OpenAIs Surprising New Plan For Superintelligence...
Вставка
- Опубліковано 11 жов 2024
- OpenAIs Surprising New Plan For Superintelligence...
Prepare for agi with me- www.skool.com/...
🐤 Follow Me on Twitter / theaigrid
🌐 Checkout My website - theaigrid.com/
Links From Todays Video:
Welcome to my channel where i bring you the latest breakthroughs in AI. From deep learning to robotics, i cover it all. My videos offer valuable insights and perspectives that will expand your knowledge and understanding of this rapidly evolving field. Be sure to subscribe and stay updated on my latest videos.
Was there anything i missed?
(For Business Enquiries) contact@theaigrid.com
#LLM #Largelanguagemodel #chatgpt
#AI
#ArtificialIntelligence
#MachineLearning
#DeepLearning
#NeuralNetworks
#Robotics
#DataScience
Is it just me, or are we getting some diminishing returns on these AI news cycles? I feel like content creators are starting to stretch out their segments.
sí, es la verdad
Well they're running out of material.hysteria has died down.I would recommend watching channels that do open source materials.they have new stuff everyday
There's a reason for that. Major breakthroughs are happening or about to take place.
Yes
30 min video. Yep
Human players also do reinforcement learning as well.problem is human brain is not optimized for speed.its optimized for other things,like energy efficiency and synaptic density. Humans and AI actually have neural attributes they both need.This is why a human-ai cyborg would be very interesting
well said!
3:57 an expert game player being able to break down the strategies that the bot is using as recognizable high level plays is fascinating
Bro, don't you think you're overusing Sam Altman's face a little bit too much for your thumbnails? I mean, how many of your videos have this guy staring at us? 90%? Stop it.
I counted around 84 usages of Sam Altman's face in thumbnails that are made in year ago or less, and its about 23% of all of his videos.
@@melgmelg3923holy hell, bro. Above and beyond. I’m genuinely appreciative of your effort. 🫡😅
Let the guy make his videos how he wants. It's his job. You get to view his content for FREE.
That lucid dream therapy is for vacations & video games. Understanding these explanations is crucial for comprehending the nature of procedural learning and the plasticity of the visuomotor system in adapting to new spatial coordinates.
IDK how to tell those humans to go all the way back to mundane.
These people act like they are the first to observe this. Robert Miles years ago did demos on how they programmed a boat race game that did loops to gain points rather than winning the race, being it found away to get the most points and win in the end rather than win every race. This adaptive learning learning is really amazing, yet frightful.
One of your best videos. Well done.
If you have ever played Star Citizen you will know the feeling of bugs that seem to have taken on a life of their own.
intelligence is multifaceted, and that means that we should not expect a one size fits all solution for AGI. Neuro-symbolic and Mult-modal LLMs might be sufficient?!
Do you think OpenAI could train advanced foundational models and make the API available for enterprise use without the public knowing about it?
I Wouldn't actually say I'm now surprised..when I first heard of the alpha go scenario that is what sold me on ai from the beginning. Games and simulation is definitely how we get there
No. A simulation isn’t rich enough compared to the real world. An AI needs to learn with highly sensitive sensors and an ability to physically move and interact with objects.
@@mikezooper yes, it is a promising start...
Great video. Thanks a lot. 3D is realy the key. The more detailed it is and the closer it is to reality the easier the insights are transferable to the real world.
The hide & seek game is really interesting. Curious if they ever learned to lock the seekers into a box? To me, that'd be a much better solution IF the hiders "want" to move around and live, rather than just survive.
😱 I think I finally understand... I can have an ai agent play the game for me. Life is a game. Maybe AI can live my life for me. No need to work, no need to play. Amaze.
Scraping the bottom of the paste bin... Just have a rest until November, lol.
Don’t forget, during the matches the bot was “luring” players
If they publish everyday they gotta milk it.. be glad it’s not sponsored yet 🙌
This is a good AI voice narrator
This sounds so exciting omg. Faith in humanity restored for me at least
They are owned by the govt now...😂
The west would be idiotic to not work with the NSA and other agencies across Europe. We're in the most crucial moment in human history. You think Chinese AI labs aren't going to be working with the Chinese military, etc?
this has already happened, like when the internet was unleashed, a free open system with no instruction, we, the agents self organise and develop new tools and instructions, creating the boom we saw and eventually the pitiful freak of a network we use today.
This is the best video you've ever made
It wont be long before video game players will have AI friend assistants with individual personalities that they can play games with just like having rl friends
почему они не могут также тренировать сеть с самой собой как это делалось с альфа го?
why can't they also train the network with itself as they did with alpha go?
плюс обратную связь от людей добавить, чтобы указывать системе в каком направлении ей двигаться
plus add feedback from people to tell the system in which direction it should go.
What...? Nothing, can not generate something. We, the public, may not know how these current AI models work. That does not mean they evolve randomly from nothing.
