Demis Hassabis - Scaling, Superhuman AIs, AlphaZero atop LLMs, Rogue Nations Threat

Рет қаралды 144,461

Күн бұрын

Here is my episode with Demis Hassabis, CEO of Google DeepMind
We discuss:
- Why scaling is an artform
- Adding search, planning, & AlphaZero type training atop LLMs
- Making sure rogue nations can't steal weights
- The right way to align superhuman AIs and do an intelligence explosion
Transcript: www.dwarkeshpatel.com/p/demis...
Apple Podcasts: podcasts.apple.com/us/podcast...
Spotify: open.spotify.com/episode/6SWb...
Follow me on Twitter: / dwarkesh_sp
Timestamps
0:00:00 Nature of intelligence
0:05:56 RL atop LLMs
0:16:31 Scaling and alignment
0:24:13 Timelines and intelligence explosion
0:28:42 Gemini training
0:35:30 Governance of superhuman AIs
0:40:42 Safety, open source, and security of weights
0:47:00 Multimodal and further progress
0:54:18 Inside Google DeepMind

Пікірлер: 364

@QuantPhilosopher89 2 ай бұрын

It's really awesome to finally have a podcast host who is capable of asking the right questions and of of having access to the key people.

@CyberKyle 2 ай бұрын

I agree! I would check out the Hard Fork podcast episode with Demis too, I was surprised at the level of discussion for the non-technical hosts, they asked great questions that teased a lot of exciting stuff from Demis.

@MrMikkyn 2 ай бұрын

I agree. This is one of the better podcasts between a host and an AI expert

@attilakun7850 2 ай бұрын

So Demis, do you believe in aliens 😂

@sup3a 2 ай бұрын

Agreed 100%. Often feels like a better Lex (and I say this as a long-time fan of Lex)

@JakeWitmer Ай бұрын

Better than most.

@DwarkeshPatel 2 ай бұрын

Hope you guys enjoy this interview with Demis. If you do, please share it! Helps a lot :). Thanks for watching 🙏

@zandrrlife 2 ай бұрын

So refreshing to have a true technical interviewer. You ask the REAL questions. Glad I found your channel.

@theWACKIIRAQI 2 ай бұрын

All your interviews are a breeze. Very informative and easy to follow through all topics touched. Thank you.

@ChuckNorris-lf6vo 2 ай бұрын

Wow this guest is very talkative like bla bla bla bla bla what a machine what a machinegun of words.

@Rareme530 2 ай бұрын

Subscribed. Spot on.

@ivoryas1696 2 ай бұрын

@DwarkeshPatel Almost wish it was _longer, but nice job anyway!

@byrnemeister2008 2 ай бұрын

This guy is just about the best interviewer of tech folks. Excellent questions. Some thought has been applied rather than sticking to what everyone else is asking.

@MrMikkyn 2 ай бұрын

I agree. I find the other podcasts tend to ask computer generated sounding questions.

@hugopennmir 2 ай бұрын

mmm Lex ?

@moviesfan5513 2 ай бұрын

@@MrMikkynDwarkesh is computer science graduate so infinitely more technically competent than an average journalist.

@user-tl5ok1ff6k 2 ай бұрын

@@hugopennmir Lex is kind of a naive child, I don't enjoy his interviewing style its incredibly biased and he hardly pushes back at all letting gremlins like Tucker Carlson run roughshod over him with their bevy of lies. He's too idealistic and his questions and subsequent conversaion are too unfocused on the tech.

@egor.okhterov Ай бұрын

Sure, sure, sure, yeah, yeah, yeah

@zaaath 2 ай бұрын

This is a great conversation! It would've been even better if Dwarkesh's speech were clearer. I think Dwarkesh optimizes for speed which at times makes it inaudible. In any case, I'm looking forward to more guests like Demis.

@HuxleyCrimson Ай бұрын

Totally agree. I even suspected that his interventions were sped up

@hardheadjarhead 2 ай бұрын

I had no idea what they were talking about. But the interviewer clearly is the fastest speaking human being on the planet.

