@CyrilSirgi يا خيي مع ربن بس فسرلي ليش بثورة ١٧ تشرين نزل الجيش كسّر العالم اللي عم بتطالب بحقوقا، واليوم ضربة كف ما صار بوج حزب الله وبعدا الطرقات كلا مسكرا!!! بدنا هيبة دولة خلي كلاب حزب الله يضربو مليار حساب قبل ما يسكرو طرقات ويقوسو شمال يمين

@TheresaSalam @OmarElYafi العهد فشل بالديقة اللي ما نزل الجيش يدعس بنيعن وقتا فاتو من جمعتين عالمناطق المسيحية/السنية وفشل وقتا ترك الحدود مفتوحة ليتسلح حزب الله وفشل وقتا ما كب كل ظالم وفاسد بالحبس وخلصت بقا ندافع عنو ونطبطبلو!

@ai_for_success definetly claude

@kimmonismus will it be available for consumers too

@slow_developer The first one, even Arc AGI makers are going the road of Neurosymbolic ai which is what all reseaechers who are against llms go

@BrentM_SpaceX hope it'll be worth it and not just hype, literally this technology is last hope for humanity it'll be able to cure all diseases

@kimmonismus Claude 4 will be literally huge🔥

@stanfordnlp @AlexisNeuhaus @polynoamial 🎯🎯

@NewSamawal they wanna kill the infidels, because afghan christians wouldn't do that shit

@Osint613 Basically natanyaho is deliberatly making the war last long to be able to strike iran nuclear, similar to how he made the war last this long to erase nasralla and syria and sinwar and haniye and eliminate all of iran's air defense😉

@polynoamial @rm_rafailov @stanfordnlp @OpenAI @ilyasut sorry but its not common sense to think that keep scaling TTC on GPT-4.5 will lead to AGI🤦🏻‍♂️

@ai_for_success Grok 3 the base model is better than Grok 2? of course it should be but why add reasoning if pretraining didn't hit a wall?🤔 trained on 100k h100 the base model itself Without reasoning should be on par with full o3

@Tilo18999694 @OmarElYafi مع الفرعون أخناتون

@kaysorin Do you think there needs to br an architectural breakthrough or just keep scaling TTC on GPT-4.5 will lead to AGI?

@matiroy I'm just following the trend of AGI and hoping it'll be real and cure diseases and help humanity, i just don't want it to be the next theranos..

@xuukyu @iScienceLuvr when it generates multiple responses it generates them internally and at the end it submits one response that it thinks its correct? or it sumbits multiple response and one of them will be correct meaning it one shots the correct response?

@polynoamial Just a question, openai said the last base model for scaling pretraining is 4.5, does this mean they will just keep scaling TTC on 4.5? will this lead to AGI or there needs to be a breakthrough in architecture?

@AudioBooksRU @mark_k @OpenAI They said 4.5 will be their last base model, so now they will keep scaling TTC on the same 4.5? do you think that will lead to AGI or there needs to be a breakthrough?

@mikeknoop @ecsquendor @MLStreetTalk Thank god for Arc AGI challenge without it we would've thought a model that still sometimes doesn't know how many "R" in strawberry is AGI, they're just selling us dreams of AGI 2 years away

@qinzytech i don't understand how they still didn't understand that current archutecture cannot lead to AGI🤦🏻‍♂️ Pretraining hit a wall, its clear as ice