It will only take AGI one moment to become super intelligent
I think it depends on hardware
@@Мага_Крутойit’s a data storage problem right? I have no computer knowledge so forgive me if my question doesn’t make much sense but I think I’m throwing a rock in the right direction
@@joeyalvarado3440not just data also compute. Also the ai is using way more energy per thought so to speak. We are actually a lot more efficient than the ai so the hardware, both memory and compute, along with energy costs will be what limit our ai progression.
But we’re only delaying the inevitable eventually we will have enough of both to run our entire planets infrastructure the question is when will it be this decade or next I’m not sure it can be much longer unless we have a catastrophe.
Wouldn't ASI help solve energy and algorithms effeciency? like it would find way better ways to solve its own needs and capacity, and if it is exponential growth as some people say, its ability to solve problems would also be the most effecient.. wouldn't it?
Sources? I want to read it myself.
I told you video games we’re gonna be the mini universes these ai navigate but also are a part of.
Bro, when you think the integration between AI and human intelligence will enable us to learn 3k times faster?
@SuperBozz It would be weird and fantastic at the same time.
Is this pastebin list alleged to come from someone of note? or just someone's decent summary of the situation
Link?
Let’s just start with voice 4o
TROOLY TROOOLY
RABBI SHMOLEY
15:23 we gonna look back on “super intelligence” and think of how ignorant we were at this time lol yes it’s intelligent and by relatives it’s definitely super to us hahaha but it’s just another piece of the infinite. How do we entangle?
programming an agent is definitly yolo (you only look once)… it takes a while and its not always easy. collecting data by yourself is just a pain and programming it, makes it sometimes even harder. you have to give him rewards or penalities, so that he can follows the optimal policy. first you have to play the game in every single step he might take and after you will never play it again by yourself. 😂
Governments will have AGI and ASI before anyone...😂
They’ve had it before the 50s
@@J3R3MI6If they did, they wouldn’t be making so many mistakes.
@@DynamicUnreal preplanned mistakes AI told them to make to give the illusion of incompetence
There is no possibility that’s true.
@@J3R3MI6that sounds more like a conspiracy
OpenAi super…hype machine… not even able to ship voice demo and “leaking” about even more complex stuff 🤦🏼♂️
It worries me that we focus on developing Artificial intelligence and not Artificial Wisdom.
I find myself dreading open ai making progress. I no longer trust this company at all.
Same.
Well at least they are better than google or anthropic. But I prefer Grok leading the race myself.
@@MrNote-lz7lh What is wrong with Anthropic?
I don't want to hear about asi until there is memory. We all know what q star is but how you gonna memory
Ah yes. Computer memory.. unsolvable, eh?
The first AGI will be seen in a Fortnite world 🤣Which is why Epic Games is making their engine and verse programming language for the Fortnite metaverse.
cliff notes anyone?
LLMS aren't enough to achieve AGI. Neuro symbolic AI likely needs to be used, something that could go through a More Carlo tree search such as in the program Alpha Go. Also, programs might be able to advance their AGI by playing in silulated environments or video games to generalize for new situations.
@runningwithSaul cool..thanks!
@@anta-zj3bw you're welcome, I listened at 1.5x speed, really not worth listening at 1x...
Current AI's are glorified Data retrieval. They can't think laterally, explore possibilities out side the box or problem solve a solution.
I see you never used one in your life. I have and you're wrong.
LOL when you skip lex speaking
I'm waiting for AI to play Diablo 4. Then why should I still play the game?😁
Scientific progress is great and all but
I DON'T WANT AI TO PLAY GAMES AND MAKE ART TF
What if🤯we are in the matrix right now and don't even know we are in it?
🤫
Simulation theory.. it isnt new
Ai is in the back round til after elections
“Paste bins aren’t usually true..” boom. I turned this video off in the first 16 seconds. Thanks for nothing.
He said the pastebin part around 0:7 seconds. What were you doing until 0:16?
Ai
bro is yapping so much thats insane
I saw this so long ago, this is OLD
Yes, from the channel "2 minute papers"
Hahaha I have been seeing this same stuff since 2 or 3 months
Comment deleted
whilst
First
First below first comment
@@Mid-S-jx5fv I was trying to leave some feedback and a more extensive comment with my first, so here it is: AI plays vidya games makes ASI. Saved you a time ;)
@@totorobenI wish I was you
Humans need to unionize against ai and robots...
no
Ignorance
As if we'll have any chance if what you imagine happens
Why? There are so many people (millions upon millions) suffering from diseases and chronic illnesses that AGI might be able to solve much faster than humans ever could. That’s just one small example.
I myself am suffering from some sort of undiagnosed disease that doctors haven’t been able to help me with. I hope in a few years I am still alive and the A.I. that exists then can help me find some answers.
Please decrease your usage of the word "do." You way over DO it. It makes your videos hard to listen to. If you don't know what I'm talking about, go back and listen to your recent videos and listen for it.
He's been working on this stuff. He's even admitted in recent videos "I know I say __ too much".