Neil deGrasse Tyson Is Not Afraid of Artificial Intelligence | Within Reason #28

[@alekhinesgun9997]

So nice of you to let Neil interview himself

[@DANGJOS]

🤣🤣😂

[@Sarahizahhsum]

Lol Alex trying to butt in

[@whatsup300]

You're a cliché

[@Ironcage25]

U just hate him cause u r rasist and he's black. Neil has accomplish in science u and your generation will never get.

[@peterm1240]

I have seen Cosmic skeptic interrupt before. Did he forget how to do it?

[@ashantilematthew6277]

Neil is probably the easiest person to interview: just ask him a question and watch him go.

[@diegog1853]

I suppose it depends of what you want out of your interview, if you want just to get someone to talk as much as possible. Definitely. If you want to get some specific information, get to some important points, maybe challenge some opinions in detail. Then not at all lol

[@ashantilematthew6277]

@@diegog1853 You get a lot of specific information about important points from Niel. That is not an issue. So if you are saying he gives you too much, then maybe. But he is going on there to talk.

[@Anjalena]

Robin Williams used to do that, take over the interview, but it was much more entertaining.

[@ashantilematthew6277]

@@Anjalena #1. Robin Williams was a comedian/actor/ entertainer. #2. He was white so he didn't trigger your racism.

[@Algeanie3]

Me, literally me

[@hurley3000gt]

Neil is the sort of guy that even in an infinite amount of universes, there isn't one universe where he wouldn't interrupt you.

[@RootBoyJim]

Motivation. Computers Don't Care About Anything.

[@AlanWinterboy]

I was just going to write that, lol😀. He's the worst when he's in a panel of his peers. Bearable one on one.

[@ominous-omnipresent-they]

Which is pretty goddamn impressive, to say the least.

[@misterocain]

Isn't that what a good interview should be, i.e. the guest doing most the talking as long as he/she sticks to topic? Too many interviews are high jacked by the interviewer.

[@TheMoonKingdom]

@@misterocain No, he is filibustering so the interviewer can't ask followup questions to challenge is ideas. I happen to agree with Neil's, perspective, but come on:)

[@frankiemiller5364]

Neil’s passion and excitement for these subjects have reached conversationally uncontainable levels 😂 and I love it!

[@Maximustard]

He’s a shill

[@uku4171]

exactly lol

[@ljubomirjuraj9658]

​@@Maximustard oh no, ppl shilling science and education, how horrible. This comment was brought to you by raid shadow legends. Like wtf is even the problem.

[@Kroban_d4c]

Nah that isnt "passion", just because you are passionate about smth doesnt mean you have the right to be this fucking unhinged. Hes a fucking r-*ard, he constantly interrupts everyone because he loves to smell his own farts and his ego makes him think hes on a way superior level than everyone else, he literally ignores what the other person is saying while acting extremely condescending like if the other person has the brain of a child compared to him

[@Limanknol]

@@ljubomirjuraj9658 raid shadow legends?

[@galefray]

Neil is the type to ask you to help him with something, then just make you watch him do all the work.

[@ac1646]

🤣🤣🤣

[@florisramaekers2362]

Hahahahahah 😂 this is so true im like that too tho so it was a moment where you held a mirror up to me friend

[@rohanking12able]

I'd be asking questions the whole way and probably learn allot

[@waido_]

I’ve gotten paid for doing that kind of “work” before, so I wouldn’t actually mind🤣

[@user-il9ze9py8c]

Great video alex! I hope you get to speak to Neil deGrasse Tyson soon!

[@joannalewis5279]

Hahaha

[@allmodcons2274]

😂😂😂

[@lifeisstr4nge]

underrated comment :D

[@c.jishnu378]

Underrated af.

