MLST called this ‘the alexnet moment for program synthesis’, and they are no fools. ARC-AGI benchmark is near saturated. AGI by 2027 might not be a meme…
you have to be utterly clueless to think Google doesn't have anything better that is cooking. They are the OG RL model builders with AlphaGo, AlphaProof
I now need to think to myself about what defines AGI. I previously thought AGI needed to be agentic but o3 isn't so it can't be classed as AGI to me. If o3 is accessible via API as well as then being in cursor or another agent framework I think that will be AGI to me
You know they don't do enough AI research when they did not mention how well the o3 model did on the ARC Prize tests. OpenAI didn't openly say that o3 is AGI because of some legal agreements with Microsoft, and because of the impact this model will have (and certainly the next), but o3 is AGI.
@@anianait Sorry I think I came off a little rude in my last reply. I feel like SORA is a bad example since they explicitly said its a research preview, it was to show that they could do it. They also never explicitly said that SORA is coming in X month .They didn't even give a general ETA. For o3 though they have said o3-,mini is coming out in January and o3-full is coming soon after.
OpenAI destroys google in a big way. Yeah yeah you can generate some videos. Who cares. Sora will catch up eventually and this will plateu very quickly.
They did amazing these 12 days
Not really. Disappointed 4/5 the times.
Have realer expectations then
Still no dall-e
MLST called this ‘the alexnet moment for program synthesis’, and they are no fools. ARC-AGI benchmark is near saturated. AGI by 2027 might not be a meme…
It stops and thinks!!!!
LEARN A LESSON, HUMANS!
Hurry before it’s too late.
Are you kidding? This is your take on it? A video generator is nothing but peanuts compared to a much smarter model
you have to be utterly clueless to think Google doesn't have anything better that is cooking. They are the OG RL model builders with AlphaGo, AlphaProof
Gemini2.0 isn't a video generator ? For my personal workflow and usecases for LLMs I am much more enthusiastic about Gemini2.0
they only announce the much smarter model tho, they did not actually ship it.
@@zooq-aigoogel is cooked it's not cooking , openai just miles ahead of any of the competition rn lol .
Largely disappointed? you don't even know what you're talking about get her out of here
WHAT A TIME TO BE ALIVE !!!!!
Google crushed Sora
I now need to think to myself about what defines AGI. I previously thought AGI needed to be agentic but o3 isn't so it can't be classed as AGI to me. If o3 is accessible via API as well as then being in cursor or another agent framework I think that will be AGI to me
Sad when you have ladies who just read scripts and don't understand tech
I would like to see the top 2 AIs compete against each other on different kinds of tasks to see which is better and smarter overall.
Thats literally what benchmarks are
@@marwin4348 So which one is better?
I hope the Green new deal types are as vocal about the energy spent on AI compute as they are for carbon emissions from a lower class family sedan.
O3. Is phd holder worker that can be hired for 2k. Per month. It is expensive than developers from many places in Asia
Google beat them to the punch pretty much every single day of shipmas...
1000 dollars for a task? looks like developer jobs are back on the menu
Shitmas? 💩
😂 i heard That
You know they don't do enough AI research when they did not mention how well the o3 model did on the ARC Prize tests. OpenAI didn't openly say that o3 is AGI because of some legal agreements with Microsoft, and because of the impact this model will have (and certainly the next), but o3 is AGI.
You gotta have a lot of charisma to make you bad news sound like thank you.
agi is here
It's not agi
@@Isaac.Sheepadoodle it is incredibly close to AGI though. Q2 2025 = AGI // 2027 = ASI
@@ravencode5505 not even
No no no and no
@@jyro5924not even close...
AGI has reach 1% yet unfortunately capacity yet
Open AI pioneering this ship model is sth else, Everyday 20 minute.
>shipmas
>doesn't ship
When your product is so bad that the marketing team can't justify its release, and decides if we do it over time maybe the consumer just won't notice.
Its shows how arrogant humans are, they dont freaking know what actually happening.
Immortality coming
Exactly lol
for the rich
Imagine having no idea how Ai works and saying largely disappointing.. Like seriously have some perspective.
Largely disappointing??? 😂 This company just solved AGI half a decade early... Show some respect lady
Humans doomed
I told yall!!
Sadly this is not good news for Nvidia. O3 shows the training was done on far less hardware resources than O1
How did you make that conclusion?
It's the exact opposite. Are you trolling?
no details were released about the o3 hardware resources used
The advancements that were made from o1 to o3 were primarily in post-training, not pre-training (the GPU compute intensive part).
Is this AI generated?
You know something off on the ladies lol
Damn son 🙀
Brace yourselves.
Disappointed?
not released = doesn't exist
They said its coming in January, I don't know what you're smoking lmao
@jld-ni3vf like they said with Sora 1 year ago ... we believe it when our hands are on it
@@anianait Sorry I think I came off a little rude in my last reply. I feel like SORA is a bad example since they explicitly said its a research preview, it was to show that they could do it.
They also never explicitly said that SORA is coming in X month .They didn't even give a general ETA.
For o3 though they have said o3-,mini is coming out in January and o3-full is coming soon after.
I love AI omg. Can't wait for more
I LOVE AI TOO ITS SO MUCH BETTER THAN EVERYONE !
It will be already till we die lol
AGI almost here!
OpenAI destroys google in a big way. Yeah yeah you can generate some videos. Who cares. Sora will catch up eventually and this will plateu very quickly.
Its so over!
o1 prevew > o1
Quantum computers > "AGI"
Except quantum is for researchers and scientists.
Early computers were for researchers and scientists too, until it wasn't.