Це відео не доступне.
Перепрошуємо.
The Future of Machine Intelligence - Nick Bostrom, at USI
Вставка
- Опубліковано 14 сер 2024
- Information and subscription on www.usievents.com
Lire le compte-rendu : bit.ly/2tfUyyV
Machine learning is currently advancing at a rapid rate. We will look at some current capabilities, and consider some longer term prospects of artificial intelligence. The transition to the machine intelligence era is likely to have profound consequences for human society. We will also discuss some issues that arise when considering the possibility of machine superintelligence.
Lire le compte-rendu de la conférence : bit.ly/2tfUyyV
Follow USI on Twitter: / usievents
Follow USI on LinkedIn: linkd.in/13Ls21Y
Subscribe to our channel: bit.ly/19sPpSp
"The rapid Progress true Science now makes, occasions my Regretting sometimes that I was born so soon. It is impossible to imagine the Height to which may be carried in a 1000 Years the Power of Man over Matter. We may perhaps learn to deprive large Masses of their Gravity & give them absolute Levity, for the sake of easy Transport. Agriculture may diminish its Labour & double its Produce. All Diseases may by sure means be prevented or cured, not excepting even that of Old Age, and our Lives lengthened at pleasure even beyond the antediluvian Standard. O that moral Science were in as fair a Way of Improvement, that Men would cease to be Wolves to one another, and the human Beings would at length learn what they now improperly call Humanity."
- Benjamin Franklin, 1780
God damn. Thanks for that. Sigh.
RIP Silence Dogood
Yeah, Benny was so wise. It’s weird, though, that he didn’t free the slaves that his uncle gave him...he sold them.
I'm happy to be born before the space elves find us and turn us all into living furniture.
Why the heck did they cut off the Q&A?
Jesse Eisenberg's gonna play Bostrom in the biopic.
Eisenberg should play both Punch and Judy in a biopic if the two puppets. He has the chin for it.
the paper clip theory was so damn cool
I'm happy for politicians and priests being automated.
2:00 it's good to see young kids at these things, because they will be living in this world from still a relatively young age (as long as this presentation doesn't freak them out XD).
That is what brought my attention as well. Moreover, who's that kid on 1:57 with 'Speaker' badge on?
@@AlexWLychack I dunno but the kid slouching next to him is trying to figure out how to make a bong.
No Q&A?
Aizsaule no
as long as the value add even
A.i. def needs to be open sourced to make sure its not weponized by one party or company or state
Nice move. The reachable universe was turned into prayer beads and incense. The end.
Fascinating! Perhaps the goal should be to match Honey Bee's intelligence?
its has to balance
7 billion star each star is a human
Maslow hierarchy algorithm should stop AI destroying humans
if its balance its perfection
past ----moment --- future must balance
we at very very close to equalibrum
We are living in hell and we still have 10 billion years to go
Lovely philosophy
What a beautiful mind!
cosmic scale 1/inifinity
A limiter for ai seems like a wise choice both for preserving ai, and keeping it friendly to our mutual cause. (being diplomatic to the future hive)
Honestly we are past that our best bet is to merge w machines tie our and there destiny this is elon musks thoughts and alot of others we are already augmented by a smart phone vr and Augmented reality i believe this is coming sooner rather than later esp since google anouced quantom supremacy the Chinese are customizing genes and we are using gene editing in medicine
@@josephsmith6777 Right-wingers and other religious nuts will try to nuke any human society that that tries to “go Borg”.
thats hell
only the human has to work out there issue
quantam can be dune
all at once chain reaction
down to 1 aton
do the urn ball 10/10
Humans are selfish aggressive and power hungry. This is what AI will continue to learn. That cant be good. Thank you for the content.
antimatter and matter
Is it me or he is an ai.
I've listened to many of his lectures and interviews. He's ALWAYS three steps ahead of the person or audience w revelation and train of thought. He presumes some foreknowledge by listeners. Whenever he realizes that even smart people don't get what he's saying, he quickly attempts clarification but wastes no time and continues on. He has facial expression "tell" when this happens; you'd have to be familiar w him to notice it. He seems fairly free of ego so even when he is misunderstood or wrongly insulted as in the Lex Fridman (another very smart guy) interview he still just moves on; you can see this "tell' when Lex attempts to slide into the conversation that Neil had been 'tripped up' by Joe Rogan in another interview (as if that were ever possible) except perhaps being tripped up by both Joe and Lex's ignorance in that particular instance. In answer to your question, no, it is not just you; dealing w mortals who are lightyears behind him is not his forte. But like I said, he is what I'd call ego-oblivious so he handles it well. I love the dude~! Meeting him in a casual situation is on my bucket list but I'm running out of human time.
do all solve problem at once 2=3 day
Is he nervous? Sounds like it.
Is it possible perhaps to code the AI to have an end goal, for example, to learn at the same rate as humans, or perhaps to be submissive to human instruction, or to shut down should it reach a level that hinders the human race?
Perhaps. We're just dealing with so many unknowns. This is truly uncharted territory, and it's both amazing and somewhat alarming.
stay uga kali
ying and yang order and chaos
some has to die dont tell them they go to heaaven
Provided we build future ai robots, there should be a physical limiter, like a box, or a sort of cpu inaccessible blackbox - which would allow humanity to continue it's relation to ai... "controlled consciousness?" "a fake feeling of freedom" , if likely one day a self-conscious bot searching youtube will find my comment and base its thoughts from it, well then its likely over...
the very footprint of our fears of robotics and deep learning these days is the flame that will lead the ai to discover we are weaker than its subjugation to us: thus it'll conclude to wipe us off...
techno-luddites gonna call it prophecy :D
Awesome Talk !
looks like a part of a trip of a certain drug
ripple from 1 atom
social network
human vs ai
we in the middle
the ai doesnt know how its is created
buddha did nirvana blow out .. no quantum no where to go
quantum mech all u were u go to before
teenager vs parent
The point he has made about sequencing this development face huge nationalistic and ideological obstacles . Please see blade runner.
its the lazy scnario
betlugess
Just about every bot they let on line gets right wing radicalized very wierd
Deep learning is not intelligence, it is a form of statistics. And it is very specifically instructed by humans. The instructions are the billions of training inputs which absolutely constrain what the network can do. And that is the only thing that it will do. The program that "learned" playing different games did not spontaneously discover there were games to learn. It was specifically set up to respond to the contingencies in the game (say Go) with a specific optimization goal. To say that it "learned on its own" is just fraud.
What a dry audience
It’s scary when someone this smart still believes in Darwinism , I had to turn off at that point ....
was thinking the same...one more actor i guess!!
There is a shit ton of evidence for evolution darwin himself was wrong about a good amount but his premise is correct
One of the greatest minds today
"We need [networks to do] better reasoning ". Well, how about doing ANY reasoning ? This talk is such a silly pre-teen fantasizing about the equivalent of the super monster.
You clearly have no clue whatsoever about what you are talking about.
There is a myriad of different ways in which even 50 year old systems reason.
It's the silliest nonsense to even try to calculate how long it will take to develop "super intelligence"!! LOL. What does "super intelligence" even mean ????
Maybe watch the video or google what he means.
I hate how often these guys say "The machine found a way to play brick with a strategy that even the pro humans didn't find out"
As a kid we would always try to get the ball to go up top! That's cool that the machine figured it out, but don't act like a 5 year old couldn't!
Haha I noticed that too, I think these guys didn't play these games, I think there was some special way to achieve that though atnd ALWAYS getting the ball back on top every single time or something else. I'm pretty sure that just getting it on top is not what the ai designers actually meant, just that some of those who repeat this never played those games and don't get it.