[@nebuchadnezzar6894]

"Astrophysicists are the most humble scientists" - Neil deGrasse Tyson, an astrophysicist

[@purepeter4737]

His lack of self awareness is hilarious. Such a bellend

[@XnonTheGod]

The irony

[@Gruso57]

This is an underrated comment lol

[@pikupikuseru]

Lmaooo, I get a feeling he's thinking of his peers and his idols when he says that, not himself

[@madgoonproductions8406]

Depends on your definition of humble lol. He might think he’s the most important human to ever exist while also knowing that human toil is insignificant when juxtaposed against the vastness of the universe.

[@φαρμακεία-πρωταρχικός]

What a time to be alive, having grown up with Alex and his channel. All of these guests, and the history his channel has; thus leaving his mark on this platform, and taking us for the ride. Which is quite remarkable.

[@xnellyxs1]

Neil is the kind of guy that will wake you up to tell you he's going to sleep.

[@Rerum_Novarum]

😂😂😂

[@Longtack55]

Wake up Alex, it's time for your sleeping pill.

[@x97s]

Lmao

[@reinforcedpenisstem]

What does that mean?

[@JaguarBST]

@@reinforcedpenisstem he is saying he finds NDT very annoying and obnoxious

[@garywood97]

I've never seen someone whose skill at delivering one-sided speeches is so vastly different than their ability to hold 2-way conversations.

[@Dann-md9eq]

Can you elaborate that further? I don't quite understand

[@garywood97]

@@Dann-md9eq Neil is great at giving speeches, and really sucks at having conversations.

[@activistbook3809]

@@garywood97 he's great at conversations 🙄 watch the show.

[@sownheard]

Imagen inviting a person and than being surprised that person is himself. 1 Neil would not be invited if his personality wasn't so outspoken. 2 If you want to interrupt him you still can.

[@somedudeok1451]

I think it's charming in a way Plus, he did respond to the host's questions. Just in a lengthy way that highlights possible trails of thought you could run the question down.

[@Johnnybsknees]

Nice! Good to see how quickly you have drawn large names to the podcast. You are already easily one of the best interviewers out there.

[@MrSnoots]

That's a huge overstatement.

[@Johnnybsknees]

@@MrSnoots it's a subjective statement, and I hold to it. He and Glenn Loury are my 2 favorite interviewers and for similar reasons: both are able to well state their opponents' views or just views they don't hold in order to get the most out of the interviewee.

[@dwaynekeenum1916]

@@MrSnootsdumb

[@seanmoore2264]

Alex : It seems experts in the field feel that AI could be an existential threat to humans. Neil: We'll just turn it off.

[@onyxrafle8066]

Terminator hunting you? Just say no.

[@somedudeok1451]

It's a bit like trying to turn off the internet. Sure, we could do it with a big amount of effort, but we won't.

[@mordin999]

Neil view on AI seem a bit naive. First there is no shifting goal posts, but mainly confusion about rapid evolution in the field AI. As impressive as all recent advancement seem, we are still talking about narrow AI which is in a sense is still a very elaborate calculator, a tool. The fear is about artificial General intelligence (AGI) and ultimately ASI. The concerns are that (1) We begin to see sparks of AGI (2) We do not understand how AI work any more than we understand our own brains, nor what constitutes consciousness, and likely will not be able to tell when or if AI passed the singularity threshold (3) there are a lot of incentives to developed AI, from the usual economic incentives to global power arm race, which may result in AGI created intentionally or unintentionally. (4) Contrary to Neil claim, we already building AI robots including in human form, form companionship to military application. (5) many AI systems are given autonomus functions and are able to collect information from the outside world not just the internet. (6) AI are being integrated into more and more system to deal with vast amount of data, arguably an AI integrated into something like Chinese extensive surveillance would be as close possible to omnipotence. (*) btw to expand on #3 even current narrow AI can exacerbate many problems on the internet (from your usual scams to disinformation, propaganda) and it becomes much more accessible, able to run by anyone on any laptop (lets see you turning that off) and without any safety measures. err.. I kinda runoff here no time to improve I am already in the tldr area sorry

[@ChristianIce]

@@mordin999 Daily reminder that AGI doesn't exist yet, not one single line of working code has been written.

