The most prestigious law school admissions discussion board in the world.
|
 |
 
|
Perplexity AI is the current GOAT
Poast new message in this thread
Date: September 15th, 2025 3:14 AM Author: chartreuse vigorous location Holy shit. For $20 a month, Perplexity Lab is better than Super Heavy Grok ($300), ChatGPT Pro ($200) or Claude Max ($150?(I forget now)).
It’s truly incredible and on another level.
I was getting really frustrated with the AIs circulating my theory. They would find a weakness, patch it, and another would pop up. I was playing whack a mole.
Tested out Perplexity for free and was impressed with the research capabilities but then I tried out the lab function after I upgraded and it basically solved all the weaknesses in my theory the same night.
My theory is not only fully operational right now, but Perplexity has my Odds API key and is scraping live odds data and making real time predictions.
None of the other AIs can do this. They just don’t have the capability and they refuse to allow me to share my key with them anyway. Perplexity took my key and ran with it.
I’ll give you bros some predictions in a minute. |
 |
Date: September 15th, 2025 3:58 AM Author: Razzle-dazzle abnormal friendly grandma doctorate so is it better because it wins u sportbet more often? what bet u make (http://www.autoadmit.com/thread.php?thread_id=5775068&forum_id=2...id.#49270416) |
|
Date: September 15th, 2025 2:35 PM Author: chartreuse vigorous location I haven't made any bets yet. Perplexity helped me out of the whack-a-mole process trying to maximize conceptual and empirical maturity across multiple AIs.
I just finished pulling NFL stats from 2018-2023 so that Grok can help backtest against that data.
The backtesting is going to help us refine the theory further.
Once we have achieved max refinement we will then pivot to predicting live games. |
Date: September 15th, 2025 4:01 AM Author: apoplectic spectacular azn prole of coursse cause gorgeous indian made it not jew CIA mossad faggot (http://www.autoadmit.com/thread.php?thread_id=5775068&forum_id=2...id.#49270418) |
|
Date: September 15th, 2025 4:29 AM Author: Very tactful shitlib Now do their respective market shares
|
|
Date: September 15th, 2025 11:13 AM Author: mint garrison I just tried it with test legal research questions which GPT 5 pro passes. Perplexity failed. (http://www.autoadmit.com/thread.php?thread_id=5775068&forum_id=2...id.#49271059) |
|
Date: September 15th, 2025 2:36 PM Author: chartreuse vigorous location They each have their strengths and weaknesses. For my theory, Perplexity is crushing it. (http://www.autoadmit.com/thread.php?thread_id=5775068&forum_id=2...id.#49271786) |
|
Date: September 15th, 2025 11:14 AM Author: multi-colored swollen stage How many hours in the Vape Lounge this month brother? (http://www.autoadmit.com/thread.php?thread_id=5775068&forum_id=2...id.#49271063) |
|
Date: September 15th, 2025 2:36 PM Author: chartreuse vigorous location Way too many. I think I've averaged 4 hours of sleep over the last six weeks. (http://www.autoadmit.com/thread.php?thread_id=5775068&forum_id=2...id.#49271792) |
|
Date: September 15th, 2025 3:35 PM Author: multi-colored swollen stage Jfc man (http://www.autoadmit.com/thread.php?thread_id=5775068&forum_id=2...id.#49272056) |
|
Date: September 15th, 2025 2:52 PM Author: cocky theater What, if any of the commercial AI's like these make it easy to "train" a model on a subset of data, like NFL stats, etc? (http://www.autoadmit.com/thread.php?thread_id=5775068&forum_id=2...id.#49271850) |
|
Date: September 15th, 2025 3:37 PM Author: concupiscible striped hyena hominid Did it predict crawford beating canelo (http://www.autoadmit.com/thread.php?thread_id=5775068&forum_id=2...id.#49272060) |
|
Date: September 15th, 2025 4:31 PM Author: chartreuse vigorous location It advised taking Crawford at +140 or better and said Canelo was overpriced. (http://www.autoadmit.com/thread.php?thread_id=5775068&forum_id=2...id.#49272329) |
|
Date: September 15th, 2025 4:58 PM Author: concupiscible striped hyena hominid 180 i lost $600 listening to chandler (http://www.autoadmit.com/thread.php?thread_id=5775068&forum_id=2...id.#49272469) |
|
Date: September 15th, 2025 11:52 PM Author: chilmata ChatGPT:
6. TL;DR (Actionable Take) • Canelo –170/–180: Overpriced relative to RACEL v3.3 model → no value. • Crawford +130 to +145: • At +130: borderline, small edge only if you rate catchweight. • At +145: clear value bet (dog + variance tail). • Stake: Cap at 3–5% bankroll (half-Kelly preferred). • Operational Note: Track CLV—if Crawford’s price shortens into +120/+125 by close, you’ve captured positive CLV. |
|
Date: September 16th, 2025 10:32 AM Author: Check out my wishlist! (http://www.autoadmit.com/thread.php?thread_id=5775068&forum_id=2...id.#49274506) |
|
Date: September 16th, 2025 12:11 PM Author: chilmata On Saturday, my theory was on Canon Blueprint v3.3.
We are now on v5.3. |