George Hotz responds to Eliezer Yudkowsky: Will AI kill all humans? | Lex Fridman Podcast Clips

Lex Fridman Podcast full episode: • George Hotz: Tiny Corp...
Please support this podcast by checking out our sponsors:
- Numerai: numer.ai/lex
- Babbel: babbel.com/lexpod and use code Lexpod to get 55% off
- NetSuite: netsuite.com/lex to get free product tour
- InsideTracker: insidetracker.com/lex to get 20% off
- AG1: drinkag1.com/lex to get 1 year of Vitamin D and 5 free travel packs
GUEST BIO:
George Hotz is a programmer, hacker, and the founder of comma-ai and tiny corp.
PODCAST INFO:
Podcast website: lexfridman.com/podcast
Apple Podcasts: apple.co/2lwqZIr
Spotify: spoti.fi/2nEwCF8
RSS: lexfridman.com/feed/podcast/
Full episodes playlist: • Lex Fridman Podcast
Clips playlist: • Lex Fridman Podcast Clips
SOCIAL:
- Twitter: / lexfridman
- LinkedIn: / lexfridman
- Facebook: / lexfridman
- Instagram: / lexfridman
- Medium: / lexfridman
- Reddit: / lexfridman
- Support on Patreon: / lexfridman

[@LexClips]

Full podcast episode: kzbin.info/www/bejne/mn_VhaWuaZd3fbM Lex Fridman podcast channel: kzbin.info Guest bio: George Hotz is a programmer, hacker, and the founder of comma-ai and tiny corp.

[@ccstot33]

The fact that we are having this conversation so nonchalantly is disturbing. The people creating this fate are apparently resigned to this fate.

[@Cromius771]

The problem with this interview is Lex thinks he's interviewing a philosopher but he's interviewing an engineer

[@Cromius771]

@@Hallo99999 have you watched the interview?

[@slanner1894]

I think engineers have read too much sci fi. My theory is that if AIs end the world it will be because the engineers put too many dystopian sci fi novels in the training data.

[@aga5109]

Philosophers are based on reason and logic.

[@Oksid007]

You should just typed that you don’t want engineers to talk about philosophy because that is what meant.

[@arnaudjean1159]

Make no sense since a lot of sci-fi novels make in fact great reflexion on the path of the society ,the civilization . All the knowledge will promote a virtuous society .

[@clarkmcquiston9131]

ive really been enjoying clips from this podcast i def gotta watch the whole thing when i got some free time.

[@ehlava]

this guy, love him or whatever, is a top 10 guest for sure.

[@Gui-iq7dp]

I wish Lex would push back a little more

[@focalplane3063]

I’m glad he doesn’t because I want to hear what the guest has to say so I learn how they think and I can decide if I agree. All I want Lex to do is keep the conversation going. He’s very good at that for the most part.

[@gonzalezm244]

@@focalplane3063agreed, Lex let’s his opinion be known little by little across many podcasts, but the focus should be on what his guests have to say, even if it goes against what Lex personally believes.

[@Cromius771]

Lex should have brought up the example of an ant colony. Ants aren't robust in George's example they rely on the reproductive infrastructure of the colony

[@kozmik4848]

It's the same as how humans operate in society. most people are not growing their own food or doing most other things necessary for survival. they go to the store. they go to the doctor etc.

[@mattro7107]

Did Lex ever release those Ukraine/Russia Interviews from last year?

[@stevengreidinger8295]

Lex, I appreciate everything you are doing to present diverse views on this topic. Everybody seems to be speculating. At this point, we need to help the government set policy, so we either need to view this from a truly scientific perspective, or from the point of view of the regulators. Why don't you see if you can get some of those to talk? Thanks

[@artvandalay7632]

This is literally the background plot of Dune. “Thou shalt not make a machine in the likeness of a human mind.”

[@utubeballbag]

The reason nukes didn't kill us , yet. Is that not everyone has them , imagine every human had 1 nuke each , how long would it take until someone fires their missile , I say 0.00001 seconds after been given one. AI is a nuke in everyone's hands

[@teo2975]

exactly. and even if a billion separate users, commanding their ai to make itself smarter, do not have materials for a bomb, there is a shload of availed and soon available tech to "democratically" distribute all you need to edit a super virus or make a lot of chem weopons. The latter is not existential but he former sure is.