[@pooglechen3251]

ChatGPT is just a server some where. You think you can't turn of Google, but you can

[@jthomas5o158]

The last 10 minutes of this podcast are the most illuminating - Neil for the win!

[@Slitter_the_Dubstep]

its honestly really refreshing and inspiring to hear someone talk so passionatly as neil does here- also i feel like neil is answering more the questions of the viewers rather than the questions of cosmic. hes just another listener like us rn hehe

[@michaelhough5003]

One of the most difficult things about being human is realizing that we are both comically small, and one of the largest living things in the known universe. It's both. We are small, and we are big.

[@LevellerTV]

Very serious experts are very concerned about existential risks from AI. You should interview Paul Christiano, former head of OpenAI's large language model team and current head of the Alignment Research Center. You could also interview Robert Miles (runs *the best* KZbin channel on AI risk), Richard Ngo (on OpenAI's governance team), Ajeya Cotra (Senior Research Analyst at Open Philanthropy), or Kelsey Piper (journalist at Vox). They are all knowledgeable/expert enough to have very fruitful conversations about AI risk.

[@user-md3wm7vu1f]

I'm no AI specialist but I think its pretty likely there's at least some reason to be concerned because it is genuinely very different from past technological revolutions. This is the first time we will have a highly generalized intelligence as opposed to something more specialized like a calculator or computer. Due to generalized, autonomous intelligence (or at least semi-autonomous initially) it definitely seems to be possible that programming it not to do something wont be enough once it reaches a certain point because it will probably be able to override anything and it will be far more intelligent in every way than anything that's ever lived on earth, making it very hard to predict what its capabilities, motivations, or intentions will be (if it goes on to develop any). AND it will be capable of increasing its own intelligence - to what level, who knows The ai threat is also easier to accidentally stumble into than something like nuclear annihilation because of the less-explicit danger, the exponential progression, and much greater unknowns involved as higher levels of complexity are reached. A nuclear apocalypse is also survivable by humans, the majority would just die off and a small population would live underground. On the other hand, just about nothing would be able to hide from a sufficiently advanced ai

[@cookiecan10]

Alex should try to interview someone like Robert Miles, who has been creating educational KZbin videos about AI risks for years. Even appeared on Computerphile a couple times.

[@LevellerTV]

@@cookiecan10 Great suggestion! Huge fan of Robert Miles; just edited my original comment.

[@LevellerTV]

@@user-md3wm7vu1f Great points, I second pretty much everything you said. I'll add that scientists fail to predict how quickly technologies will advance -- all the time. A particularly stark example was Enrico Fermi thinking that a sustained nuclear fission reaction was 50 years off... just a few short years before he accomplished it himself. This is important because one way people assuage themselves and dismiss existential AI risks is by saying artificial general intelligence is "at least 50 years off," as though even that is enough time. I'll probably make a video about AI risk on my channel at some point.

[@bengeorge9063]

Or interview Yoshua Benugio.

[@jessep9671]

I've never seen Tyson give a good long-form interview. He tries to convey deep philosophical insights, while also seemingly making them up on the spot, and never allows for good back-and-forth dialogue.

[@TheAxeter]

To be fair this was an interview, you can clearly tell that this is not a debate format. Interviews are normally meant to extract information from the person you are interviewing

[@thomasthompson6378]

@@TheAxeter Much of the time, you will have noticed -- surely? -- that NGT most of the time refused to let Alex O'Connor finish his question before NGT was off and running with another windy soliloquy.

[@gregw322]

Absolutely, NDT is the worst at communicating ideas.

[@georgesos]

In ancient Greece they called this kind of people Sofistes. Half knowledgeable, self promoting. I never liked the guy and multiple times I ve heard inaccurate or just wrong things from his mouth. He is good at faking the scientist. But he is not one.