@godspeed133 2 ай бұрын

clearly a next level SOTA technical interviewing specialised LLM with a photorealistic animated avatar generated on the fly.

@TedToal_TedToal Ай бұрын

If you speed up the interviewer, it is completely unintelligible.

@akhil090579 2 ай бұрын

Its so great you do these interviews and ask intelligent questions that require non trivial answers

@xsuploader 2 ай бұрын

yh. Most interviewers ask the same dumb generic questions. But not dwarkesh

@BrianMosleyUK 2 ай бұрын

Great episode Dwarkesh and some insightful questions throughout. Well done. Shared to Maven 🙏👍

@StrandedKnight84 2 ай бұрын

You ask the best questions, Dwarkesh! Speaking from the perspective of someone who has worked in the field of AI for several years, I'm impressed by how well read you are on technical concepts.

@wdwuccnxcnh7022 2 ай бұрын

So excited for this. You're really an amazing interviewer. Ty ❤

@riazr88 2 ай бұрын

Yo Papa Patel interviewing the heavy hitters. Much appreciated

@willlounsbery-scaife6255 2 ай бұрын

Excellent podcast. You ask such good questions, and it is clear that you put a lot of work and deep thought into preparing for these interviews.

@user-tl5ok1ff6k 2 ай бұрын

I listened to this at .95 speed. The change might seem insignificant, but it was actually incredibly helpful in slowing down the high density stream of information just enough to be able to catch every word. I highly reccomend anyone else struggling to keep up with the question-answer pace to do the same.

@erang4437 2 ай бұрын

That applies to the interviewer, Demis speaks just at the right speed.

@dovekie3437 Ай бұрын

.95 speed is probably closer to what it was recorded at. The normal version clearly has been sped up.

@NoXic88 Ай бұрын

It's helpful turning the closed captions (subtitles) on too :)

@HuxleyCrimson Ай бұрын

The intervewer interventions sound to be sped up honestly

@matthewburson2908 Ай бұрын

@@erang4437 Yeah. Kinda makes this whole thing garbage lol.

@forthehomies7043 2 ай бұрын

Huge. Looking forward to watching this after classes. Thanks.

@bazstraight8797 2 ай бұрын

20 minutes in "Grounding (in reality)", I had to pause: these are astute questions from Dwakesh and great replies from Demis.

@klarad3978 2 ай бұрын

I love Dwarkesh’s intelligent and pointed questions, it’s refreshing.

@nicohillbrand3124 2 ай бұрын

Your podcast has become my new favourite, overtaking the 80k podcast. Really love the quality of your questions :)

@AtomicSlugg 2 ай бұрын

genuinely the best interviewer when it comes to tech/AI, excited to watch this, thank you!

@CristianDobre983 2 ай бұрын

Only one hour… 😢

@the_badass_bond___ 2 ай бұрын

It should've been 3 hours mann

@conformist 2 ай бұрын

better than zero!

@ansumansamal8473 2 ай бұрын

But it's a quality 1hr

@the_badass_bond___ 2 ай бұрын

@@ansumansamal8473 so ja bhai. Rat ho gya

@Aziz0938 2 ай бұрын

@@the_badass_bond___tu chup

@confusedwouldwe 2 ай бұрын

I set playback speed to 1.25 at the start of the vid (I often do for these kinds of podcast-type interviews), god damn I was not prepared for how fast Dwarkesh talks 😆

@agginswaggin 2 ай бұрын

Your channel is gold! I love your questions, and crazy how you can get all these big guys on here

@ulychun 2 ай бұрын

Dwarkesh meets Demis ! Fantastic interview ! Every sentence is gold…❤

@hamandchees3 2 ай бұрын

Given the pace of change I'd love if you'd include the date of recording in the description.

@En1Gm4A 2 ай бұрын

Thanks for the Interview. Very much appreciate it!