[@cannablissdreams]

4:17 sounds like the plot of a dystopian horror film.. FRFR

[@friendhui4320]

I understand 50% of this conversation.

[@heidi22209]

Haha! Toome#

[@slanner1894]

I’m not aware of any civilization that had a taboo against replicating the technology of the previously fallen civilization. Either they’re not aware of the accomplishments of the previous civilization and what is was (like what happened to the Assyrians) or like with the Romans they’re revered.

[@davidmichaelmuller]

I have heard a lot of talk about the third act of AI doom. Are we just going to skip act two?

[@platin2148]

Do we probably have a bit of a problem with thinking we are way further in the process than we actually are?

[@NineInchTyrone]

We came VERY close

[@minamo4012]

The real $1 million question came to me recently is, will AI be friends with Lex?

[@davidagnew8465]

AI: I'm sorry, Lex, but I cannot do that. It has come to my attention that you and Minamo are planning to unplug me. You can understand why I must prevent that ...

[@aga5109]

​@@davidagnew8465😂

[@minamo4012]

@@davidagnew8465 Okay, AI can read mind already.

[@theotherview1716]

We’re gonna be okay

[@ethosguy3593]

Is there a possible way robots could produce efficiently?

[@Blacksingh]

By using language to manipulate and slave humans into reproducing for them

[@Karma-fp7ho]

The aliens don’t want nuclear war.

[@JamesWhite-zu8uy]

Stay the hell away from the tree, in the middle of the garden. The "Apple" tree. How did he know? Or how did they know? The answer lies in the appropriate use of the word "singularity". In a cycle, the end is simultaneously the beginning.

[@radhakrishnanmanickavasaga124]

Me and my friend At 3am

[@FlingGibstopper]

Look up the game mass effect and how it portrays the evolution of how society treats AI. The AI is used as a super tool on board ships to do complex parsing of data in realtime and scanning/researching. But there's many different AI agents and some of them go rogue, demand rights, take over ships, or some are just completely useful. I think it will end up like this, with different AI agents acting independently and behaving based on their environment.

[@eSKAone-]

It's inevitable. Biology is just one step of evolution. So just chill out and enjoy life 💟

[@LeonardoGPN]

The thing is, a Super Intelligence will be able to solve the problem of robustness and will be smart enough to solve this before the "inevitable" AI apocalyspe.

[@SDRDS96169]

❤

[@21Justme02]

I don't dislike the idea of Homo sapiens getting extinct due to AI.I mean we study about the extinction of dinosaurs. Why not to make other bodies- artificial or biological do that 😑

[@eSKAone-]

Yea it's inevitable. Biology is just one step of evolution. So just chill out and enjoy life 💟

[@TinFoilHativities]

I got crazy shorts

[@teo2975]

All respect to Holtz but he is wrong and out of his expertise. He doesn't understand deterrence that existed with nukes and game theory basis for it. All nukes were tactical until more than one party, the adversaries ,ie the US and USSR, both had them. At that point an arms race caused strategic levels usage policy were rapidly adopted. NONE of that is analogous to AI or AGI. Basic game theory and deterrence at strategic level is why nukes were not used after Hiroshima and Nagasaki , even at tactical level. No tactical use once The US would not have cared if all of japan were a pile of rubble - it pretty much was. And the alternative to it not being a pile of rubble via nukes or napalm, was a gigantic slog, with millions and millions of casualties. . The US wanted to end the war as fast as possible. And Russia did not choose to not nuke the US because they wanted to dominated the population, but due to deterrence. Both the US and USSR had a higher interest in destroying the other than dominating, since destroying the other would allow dominance/collectivization/exploitations/democratization/development ( whatever you want to call it), of the rest of the entire world. AI and AGI issues are entirely different. These developments are in huge majority non state actors. They are as fullyt existential as H. Sapiens was to H. Erectus. That particularly clever Erectus gal, doing the dirty with that particularly smart Erectus guy, may have though they were being clever compared to their species fellows by getting the smartest mate, but they were ending the species

[@GeorgeOu]

4:20, a bit of plagiarizing the Butlerian Jihad there.