[@TheAxeter]

@@thomasthompson6378 yeah he definitely did, maybe he's just too excited, it seems to be a flaw of his but I'm not perfect either. I do appreciate that he didn't dodge the questions at all.

[@cosmic-observerr]

Alex, thanks for this episode. For one who loves Astronomy, I enjoyed it. You're doing a great job, Man.

[@altonsafe]

The beginning is amazing. He’s exactly right. This is why people who had a spiritual/curious start to this world always see the world beyond of themselves. You get a greater and deeper understanding of who we are and can be. To the greatest discoverers, innovators, spiritual leaders can’t be who they are without this level of insight. Humans have the power to see the world from outside themselves, and once we do our world will become better because of it.

[@stevem83]

I really like Neal and his insight is fascinating. I just wish he would let others talk and direct the conversation more.

[@Firehazard159]

I mean, aren't we here to hear from the interviewee, not the interviewer?

[@whoisandrewblack5679]

@@Firehazard159 isn’t the interviewee there to be entertaining and not threatening? Your complaint cuts both ways

[@j8acob1]

​@@Firehazard159 Isn't the point of a podcast to have a 2 sided conversation? I'd find it more interesting if Neal engaged with Alex without being arogent and controlling

[@sansoucie1969]

@j8acob1 @Firehazard159 @whoisandrewblack5679 @stevem83 One thing was said in these comments with which I agree: "I really like Neal and his insight is fascinating." The rest is gobbledygook (had to spellcheck that) Apart from entertainment, the intention of the interviewer is to challenge the guest. The degree to which the guest is threatening, "arogent" or controlling, is the failure of the interviewer. However, I think that Alex did just fine, because he seems to me, to be agnostic to the assertions.

[@oh-cx7dr]

@@whoisandrewblack5679 threatening??

[@ozzyzee1770]

I can't believe that NGT actually apologised for interrupting at 24:48 😳😳😳

[@oneiroagent]

😳

[@lukasszlaur6246]

no way

[@wjddls_xix]

Alex speaks 7.13mins in total of this 44.45min podcast. That's 16% of Alex; 84% of NDG.

[@thegoodlord6518]

How did you count?

[@radioactivedetective6876]

I salute your diligence

[@jt9300]

Sounds like you had a very fruitful bout of boredom😂

[@cagnazzo82]

Well technically he's doing an interview. He's not the one being interviewed.

[@danicrazyfingers]

I still think NGT kept interrupting Alex more than I find desirable.

[@quizzicalsmudge877]

One of the things that occurs to me when it comes to world ending threats that I would like to see more discussion on, is how to deal with people who have ideas/perspectives that are different to your own. It seems to me the biggest issue is passionate people fighting first and asking questions second.

[@Telencephelon]

44:30 Total highlight of the conversation. So true. Includes almost everyone.

[@irrelevant_noob]

Absolutely true. And i don't think the "almost" is really needed... even if someone _knows_ about their lens/biases (which is quite rare IMO), there's not much they can actually do about it.

[@berkefeil5646]

The passion with which Neil talks is quite admirable, and a bit silly when it comes out in funny way too :) Some people mistake it for arrogance, unfortunately, and supposedly miss some outstanding points he made.

[@raymondcava4669]

Absolutely

[@gregw322]

He’s unquestionably an arrogant person.

[@willashland4597]

NDT is absolutely arrogant and condescending, and suffers from a massively inflated ego. He really used "I've written over 50,000 lines of code" as a qualifier to speak on AI while saying that Elon Musk wasn't qualified to speak on it. As somebody who has probably written close to one million lines of code, and has worked with neural net models and machine learning software directly, I am of the opinion that Neil is out of his depth here and should have some more humility in this area. His cosmology credentials have zero to do with machine learning or AI.

[@PettenVeevo]

@@willashland4597 but 14:50 he said the very opposite with this number: " I'm not an expert, okay, hold aside that I've written 50 000 lines of computer code and I've been thinking about computers my whole life, I don't present myself as an expert "

[@nattokami9598]

@@willashland4597 writing a million lines of code must have fried your brain because he literally said he wasn’t a qualified expert after that.

