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

Подписаться 178 тыс.

Просмотров 157 тыс.

50% 1

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

Наука

Опубликовано:

13 июн 2024

Ссылка:

Скачать:

Готовим ссылку...

Добавить в:

Мой плейлист

Посмотреть позже

Комментарии : 384

@QuantPhilosopher89 3 месяца назад

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 3 месяца назад

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 3 месяца назад

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

@attilakun7850 3 месяца назад

So Demis, do you believe in aliens 😂

@sup3a 3 месяца назад

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

@JakeWitmer 2 месяца назад

Better than most.

@DwarkeshPatel 3 месяца назад

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

@zandrrlife 3 месяца назад

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

@theWACKIIRAQI 3 месяца назад

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

@ChuckNorris-lf6vo 3 месяца назад

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

@Rareme530 3 месяца назад

Subscribed. Spot on.

@ivoryas1696 3 месяца назад

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

@byrnemeister2008 3 месяца назад

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 3 месяца назад

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

@hugopennmir 3 месяца назад

mmm Lex ?

@moviesfan5513 3 месяца назад

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

@user-tl5ok1ff6k 3 месяца назад

@@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 3 месяца назад

Sure, sure, sure, yeah, yeah, yeah

@zaaath 3 месяца назад

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 3 месяца назад

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

@hardheadjarhead 3 месяца назад

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

@godspeed133 3 месяца назад

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

@TedToal_TedToal 3 месяца назад

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

@marka1000 Месяц назад

Inference engine go brrrrrrr

@akhil090579 3 месяца назад

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

@xsuploader 3 месяца назад

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

@user-tl5ok1ff6k 3 месяца назад

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 3 месяца назад

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

@dovekie3437 3 месяца назад

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

@NoXic88 3 месяца назад

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

@HuxleyCrimson 3 месяца назад

The intervewer interventions sound to be sped up honestly

@matthewburson2908 2 месяца назад

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

@AtomicSlugg 3 месяца назад

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

@StrandedKnight84 3 месяца назад

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.

@forthehomies7043 3 месяца назад

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

@wdwuccnxcnh7022 3 месяца назад

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

@BrianMosleyUK 3 месяца назад

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

@bazstraight8797 3 месяца назад

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

@willlounsbery-scaife6255 3 месяца назад

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.

@riazr88 3 месяца назад

Yo Papa Patel interviewing the heavy hitters. Much appreciated

@klarad3978 3 месяца назад

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

@videowatching9576 3 месяца назад

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

@bobby288m 3 месяца назад

my money is on amphetamines

@HuxleyCrimson 3 месяца назад

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

@Jonathan2x6 2 месяца назад

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.

@ulychun 3 месяца назад

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

@nicohillbrand3124 3 месяца назад

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

@hamandchees3 3 месяца назад

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

@agginswaggin 3 месяца назад

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

@En1Gm4A 3 месяца назад

Thanks for the Interview. Very much appreciate it!

@confusedwouldwe 3 месяца назад

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 😆

@yoyo-jc5qg 3 месяца назад

Cool, 1.5x speed without the 1.5x playback speed

@umberto488 3 месяца назад

Lots of brain power

@user-kh9hq6iw1o 3 месяца назад

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

@derekjordan1958 3 месяца назад

the youtube url ends with AI

@VesperanceRising 3 месяца назад

Awesome

@LiquidRR 3 месяца назад

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

@noone-ld7pt 3 месяца назад

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 3 месяца назад

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

@CristianDobre983 3 месяца назад

Only one hour… 😢

@the_badass_bond___ 3 месяца назад

It should've been 3 hours mann

@conformist 3 месяца назад

better than zero!

@ansumansamal8473 3 месяца назад

But it's a quality 1hr

@the_badass_bond___ 3 месяца назад

@@ansumansamal8473 so ja bhai. Rat ho gya

@Aziz0938 3 месяца назад

@@the_badass_bond___tu chup

@bastost 3 месяца назад

This is becoming one of my favorite podcasts.

@Josh-dp1uw 3 месяца назад

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

@mattwesney 3 месяца назад

What a fantastic interview and interviewer! Kudos

@fredrik241 3 месяца назад

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 3 месяца назад

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!"

@michaelneumaier5912 2 месяца назад

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.

@propilot7428 3 месяца назад

You ask really good questions! Very good work 👌

@asherifs9383 2 месяца назад

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.

@trevorsmissaert5687 3 месяца назад

great questions for sure. Very good interview.

@treesandgeeking 3 месяца назад

Great stuff Dwarkesh ❤

@En1Gm4A 3 месяца назад

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

@vish202uk5 3 месяца назад

Truly inspiring. Fantastic talk.

@andersfant4997 3 месяца назад

Great interview👍

@parisfrancepp 3 месяца назад

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

@adriansomor 3 месяца назад

Really cool videos, thanks for doing this!

@sioncamara7 3 месяца назад

10 minutes in and this is excellent!

@PythonAndy 3 месяца назад

Very good interview, well done 🎉

@stevenahrens1655 3 месяца назад

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

@collins4359 3 месяца назад

love demis. he's very open.