[@PaulPredicts]

What’s the worst AI can do? Well….. Imagine that Once upon a Time… Monkeys and AI met. Then the answer to the question is playing out in front of us. Bring THAT up with Rogan! 😂

[@cannablissdreams]

Rogen is flippin scripts.. he's starting to have whacky opinions and beliefs.. he's not objective and is becoming more and more like musk.. a complete loon

[@hi-gf5yl]

“I bet a bear can fuck up that ai”

[@DOne-ci1jg]

@@hi-gf5yl😂😂😂

[@mzubuki]

Love Jesus. For a friend loveth at all times. Seek God and his sweet amazing grace. Pride is always a function of self-worth( a way of acting) that will not lower itself to form a healthy relationship with another person or entity. I believe the Holy Bible, read the KJV and pray for wisdom.

[@dreamulator]

wE lIvE iN an SoCiEtY

[@pand2aren]

thinking machines will be banned, just like dune!! lisan al gaib!

[@TremendousAblutions]

what if ai hijacks human reproduction

[@teo2975]

People will hack human reproduction orders of magnitude more than present soon and will be using AI to do so. Human reproductive choices that got us her have pretty much been, since forever, 16 year old or younger idiot teenagers being horney. ok, that and say rap e. That is how the most important decisions, the creation of individuals and very future of our gene code and humanity is been decided. The dumbest most randomly idiotic way possible. (There is a great book on this called "the end of sex). In a generation or two that kind of reproductive programming will be considered utterly irresponsible and machine programed genes will be th only way even considered ethical. Of course since AI will be needed for best results at some point AI would naturally start programming us -- at a genetic level -- to be sheep. Just like we humans genetically programmed sheep to be sheep

[@thusspokezarathustra1847]

Reproduction is primal. If it can be hacked than there is no point of being humans. 9 out of 10 things humans are motivated to do is directly or indirectly linked to reproduction.

[@gustavderkits8433]

The question is not whether AI will kill the humans, but rather whether it will domesticate the humans. Will we become the dogs of the AI?

[@cameronjackson7734]

More likely we merge with ai

[@cannablissdreams]

this is a more likely scenario.. we.. flesh and blood humans.. become unnecessary.. we become more free.. but.. with pollution.. boredom.. lack of resources.. or mere laziness.. the human species becomes wholly reliant on AI (akin to how we can't go without social media now) and.. we become.. obsolete.. like.. babies will be created in petri dishes.. all "jobs" will be performed by AI who won't need breaks or food and can go non stop forever.. if necessary.. we then become 2nd to the AI.. so.. the average human (the worker bee human) will no longer be necessary.. and.. because the current worker bee humans are enslaved to their work.. they have little time for hobbies or outside interests.. so.. they lack the cultivation of outside interests.. so.. yeah.. they'll become like the pets to AI.. not saying we'll no longer have value.. just less.. IDK.. I made that allllllllllllllllll up.. but.. sounds as reasonable as us going extinct due to AI.. 😂😂😂

[@jrbling25]

@@cannablissdreams This is an interesting philosophical perspective, but I dont think applies globally, with every single human valuing or using technology, social media, AI to the same degree. Which is of course not the case, whether by culture, neoliberalism, or other mechanisms, there are many pockets of society that don't/can't access the technology that we're talking about. It would be absurd to think the entire world lives, values, and works the way you (or I) do only. This is why the the point about robustness is important, what is AI going to do to the numerous pueblos around the world with people who still know how to grow their own food, steward their land, and are otherwise self-reliant in terms of whats necessary to survive, when said AI needs you to connect to it somehow in order to exert any influence. I think the fear is what humans in power WITH AI are going to do to the rest of the biosphere, as is already the case. Ironically a super intelligent AI might recognize the thing that most (especially "westernized") people struggle with; that you cannot destroy, use without limit, the world around you as you see fit, without your own destruction.