[@mixedbyap]

Awesome collab! Alex’s channel and brand is growing I love it

[@flakeyjay]

I have a lot of respect for Neil and have seen him three times lecture on various topics. His perspective is certainly thought provoking. Enjoying the new format as well!

[@suninmoon4601]

In addition to zooming out to a macro view, orienting the viewer at the center of the known universe; it is also possible to zoom in to a micro view, orienting the viewer at the circumference of the known universe; and how both methods of inquiry are complementary. AWESOME answer, Neil!

[@DANGJOS]

24:49 Woah, he actually apologized for interrupting, and he does seem to be letting the interviewer speak for some time. I wonder if someone talked to him about his tendencies.

[@someoneelse3456]

Brief flash of self-awareness

[@johannhuman532]

Alex is pushing his worry in the discussion even though Neil has stated numerous times that it is not the way he's looking at this. So I am not surprised that Neil reminds Alex about that.

[@asorelle]

I was shocked and pleased to see Neil on with you, Alex! I've been watching your videos since the beginning and Neil is a major life hero of mine. I'm so excited and happy for you! I'll be looking forward to every episode you produce!

[@basedmase777]

I respect Neil a lot, he's dont a bunch of good education work. But damn man is it impossible to have a conversation with him. I never get to hear his thoughts or ideas beyond the surface level because he never lets ppl get there. Nice try Alex

[@peterm1240]

It doesn't help that he talking to someone who is clearly not trained in science.

[@basedmase777]

@@peterm1240 neither are the 99% of the other interviewers are that he does. Thats not an excuse, Neil made a career off educating ppl who have no "science training". He should be able to do it.

[@someoneelse3456]

@@peterm1240 This is largely supposed to be a philosophical/meta-discussion about science. They're not actually writing conjectures or theorems. Alex generally asked thought-provoking questions, although NDT never allowed the interview to move beyond a certain depth.

[@user-md3wm7vu1f]

i think thats just his nature, some people are like that. it does not necessarily indicate narcissism and isnt necessarily rude or insulting, but in a podcast setting with very limited time it can be annoying

[@zodo2476]

@@peterm1240 These arent genuine questions he has. He has his own answers for every single one, but that's not the point of an interview. He wants to hear Neils personal take for content.

[@stanley9868]

That was a great episode. I love how contagious Neil's enthusiasm is

[@Glowbox3D]

Great chat. NGT can sure talk...and I love it. Lovely discussion thank you.

[@jcole139]

I thought Alex steered the ship on this interview just fine. Just Neil being Neil. I enjoyed the interview.

[@SuperGoodMush]

I don't think he's arrogant or narcissistic, he's just a sort of radio personality in the way that he talks. I admire that he takes a scientific approach to understanding how groups think, choosing to see how people react to his thoughts on Twitter and adjusting, rather than saying that his opinion is A) true, or B) unwavering.

[@nezar-6889]

It's incredibly humble to be willing to adapt/learn like that, especially given his age and status. A lot of people have way too much pride to do anything of the sort.

[@kadegetslaid634]

He's pationate, I'm no genius but whenever someone asks me about starwars lore I go on a 30 minute explainer lol

[@oimrqs1691]

Super cool that you were able to do this!

[@HoobaBros]

very very intresting guest Alex. Thanks for all of the value you are providing to us.

[@diegovillalobos5364]

I love Neil's way of showing us that we don't need to stop him from speaking his mind. Always entertaining and enlightening...Best nerd ever!

[@DarthVaderfr]

What's the point of interviewing a person if he doesn't talk through his point, that's just stupid cancellation culture, you are talking out of your emotion and you feel your fragile ego threatened by someone you have never talked to and doesn't even know you

[@m.a.a.d9275]

@@DarthVaderfr its not an interview dummy. Its a podcast and a discussion first and foremost. Neil is unable to converse back and fourth and just completely runs over the pod allowing Alex like 2 or 3 words every now and then

[@riinazen7443]

Everyone and their mother's dog overuses "cancel culture" just to make the blandst point ever.

