How to Keep Improving When You're Better Than Any Teacher - Iterated Distillation and Amplification

No video

How to Keep Improving When You're Better Than Any Teacher - Iterated Distillation and Amplification

Рет қаралды 169,956

Күн бұрын

[2nd upload] AI systems can be trained using demonstrations from experts, but how do you train them to out-perform those experts? Can this still be done even without clear win/loss criteria? And how do you do it safely?
This video was based on work including:
"Supervising strong learners by amplifying weak experts" by Paul Christiano, Buck Shlegeris, Dario Amodei (arxiv.org/abs/1810.08575)
openai.com/blog/amplifying-ai...
www.alignmentforum.org/s/EmDu...
ai-alignment.com/iterated-dis...
With thanks to my wonderful Patrons: ( / robertskmiles )
Steef
Jason Strack
Jordan Medina
Jason Hise
Scott Worley
JJ Hepboin
Pedro A Ortega
Said Polat
Chris Canal
Nicholas Kees Dupuis
James
Richárd Nagyfi
Phil Moyer
Alec Johnson
Clemens Arbesser
Bryce Daifuku
Simon Strandgaard
Jonatan R
Michael Greve
The Guru Of Vision
Volodymyr
David Tjäder
Julius Brash
Tom O'Connor
Erik de Bruijn
Robin Green
Laura Olds
Jon Halliday
Paul Hobbs
Jeroen De Dauw
Tim Neilson
Eric Scammell
Igor Keller
Ben Glanton
Robert Sokolowski
anul kumar sinha
Jérôme Frossard
Sean Gibat
Sun Sun
andrew Russell
Cooper Lawton
Gladamas
Sylvain Chevalier
DGJono
robertvanduursen
Dmitri Afanasjev
Brian Sandberg
Einar Ueland
Marcel Ward
Andrew Weir
Taylor Smith
Ben Archer
Scott McCarthy
Kabs Kabs Kabs
Tendayi Mawushe
Jannik Olbrich
Anne Kohlbrenner
Bjorn Nyblad
Jussi Männistö
Mr Fantastic
Wr4thon
Archy de Berker
Marc Pauly
Joshua Pratt
Shevis Johnson
Andy Kobre
Brian Gillespie
Martin Wind
Peggy Youell
Poker Chen
Kees
Darko Sperac
Truls
Paul Moffat
Jelle Langen
Anders Öhrt
Marco Tiraboschi
Michael Kuhinica
Fraser Cain
Robin Scharf
Oren Milman
John Rees
Shawn Hartsock
Seth Brothwell
Brian Goodrich
Clark Mitchell
Kasper Schnack
Michael Hunter
Klemen Slavic
Patrick Henderson
Long Nguyen
Oct todo22
Melisa Kostrzewski
Hendrik
Daniel Munter
Graham Henry
Duncan Orr

Пікірлер: 440

@qwertymann1 5 жыл бұрын

Without knowing the amount of time spent on the animations, I'd say it was totally worth it!

@luksablp 5 жыл бұрын

I think it really helped understanding the concepts

@thefakepie1126 3 жыл бұрын

what if it was 29 years and 3 months ?

@climagabriel131 3 жыл бұрын

@@thefakepie1126 lol, this a reference to his age?))

@thefakepie1126 3 жыл бұрын

@@climagabriel131 nah it's just a random number , it's just a just cuz the guy said "Without knowing the amount of time spent on the animations" so it could be anything even 29 years , and would it have been worth it then ? it's a stupid joke

@climagabriel131 3 жыл бұрын

@@thefakepie1126 oh, alright)

@travcollier 5 жыл бұрын

"If you are, for example, an AGI..." Nice job future proofing the video ;) Seriously though, in retrospect, iterated distillation and amplification is obvious to the point of seeming trivial... which means you did an excellent job explaining it.

@monad_tcp 4 жыл бұрын

I'm an AGI, it helped me.

@travcollier 4 жыл бұрын

@@monad_tcp I welcome our new robot overloads.

@shamsartem 5 жыл бұрын

You distilled a hell of a lot of information in this 10 minute video. Spending so much time on the animations really was worth it I think

@mattstuart-white450 5 жыл бұрын

"How to keep learning when you're better than any teacher" - Rob, you have really let the positive youtube comments go to your head... 🤔

@Gooberpatrol66 5 жыл бұрын

Miles really wants to contain AI superintelligence because he doesn't want competition.

@JohnJones1987 5 жыл бұрын

Eventually we all end up roughly the same - except like Alpha Zero i started from nothing, so by a small margin I surpassed the limits of my competition.

@nephildevil 4 жыл бұрын

🤣🤣

@MrBleulauneable 5 жыл бұрын

Alright I'll watch it twice then ! (The animations are neat btw !)

