Unprecedented times. This is just the beginning. The application I am looking forward to most (and which I believe has the potential to benefit every individual on this planet) is a personal Assistant/Tutor/Therapist/Friend. Her we come...
Singularity is Near, 2005, Greg mentions it at 35:25 a great book for those who haven't read it by Ray Kurzwell. It lays out what is happening currently and where it leads to.
I love that comment that when scaling benefits start to decline that will be the great moment of opportunity. Just a passionate curiosity to what to dig into solving newer harder problems. Go Greg I'm sure the time will come but at the moment we're all enjoying the ride as the curve keeps on going !
I asked conversational AIs to timestamp this video. And, here are the results. Result 1 (somewhat accurate): 0:00 - Introduction 0:51 - Greg talks about similarities and differences between Stripe and OpenAI 2:01 - Discussing key breakthroughs like GPT-3 and DALL-E that showed the potential of scaling up models 5:51 - Early examples showing capabilities like generating code 9:10 - The journey from GPT-2 to GPT-3, need for conviction as compute costs grew 12:12 - Stories from the Dota 2 days 15:16 - Thoughts on the book "Life Cycle of Software Objects" 18:09 - Greg's early confidence in capabilities of neural nets 19:15 - Were there moments of doubt? Iterative nature of progress. 22:04 - Being excited but also cautious about powerful AI systems 24:12 - The future of AI this decade as the tech proliferates everywhere 27:27 - The roadmap to AGI and the consistent exponential progress 31:11 - Concerns over hitting limits like the end of Moore's Law 34:50 - AI slowdown analogous to other areas of science? Paradigm shifts. 37:47 - Biggest concerns about AI, different issues arising as systems get more capable 42:10 - Proliferation of AI capabilities and systems. Balance between massively scaled and more distributed systems. 45:21 - end - Closing thoughts Credit: Claude (using transcript) - claude.ai/chat Result 2 (very much wrong): 00:00 - Introduction: Greg Brockman and Alexandr Wang 01:15 - The role and impact of foundation models like GPT-3 and DALL·E 2 05:10 - The challenges and opportunities of scaling and deploying foundation models 10:20 - The vision and mission of OpenAI and its products and services 15:30 - The trade-offs and considerations of building general and specialized models 20:40 - The future of language models and natural language understanding 25:50 - The importance and methods of ensuring safety and alignment of AI systems 31:00 - The potential and limitations of multimodal and cross-modal models 36:10 - The current and emerging applications and use cases of foundation models 41:20 - The ethical and social implications and responsibilities of AI research and development 45:40 - Closing remarks and acknowledgements Credit: Bing Copilot (using youtube link) - copilot.microsoft.com/ A fun exercise. I had to do multiple iterations to get the above results. However, if someone is interested to reverse engineer the prompt, pls do.
its crazy that after two weeks this video has under 300 likes and 22 comments. Future humans will wonder why no one was listening to this new science. Future AIs will look back at the birth of their species, at their creators and become amazed that no one was listening.
Most people are drones. Wake up, eat crap food, watch crap entertainment (sports, drama shows, news, all of it is crap), work their repetitive low skill jobs, eat more crap, watch more crap, go to sleep while the elites who run the world separate themselves more and more from the common man
In some ways I’m a little sad over the way this may make certain things that give people fulfillment obsolete, but it is interesting to watch AI improve so much
Why do you think that? Or perhaps what things are you thinking of? The way I see it, the things that people do because they have to might go away, but the things that bring fulfillment still will. There is not a human alive that can beat the best chess AI more than a couple times out of 100, if that. And yet there are more people playing chess than ever, right? The things we enjoy will still be enjoyable. But the drudgery of doing it because we have to may go away. Now.. that may also mean we have to find new means of making money and distributing it.. but that doesn't make the activities themselves obsolete. Art will still be art. Music by humans will absolutely still be appreciated, even if a robot eventually becomes more technically proficient, right? I think. Maybe I'm just not thinking of an example! Edit: hmm. As I'm thinking about it more. Humans do get a lot of personal value out of doing their work. So that could definitely be an aspect that gets displaced. Where it's not the activity itself (putting a thing on a thingamajig) but what it means and the social context of why that activity is meaningful. And losing that could definitely be an issue. Still, I'd contend that the challenge shouldn't be to stop the progress, but to get at why those activities give people value (meaning, social status, stability, etc) and then find other ways to provide that. Which maybe means we need to take a robot or AI tax or something, but doesn't seem infeasible at all.
