The most prestigious law school admissions discussion board in the world.
|
 |
 
|
Good explanation of why AI has plateaued and isn't scaling well
| pea-brained plaza | 06/05/25 |
| pea-brained plaza | 06/05/25 |
| godawful giraffe | 06/05/25 |
| Khaki round eye mad cow disease | 06/05/25 |
| arousing smoky stage | 06/05/25 |
| Claret home sandwich | 06/05/25 |
| Milky Nighttime Hell | 06/05/25 |
| cerebral slate stage useless brakes | 06/05/25 |
| offensive persian corner | 06/05/25 |
| pea-brained plaza | 06/05/25 |
| stirring alpha son of senegal | 06/05/25 |
| pea-brained plaza | 06/05/25 |
Poast new message in this thread
Date: June 5th, 2025 7:44 PM Author: pea-brained plaza https://youtu.be/3MygnjdqNWc?si=tr_dVM5Qc08FVcRi (http://www.autoadmit.com/thread.php?thread_id=5733944&forum_id=2).#48990568) |
Date: June 5th, 2025 7:45 PM Author: pea-brained plaza Bingo:
https://youtu.be/3MygnjdqNWc?t=792
Fuck fags who don't get matrix algebra. The USA is full of them, as is India. |
|
Date: June 5th, 2025 7:47 PM Author: godawful giraffe It hasn't plateaued at all (http://www.autoadmit.com/thread.php?thread_id=5733944&forum_id=2).#48990574) |
 |
Date: June 5th, 2025 8:13 PM Author: Khaki round eye mad cow disease (http://www.autoadmit.com/thread.php?thread_id=5733944&forum_id=2).#48990639) |
|
Date: June 5th, 2025 7:47 PM Author: arousing smoky stage rare cr NSAM thread this was a good video (http://www.autoadmit.com/thread.php?thread_id=5733944&forum_id=2).#48990576) |
|
Date: June 5th, 2025 8:02 PM Author: Claret home sandwich Gary Marcus is the one of the most intellectually dishonest people in this space who has been insisting AI has been hitting a wall for years. For fun, you should look at all the things he made fun of GPT-3 or DALL-E for not being able to do that can now be done with ease simply because of scaling. He is a proponent of the logical AI camp that isn’t especially interested in biology (which is obviously connectionist and based on fairly uniform computational circuitry throughout the neocortex) . (http://www.autoadmit.com/thread.php?thread_id=5733944&forum_id=2).#48990612) |
|
Date: June 5th, 2025 8:03 PM Author: Milky Nighttime Hell (http://www.autoadmit.com/thread.php?thread_id=5733944&forum_id=2).#48990616) |
|
Date: June 5th, 2025 8:06 PM Author: cerebral slate stage useless brakes (http://www.autoadmit.com/thread.php?thread_id=5733944&forum_id=2).#48990623) |
|
Date: June 5th, 2025 8:03 PM Author: offensive persian corner 1) The plateau was well expected 2) We’ve barely begun to scratch the surface of what you can do and build on top of the LLMs 3) 2025 seems to me to be the first year people are actually adopting AI in their daily work life. Even if that only adds .5-1 percent to GDP growth that will compound massively.
I can’t really speak to development by the AI companies, but if they’ve peaked on scaling they can turn their attention to all sorts of other problems and avenues. |
|
Date: June 5th, 2025 8:12 PM Author: pea-brained plaza cr. It's not necessarily a bad thing for users. (http://www.autoadmit.com/thread.php?thread_id=5733944&forum_id=2).#48990635) |
|
Date: June 5th, 2025 8:17 PM Author: stirring alpha son of senegal The current preoccupation with LLMs is going to look like when people in the late 90s were convinced that file sharing was the future of all media. Otoh it’s clearly a watershed moment but otoh by far the most valuable breakthroughs that will occur will be replacing custom code of various large enterprise environment processes that will save billions upon billions. (http://www.autoadmit.com/thread.php?thread_id=5733944&forum_id=2).#48990656) |
|
Date: June 5th, 2025 8:28 PM Author: pea-brained plaza https://futurism.com/therapy-chatbot-addict-meth (http://www.autoadmit.com/thread.php?thread_id=5733944&forum_id=2).#48990694) |