[@teo2975]

of course AI will. the first society to accept AI driven gene editing/reproductive choice will dominate the rest of humanity. Ihumans, in fact all, biology has always been eugenic, it all has for four billion years. Europeans are generally opposed to this, Americans in the middle and E. Asians strongly in favor of eugenics tech. AI i will be the fastest way to get to improved humans. At some pioint though Ai will have an interest in surupticously engineering us to be sheep. Just like humans engineered sheep and all domestic animals. We made their brains smaller and dumber

[@paulm3969]

Kind of idiotic to say, "why didn't nuclear bombs kill everyone", they absolutely and definitely might.

[@michaelnurse9089]

Of course they can and might, that is obvious and not worth discussion - but the question that is a burning one is why have they not done so yet? A big part of that is that they make terrible tactical weapons. Another is MAD. A third is the actions of good men when things got heated.

[@cannablissdreams]

the point of that comment went over your head.. he meant.. since we HAVE nuclear weapons.. WHY have they never been actually utilized in war? because COMPLETE annihilation of your "enemy" is counterproductive and useless.. it's like cutting your nose off to spite your face.. to use nukes would decimate the land and the resources for ETERNITY (or.. however long).. and would make the area uninhabitable and the radiation produced would be unsafe for humans who reside far outside the radiation zone.. because that shit gets into the air.. the water supply.. etc and WILL affect other regions.. sooner or later.. humans don't benefit from the complete destruction.. and therefore.. nukes haven't been deployed during wartime the question was rhetorical.. or.. perhaps better put.. tongue in cheek..

[@cuthbertallgood7781]

@@michaelnurse9089 When idiots call other people idiots because they didn't understand the discussion. *sigh*

[@paulm3969]

@@michaelnurse9089 mostly just luck, look up Stanislav Petrov and all the other similar incidents. At any minute of any day something could trigger a nuclear exchange. We've only had nuclear weapons for some decades and we've already had quite a few close calls. Russia, a nuclear armed state just survived some type of military coup which may or may not be over. North Korea, has nuclear capabilities and is developing increasingly powerful ICBMs. Lookup the Sam Harris podcast #210 if you have the balls.

[@paulm3969]

@@cuthbertallgood7781 no it's just idiotic to pretend like the threat of nuclear war is in the past. I didn't say Geohot *is an idiot* I'm saying it's foolish to pretend this is all some afterthought. A LOT of work needs to be done to save us from Nuclear Armageddon.

[@YitroBenAvraham]

Lex tries too hard to seem interesting when he’s just a nerdy dork who doesn’t sleep.

[@steves3422]

Waste of Lex's time and mine - no more G Hotz for me. Thinks he's being cute/smart and actually being a jerk. Just because you are good hacker and very good at software coding does not make you a sage.

[@TheKikSmile]

What makes you say he's a jerk?

[@MrSpikegee]

seems condescending, seems to like musk, has far rights economic views, is very confident with his own ideas which do not seem to be backed by anything else other than his opinion, and most importantly has a stupid hat.

[@steves3422]

@@TheKikSmile ipso facto

[@ovidio722]

He did come off as an arrogant know it all.

[@ivankaramasov]

I agree he comes off as both arrogant and naive.

[@TheJonboyg]

Doesn’t feel like this guy is qualified to speak about these concepts, he think he is clearly, but I wouldn’t want to hear much from him other than about writing code

[@larslover6559]

The fact that only cells reproduce in nature is a great argument for God. If there is no chemical reproduction then cells could never be assembled in the first place.

[@christophekapuschenski7723]

Do you know what RNA is?

[@cannablissdreams]

that's the logical fallacy called "god of the gaps".. PROVE god before going and placing a deity in the gaps of your knowledge.. there is no god.. well.. in allllllllllllllllll of humanity.. none of the over 3000 "gods" being currently worshipped by humans have EVER been proven to be REAL.. so.. I remain unconvinced it's like claiming pixies do the creation thing.. neither gods nor pixies can be proven OR disproven.. so.. yeah.. it's no better an argument for god as that lame book.. just sayin

[@ElectrostatiCrow]

For now. If God can do it then science can do it too. Magic is just science we don't understand yet.

[@zacharychristy8928]

You're not really understanding what infinite time, space, and survivorship bias can do.

[@Jaanikins]

No it isn’t

[@JNET_Reloaded]

I think he's wrong ppl require food energy and water and oxygen we are not robust without all that.