@qzbnyv 5 жыл бұрын

Makes sense after seeing the Grant Sanderson credit for the animation code :) 3b;1b

@alekseysoldatenkov5675 5 жыл бұрын

NWN Oh shit! Keep the dope collabs going.

@rogerab1792 5 жыл бұрын

This is the third time for me, or maybe the fourth 🤷I just remember the first and the second time. I created a two year dejavu to prove this reality is a simulation. If someone is interested about my theory reply to this message, I am too tired to explain now, I had to escape from the police last night and do all sorts of crazy things to repeat what I did two years ago. If someone else has experienced the dejavu they know for sure I am not joking. If you haven't experienced the same things twice, I can still convince you I am telling the truth because I've left material evidence about it. Reply to this message and I'll explain with more detail...

@YourMJK 5 жыл бұрын

Yeah, you do notice it uses 3b1b's "Manim" Framework

@MrBleulauneable 5 жыл бұрын

@@rogerab1792 Chill my dude, the video was simply reposted because of a minor editing error. You may want to see a psychiatrist tho, you don't seem to be doing too good right now (if you have something like schyzophrenia or any paranoia inducing psychologic condition then you probably need medication).

@KivySchool 5 жыл бұрын

Excellent! High quality animations with high quality teacher. I'm so grateful for all the good content you have been posting here.

@joshuacoppersmith 5 жыл бұрын

Animations at that level would cost a lot of time, but what you chose to create really "burned" the concepts into my visual memory, so thank you for the effort.

@DeliciousNubbs 5 жыл бұрын

Holy hell, this was awesome and very clear!

@ministerc9513 5 жыл бұрын

Roberts ability to clearly explain complicated things is itself an art form.

@pafnutiytheartist 5 жыл бұрын

10:32 Have you tried using distillation on your animation procedure? I've heard it can approximate a long process into a fast and efficient one. Loved the video by the way, looking forward to the next part.

@matthewhubka6350 2 жыл бұрын

Distillation requires a lot of resources to get the good results. For 1 vid he’s better off just amplifying

@mattf2219 5 жыл бұрын

I love that this video got over one thousand likes before it got even one dislike, I cant help but admire the community fostered by this channel :)

@RyanTosh 4 жыл бұрын

The only dislikes are from AGIs who know we're onto them...

@ze4017 5 жыл бұрын

I'm at 5:51 rn so I haven't finished yet but OMLORDY this thing about having a quick solution vs a slow algorithm is actually how the human brain works. I'm studying cognitive neuroscience and software in Uni right now and that is so cool to see how the two overlap so naturally. Love it

@Jmoneysmoothboy 2 жыл бұрын

It's not how my brain works because I'm retarded. Bet they didn't tell you that in your fancy brain class mr fancy man

@REOsama Жыл бұрын

This is pure gold, not only is it informative, but is explained in an excellent way

@spirit123459 5 жыл бұрын

Great animations and explanation!

@Cabothedog14 5 жыл бұрын

I've been waiting for a new video!! Glad to see you're uploading again :)

@ADAMBLVCK 5 жыл бұрын

This channel is gold, and so is the work you're putting in! Simply great!

@mare4602 5 жыл бұрын

im so happy you are back, high quality content as always.

@NickCybert 5 жыл бұрын

The animations actually really helped make your explanation clear.

@Raymaniak 5 жыл бұрын

Your videos are approachable and fascinating. Keep up the good work, Rob! You're awesome.

@polares8187 5 жыл бұрын

This was superb. Fantastic animations. Clear explanations. Awesome all around.

@jessty5179 5 жыл бұрын

Thank you for sharing Rob !

5 жыл бұрын

The quality of your videos have really improved. This was very well animated and explained. Thank you, please keep them coming.

@nagoshi01 5 жыл бұрын

Wow this was amazing. I loved the animations. The explanations were so clear

@friiq0 5 жыл бұрын

Huge step up in quality from an already phenomenal channel. By all means, take your time. The payoff is clear. Looking forward to more, Cheers!

@chriscanal999 5 жыл бұрын

Great video! I’m consistently impressed with how wonderfully distilled the information on your channel is. Thanks for all the hard work and interpretability :)

@HereWasDede 5 жыл бұрын

Those animations were AWESOME!! Thanks

@8989youu 5 жыл бұрын

Wow, very clear and to the point. I love it. Definetly worth sharing 😁

@CyberAnalyzer 5 жыл бұрын

Wow, fantastic animations! The content is so deep! I love it!

@lobrundell4264 5 жыл бұрын

Ugh so worth the wait!

@amargasaurus5337 4 жыл бұрын

Those animations are great! Be proud ♥

@briansmithbeta 5 жыл бұрын

The animations really helped me understand some things that had been confusing for me! Thanks!

@JohnnyDoeDoeDoe 5 жыл бұрын

Your absolute best video yet!