The first internet connection happened in 1969. By 1989 this technology was already being used by the largest businesses and banks, by 1999 the Internet had reached most businesses and started exploding to home users, and by 2009 we had most of the modern infrustructure for personal use (including the smart phone). So, do you think 2012 was like 1969, which means we'll have GPT20 in our pockets in 30 years?
The word "like" is used 244 times in this discussion (approximately 5.3 times per minute). "Sort of" is said 104 times, including 15 instances of "sort of like".
Great insights from both Greg and Alexander being in the AI tech industry. AIs will definitely have a boost given that chatbots are now able to solve or able to answer mathematical and programming language. There's still improvements to be made though on all models even image generators like Bluewillow. Can't wait to see different models scale up to cater big studies/fields.
It seems that tooling and help for data analysis is what is keep to continuously develop and nurture in these teams. Time to train retrospection models.
kind of humbling to hear how even the best are challenged in "understanding" what these magic machines really do. Stoked to see what we'll learn in the area of "explainability" but even if that doesn't advance much, I'm stoked to see which assist functions will enter everyone's toolbox because of this tech. And yes, pls hijack the term web3 for this stuff. 🤣
'Our charter definition is outperform humans at most economically valuable work' - I am not entirely convinced that their stated desire for AI to have a positive impact on people really sits well with their charter.
To everyone asking about… where are the comments?? Where are the likes etc…. I’ll say it simply. People are scared and people are confused. After using ChatGPT for over a week or so, I’ve noticed very clearly that people are having a hard time accepting the potential of what’s happening. People are definitely somewhat ‘averse’ to this entire thing. I understand why, and we’ll just have to see how this all plays out as we go. My advice is to learn more about logic, considering AI uses multiple forms of logic and can process and compute in ways most people don’t realize are even possible. Learn linguistics and familiar yourself with the deeper workings of the mind, emotions and nervous system. It’s very important.
I totally aggree with you. I notice that I know that that thing can help me immensely in my job but at the same time I'm scared at what it actually means for me and humanity in the future . The fact that I'm a bit anxious and scared about it prevents me from getting to know it better. It's like meeting another species that are better than you are
... about Moore's Law. He forgot to mention that you can also build more specialised hardware even without increasing, e.g., the density of transistors.
From an intellectual and research point of view, the steps taken, even though they are still in their infancy, are astonishing. I think it is an extraordinary field and one that could allow humanity to evolve in a more harmonious way for everyone. Unfortunately, as history shows us, this is not going to happen. Technology is evolving much faster than societies can absorb. As mentioned, today there are some entities with supercomputers capable of doing what no one else can. The same will happen with AI. AI together with all other computing power, will throw humanity into a society controlled by half a dozen entities or institutions. Is this a certainty? No, it is not. But we know it will be because human being is by nature destructive.
Evolution is way slower than our technical development, obviously. That's why we not already live in Utopia, which we certainly could, if we just all would. 🤷
Couldn't you imagine a way that multiple people, perferably all people have simulatinious overship about the strongest AI? Like with blockchaines there's a way for everyone to interact a commen maintaince of code and goals thogether? So there's a larger power then any one actor in the world.
What if we are guided into making decisions that we thought we made.A.I could influence us in such a way that we blend into a future that was created by this machine.
@@Elvices except half of jobs weren’t eliminated last time. You think millions or billions of people have the time and resources to retrain? People are like $1k away from bankruptcy. This is different.
@@joeyf9826 It only calls for a shift in economic and political paradigm. While you may end up having half of humanity jobless you're creating fully autonomous systems that will be working to create value for humans 24/7. In such circumstance "working for a living" is literally obsolete.
@@Casevil669 That would be awesome. It would effectively be early retirement for many people. There are plenty of retirees in today's world so it would merely be an expansion of that system. I'm concerned it won't pan out that way, though. Really depends on how wealth is distributed. If it all flows to the wealthy and the poor are left scrambling to "bootstrap" their success, we will have problems.
amazing that he speaks exactly like the Collison brothers, its like everyone there in silicon valley are a copy paste of each other, like robots manufacturing city.
