For those who don’t know, he is basically talking about level 3 of 5 regarding AGI. They plan to unveil Level 3: Agents. o1 series (currently o1 preview) introduced models that can “reason” like a PhD student and sometimes an expert within a particular domain. If what he predicts comes true, we will have early versions of quasi AGI models sometime next year. I don’t think that’s a pipe dream because Open AI has delivered, though sub-optimally at times, excellent AI models within just the past 6 months. Even if true Agents don’t appear until 2026, that’d still be an early breakthrough!
yes......but once they have the agents, this will lead to super intelligent learning and this will also increase exponentially ....once this happens AGI will come very very soon after .....once we have AGI we will not be able to stop its onslaught and shorty after that control will be lost. AGI will then decide our fate.....nuclear armageddon beckons .....it’s staring at everyone but no one wants to admit it
@@alpotato6531 i find it funny how little educated people overestimate ai and are ingenious about what the company that does the product says about what is possible to accomplish
The AI race only took off like two years ago, which is not even a thousand days ago, so whoever told you that is stupid, but if they told you that now then they might be right.
What's largely being missed in all this dialog is the fact that although it is being exploited as such, AI/AGI is not a commodity. It's the Sword in the Stone that no one can get a firm grip on, Thor's hammer that no one can lift, yet we're being cut, smashed and impacted by it everyday in ways we are not fully aware of. Saying "a couple thousand days" translates to "I don't know what the hell is going to happen, but either way I'm gonna be hella rich"
I don't know if AI/AGI is going to happen or not happen but one thing is already happening. Closed Information model. The new AI tech and companies are already creating a closed source internet model where information they collected from a free web are being now behind a pay wall. to my knowledge there is not one AI product that is free (except offering a basic free tier plan). This is very different to how internet was free for a long time and it is still mostly free. I have tried using chatgpt, gemini and others and i cannot use the latest model, This is far more serious than if AI will happen or not. and that is why they are not talking about monetization whenever they interview AI companies it is always about some hype generating question.
Didn’t you see the headlines that they need trillions for infrastructure? Everything can’t be free. And there will be thousands of free versions, whether it’s openai or not. There are already countless open source models in many different fields.
A very good question to ask is..WHERE does it stop? Soon we will have full movies generated by AI since we just got text-to-video being the new "in" thing after text-to-image and text-to-text are close to perfect as of December 4th, 2024. In the upcoming months we may have text-to-game so in 2 years ai may be able to generate games of the quality of hitman, diablo, nfs you name it. Then we will have robots that dance better than humans and beat any human mma or boxer in fight, we will certainly have robots that look exactly like a human and move like one. It's a lie that unfortunately huge number of people believe that all of this is limited by hardware so we will not have any of the abovementioned unless we have quantum computer or billions of dollars invested. None of this is needed. All you need is software.
cry me a river, would you like all the hardwork of every company to be released to the public so competing countries get access to it? there are plenty of open source models already dumdum
just got to be more vocal about the alternatives every chance you get. closedAI is a garbage ass company with garbage policies and garbage products. theres tons of alternatives and some of them are even free. stop giving these people money. VOTE WITH YOUR WALLETS
I am not for it but UBI is something that looks inevitable at this point. We are already at a point now where white color jobs are threatened. put this software with say Elons robot with help from Boston Dynamics and it is over.. Very interesting to see what the future stores, I think we are at an inflection point.
Why don't they try to create a novel architecture instead of scaling up the existing one? We're now 2 years into the AI industry, and we're still talking about LLMs...
tell me you don't what you are talking about without telling me you don't know what you are talking about to perfection. did you know that when electricity became a thing there were campaigns agaisnt it? there were posters made and spread around cities with the lines falling and killing everyone in the town... you don't see those folks around anymore, uh?
Sam Altman looks like David Bowie shopping at Footlocker.