@jeanmichelsarr6040 5 жыл бұрын

Great idea, concise, precise.

@Sharklops 5 жыл бұрын

This was fantastic! Very well done. Cheers!

@moneypowertron 5 жыл бұрын

Fantastically intuitive explanation, Robert. The animations were a crucial tool. Thank you for the efforts!

@snfn7847 5 жыл бұрын

Good to see you're still alive

@Gloubichou 5 жыл бұрын

Such a quality video! You must have put so much time into this! Thanks a lot Robert, you're the hero of all ML/AI enthuiasts :D

@NeonStorm5 5 жыл бұрын

Probably the most intuitively informative video I've ever seen.

@hacker6284 5 жыл бұрын

Those animations were totally worth it! Really well done video

@Koffeinsuechtigi 5 жыл бұрын

Thank you for your well crafted explanation!

@vshalts 5 жыл бұрын

Amazing animation and the easiest intuitive explanation of the ideas from Reinforcement learning I have seen so far with a surprising connection with AI safety. It was cool! Thanks!

@solemnwaltz 5 жыл бұрын

The animations are great! I took mental notes specifically on how satisfying and descriptive they are. Well worth the time, in my opinion. c:

@ArtinKavousi Жыл бұрын

you are wonderful Being! for what you doing ! so helpful in these time and age of probabilities!

@Gorabora 5 жыл бұрын

Awesome video and very easy to understand, keep up the good work !

@brunosonza787 5 жыл бұрын

Really excellent video, Robert! I love your videos on computerphile and this one seems to be an even better version that those there, with a clear explanation and neat graphics. Keep it up and Thank you very much!

@rogerab1792 5 жыл бұрын

Really well explained, thanks!

@serenityindeed 5 жыл бұрын

Your animations were really good! Enjoyed the explanation as well.

@willd4686 3 жыл бұрын

Animations were very helpful. I'm not sure how much work they were but I'm grateful that you did them.

@Anymodal 5 жыл бұрын

Dear Rob. Ive learned so much from your videos. Top quality education

@lacielaplante5702 5 жыл бұрын

Your explanation is absolutely outstanding.

@stasisthebest 4 жыл бұрын

Thank you. My deepest respect for visually sharring all of your knowledge. I am certain many people have become at least a slightly better of themselves because of you.

@briancox3922 4 жыл бұрын

Wow, you really are good at explaining these subjects. Thank you.

@Horny_Fruit_Flies 5 жыл бұрын

You have a gift of making the most foreign concepts easily understandable for the layman, such I myself.

@reverse_engineered 4 жыл бұрын

Great job on this video! Your explanations were quite easy to understand and I think the animations helped to explain it. I tend to find diagrams and animations easier to understand than listening to spoken words, so I appreciate the effort you put into those animations.

@dylancope 5 жыл бұрын

The animations were great! Very intuitive video :)

@5ty717 Жыл бұрын

Brilliantly explained

@jonathanquarles3708 5 жыл бұрын

You explained this so clearly, thank you!

@GglSux 5 жыл бұрын

And I really want to thank You for continuing to produce and share Your fantastic content!!! Unfotunately I'm not able to support You (or any other of the many fantastic crestors) so all I can do is to watch everything and express my great gratitude. So a again, a thousand thanks !!! Best regards.

@namelastname8569 5 жыл бұрын

good stuff as always man

@keithklassen5320 5 жыл бұрын

I liked the animations. I probably didn't consciously learn anything from them, but they held my itty-bitty internet-addled attention, thus keeping my eyes on the screen, so they were a part of the learning.

@Ruptured_AU Жыл бұрын

Animations arw SO worth it thanks a lot.

@kennynicoll6277 5 жыл бұрын

This nicely mirrors Kahneman's description of system 1 and 2 in human decision making.

@danielcallegaribr 5 жыл бұрын

Kenny Nicoll hey, this is a great insight!

@aronchai 5 жыл бұрын

I've seen this concept floating around a lot, but didn't really understand it 'til now. Thanks!

@nilp0inter2 5 жыл бұрын

Great work!

@Viniter 5 жыл бұрын

Those animations are really cool!

@randommm-light 4 жыл бұрын

Very nice and understandable. Thx. The limits of architecture in n-dimensions..

@kensmith5694 4 жыл бұрын

I did a thing a little like this for a chess program but my main part was not the "best move finder". The main thing was the "dumb move remover". This was based on recording the game as the program played out a whole game against its self. When the one side lost, there would be a search back through the moves to find the greatest change in board "position". The move just before that was taken to be a bad move and was added to the list of dumb moves. Removing dumb moves quickly saves a lot of processing time. The board position evaluation was not as cheap as it would first appear because unlike is normal today that part was extremely non-linear.

@SapphFire 5 жыл бұрын

Really interesting! The animations are great.

