The most prestigious law school admissions discussion board in the world.
|
 |
 
|
ChatGPT 5.2 released moments ago
| .,.,...,.,.;,.,,,:,.,.,::,....,:,..;,.., | 12/11/25 |
| .,.,...,.,.;,.,,,:,.,.,::,....,:,..;,.., | 12/11/25 |
| spaceport2525 | 12/11/25 |
| stool | 12/11/25 |
| merry screenmas | 12/11/25 |
| ,.,....,..,.,.,,,,..,..,.,..,.,.,.,... | 12/11/25 |
| mouse utopia | 12/11/25 |
| ,.,....,..,.,.,,,,..,..,.,..,.,.,.,... | 12/11/25 |
| merry screenmas | 12/11/25 |
Poast new message in this thread
Date: December 11th, 2025 1:28 PM Author: .,.,...,.,.;,.,,,:,.,.,::,....,:,..;,.., https://x.com/sama/status/1999184337460428962?s=46 (http://www.autoadmit.com/thread.php?thread_id=5809123&forum_id=2Elisa#49502204) |
Date: December 11th, 2025 1:34 PM Author: .,.,...,.,.;,.,,,:,.,.,::,....,:,..;,.., https://x.com/sama/status/1999185784012947900?s=46 (http://www.autoadmit.com/thread.php?thread_id=5809123&forum_id=2Elisa#49502214) |
|
Date: December 11th, 2025 1:37 PM Author: spaceport2525 Yeah but does it pass the FAGGOT test? (http://www.autoadmit.com/thread.php?thread_id=5809123&forum_id=2Elisa#49502220) |
|
Date: December 11th, 2025 1:38 PM Author: stool What about ShitLawBench (http://www.autoadmit.com/thread.php?thread_id=5809123&forum_id=2Elisa#49502225) |
 |
Date: December 11th, 2025 2:21 PM Author: merry screenmas (http://www.autoadmit.com/thread.php?thread_id=5809123&forum_id=2Elisa#49502337) |
|
Date: December 11th, 2025 2:16 PM Author: ,.,....,..,.,.,,,,..,..,.,..,.,.,.,... Nice, we are seeing non-trivial model improvements on a monthly basis still. Inching closer to replacing all human labor (http://www.autoadmit.com/thread.php?thread_id=5809123&forum_id=2Elisa#49502322) |
|
Date: December 11th, 2025 2:21 PM Author: mouse utopia (Narrator: the improvements are in fact trivial) (http://www.autoadmit.com/thread.php?thread_id=5809123&forum_id=2Elisa#49502343) |
|
Date: December 11th, 2025 2:31 PM Author: ,.,....,..,.,.,,,,..,..,.,..,.,.,.,... It’s not likely to have a major impact on most tasks but these improvements over small time horizons aren’t negligible. It’s conceivable we could have SWE agents doing large scale, autonomous AI research fairly soon (<5 years), assuming there isn’t a major deceleration in progress in the near future
https://metr.org/blog/2025-03-19-measuring-ai-ability-to-complete-long-tasks/ |
|
Date: December 11th, 2025 2:22 PM Author: merry screenmas client email this morning
"disregard my previous email I had chatgpt do it"
tp |