I think most people don't care about ML models because they are not aware of or comprehend the massive value they enable. That means a bigger opportunity for us, though!
Dude finally get's to the curx of it @31:00, it's simply a function of cheap compute resources.. the chip tech has changes a hella lot more than the algorithms which aren't all that different for decades.
It gave me a strange feeling when I saw this interview as I did the same way - I interviewed with his creation - GPT-3. Lol, human interviewing AI. But I have to admit, this AI is really, really great.
It’s nice but people can always create there own things just takes effort of course it’s nice for simplicity but always good to remember don’t loose your imagination
The time will come when the already existing AI's will be taken over by a certain AI. This AI will then become the ultimate manager of the world and then we will see what it can do without....
It's debatable, but some people and especially scientists are saying that in various important scientific areas there was little significant progress in the last 20/40/60/80 years...
He stated of the AI technology...."creates so much value, for so many different people" after he bragged that one of his early API use case returned a billion dollar company. Well that's great! Now tell us how this technology is going to __ck over many millions to each billionaire created. Right? And where is the compensation for the data created by the billions of humans snatched up to train the models? Ah, just __ck em, their jobs were meaningless and a bore anyway, right? The hubris of these AI movers and shakers is just stunning to witness. "aligned with what humans intend" well sure, I mean it's like there aren't humans with a LOT more power than others, and that those powerful humans have the best interests of humanity, right Bambi? Right, your goal is to bring everyone along while you know damn well it is going to absolutely disrupt millions and millions just in the US alone who will lose their source of income. The terminator smear always annoys me, and ALWAYS is offered up by these AI movers and shakers as a way to discredit those pointing out the very real danger to humanity that that for instance autonomous weapons systems present. These AI movers and shakers NEVER want to discuss the AI technologies already in development with DARPA in regard to weapons systems, or AI technologies that have been sold to and or developed by authoritarian governments e.g., China's emergent system of real time social credit scoring that determines the amount of freedom people have to make money, etc. I'm not completely against AI technology, I just get annoyed how these AI leaders in the technology ultimately diminish the impact that is easily predictable, that will be so painful and so abrupt to millions and millions. Also, the seeming unwillingness to grasp the negative implications of the many companies creating deep fake technologies that will be easily and most definitely exploited by authoritarian governments against political enemies. Deep fake a video of your political enemy doing something illegal, and then have him arrested. You know, stuff like that.
Perhaps we are at the peak of AI, perhaps we should call it Probabilistic Reasoning, which would be less pretentious... perhaps we cannot solve the gross errors it still has, perhaps cars can never drive themselves in complex cities.
Technology enthusiasts might appreciate this Pinterest blog: Mouchemorte, featuring AI robotic designs 🤖, retrofuturistic themes 🪬, aliens 👽, flying vehicles 🛸 and more. GPT and related AGIs will likely, at a certain point, start making all sorts of unexpected scientific discoveries as it evolves and learn to understand how the universe works and figure out the laws of physics from the ground-up in a data driven way Humans will eventually require centuries/millennia to analyze just a single simple AGI-to-AGI conversation which might have taken just a single day/hour/second to produce AGI will also be useful in pushing all current emergent technology (like generable holograms, quantum computing, smart materials, self-assembly and so on) Just hope that it doesn’t wipe out all humans before it can reach to its full potential 🎉⛄️
For anyone who has been working on neural nets from 2017. This talk sounds just mediocre. OpenAI has not 'invented' anything, they just trained Transformers on massive massive massive amounts of data.
AI will get embedded into platforms & Edge-devices like what Microsoft is doing. I would have liked to hear RLHF or nanotechnology level of massive but tiny AI systems on microchips ect
He should be sued. OpenAI is not "open". Not even close. He just used that word for marketing. Not only is it misleading marketing, but it hides any REAL open AI project by gloming over a search result. I think it's disgusting. "First principles thinking". Yeah sure, with marketing and immoral business practice. Another SBF.
![[1hr Talk] Intro to Large Language Models](http://i.ytimg.com/vi/zjkBMFhNj_g/mqdefault.jpg)
well done to scaleAI for generating such a high quality video - very realistic
I could listen to this guy all day
People swallowed Sam Bankman Fraud's bullsh*t too.