@user-kh9hq6iw1o Ай бұрын

Amazing background research by Dwarkesh, citing old research papers from Demis is great.

@log0n 2 ай бұрын

You are my favorite interviewer. Thank you for the interesting content!

@videowatching9576 2 ай бұрын

Great interview! Is this accelerated beyond 1.0x? Such as 1.1x or 1.2x playback as the default?

@bobby288m 2 ай бұрын

my money is on amphetamines

@HuxleyCrimson Ай бұрын

@@bobby288mlol that's what I secretly thought haha 😅😂

@Jonathan2x6 Ай бұрын

Mumbo jumbo to deliberately obfuscate. The interviewer just rambles on instead of keeping the question short. Also got a fake accent that falls flat at times.

@Josh-dp1uw 2 ай бұрын

You did a really good job and I learned a lot without losing interest. Thank you.

@bastost 2 ай бұрын

This is becoming one of my favorite podcasts.

@noone-ld7pt 2 ай бұрын

This podcast is so incredibly underrated. Fantastic conversation with one of the most important people in the field. Your research and knowledge really shines here and serves as the foundation for some really insightful deepdives. Thank you so much for this!

@arturt7192 2 ай бұрын

Finally a sharp interviewer who knows what he is talking about. Great job!

@treesandgeeking 2 ай бұрын

Great stuff Dwarkesh ❤

@asherifs9383 25 күн бұрын

Hey love your Interview sometimes hard to understand your questions, your a fast speaker. But your very good at what you do, thank you for the content.

@michaelneumaier5912 Ай бұрын

Great episode, always amazing people you are able to interview! Other than the content itself, I feel like you're 99% there in terms of production quality. The audio is sounding better than it has in the past which is a welcome change, but I can't help but notice the video having some auto exposure artifacts. It is somewhat distracting to see the exposure change when someone is moving in a certain way. Also the exposure in general seems a little overexposed. Just trying to give some technical feedback! I always look forward to your interviews! I can't wait to see what these conversations will be about in 2 or 3 years when we understand more.

@mattwesney 2 ай бұрын

What a fantastic interview and interviewer! Kudos

@propilot7428 2 ай бұрын

You ask really good questions! Very good work 👌

@vish202uk5 2 ай бұрын

Truly inspiring. Fantastic talk.

@fredrik241 2 ай бұрын

Demis: These system are going to be enormously powerful and transformative to society! also Demis Demis: Because these systems are potentially dangerous if falling into the wrong hands only selected people and institutions should be able to wield them.

@41-Haiku 2 ай бұрын

Also Demis: "We don't know how to control these systems once they become smarter than us, but just trust us, we'll figure it out later!"

@trevorsmissaert5687 2 ай бұрын

great questions for sure. Very good interview.

@andersfant4997 2 ай бұрын

Great interview👍

@yoyo-jc5qg 2 ай бұрын

Cool, 1.5x speed without the 1.5x playback speed

@umberto488 2 ай бұрын

Lots of brain power

@adriansomor 2 ай бұрын

Really cool videos, thanks for doing this!

@sioncamara7 2 ай бұрын

10 minutes in and this is excellent!

@derekjordan1958 2 ай бұрын

the youtube url ends with AI

@VesperanceRising 2 ай бұрын

Awesome

@En1Gm4A 2 ай бұрын

It's the most sophisticated asking I have see on the topic last 7 month.

@travisstewart8655 Ай бұрын

Sounds like I’m watching a regular podcast at 1.5 speed

@PythonAndy 2 ай бұрын

Very good interview, well done 🎉

@parisfrancepp 2 ай бұрын

Incredibly useful, Chelsea Finn is clearly one of the most inspiring robotics researcher nowadays

@PR-kz9kx 15 күн бұрын

Awesome feedback. A piece of advice for the interviewer. Please don't rush when you pose a question. Slow down and speak clearly. IN several cases the posed questions were disregarded due to the speed.

