\
  The most prestigious law school admissions discussion board in the world.
BackRefresh Options Favorite

So OpenAI already lost? It's the Blockbuster of gpt?

Crazy how quickly that turned around
Pearl Lay
  12/03/25
...
Heady nowag
  12/04/25
...
Pearl Lay
  12/03/25
Early employees will still make millions though right? Askin...
Curious chapel
  12/03/25
"they expect one of us in the wreckage, brother"
Pearl Lay
  12/03/25
Google has access to better researchers and hardware and is ...
salmon cracking plaza
  12/03/25
Weird shtick. Even gemini 3 is still lagging behind
costumed smoky alpha base
  12/03/25
weird schtick, gemini 3 is widely recognized as best in clas...
Pearl Lay
  12/03/25
Gemini 3's memory/context awareness is pure shit
Heady nowag
  12/04/25
It's all the same crap
Vengeful Supple Gaping Public Bath
  03/29/26
[THIS THREAD SPONSORED BY ALPHABET INC.]
Odious violent wrinkle boltzmann
  12/03/25
no that would be an ethical conflict 😇
Pearl Lay
  12/03/25
...
hairraiser azure ratface
  12/03/25
MSFT lost the gen again 😔
Submissive pocket flask
  12/03/25
thoughts and prayers
Pearl Lay
  12/04/25
https://x.com/BenjaminDEKR/status/1996059933998285094
crawly khaki house partner
  12/03/25
...
Pearl Lay
  03/29/26
I prefer OpenAI's codex over claude now, noticing claude cod...
vivacious magenta nibblets
  03/29/26
its way better ime i dont even know how the two could be ...
Dun irradiated station
  03/29/26
codex 5.4 is miles better right now than claude code. It's s...
costumed smoky alpha base
  03/29/26
come on man no one expected them to be able to compete with ...
hairraiser azure ratface
  03/29/26


Poast new message in this thread



Reply Favorite

Date: December 3rd, 2025 8:55 AM
Author: Pearl Lay

Crazy how quickly that turned around

(http://www.autoadmit.com/thread.php?thread_id=5805605&forum_id=2\u0026mark_id=5310916#49479403)



Reply Favorite

Date: December 4th, 2025 12:05 AM
Author: Heady nowag



(http://www.autoadmit.com/thread.php?thread_id=5805605&forum_id=2\u0026mark_id=5310916#49482418)



Reply Favorite

Date: December 3rd, 2025 10:39 AM
Author: Pearl Lay



(http://www.autoadmit.com/thread.php?thread_id=5805605&forum_id=2\u0026mark_id=5310916#49479635)



Reply Favorite

Date: December 3rd, 2025 10:51 AM
Author: Curious chapel

Early employees will still make millions though right? Asking for a friend

(http://www.autoadmit.com/thread.php?thread_id=5805605&forum_id=2\u0026mark_id=5310916#49479655)



Reply Favorite

Date: December 3rd, 2025 11:58 PM
Author: Pearl Lay

"they expect one of us in the wreckage, brother"

(http://www.autoadmit.com/thread.php?thread_id=5805605&forum_id=2\u0026mark_id=5310916#49482403)



Reply Favorite

Date: December 3rd, 2025 11:00 AM
Author: salmon cracking plaza

Google has access to better researchers and hardware and is not nearly as constrained by money. They also have the ability to shove Gemini in all of their products and make it the obvious choice (since it’s clearly better than ChatGPT for most things). The first mover advantage doesn’t count for much when your competitor can churn out a better, cheaper product and put it in everyone’s browser and normal search.

Thankfully antitrust policy is weak and the better product from Google will be allowed to win.