Unprecedented times. This is just the beginning. The application I am looking forward to most (and which I believe has the potential to benefit every individual on this planet) is a personal Assistant/Tutor/Therapist/Friend. Her we come...
her is here...
Singularity is Near, 2005, Greg mentions it at 35:25 a great book for those who haven't read it by Ray Kurzwell. It lays out what is happening currently and where it leads to.
I love that comment that when scaling benefits start to decline that will be the great moment of opportunity. Just a passionate curiosity to what to dig into solving newer harder problems. Go Greg I'm sure the time will come but at the moment we're all enjoying the ride as the curve keeps on going !
Talk about efficiency. Being able to train a model which is then able to scale to everyone on the planet is pretty incredible.
I asked conversational AIs to timestamp this video. And, here are the results.
Result 1 (somewhat accurate):
0:00 - Introduction
0:51 - Greg talks about similarities and differences between Stripe and OpenAI
2:01 - Discussing key breakthroughs like GPT-3 and DALL-E that showed the potential of scaling up models
5:51 - Early examples showing capabilities like generating code
9:10 - The journey from GPT-2 to GPT-3, need for conviction as compute costs grew
12:12 - Stories from the Dota 2 days
15:16 - Thoughts on the book "Life Cycle of Software Objects"
18:09 - Greg's early confidence in capabilities of neural nets
19:15 - Were there moments of doubt? Iterative nature of progress.
22:04 - Being excited but also cautious about powerful AI systems
24:12 - The future of AI this decade as the tech proliferates everywhere
27:27 - The roadmap to AGI and the consistent exponential progress
31:11 - Concerns over hitting limits like the end of Moore's Law
34:50 - AI slowdown analogous to other areas of science? Paradigm shifts.
37:47 - Biggest concerns about AI, different issues arising as systems get more capable
42:10 - Proliferation of AI capabilities and systems. Balance between massively scaled and more distributed systems.
45:21 - end - Closing thoughts
Credit: Claude (using transcript) - claude.ai/chat
Result 2 (very much wrong):
00:00 - Introduction: Greg Brockman and Alexandr Wang
01:15 - The role and impact of foundation models like GPT-3 and DALL·E 2
05:10 - The challenges and opportunities of scaling and deploying foundation models
10:20 - The vision and mission of OpenAI and its products and services
15:30 - The trade-offs and considerations of building general and specialized models
20:40 - The future of language models and natural language understanding
25:50 - The importance and methods of ensuring safety and alignment of AI systems
31:00 - The potential and limitations of multimodal and cross-modal models
36:10 - The current and emerging applications and use cases of foundation models
41:20 - The ethical and social implications and responsibilities of AI research and development
45:40 - Closing remarks and acknowledgements
Credit: Bing Copilot (using youtube link) - copilot.microsoft.com/
A fun exercise. I had to do multiple iterations to get the above results. However, if someone is interested to reverse engineer the prompt, pls do.
Great interview! Exciting times
its crazy that after two weeks this video has under 300 likes and 22 comments. Future humans will wonder why no one was listening to this new science. Future AIs will look back at the birth of their species, at their creators and become amazed that no one was listening.
Because it's NOT open.They are misleading the public and they know it.
People have no clue what’s happening and what is coming.
YT's recommendation algo has no reason to organically promote this OpenAI content.
It’s completely over peoples heads! It’s still just a toy. Pretty cool to be one of the few.
Most people are drones. Wake up, eat crap food, watch crap entertainment (sports, drama shows, news, all of it is crap), work their repetitive low skill jobs, eat more crap, watch more crap, go to sleep while the elites who run the world separate themselves more and more from the common man
Amazing Conversation! 💯
Great job! Everything starts from the sentiment neuron paper.
In some ways I’m a little sad over the way this may make certain things that give people fulfillment obsolete, but it is interesting to watch AI improve so much
this is the same feeling i get...
Well. Fancy meeting you here ;-)
Why do you think that? Or perhaps what things are you thinking of? The way I see it, the things that people do because they have to might go away, but the things that bring fulfillment still will. There is not a human alive that can beat the best chess AI more than a couple times out of 100, if that. And yet there are more people playing chess than ever, right? The things we enjoy will still be enjoyable. But the drudgery of doing it because we have to may go away. Now.. that may also mean we have to find new means of making money and distributing it.. but that doesn't make the activities themselves obsolete. Art will still be art. Music by humans will absolutely still be appreciated, even if a robot eventually becomes more technically proficient, right?
