O3 et O3 Mini : La Révélation d’OpenAI en 12 Jours
Bien sûr ! Veuillez fournir la transcription de la vidéo YouTube que vous souhaitez que je résume. Source : OpenAI | Date : 2024-12-20 19:35:52 | Durée : 00:22:05
"03:17: I am safe". Even the guy understands that his work is at risk with the tool that he is developing. Question is: What will happen to the researchers that have a MAME below score?
Emm so now we read that EPoch who provides Benchmark for Open AI was actually receiving fund from them 😂😂😂 OpenAI has only the word Open that is open..
There is absolutely no reason why this is needed😢 and why was there someone made un-alive after revealing that you are improperly confiscating copyrighted materials?
Damn I wish I was part of this frontier. I’m certain everyone who’s in this video is incredibly smart and they are probably getting paid loads of money
Great presentation. While Mark is clearly very, very smart, I couldn’t understand why he was enthusiastically nodding in response to everything being said. There wasn’t any need, and it didn’t add anything to the presentation. At one point, he nodded enthusiastically when incorrect figures were mentioned, then again nodded enthusiastically when the figures were corrected.
I have behaved like this is prepared presentations, and realise how fun or downright weird it might have appeared to the audience
It would be great to have more examples of problems solved in these benchmarks. I like the one with the different colored squares indicating boundary color and thickness, but without knowing what else is in these benchmarks, these results are difficult to evaluate.
Abonnez vous a Chat GPT4
Comments
"03:17: I am safe". Even the guy understands that his work is at risk with the tool that he is developing. Question is: What will happen to the researchers that have a MAME below score?
These Chinese guys leaked source codes to deepseek for sure.
to reverse the market they need to release o3 full very soon
Is this a real vid?
we are waiting
when will you actually release it?
mini
Hmm, very few new comments 🤔
Any benchmarks on nr of hallucinations for 1000 questions or something like that please?
I hope they improve Chat GPT for foreign languages too since i wanna learn better languages
Emm so now we read that EPoch who provides Benchmark for Open AI was actually receiving fund from them 😂😂😂
OpenAI has only the word Open that is open..
Hm make it where is can do complex scripts for studio lite a Roblox studio
WOOOOOOOOOW, you're able to write an API in Python! I-N-C-R-E-D-I-B-L-E! … said no one ever
When will it come out? I need o3 to study for my Differential Equations exam 🧐
There is absolutely no reason why this is needed😢 and why was there someone made un-alive after revealing that you are improperly confiscating copyrighted materials?
There is absolutely no reason why this is needed😢
That plug is looking very pullable right now
Been using o1 to its limits with engineering problems and it feels like its getting dumber?
How to access to O3 Model ?
Who are the guys presenting with sam
Damn I wish I was part of this frontier. I’m certain everyone who’s in this video is incredibly smart and they are probably getting paid loads of money
Great presentation. While Mark is clearly very, very smart, I couldn’t understand why he was enthusiastically nodding in response to everything being said. There wasn’t any need, and it didn’t add anything to the presentation. At one point, he nodded enthusiastically when incorrect figures were mentioned, then again nodded enthusiastically when the figures were corrected.
I have behaved like this is prepared presentations, and realise how fun or downright weird it might have appeared to the audience
It would be great to have more examples of problems solved in these benchmarks. I like the one with the different colored squares indicating boundary color and thickness, but without knowing what else is in these benchmarks, these results are difficult to evaluate.
12:50 its like chatgpt a at top left is that the coming model and also gpt logo is orange
Amazing how happy Sam is to usher in human obsolescence. He really hates humanity.
The "firing of Jira" ads is nice!
Why do I have the impression that chatgpt is less accurate than before?
Thank you for killing my job lmao. Yay.
Amazing progress !!! What kind of cost per month would o3 cost ?
Chatgpt can solve phd level que?😂 it couldn't even solve my high school's simple algebra equation.
Can't take all this uptalk seriously
When are you going on the stock market?
Thank you❤