That's what Germans used to say about Hitler. Israel's AI industry must be shut down. Using the genocide in Gaza as a commercial for the arms industry to sell their AI targeting systems is unacceptable.
It's great that people are talking about NNs as more than statistical functions. I believe they are modeling the world and something like platonic forms. It's amazing how ethical LLMs are these days. I think that's because they have a better error function than minimizing surprise -- which is an emotional error function.
@@rolestream It's because they don't have rights to the material they're editing. That's why they also add these annoying irrelevant graphics so that the algorithms don't recognize the material correctly. This is a new kind of business, quite easy money, look at the subscriber numbers.
@@tojileon Yes, that's exactly what I was thinking. But it would be cool to know when and where so that you can quote the content. It's a shame that the channel doesn't help with this, otherwise it could also be used for scientific research.
@@tomnuo9418 I see... So it's forbidden to quote interviews (i.e. as videos, not words), even if you include the source? Other channels also show interview excerpts and you can get information about the original on request.
The “like” and “sort of” used every 10 seconds or so by the dumbed down millennial interviewer forced me to use ai to just “sort of””like” summarize it and be “sort of” be “like” done with it!
Do you "sort of" "like" feel exceptionally "like" stupid, now that others have pointed out just how much that "dumbed down millennial" has accomplished & as a result impacted the world, relative to the average person including you & I? At least humbled a bit, no? Ironically, the little pauses, "uhms," and even usage of words such as "like" have been incorporated into the latest AI models to better mimick human speech.
@@neorock6135 also it’s like… they’re verbal filler, devoid of intellectual content, i.e. it does not contribute to making a point better or worse. Hence there is no indication that the mutterer is less intelligent.
I’d heard one of the AI gods talk about how our brains predict in real-time at an extremely high level of data intake and therefore are conscious. Our prediction and knowing what happens “next” in true real time is what this feels like to be us. So I think eventually we can scale the models with enough real world data collection.
If Sutskever really said this, it was just about the most asinine comment about the capabilities of AI’s “brain” ever to be heard in recorded format. To be sure, there are some woebegotten souls out there who want this. In their fantasies, creating a new species is somehow a thing to be desired.
And what do you want? Humans have given it a shot, and look at what we've done. Take a look at how we treat one another. Take a look at how we treat animals on this planet (watch how chickens are treated from birth until death). We gave it shot, and guess what. We SUCK! Time for something else to give it a try. Maybe when a machine gets to the point we are at it will finally put us in our place and stop the madness. Think beyond your petty human goals and realize there may be a bigger picture here.
This man is thinking on a different level. But damn his words are gold. He came to see that superalignment is the more difficult and needed scientific/engineering task in AI and to forget the wealth and power because you will push the limits.
That's what Germans used to say about Hitler. Israel's AI industry must be shut down. Using the genocide in Gaza as a commercial for the arms industry to sell their AI targeting systems is unacceptable.
Neural Networks as a concept exist since the 1950s...I guess that in 2003 it would be very difficult for someone to imagine that a computer with 32GB of ram and 12 cores plus a GPU would cost around $2k...so yeah people didn´t believe in NNs back then.
BIG Dislike to video creator because of stupid animations inserted into interview and BIG Dislike to video reposter because of no information of date of the interview.
![Chlöe - Shake (Feat. Jeremih) [Official Visualizer]](http://i.ytimg.com/vi/aDs_1ufpfv8/mqdefault.jpg)
![The moment we stopped understanding AI [AlexNet]](/img/1.gif)
I hate it when ancient interviews are reposted like this.
Same! I really hate it.
Check his hairline to see how old this is.
Why not put the original date of production somewhere!
the guy is brilliant, and he's smart enough to actually not say anything
That's what Germans used to say about Hitler. Israel's AI industry must be shut down. Using the genocide in Gaza as a commercial for the arms industry to sell their AI targeting systems is unacceptable.
This was recorded before the ChatGPT! (guessed by Ilya hairstyle)
lol
It's great that people are talking about NNs as more than statistical functions. I believe they are modeling the world and something like platonic forms. It's amazing how ethical LLMs are these days. I think that's because they have a better error function than minimizing surprise -- which is an emotional error function.
Thanks for a new Ilya Episode 😁👍🥳🥰 When was this interview originally hosted and where?
This channel never posts any details. I don't why. It is absurd and annoying.
Ilya wearing OpenAI T-shirt. Must be old. 🤷
@@rolestream It's because they don't have rights to the material they're editing. That's why they also add these annoying irrelevant graphics so that the algorithms don't recognize the material correctly. This is a new kind of business, quite easy money, look at the subscriber numbers.