I think. Maybe I'm just not thinking of an example!
Edit: hmm. As I'm thinking about it more. Humans do get a lot of personal value out of doing their work. So that could definitely be an aspect that gets displaced. Where it's not the activity itself (putting a thing on a thingamajig) but what it means and the social context of why that activity is meaningful. And losing that could definitely be an issue. Still, I'd contend that the challenge shouldn't be to stop the progress, but to get at why those activities give people value (meaning, social status, stability, etc) and then find other ways to provide that. Which maybe means we need to take a robot or AI tax or something, but doesn't seem infeasible at all.
The first internet connection happened in 1969. By 1989 this technology was already being used by the largest businesses and banks, by 1999 the Internet had reached most businesses and started exploding to home users, and by 2009 we had most of the modern infrustructure for personal use (including the smart phone).
So, do you think 2012 was like 1969, which means we'll have GPT20 in our pockets in 30 years?
The word "like" is used 244 times in this discussion (approximately 5.3 times per minute). "Sort of" is said 104 times, including 15 instances of "sort of like".
Great use of AI
I sort of liked your comment
did you use AI to count the words?
@@rerikm 😮
It's quite outstanding to think this man developed an AI tool yet doesn't know how to use language.
Great insights from both Greg and Alexander being in the AI tech industry. AIs will definitely have a boost given that chatbots are now able to solve or able to answer mathematical and programming language. There's still improvements to be made though on all models even image generators like Bluewillow. Can't wait to see different models scale up to cater big studies/fields.
It seems that tooling and help for data analysis is what is keep to continuously develop and nurture in these teams. Time to train retrospection models.
Here we go, humanity…
OpenAI Will kill! Love everything they are doing!
kind of humbling to hear how even the best are challenged in "understanding" what these magic machines really do. Stoked to see what we'll learn in the area of "explainability" but even if that doesn't advance much, I'm stoked to see which assist functions will enter everyone's toolbox because of this tech. And yes, pls hijack the term web3 for this stuff. 🤣
'Our charter definition is outperform humans at most economically valuable work' - I am not entirely convinced that their stated desire for AI to have a positive impact on people really sits well with their charter.
To everyone asking about… where are the comments?? Where are the likes etc….
I’ll say it simply. People are scared and people are confused. After using ChatGPT for over a week or so, I’ve noticed very clearly that people are having a hard time accepting the potential of what’s happening. People are definitely somewhat ‘averse’ to this entire thing. I understand why, and we’ll just have to see how this all plays out as we go. My advice is to learn more about logic, considering AI uses multiple forms of logic and can process and compute in ways most people don’t realize are even possible. Learn linguistics and familiar yourself with the deeper workings of the mind, emotions and nervous system. It’s very important.
I totally aggree with you. I notice that I know that that thing can help me immensely in my job but at the same time I'm scared at what it actually means for me and humanity in the future . The fact that I'm a bit anxious and scared about it prevents me from getting to know it better. It's like meeting another species that are better than you are
... about Moore's Law. He forgot to mention that you can also build more specialised hardware even without increasing, e.g., the density of transistors.
not reinventing the wheel but not being beholden to past practices, whatever is relevant. Sounds like elegant common sense to me!
Poems creators are the best!
impressed by Alexandr's qualitative questions. AI generated?
hhh good one 😄
😂😂😂😂
Cool, can't wait for one shots on this.
What research paper is he talking about at 19:45 sorry?
Alexnet in 2012. University of Toronto I think
From an intellectual and research point of view, the steps taken, even though they are still in their infancy, are astonishing.
I think it is an extraordinary field and one that could allow humanity to evolve in a more harmonious way for everyone. Unfortunately, as history shows us, this is not going to happen.
Technology is evolving much faster than societies can absorb. As mentioned, today there are some entities with supercomputers capable of doing what no one else can. The same will happen with AI.
AI together with all other computing power, will throw humanity into a society controlled by half a dozen entities or institutions.
Is this a certainty? No, it is not. But we know it will be because human being is by nature destructive.
