The most prestigious law school admissions discussion board in the world.
|
 |
 
|
8/15/25 AI thread
Poast new message in this thread
Date: August 15th, 2025 11:20 AM Author: lavender rough-skinned corn cake gpt-5 beats pokemon gen 1 3x as fast as o3
https://www.reddit.com/r/singularity/comments/1mq2irv/gpt5_just_finished_pokemon_red/
google releases 270m parameter gemma 3 local model. people claim it is fast and produces good results on a phone
https://x.com/fchollet/status/1956059444523286870 https://x.com/code_star/status/1956033343465906379
gpt5-reasoning gets perfect scores on some medical test. hard to know what this means if anything
https://x.com/omarsar0/status/1956003145349521780
apparently closed-source models are much, much more efficient with token output than open source models are (since they're being engineered to be that way). pretty interesting. means that open source models will eventually be much more efficient, which is what people running local models want anyway
https://x.com/NousResearch/status/1956090990005248341
free instruction on building deep research agent architectures
https://x.com/LangChainAI/status/1956027411302375631 https://x.com/hwchase17/status/1956036358709108979
"agentic RAG" system that is claimed to be able to reason and decision-tree its way through prompts requesting local data. it "learns" from patterns in your queries and only chunks data at query time. this is actually really interesting. tsinah you should check this out
https://x.com/philipvollet/status/1955945448860008655 |
Date: August 15th, 2025 3:24 PM Author: lavender rough-skinned corn cake these people claim that you can recover pre-LORA fine-tuning weights of models
pretty crazy if true
https://x.com/jxmnop/status/1956382800707240297 |
|
Date: August 15th, 2025 3:25 PM Author: lavender rough-skinned corn cake BREAKING: Sam Altman just revealed who GPT-5 is really built for:
> India is our 2nd largest market > It may become our LARGEST > We’ve taken a lot of feedback from users in india what they’d like from us… > more affordable access > and we’ve put that in GPT-5
Now it all makes sense..
https://x.com/ns123abc/status/1956404043783270853
It's Over |
 |
Date: August 15th, 2025 3:26 PM Author: passionate sapphire senate roast beef congrats tommy! (http://www.autoadmit.com/thread.php?thread_id=5762709&forum_id=2E#49188259) |
|
Date: August 16th, 2025 11:10 AM Author: lavender rough-skinned corn cake These are two excellent poasts on forms of continual learning and the possible ways to use RL to continue to scale models. I agree with almost everything this guy says. I am starting to change my mind about the impact of large frontier model training/inference infrastructure vs many smaller local models. If it does turn out to be cr to train models with RL for different specialized agentic tasks/fields, having the compute to generate vast amounts of synthetic data is going to be really important, because the data will be so large and complex and the time needed to train on these tasks will be so long
https://www.interconnects.ai/p/contra-dwarkesh-on-continual-learning
https://www.interconnects.ai/p/what-comes-next-with-reinforcement
This guy is mfcr btw:
https://x.com/natolambert/status/1956394775705239726
Good article with various cr and interesting thoughts about what exactly LLMs are missing and how we might get around their limitations
https://secondthoughts.ai/p/thoughts-about-agi-and-gpt-5?triedRedirect=true
OpenAI now making GPT-5 "friendlier" because Redditors had a lot of tantrums. Lol
https://x.com/OpenAI/status/1956461718097494196
Anthropic giving its models the ability to cut worthless problem/mentally ill users off. They're giving the excuse of "model welfare" and people are actually buying it. Lmao
https://x.com/AnthropicAI/status/1956441209964310583 |
|
Date: August 18th, 2025 1:54 PM Author: important internet hot takes https://academy.langchain.com/collections
i think i already linked these above but highly recommend them as a free resource for tinkering around with setting up your own agent architectures |
|
Date: August 18th, 2025 2:19 PM Author: important internet hot takes Teknium (e/λ)
@Teknium1 The future looks like it may just be a billion RL environments
https://x.com/Teknium1/status/1957257677249319422
i agree! in fact, that is what the physical world already is
this sounds really fun. i'm gonna apply to some of these positions |
|
Date: August 18th, 2025 3:25 PM Author: important internet hot takes https://research.nvidia.com/labs/adlr/NVIDIA-Nemotron-Nano-2/
nvidia just released an open-source model that they claim performs better than qwen-3. notably, they also released all their pretraining data too. which is pretty wild |
|
Date: August 19th, 2025 8:24 AM Author: important internet hot takes https://x.com/TheZvi/status/1957438140714504696
good post explaining misconceptions about GPT-5 and progress on models in general |
|
Date: August 19th, 2025 3:44 PM Author: important internet hot takes https://x.com/kyliebytes/status/1957849347934286203
let's check in with Informed Investors Rationally Distributing Capital |
|
Date: August 19th, 2025 3:45 PM Author: important internet hot takes https://www.wired.com/story/openai-valuation-500-billion-skepticism/
grok how do i "short" this somehow |
|
Date: August 19th, 2025 3:46 PM Author: KarIstack (Retired) (🧐) a softbank investment is the equivalent of a jim cramer prediction or rudolph observation. they are fucking done here (http://www.autoadmit.com/thread.php?thread_id=5762709&forum_id=2E#49197942) |
|
Date: August 19th, 2025 3:49 PM Author: important internet hot takes https://x.com/sama/status/1957849495733166587
these guys are Finished imho |
|
Date: August 19th, 2025 3:51 PM Author: KarIstack (Retired) (🧐) just think of all the subsaharan africans who will subscribe to chadgbt for $5/mo, brother. (http://www.autoadmit.com/thread.php?thread_id=5762709&forum_id=2E#49197959) |