Author: phenotype respecter i'm serious. academic papers were/are proto-AI slop
they're deliberately and superfluously stacked with abstractions to the extent that it detracts from the clarity of the substance being presented |
|
Date: August 17th, 2026 1:06 PM Author: agenticman (http://www.autoadmit.com/thread.php?thread_id=5894068&forum_id=2/en-en#50073370) |
|
Date: August 17th, 2026 1:10 PM Author: phenotype respecter part of the reason why LLMs write the way they do is because they were trained on so many academic papers and academic papers were given disproportionate weight in their training
i don't even read academic papers anymore. i put them in LLMs and have them summarize them for me in clear language. anything that's unclear/needs more elucidation i ask the LLM to elaborate on
i assume everyone else with a clue is doing the same thing |
|
Date: August 17th, 2026 1:07 PM Author: Natalie Harp's husband (http://www.autoadmit.com/thread.php?thread_id=5894068&forum_id=2/en-en#50073373) |