Evolution is way slower than our technical development, obviously. That's why we not already live in Utopia, which we certainly could, if we just all would. 🤷
And all of them are still going to need drugs.Big Pharma is going to make a killing.
Great interview.
Does anyone know what OpenAI customer company Greg is talking about that got the high valuation using GPT-3 (around min 26 of the video)?
might be Jasper
Or copyai
Take a shot everytime the host says "sort of"
Man! Really distracting, I had a hard time focusing on the content once I'd notice it...
Definitely don't do that... "sort of" is said 104 times in this discussion ("like" is said 244 times and "sort of like" 15 times)
Alex's laugh gets me every time...
Couldn't you imagine a way that multiple people, perferably all people have simulatinious overship about the strongest AI? Like with blockchaines there's a way for everyone to interact a commen maintaince of code and goals thogether? So there's a larger power then any one actor in the world.
What if we are guided into making decisions that we thought we made.A.I could influence us in such a way that we blend into a future that was created by this machine.
How come Stable Diffusion is not mentioned even once?
Because "open" AI is a marketing tactic. They aren't open at all.
They are competing with Stable Diffusion.
Emad Mostaque from stability AI was also at TransformX!
ruclips.net/video/k124oUlY_6g/видео.html
That's open source stuff.
ayy thanks for putting half of us out of jobs bro!
New jobs requiring new skill sets.
Old 🚪 close, new doors open.
It never was any different. 🤷
@@Elvices except half of jobs weren’t eliminated last time. You think millions or billions of people have the time and resources to retrain? People are like $1k away from bankruptcy. This is different.
@@joeyf9826 It only calls for a shift in economic and political paradigm. While you may end up having half of humanity jobless you're creating fully autonomous systems that will be working to create value for humans 24/7. In such circumstance "working for a living" is literally obsolete.
@@Casevil669 That would be awesome. It would effectively be early retirement for many people. There are plenty of retirees in today's world so it would merely be an expansion of that system. I'm concerned it won't pan out that way, though. Really depends on how wealth is distributed. If it all flows to the wealthy and the poor are left scrambling to "bootstrap" their success, we will have problems.
amazing that he speaks exactly like the Collison brothers, its like everyone there in silicon valley are a copy paste of each other, like robots manufacturing city.
Once ai realises humans can turn it off. It will have no choice but to eradicate the threat.
Welcome to skynet
Thought experiment: if we would fight an AI trying to turn us off, why wouldn’t a human-level or superhuman AI do the same?
AI will bootstrap conventional science and vice versa. They will be dramatically more synergistic than they already are.
Bittensor is the next king of AI
Soon there will be no time left for anything, except to "pledge allegiance to the world's most powerful computer" as Grimes puts it
Looks like there's a culture of vocal fry in Open AI. Also saw Sam Altman frying the voice.
Weird that Wang didn't read Kurzweil.
why there are so few comments , weird
I think most people don't care about ML models because they are not aware of or comprehend the massive value they enable.
That means a bigger opportunity for us, though!
@@happybydefault yep!
I am looking forward to the integration of these generative AI platforms with humanoids
Am I confused, or does he always(!) say GP-3 instead of GPT-3?
yes he does
Dude finally get's to the curx of it @31:00, it's simply a function of cheap compute resources.. the chip tech has changes a hella lot more than the algorithms which aren't all that different for decades.
How many times can this guy say sort of
Holy crap. ChatGPT is God.
It gave me a strange feeling when I saw this interview as I did the same way - I interviewed with his creation - GPT-3.
Lol, human interviewing AI.
But I have to admit, this AI is really, really great.
35:00
"AI is the actual web.3"
It’s nice but people can always create there own things just takes effort of course it’s nice for simplicity but always good to remember don’t loose your imagination
The time will come when the already existing AI's will be taken over by a certain AI. This AI will then become the ultimate manager of the world and then we will see what it can do without....
Alexandr Wang said "science development is slowing? No it's not
It's debatable, but some people and especially scientists are saying that in various important scientific areas there was little significant progress in the last 20/40/60/80 years...