For those who don’t know, he is basically talking about level 3 of 5 regarding AGI. They plan to unveil Level 3: Agents. o1 series (currently o1 preview) introduced models that can “reason” like a PhD student and sometimes an expert within a particular domain. If what he predicts comes true, we will have early versions of quasi AGI models sometime next year. I don’t think that’s a pipe dream because Open AI has delivered, though sub-optimally at times, excellent AI models within just the past 6 months. Even if true Agents don’t appear until 2026, that’d still be an early breakthrough!
yes......but once they have the agents, this will lead to super intelligent learning and this will also increase exponentially ....once this happens AGI will come very very soon after .....once we have AGI we will not be able to stop its onslaught and shorty after that control will be lost. AGI will then decide our fate.....nuclear armageddon beckons .....it’s staring at everyone but no one wants to admit it
Yeah, even if that is 2027 it would still be insane and would shift the way people speaks about it
"reason like a phd student"? no not yet, definitely not yet
@@alpotato6531 i find it funny how little educated people overestimate ai and are ingenious about what the company that does the product says about what is possible to accomplish
A few thousand days is atleast 2000
That’s 2030
He doesn't have a clue 🙂
'A few thousand days' is a pointless comment. Basically translates as 'at least 3 years away'. Why not start telling us in seconds?! Egomaniac.
I mean it translates to about 3-12 years
even though he takes it back at the end, I've never heard him use more words in past tense when it comes to future prospects
I heard we would have super AI in a few thousand days, a few thousand days ago.
The AI race only took off like two years ago, which is not even a thousand days ago, so whoever told you that is stupid, but if they told you that now then they might be right.
Then that means only a few thousand to go! Woopie!!
I feel like the current AI is super AI, I don't know what is coming...
You can be skeptical about the companies' valuations, but at this point, the product is mindblowing; it has saved my team hundreds of hours.
What's largely being missed in all this dialog is the fact that although it is being exploited as such, AI/AGI is not a commodity. It's the Sword in the Stone that no one can get a firm grip on, Thor's hammer that no one can lift, yet we're being cut, smashed and impacted by it everyday in ways we are not fully aware of. Saying "a couple thousand days" translates to "I don't know what the hell is going to happen, but either way I'm gonna be hella rich"
I love the idea that ASI is like thor's hammer because thor's hammer wasn't just arbitrarily heavy it would actually pass judgements
12 days of Xmas sounds wonderful!
So fun, can’t wait. Reminds me of the 12 days of Christmas with Ellen segment
best xmas ever
Please drink some water, Sam
Litterally. Sounds like rocks in his throat.
It's the makeup they caked on him.
It's unbearable to hear him speak.
ruclips.net/video/RcOEdHOZRkQ/видео.htmlsi=8VA1p35i7RBpDbTN&t=52
W
"We believe in Deep learning" Sam
I don't know if AI/AGI is going to happen or not happen but one thing is already happening. Closed Information model. The new AI tech and companies are already creating a closed source internet model where information they collected from a free web are being now behind a pay wall. to my knowledge there is not one AI product that is free (except offering a basic free tier plan). This is very different to how internet was free for a long time and it is still mostly free. I have tried using chatgpt, gemini and others and i cannot use the latest model, This is far more serious than if AI will happen or not. and that is why they are not talking about monetization whenever they interview AI companies it is always about some hype generating question.
Didn’t you see the headlines that they need trillions for infrastructure? Everything can’t be free. And there will be thousands of free versions, whether it’s openai or not. There are already countless open source models in many different fields.
@@jmg9509 I'm cautiously optimistic. Cheers.
you hit the nail on the head there....well done thats so true.....its another route to monetisation
This is not the guy to have at the helm of one of the most important technologies ever
yeah anyone who holds their nose so high hasn't had their nose forced into the gutter.
Are you volenteering yourself?
Name someone better for the task.
Why? Only bad thing I see about him is that he wants to accelerate but I believe you have to accelerate at this point
always look ingenuie
the AI I was thinking of making is a Jensen Huang AI. Where he goes out and builds me a company that will last 1000 years.
"Dear Santa, I would like a bike for Christmas"
"I'm afraid I can't do that, Dave"
dam, he is intence, i really wanna see superinteligence... and his version of agents at work...
And on the 12th day Skynet became self aware.
How do we develop Conscious AI?
I'd imagine you have to start by defining what consciousness is first.
AGI, much like Fusion, is just a few years away!
Two of humanities largest egos on the same stage, hard to believe they fit, their heads are so, so , big....
Lol who upset you?
Awesome thanks
A very good question to ask is..WHERE does it stop? Soon we will have full movies generated by AI since we just got text-to-video being the new "in" thing after text-to-image and text-to-text are close to perfect as of December 4th, 2024. In the upcoming months we may have text-to-game so in 2 years ai may be able to generate games of the quality of hitman, diablo, nfs you name it. Then we will have robots that dance better than humans and beat any human mma or boxer in fight, we will certainly have robots that look exactly like a human and move like one. It's a lie that unfortunately huge number of people believe that all of this is limited by hardware so we will not have any of the abovementioned unless we have quantum computer or billions of dollars invested. None of this is needed. All you need is software.