(http://www.autoadmit.com/thread.php?thread_id=5805605&forum_id=2\u0026mark_id=5310916#49479671)



Reply Favorite

Date: December 3rd, 2025 12:43 PM
Author: costumed smoky alpha base

Weird shtick. Even gemini 3 is still lagging behind

(http://www.autoadmit.com/thread.php?thread_id=5805605&forum_id=2\u0026mark_id=5310916#49479975)



Reply Favorite

Date: December 3rd, 2025 11:59 PM
Author: Pearl Lay

weird schtick, gemini 3 is widely recognized as best in class and gemini 4 will be out before we know it

(http://www.autoadmit.com/thread.php?thread_id=5805605&forum_id=2\u0026mark_id=5310916#49482404)



Reply Favorite

Date: December 4th, 2025 12:05 AM
Author: Heady nowag

Gemini 3's memory/context awareness is pure shit

(http://www.autoadmit.com/thread.php?thread_id=5805605&forum_id=2\u0026mark_id=5310916#49482419)



Reply Favorite

Date: March 29th, 2026 9:26 AM
Author: Vengeful Supple Gaping Public Bath

It's all the same crap

(http://www.autoadmit.com/thread.php?thread_id=5805605&forum_id=2\u0026mark_id=5310916#49778066)



Reply Favorite

Date: December 3rd, 2025 12:25 PM
Author: Odious violent wrinkle boltzmann

[THIS THREAD SPONSORED BY ALPHABET INC.]

(http://www.autoadmit.com/thread.php?thread_id=5805605&forum_id=2\u0026mark_id=5310916#49479880)



Reply Favorite

Date: December 3rd, 2025 11:59 PM
Author: Pearl Lay

no that would be an ethical conflict 😇

(http://www.autoadmit.com/thread.php?thread_id=5805605&forum_id=2\u0026mark_id=5310916#49482405)



Reply Favorite

Date: December 3rd, 2025 12:44 PM
Author: hairraiser azure ratface



(http://www.autoadmit.com/thread.php?thread_id=5805605&forum_id=2\u0026mark_id=5310916#49479979)



Reply Favorite

Date: December 3rd, 2025 12:46 PM
Author: Submissive pocket flask

MSFT lost the gen again 😔

(http://www.autoadmit.com/thread.php?thread_id=5805605&forum_id=2\u0026mark_id=5310916#49479985)



Reply Favorite

Date: December 4th, 2025 12:00 AM
Author: Pearl Lay

thoughts and prayers

(http://www.autoadmit.com/thread.php?thread_id=5805605&forum_id=2\u0026mark_id=5310916#49482407)



Reply Favorite

Date: December 3rd, 2025 12:46 PM
Author: crawly khaki house partner

https://x.com/BenjaminDEKR/status/1996059933998285094

(http://www.autoadmit.com/thread.php?thread_id=5805605&forum_id=2\u0026mark_id=5310916#49479986)



Reply Favorite

Date: March 29th, 2026 7:39 AM
Author: Pearl Lay



(http://www.autoadmit.com/thread.php?thread_id=5805605&forum_id=2\u0026mark_id=5310916#49777979)



Reply Favorite

Date: March 29th, 2026 8:46 AM
Author: vivacious magenta nibblets

I prefer OpenAI's codex over claude now, noticing claude code starts to ignore instructions sometimes and does it own thing and you have to give it more detailed instructions.

(http://www.autoadmit.com/thread.php?thread_id=5805605&forum_id=2\u0026mark_id=5310916#49778040)



Reply Favorite

Date: March 29th, 2026 9:25 AM
Author: Dun irradiated station

its way better ime

i dont even know how the two could be compared

(http://www.autoadmit.com/thread.php?thread_id=5805605&forum_id=2\u0026mark_id=5310916#49778065)



Reply Favorite

Date: March 29th, 2026 9:53 AM
Author: costumed smoky alpha base

codex 5.4 is miles better right now than claude code. It's so surgical can get everything mostly right on one pass. in the metrics its not showing as being THAT much better but the amount better makes a difference plus there are probably things not being measured. The amount better on the AGI reasoning test, which is one of the bigger gaps, probably makes a big difference because it shows how good it is at solving novel problems adversarial to AI systems.

(http://www.autoadmit.com/thread.php?thread_id=5805605&forum_id=2\u0026mark_id=5310916#49778104)



Reply Favorite

Date: March 29th, 2026 9:25 AM
Author: hairraiser azure ratface

come on man no one expected them to be able to compete with Google

(http://www.autoadmit.com/thread.php?thread_id=5805605&forum_id=2\u0026mark_id=5310916#49778064)