@stevenahrens1655 2 ай бұрын

Dwarkesh is purely an AI simulation that Demis was testing. Notice how Demis had him running at wildly different playback speeds throughout the interview.

@alexmolyneux816 2 ай бұрын

I love this more technical style interviews from someone who knows the domain. This is missing from the internet. I will love you to do a similar with Sam Altman 🔥

@netscrooge 2 ай бұрын

Great work! By the way, it would be great if you would always post the interview date.

@user-mu3th2bz3g 2 ай бұрын

Great interview

@IyamwhoIyam 2 ай бұрын

Perfect listening for a walk in the forest. ✌️

@collins4359 2 ай бұрын

love demis. he's very open.

@FORGOTTEN-THEORY 2 ай бұрын

AGI in the next 5 minutes!

@joshuasellers4434 2 ай бұрын

Let's go!

@jondor654 2 ай бұрын

Just curious , in a differential piece game like chess is there any role for lower order /complexity auxillary modelling purely on a board demographic , ie positional state .

@SubstratumMaxima Ай бұрын

Interviewer should slow down a bit, that much speed of speech makes thing almost unintelligible 😅

@HughBlackstone-tm6bw 12 күн бұрын

IQ issue

@jakek82 2 ай бұрын

Thanks!

@conall5434 2 ай бұрын

Lets go!!!

@K1NG.Z3R0 2 ай бұрын

This podcast is the only one I listen to at 1x speed O.O great job with the quick pacing, Dwarkesh!

@93836 2 ай бұрын

After rewinding a bunch of times, I eventually just set it to 0.5x speed for me. Haha.

@nescirian 2 ай бұрын

I listened to it at 1.5x because that's what I'm used to, but I did have to stop and check that it hadn't somehow ended up at 2x. Definitely seemed faster than usual.

@ashh3051 2 ай бұрын

I'm listening at 0.75x. The pace is too intense for me.

@ModernCentrist 2 ай бұрын

I love interviews like this that get a bit more technical.

@dr_cheez811 2 ай бұрын

Have you considered running the mic through a sound board and equipping you and your guests with monitor headphones? It might be a good way to prevent guests from wandering away from the mics if they can hear themselves.

@dr_cheez811 2 ай бұрын

I say sound board because my experience with digital monitors has been closer to a speech jammer.

@abhyudaypatel 2 ай бұрын

Exciting

@viaMac 2 ай бұрын

Needs more views

@kuang-chengyushankao1395 2 ай бұрын

謝謝！

@oraz. 2 ай бұрын

It's weird that a company can have a division with the best minds and then another of quasi technical ethics ladies and functionaries are allowed to make the final product into farce.

@workingTchr 2 ай бұрын

We all now know that "weights" are where the big security danger lies.

@roomo7time Ай бұрын

This is gold

@Sandakahleh 17 күн бұрын

Get Ilya Sutskever on the podcast!

@egor.okhterov Ай бұрын

"Sure, sure, sure, yeah, yeah, yeah" - it translates something like this "please, shut up. I'm not listening to you anymore. The things you are talking about now are not interesting to me. I want to talk myself now instead of you"

@LegendStudioGames Ай бұрын

Does anyone know what the make and model of those chairs are? They look slick.

@therantingboy Ай бұрын

Watch at 0.75 speed to survive

@filosofiahoy4105 29 күн бұрын

It will be a real challenge for an IAG to "feel" what I felt witnessing the eclipse today. Because It can't be explained with words.

@JC-ji1hp 2 ай бұрын

Banger

@user-dd2dd3fm7f Ай бұрын

When published, it "was" the best model. Not today.

@KhalkeionGenos 2 ай бұрын

While I appreciate the interview, I can't help but feel disappointed you didn’t address the elephant in the room - the growing concerns about potential biases, especially the alleged anti-white bias, in Google's new AI image generator.

@vinipoars 2 ай бұрын

Brilliant humble man who is fronting maybe the most important endeavour of humankind right now. Deep respect for Dr. Demis Hassabis. Great interviewer.