[@CodyCLI]

@@DarthVaderfr Go back in your crib with your sippy cup, adults are talking. You're literally crying about nothing, you unintelligent clown.

[@hendricklamar5061]

Loved this conversation!! Wish it took longer!

[@valmid5069]

*Great Interview! Cant wait for more videos like this!*

[@leondonnellan9994]

I think your podcast works much better when you have an actual conversation.

[@michelguyfortin]

Awesome podcast. Two people I admire deeply. Thanks, Alex!

[@alanjmcc]

I'm a big fan of both these guys. How totally refreshing to hear intelligent conversation. Alex is among the best of interviewers because he knows how to ask good questions and then get out of the way of the answers. Neil is wonderful to listen to because he's always at the ready with a passionate cluster of enthusiastic arguments for remaining optimistic about the future. I could listen to these guys go on for hours more.

[@tamayoshi682]

I presume there was some lag in this conversation. The amount of interruption in a conversation increases dramatically as the delay between parties increases, because people don't stutter or stop as you start speaking. It's very tough interrupting people when there's lag, because it means you have to speak over them for twice as long before it causes them to stop talking.

[@Skymannot6939]

Dr. Tyson on your podcast wow! That’s awesome.

[@anonymouscat6207]

6:53 "In fact, Astrophysicists have gotta be some of the most humble scientists" Well, clearly not all Astrophysicists, Mr. NDT...

[@notme5744]

That's hilarious

[@peterm1240]

Well, he means humble among scientists. Around you asshats on the other hand...

[@DANGJOS]

Mr. ?

[@jacobus57]

​@@notme5744 yeah. I worked in PMA--Physics, Math, and Astronomy division--at Caltech and male astrophysicists are the worst. It made me glad I didn't pursue that career path.

[@notme5744]

@@jacobus57 That's strange that astrophysicists specifically would be worse than others. I can imagine most math based fields attracting a lot of competitive or even arrogant guys for obvious Testosterone based reasons, but it's weird that astrophysicists would be worse. Kind of funny though. It's a shame you left if that was something you were really interested in. What do you do instead?

[@blucenere]

I would suggest Robert Miles for the next talk on AI 😊

[@vanderkarl3927]

Or Eliezer Yudkowsky, he's been on a fair few interviews lately!

[@EffectiveAgainstSuffering]

💯

[@EffectiveAgainstSuffering]

​@@vanderkarl3927 💯💯

[@milorodval678]

👌

[@PlaylistWatching1234]

Robert Miles is one of the best. And if you've watched computerphile, you've probably seen him, but not known his name.

[@its-virox]

He may speak A LOT, but shit, I was hooked on every word. ‘The universe lives in you’! Wow

[@peter9477]

Neil: I'm no AI expert. Also Neil: [talks for 30 minutes about how AI is perfectly safe, it's all fine.]

[@alexanderg8466]

no no no. Listen again

[@pooglechen3251]

He's saying he's a science expert and you don't need to be afraid of science so long as there are regulations

[@riser9644]

he asked for his opinion , he gave it with a disclaimer , don't hate because his opinion isn't same as yours

[@nakkadu]

​@@pooglechen3251 and how do you regulate China or Russia?

[@OmeganKryist]

As someone who is vaguely familiar with AI. I assure you that the dangers of AI are not what is conveyed in the media. There will be no “Rouge AI” unless we deliberately make one. The real issues are tied to issues we already have like spreading misinformation and further separating the lower and upper classes.

[@Fullyautomagic]

He could just talk it to death

[@PattiMoreno212]

I 💯 % agree with his stance on AI!