@dialetheic 3 месяца назад

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

@93836 3 месяца назад

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

@nescirian 3 месяца назад

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 3 месяца назад

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

@hassb00zy 18 дней назад

my optimism is through the roof ❤

@netscrooge 3 месяца назад

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

@travisstewart8655 2 месяца назад

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

@user-mu3th2bz3g 3 месяца назад

Great interview

@alexmolyneux816 3 месяца назад

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 🔥

@ModernCentrist 3 месяца назад

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

@IyamwhoIyam 3 месяца назад

Perfect listening for a walk in the forest. ✌️

@FORGOTTEN-THEORY 3 месяца назад

AGI in the next 5 minutes!

@LegendStudioGames 3 месяца назад

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

@dr_cheez811 3 месяца назад

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 3 месяца назад

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

@jondor654 3 месяца назад

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 .

@PR-kz9kx Месяц назад

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.

@jakek82 3 месяца назад

Thanks!

@oraz. 3 месяца назад

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.

@dr.mikeybee 2 месяца назад

Are you using elastic weight consolidation and curricular training?

@workingTchr 3 месяца назад

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

@joshuasellers4434 3 месяца назад

Let's go!

@Sandakahleh Месяц назад

Get Ilya Sutskever on the podcast!

@bmoney6482 3 месяца назад

Wish it was longer

@b_two 3 месяца назад

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

@viaMac 3 месяца назад

Needs more views

@JakeWitmer 2 месяца назад

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).

@egor.okhterov 3 месяца назад

"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"

@nicklausbrain 3 месяца назад

What makes it best model in the world exactly?

@dr.mikeybee 2 месяца назад

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

@filosofiahoy4105 2 месяца назад

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.

@afterthesmash 17 дней назад

Once the models are grounded, multimodal, with decent conceptual generalization, eliminating capabilities by selectively suppressing training data can't be guaranteed to work. The two criteria strike me as directly adversarial: to the degree that selective suppression of training data works, is also the degree to which the model isn't very powerful at routing around damage (selective data lobotomy).

@conall5434 3 месяца назад

Lets go!!!

@hassb00zy 18 дней назад

actually next week ❤

@therantingboy 2 месяца назад

Watch at 0.75 speed to survive

@vinipoars 3 месяца назад

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

@Glowbox3D 3 месяца назад

Let's get this right!

@user-dd2dd3fm7f 3 месяца назад

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

@MrRonjivRoy 3 месяца назад

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.

@abhyudaypatel 3 месяца назад

Exciting

@gysbertbremer8284 3 месяца назад

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

@mahon257 3 месяца назад

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.

@JC-ji1hp 3 месяца назад

Banger

@mrpicky1868 3 месяца назад

by what metric are you calling Gemini best model? by google advertisement statements?

@skierpage 3 месяца назад

By the incredible Google video where Gemini was watching the presenter and cracking jokes while he drew a rubber ducky and put a map on the table. Amazing stuff, shame it was faked.

@ShawnFumo 3 месяца назад

In fairness, I think 1.5 Pro can be considered better in terms of dealing with long context. I got access today and it is pretty impressive so far. Like with a half hour video tour of a museum, I asked which exhibit had an animal digging near water and who gifted it to the museum. And it got it even with the length and the fact the name placard, animals, and gift placard weren’t all on screen at the same time.

@ShawnFumo 3 месяца назад

Google has been shooting themselves in the foot a lot. Instead of doing that stupid fake marketing video, they should have waited for 1.5, when they could actually show something very impressive. 1.0 Ultra is basically just GPT-4 with pluses and minuses (overall more minuses). 1.5 is something different altogether. We haven’t seen a model until this one that can actually reason well over a very large context. Hopefully they’ll be able to roll it out to more people without a huge delay. Am hoping it is a good sign I got past the waitlist, since I’m not a famous influencer or anything like that.

@oraz. 3 месяца назад

@@ShawnFumoThere are still many answers that will be manipulated or not allowed. I guess it's up to people to decide if they are fine with that.

@roomo7time 3 месяца назад

This is gold

@JakeWitmer 2 месяца назад

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.

@kavinvikram2960 3 месяца назад

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

@41-Haiku 3 месяца назад

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!

@akratlapidus2390 3 месяца назад

Very interesting interview but the acceleration effect was annoying.

@ryzikx 3 месяца назад

very hella swag podcast

@DAG_42 Месяц назад

Has Hassabis considered these models may inadvertently develop an "identity"? We've seen cases like early oopsies Bing Chat that talks like it has personhood, but I expect that was closer to chatbot than human. What if it went much further and deeper?

@kuang-chengyushankao1395 3 месяца назад

謝謝！

@filosofiahoy4105 2 месяца назад

Yeah ok, but a question ¿general intelligence is the top intelligence that can be achieved by a general intelligent machine? It's very naïve to think that a AGI robot didn't scale at exponential level immediately after reach it's first stage. Cracking up any conceivable and possible way of human think concept. What'd occurs in AGI1.2 or AGI2.1? etc. OF course this is the very end of human cerebral capabilities ERA.