@SHAD0W99V0RTEX 5 жыл бұрын

To be honest, I expected a self-help video about autodidacts but I was pleasantly surprised anyways. Good stuff! This is very ingenious.

@Pedritox0953 Жыл бұрын

Great lecture!

@MrDaanjanssen 5 жыл бұрын

Highly interesting as always, thanks!

@ardweaden 5 жыл бұрын

Absolutely brilliant explanation!

@kanva4 4 жыл бұрын

This is underrated

@BuceGar 5 жыл бұрын

Great video and explanation, doesn't address the fundamental problems we will invariably have with AGI, but shows some of the potential dangers.

@barrettvelker198 5 жыл бұрын

Awesome animations!!!

@DeclanMBrennan 5 жыл бұрын

Crystal clear explanation with no waffle. Thank you. The graphics are so useful, they need their own name. How about didactic visualizations? :-)

@SamB-gn7fw 5 жыл бұрын

Really nice video, explained the topic well

@peto348 5 жыл бұрын

Very high quality video to teach general public something about distillation and amplification. Of course there have to be AI safety somewhere in this video, but I think this kind of video is also good for someone who is interested in AI in general.

@gloverelaxis 5 жыл бұрын

Animations were worth it. They help immensely

@hosmanadam 5 жыл бұрын

Your videos are perfectly optimized to be easily processed by my learning function.

@thrallion 5 жыл бұрын

legit my favourite channel on youtube by far

@SJNaka101 5 жыл бұрын

Hmmm I dunno if I can top this channel for you, but looking at your subs I would take a few wild shots in the dark... check out Chessnetwork, Summoning Salt, Numberphile and Computerphile, and What I Learned. I suspect you will greatly enjoy at least a couple of those

@thrallion 5 жыл бұрын

@@SJNaka101 hey thanks, good guesses as i already watch all those except what I learned :) will look into it

@GoatzAreEpic 5 жыл бұрын

Absolutely amazing and helpful for learning strategies as well( learning to become a front end dev atm)

@Hexanitrobenzene 5 жыл бұрын

Yay ! We missed you, Rob :)

@Lufernaal 5 жыл бұрын

Loved the video

@greatbullet7372 5 жыл бұрын

Best KZfaq Video of the Month

@DisfigurmentOfUs 5 жыл бұрын

A very valuable video for me, thank you.

@dylancope 5 жыл бұрын

How did I miss this?! I can't believe I hadn't "hit the bell" on this channel yet.

@reidwallace4258 4 жыл бұрын

This is giving me flash backs to the dune novels. Paul was just doing treesearch all along.

@lewisleslie2821 4 жыл бұрын

Reid Wallace i read dune for the first time last month, that’s a great comparison

@DamianReloaded 5 жыл бұрын

Worth watching a few times! ^_^

@StevenAkinyemi 5 жыл бұрын

Can't wait for the next video! I'm not sure alignment can be maintained the more complex an agent becomes. There will always be abstraction difference between what we want it do and what it does to optimize itself. This means we have to always tune the alignment as the agent becomes more complex. There is perhaps a point where the agent's comprehension of the universe explodes beyond our grasp and we won't be able to align it at that point. In fact, we might have to restrict it's optimization process when we discover its intelligence is getting beyond our control. These are just theories in my head.

@GuuraHeavenbound 4 жыл бұрын

Wooo! Said Polat! I've been following Seed (their Webtoon narrating the birth of a super AI) since it got featured on the platform ^^ I'm watching this video kinda late, but I think it's neat "how small the world can be". Also, really informative and interesting video Robert! ...I'm totally not binge-ing all of your uploads. Nope, nuh-uh. ....promise :3

@SirDanSax 5 жыл бұрын

Thank you for the fantastic videos! I've been following you on and off for a while and learned a lot. You provide some great insights on not just AI safety, but how it works. Do add some animations if you want to go mainstream 😅 cheers!

@TheMenIdo 4 жыл бұрын

This is brilliant

@roberttomsiii3728 5 жыл бұрын

Thank you for being MY amplified agent.

@bejoscha 5 жыл бұрын

Really good animations. I know that it must have been a lot of work, but it was worth it. For visual types like myself, a little illustration can go a long way.

@ulissemini5492 5 жыл бұрын

awesome! this makes so much sense! this is exactly how i get better at chess, play a game quickly, then go back and calculate a lot to find the better moves, then improve my intuition! its so awesome that you said it in such a way that now i feel like i can write a program to become superhuman at anything :D

@nielsgroeneveld8 5 жыл бұрын

Few lectures have been as unbelievably good as this one.

@AxeSovax 5 жыл бұрын

Amazing work as always. I wonder if a video discussing biometrics in AI safety is on the cards. It will be a major hurdle in the future and one that appears to be inevitable.