@@tojileon Yes, that's exactly what I was thinking. But it would be cool to know when and where so that you can quote the content. It's a shame that the channel doesn't help with this, otherwise it could also be used for scientific research.
@@tomnuo9418 I see... So it's forbidden to quote interviews (i.e. as videos, not words), even if you include the source?
Other channels also show interview excerpts and you can get information about the original on request.
That host sucks
I think true AGI will arive with fully developed Morphological neural chips and not before ^^
The “like” and “sort of” used every 10 seconds or so by the dumbed down millennial interviewer forced me to use ai to just “sort of””like” summarize it and be “sort of” be “like” done with it!
That dumbed down millennial is the founder of Scale AI (provides training data for LLMs) and has a net worth of $2 billion.
@@Jackson_Zhengthank you! I hate people judging others on how they talk rather than what they are saying.
He is young and has achieved more than 99.9999% of us... in 10 lifetimes.
Do you "sort of" "like" feel exceptionally "like" stupid, now that others have pointed out just how much that "dumbed down millennial" has accomplished & as a result impacted the world, relative to the average person including you & I?
At least humbled a bit, no?
Ironically, the little pauses, "uhms," and even usage of words such as "like" have been incorporated into the latest AI models to better mimick human speech.
@@neorock6135 also it’s like… they’re verbal filler, devoid of intellectual content, i.e. it does not contribute to making a point better or worse. Hence there is no indication that the mutterer is less intelligent.
A perfect guide on how to become Machine Learning Expert.
I’d heard one of the AI gods talk about how our brains predict in real-time at an extremely high level of data intake and therefore are conscious. Our prediction and knowing what happens “next” in true real time is what this feels like to be us. So I think eventually we can scale the models with enough real world data collection.
If Sutskever really said this, it was just about the most asinine comment about the capabilities of AI’s “brain” ever to be heard in recorded format. To be sure, there are some woebegotten souls out there who want this. In their fantasies, creating a new species is somehow a thing to be desired.
And what do you want? Humans have given it a shot, and look at what we've done. Take a look at how we treat one another. Take a look at how we treat animals on this planet (watch how chickens are treated from birth until death). We gave it shot, and guess what. We SUCK! Time for something else to give it a try. Maybe when a machine gets to the point we are at it will finally put us in our place and stop the madness. Think beyond your petty human goals and realize there may be a bigger picture here.
Old interview
This man is thinking on a different level. But damn his words are gold. He came to see that superalignment is the more difficult and needed scientific/engineering task in AI and to forget the wealth and power because you will push the limits.
That's what Germans used to say about Hitler. Israel's AI industry must be shut down. Using the genocide in Gaza as a commercial for the arms industry to sell their AI targeting systems is unacceptable.
I remember that Elon started Open AI?
Neural Networks as a concept exist since the 1950s...I guess that in 2003 it would be very difficult for someone to imagine that a computer with 32GB of ram and 12 cores plus a GPU would cost around $2k...so yeah people didn´t believe in NNs back then.
To get the data for automated lawyers, may he people should sue each other more often😅.
Next word prediction is the chisel, not the model. :)
0:04 Ilya who’s the well adjusted, healthy and emotionally intelligent human?
he talks about generalization. Thats is ingenuity... AI doesn't have it.
Cool
Wow
wait is this turning sam vs. ilya thing like apple vs. microsoft :)
Not true what he says, we can if course understand how we do it, wtf...
...stupid animations again...
what the date of interview ?
How old is this video? He's wearing the OpenAI shirt
I hope AI can give this man his hair back
You overestimate AI
Everyone will have a slave genie what could go wrong...
how old is this?
How old is this?
Shame on you reposting old interviews and not disclosing it
Its will think until not , this is a Hype , and this one is a promoter😅
Both dudes are using lots or words but not quite saying much?
Yet another interviewer who most annoyingly injects the word "like" into ever sentence - super interesting - not
so why are you slowing down the development if you expert?
Huh?!
@@umberto488 yeah, why???!!!
@@pacanosiuhe’s not. He started his own company. Wtf r u talking about?
@@shrek22 For you to understand, I would have to create it and show it to you
@@pacanosiuPac you're not making sense.
LOSE THE STUPID ANIMATIONS
Are I will even AI is lied
BIG Dislike to video creator because of stupid animations inserted into interview and BIG Dislike to video reposter because of no information of date of the interview.
Old shit
I'm sure the interviewer is really smart, but does he really needs to talk like a 13 year old girl?