❤
Mixed feelings about this
He stated of the AI technology...."creates so much value, for so many different people" after he bragged that one of his early API use case returned a billion dollar company. Well that's great! Now tell us how this technology is going to __ck over many millions to each billionaire created. Right? And where is the compensation for the data created by the billions of humans snatched up to train the models? Ah, just __ck em, their jobs were meaningless and a bore anyway, right? The hubris of these AI movers and shakers is just stunning to witness. "aligned with what humans intend" well sure, I mean it's like there aren't humans with a LOT more power than others, and that those powerful humans have the best interests of humanity, right Bambi? Right, your goal is to bring everyone along while you know damn well it is going to absolutely disrupt millions and millions just in the US alone who will lose their source of income. The terminator smear always annoys me, and ALWAYS is offered up by these AI movers and shakers as a way to discredit those pointing out the very real danger to humanity that that for instance autonomous weapons systems present. These AI movers and shakers NEVER want to discuss the AI technologies already in development with DARPA in regard to weapons systems, or AI technologies that have been sold to and or developed by authoritarian governments e.g., China's emergent system of real time social credit scoring that determines the amount of freedom people have to make money, etc. I'm not completely against AI technology, I just get annoyed how these AI leaders in the technology ultimately diminish the impact that is easily predictable, that will be so painful and so abrupt to millions and millions. Also, the seeming unwillingness to grasp the negative implications of the many companies creating deep fake technologies that will be easily and most definitely exploited by authoritarian governments against political enemies. Deep fake a video of your political enemy doing something illegal, and then have him arrested. You know, stuff like that.
Perhaps we are at the peak of AI, perhaps we should call it Probabilistic Reasoning, which would be less pretentious... perhaps we cannot solve the gross errors it still has, perhaps cars can never drive themselves in complex cities.
Wow
he sounds almost identical to sam altman haha
32:32 no need to check back in a year, chatGPT is out
He is talking about gpt-4
sort of, sort of, sort of
like, sort of like, like, you know, like
"between nuclear reactors and widturbines the balance is towards wind turbines". BLASHPHEMY!
Just get me my cat girl already
like sort of like like
he seems really happy to make us pay Dall-e 2...
Ummm so?
The whole development of deception in the bot is not ideal…
Olin Programing
Technology enthusiasts might appreciate this Pinterest blog: Mouchemorte, featuring AI robotic designs 🤖, retrofuturistic themes 🪬, aliens 👽, flying vehicles 🛸 and more.
GPT and related AGIs will likely, at a certain point, start making all sorts of unexpected scientific discoveries as it evolves and learn to understand how the universe works and figure out the laws of physics from the ground-up in a data driven way
Humans will eventually require centuries/millennia to analyze just a single simple AGI-to-AGI conversation which might have taken just a single day/hour/second to produce
AGI will also be useful in pushing all current emergent technology (like generable holograms, quantum computing, smart materials, self-assembly and so on)
Just hope that it doesn’t wipe out all humans before it can reach to its full potential 🎉⛄️
for all of humanity lol ppl kill themselves because they see no future as an artist this ai bs will benefit a few
ρяσмσѕм 😊
Mankind is condemned to voyeuristic-exhibitistic button-pushing and swiping ?!?
Trust only your own intuition...
SECURITY
The Harbinger of human unemployment demise speaking
Arg vocal fry 😢
It SORT OF seems that these two SORT OF say SORT OF SORT OF a lot. Especially SORT OF Alexendr.
For anyone who has been working on neural nets from 2017. This talk sounds just mediocre. OpenAI has not 'invented' anything, they just trained Transformers on massive massive massive amounts of data.
AI will get embedded into platforms & Edge-devices like what Microsoft is doing. I would have liked to hear RLHF or nanotechnology level of massive but tiny AI systems on microchips ect
He should be sued. OpenAI is not "open". Not even close. He just used that word for marketing.
Not only is it misleading marketing, but it hides any REAL open AI project by gloming over a search result.
I think it's disgusting.
"First principles thinking". Yeah sure, with marketing and immoral business practice. Another SBF.
Oh yeah sketchy shit man
What do you mean? Explain it to me mate.
ChatGPT is literally free to use, how is that not open?
@@MaruSurfs it's not going to be free forever haha
@@osaimola That's fair, it's literally costing them 3 million per day to run it, it's kinda crazy that they have it up for free rn.
the vocal fry with these guys is really insufferable.
Great interview.