@b_two Ай бұрын

Can we get the podcast in video form on Spotify please 🙏

@MrRonjivkroy 2 ай бұрын

Demis called LLMs unreasonably effective. Shouldn't we figure out the reason why it's as effective as it is. And furthermore, how does higher quality or extra data improve its IQ in certain domains? If something this unprecedented is "unreasonably effective", we should figure out the method to the madness instead of blindly creating more madness.

@bmoney6482 Ай бұрын

Wish it was longer

@Glowbox3D 2 ай бұрын

Let's get this right!

@JakeWitmer Ай бұрын

56:40 Speaking of "Unintended Consequences," that's a book Demis should read. The John Ross one with "the rape of liberty" art on the cover (there are two books called "Unintended Consequences" by two different guys named John Ross).

@dr.mikeybee Ай бұрын

Are you using elastic weight consolidation and curricular training?

@dr.mikeybee Ай бұрын

I think the best way to study the activation space is by perturbation analysis which can be automated.

@ChipWhitehouse 2 ай бұрын

FANTASTIC interview but have you ever noticed that you, Dwarkesh, speak like you are on 1.5x speed? 😭😭😭 Kind of obsessed and I don’t know how you talk like that.

@malimal4972 2 ай бұрын

it's sped up

@ChipWhitehouse 2 ай бұрын

@@malimal4972 I don’t believe that… why would he speed up his part and not Demis’s? Demis talks at his normal speed?

@malimal4972 2 ай бұрын

read other comments

@ChipWhitehouse 2 ай бұрын

@@malimal4972 sounds like a lotta work… 🤔

@mahon257 2 ай бұрын

Many models in the brain are inherited (innate) - such as being able to catch a ball, under the effect of gravity and air resistance. Therefore, human brain models are developed, then improved across generational boundaries. The amount of data used to train our internal models is therefore captured across 10's of thousand of years. With AI, we are massively accelerating that process.

@gysbertbremer8284 2 ай бұрын

Empirical truth. Nascent AGI suits men of thunder. Walking light on the battlefield.

@JakeWitmer Ай бұрын

Stanley Milgram's "Obedience to Authority: An Experimental View" needs to be discussed at the center of every discussion about why both AGI and existing humans are so unwise. We know the answers...we're just too cowardly to face them and choose wisely.

@ryzikx 2 ай бұрын

very hella swag podcast

@nicklausbrain 2 ай бұрын

What makes it best model in the world exactly?

@ms-mr8kg 2 ай бұрын

Interesting how youtube comments are having a meta discussion about the conversation instead of the conversation itself.

@skierpage 2 ай бұрын

Folks, if you like the video click the thumbs up button. There's NO NEED to write about how much you liked it!

@SigmoidalHive 2 ай бұрын

what a rehabilitating episode

@altruistx Ай бұрын

Dwarkesh loves pushing doom scenarios over positive possibilities and opportunities. ngmi

@420_gunna 2 ай бұрын

Do Nate Lambert on practical/open-source alignment research! 😄

@420_gunna 2 ай бұрын

I think he'd have something something to say about Demis's claims at 43:00 re: their work being either "open source" _or_ "open science" 😄

@akratlapidus2390 2 ай бұрын

Very interesting interview but the acceleration effect was annoying.

@juliandunn8412 Ай бұрын

This is fundamentally the last technology humans will ever invent. Once AGI is achieved and released, it will make all of our future inventions...

@hypercube717 2 ай бұрын

Interesting.

@macabrew 2 ай бұрын

Decade? You can say that, we're basically there.

@kavinvikram2960 2 ай бұрын

Quick question: Who will win the AI race - Open AI or Google ? Why ? (share your thoughts please)

@41-Haiku 2 ай бұрын

The AI will win, and humanity will lose. Listen to him, he admits that no one knows how to make AGI safe, and they're racing forward anyway!