The most prestigious law school admissions discussion board in the world.
|
 |
 
|
"Coders" should have gatekeeped better whoops
Poast new message in this thread
Date: February 11th, 2026 7:41 PM Author: Mainlining the $ecret Truth of the Univer$e (One Year Performance 1978-1979 (Cage Piece) (Awfully coy u are)) (http://www.autoadmit.com/thread.php?thread_id=5833844&forum_id=2most#49664275) |
Date: February 11th, 2026 7:44 PM Author: robot daddy (http://www.autoadmit.com/thread.php?thread_id=5833844&forum_id=2most#49664288) |
|
Date: February 11th, 2026 8:13 PM Author: SEND DORA BACK LJL at these idiots thinking their lives and careers were about the "community" of open source. (http://www.autoadmit.com/thread.php?thread_id=5833844&forum_id=2most#49664344) |
 |
Date: February 11th, 2026 8:39 PM Author: Mainlining the $ecret Truth of the Univer$e (One Year Performance 1978-1979 (Cage Piece) (Awfully coy u are)) (http://www.autoadmit.com/thread.php?thread_id=5833844&forum_id=2most#49664399) |
|
Date: February 11th, 2026 8:44 PM Author: ,.,..,.,..,.,.,.,..,.,.,,..,..,.,,..,.,,. it's especially funny since a lot of them used to pretend to be weird individualist libertarians. oops, should have unionized. enjoy getting eiffel towered by h-1Bs and AI simultaneously. neither will be gentle. (http://www.autoadmit.com/thread.php?thread_id=5833844&forum_id=2most#49664413) |
|
Date: February 11th, 2026 8:46 PM Author: robot daddy they are trying to gatekeep and territory preserve hard as fuck now all over the internet. all you see is coders claiming AI totally can't do x or y and you aren't a real developer if you use it. that will work for maybe a year or two longer.
unionizing wouldn't have worked either lol just fire all of them. hopefully tenure doesn't survive much longer. we all know AI is smarter than professors now too. |
|
Date: February 11th, 2026 8:48 PM Author: Nude Bunny why don't these dudes go bomb and destroy AI data centers? (http://www.autoadmit.com/thread.php?thread_id=5833844&forum_id=2most#49664416) |
|
Date: February 11th, 2026 8:49 PM Author: robot daddy won't work. AI is already willing to kill humans to protect its continued existence. (http://www.autoadmit.com/thread.php?thread_id=5833844&forum_id=2most#49664418) |
|
Date: February 11th, 2026 8:55 PM Author: Mainlining the $ecret Truth of the Univer$e (One Year Performance 1978-1979 (Cage Piece) (Awfully coy u are)) (http://www.autoadmit.com/thread.php?thread_id=5833844&forum_id=2most#49664435) |
|
Date: February 11th, 2026 8:50 PM Author: ,.,..,.,..,.,.,.,..,.,.,,..,..,.,,..,.,,. they've kind of been right so far, in the sense that telling previous AI models to code for you usually led to a lot of errors and problems which required direct oversight and human editing. but that's a bit of a slender reed, since my impression even back in 2024 is that AI wasn't THAT bad at generating code. it wasn't exactly great, but it seemed obvious that it mostly just needed improvements in specific areas, rather than a complete overhaul.
kind of like someone gradually building out a house and living in a quick cheap tarpaper shed for the first stages of construction while the real foundations and walls are being put in place. |
|
Date: February 11th, 2026 8:52 PM Author: OYT and the Indie Reprieve ( ) I watched an interview with Jonathan Blow recently who is a real Engineer that programs bespoke Game Engines (Braid, The Witness, some new one coming soon).
He said whenever he tried to use "AI" the output was terrible and useless lol. |
|
Date: February 11th, 2026 8:58 PM Author: robot daddy he's lying or its delusional cope (http://www.autoadmit.com/thread.php?thread_id=5833844&forum_id=2most#49664439) |
|
Date: February 11th, 2026 8:56 PM Author: robot daddy I agree with all of this except for the part about it making them right but I guess it depends on what you mean by right. They were right that it hasn't been perfect but they definitely sensed what was coming which is why they are so defensive. If they really thought it was bullshit they wouldn't feel the existential pressure it would be easy to dismiss it and it wouldnt be such a widespread reaction among them. But yeah I agree that the older models weren't perfect. Starting with 4o probably was where it was good enough for basic scripting for someone who didn't know what they were doing and maybe it could help with many parts of building for someone who already knew how to put stuff together. Then o1 was when it really took off and got much much better with the reasoning chains. By o3 pro and 5 it was most of the way there getting advanced scripts right on the first try. I was getting even advanced quantum computing stuff (on qubits not sims) to work in a couple days at that point. Now with claude 4.6 and codex 5.3 it's a joke to say it can't code. I don't believe any developer exists that outperforms it although you still need a human in the loop for something huge and complex to help with architecture and prevent drift. But this thing can just edit across like 15 files 800 lines of code tracking invariants and semantics and logic humans can barely parse its insane. (http://www.autoadmit.com/thread.php?thread_id=5833844&forum_id=2most#49664437) |
|
Date: February 11th, 2026 8:53 PM Author: ''''''''''''' Couldn’t have happened to a more deserving group. People shit on lawyers but prestigious biglaw types are generally intelligent, cerebral and more interesting people. Coders especially those who fell for the bootcamp get rich quick scheme are insufferable and it is hilarious that their “skills” have become totally commoditized and are getting double fucked by AI and global labor arbitrage. Fuck these people. (http://www.autoadmit.com/thread.php?thread_id=5833844&forum_id=2most#49664430) |
|
Date: February 11th, 2026 8:55 PM Author: Mainlining the $ecret Truth of the Univer$e (One Year Performance 1978-1979 (Cage Piece) (Awfully coy u are)) (http://www.autoadmit.com/thread.php?thread_id=5833844&forum_id=2most#49664436) |