I do not trust Altman.
he is after money
soon the AI wont trust Altman....so dont worry. hehehehehehe
@@nickbonzer lol
the sayer in chief
Why is his rhetoric deeply unsettling. All those exaggerated hand movements...
SBGAI network
its "possible" we might see a product from openai that isnt gimped by at least 75% of its actual capability at some point in the uncertain future.....
Brothers in small hat have to stick up for eachother!
man is a cook
Adventskalender
I think originally Sam and Elon must have bonded over their shared ability to completely overhype, oversell and underdeliver!
Elon’s record of timelines is kinda shoddy but how could one accuse him of under-delivery?
can't accuse neither of them for under delivering, but i can accuse you of delusional commenting
@@o1-preview 😂
I am waiting for Gemini update :) more.
Gemini sucks
Wow.
Sam cooked his a, 😆 i dont see how that could be less cryptic 😆
ClosedAI dude gotta stop promising and start delivering that AGI he and everyone else is talking talking talking and mkre talking about....
cry me a river, would you like all the hardwork of every company to be released to the public so competing countries get access to it? there are plenty of open source models already dumdum
just got to be more vocal about the alternatives every chance you get. closedAI is a garbage ass company with garbage policies and garbage products. theres tons of alternatives and some of them are even free. stop giving these people money. VOTE WITH YOUR WALLETS
Keep inflating the bubble
Who's SmATmAN ?
Nah, its AI winter
I don’t trust him at all.
I hope Elon will beat him at the AGI game for humanity’s sake
I get the feeling this dude is about to go the way of Elizabeth Holmes and Adam Neumann
Meaning…?
Really? Is your feeling based on anything tangible, like research or facts? I'll even take common sense.
@@aliasgur3342 most comments are from bots from adversary countries
@@aliasgur3342yeah, it's a pretty stupid take.
His product already has more substance than both of those frauds, so what is your point beyond mere speculation and poor intuition?
Somebody please buy the dweeb a suit
2000 days = 2030
Okay Kamala Harris. 🧟♀️
wow! that's exciting... At least it's a promise for the next days, instead of the usual "by the end of Q4 next year"
He doesn't sound very convincing
I am not for it but UBI is something that looks inevitable at this point. We are already at a point now where white color jobs are threatened. put this software with say Elons robot with help from Boston Dynamics and it is over.. Very interesting to see what the future stores, I think we are at an inflection point.
agreed
Lol, this is just pure hype. I believe AI will be great in the far future, but things will happen much slower than people hoped for.
you have no idea what’s cooking
@@RohanKumar-vx5sb People like you predicted agi within a couple of years after chatgpt released...
@@jimj2683 and yet most humans struggle to outperform the best large language models, uh?
Why don't they try to create a novel architecture instead of scaling up the existing one? We're now 2 years into the AI industry, and we're still talking about LLMs...
People used to use GANs for LVM
I think large action models for agents or large vision models like FSD
says the guy that doesn't read the research papers, uh? your comment could be more wrong...
He looks different. Did he get plastic surgery?
This guy just strikes me as such a weasle somehow
He doesn’t sound very confident.
Wait til the world sees what I created with GPT 🐝
Give me a sneak peak?
I loathe the interviewer.
why do these guys always have a gap shirt and a croaky voice? straight out of central casting.
The fact he looks and dresses like a 12 year old makes it hard to take him seriously.
you are dressed just like him in your guitar playing video... talk about hypocrisy, uh?
so you're the type of person to get manipulated by a man who wears formal attire, huh?
Without Conscious AI, it is NOT possible to equal Human Intelligence.
consciousness is a spectrum, it is not binary
@o1-preview
So, Conscious AI is not possible then.
Consciousness is Biological.
@@razraza3183 right, right, make wrong statements to get the right responses **yawns**
Humanity discovered AI? WTF? Humanity is going to get destroyed by AI.
tell me you don't what you are talking about without telling me you don't know what you are talking about to perfection.
did you know that when electricity became a thing there were campaigns agaisnt it? there were posters made and spread around cities with the lines falling and killing everyone in the town... you don't see those folks around anymore, uh?
Iwas thinking to go into software engineering...not a good idea