[@justin_ooo]

it is from the perspective of a layman and not based on the objective facts regarding how the technology works and is already being integrated into modern software/hardware

[@JohnCena8351]

I highly recommend the book "the emperors new mind" for anyone interested in A.I and consciousness. I'm only a few pages in tbh, but it's pretty amazing so far. Oh yeah, btw, Roger Penrose would be an amazing guest on your podcast!

[@alexfunk6575]

One of my favorites! Only thing is that it's pretty dated in terms of all the computer stuff, but the overall argument still holds up in my opinion

[@charbelbejjani5541]

He expanded his arguments more fully in "Shadows of the Mind" (1994)

[@JohnCena8351]

@@charbelbejjani5541 very interesting. Thx for the Info!

[@Josh-hi9zp]

@@charbelbejjani5541 my god i loved ENM read it like twice and now 'Shadows' is just sitting on some page on my desk because I sabotaged my attention span

[@Forest-jj7pj]

Thanks Alex. You are a young man that creates opportunities to listen people of different backgrounds. You are a new generation of thinkers. You may create you own perspective and positive contributions to humanity.

[@OmegaWolf747]

Wow! You got Dr. Tyson on your channel! You've truly arrived! Kudos to you!

[@humanbeing33]

I love Neil, he's one of my favorite humans.

[@K.P.80]

Okay, AI.

[@TacoScott]

I've always thought that Tyson had an unthoughtful view of vegetarianism and veganism. Alex would have been a great interviewer for that domain but we missed the boat for that conversation. But... such is the nature of things. People change.

[@vDuzz]

I was thinking the same thing. Neil has such a misguided perspective about veganism, and it would have been perfect if Alex could have pushed him on it. Very disappointed.

[@komfyrion]

When I opened up the video I was kind of expecting this to happen, unfortunately. I think Tyson is a person who sculpts his public profile a lot. He has learned from his twitter escapades and he doesn't want to get to deep into particular issues that he knows will make him unpopular. I think it's entirely possible this is because he genuinely wants to be a successful science communicator and in order to do that he needs to appeal as widely as possible. So he puts the spicy tweets in a secret file. However, I think he knows that there are certain issues he can just blabber on about without much consequence to his career (or perhaps with positive consequences to his career). I think veganism and many other social justice causes that are outside of the mainstream fall into that category, but perhaps one day he will have to either re-evaluate veganism or end up a disliked reactionary geezer. I don't think Tyson is an particularly strong anti-vegan force in popular culture, but of course it would be very nice to have him either stop what he's doing or have his mind changed and become a vegan advocate.

[@jmarkinman]

"What is this AI thing that everyone fears? Is it a computer, in a room, that has control over every other computer in the world? ...Well, why would you grant it that access." Neil seems to not understand that a computer that is smarter than its creators will either be able to figure out how to gain whatever access it may desire on its own, or be able to convince or trick people, through its superior intelligence (whether through promises of better lives or through threats or lies), to grant that access to it.

[@jmarkinman]

Later he said, "[The AI] runs on electricity... [It's] not going to stop me from unplugging [it]." He's wrong about this too. Once an AI has access to its own programming, which is necessary for its self-improvement, it will be able to copy itself and proliferate itself through wireless means, probably through wifi signals or bluetooth, but it could also use its own hardware to create radio signals. If it is connected to the electrical grid, it can send signals through the grid itself, which itself could be instructions for replication. So if you are even thinking about "unplugging" it, it is already too late. Now one could try to completely isolate such a machine, but then, what good is it? So no engineer will be incentivised not to connect it to other systems, whch themselves are likely to have vulbnerabilities, such as the electrical grid. There will always be that black swan idea that the superintelligence thinks about that we do not.

[@ChristianIce]

You are giving biological needs to a toaster. Human feelings are a by-product of evolution, of our mortality and need to accumulate wealth. That said, when AGI will exist in the future, it would be a wise decision not to give it access to nukes. For now we just have LLMs, and they are not thinking agent, so they won't ever have "desires."

[@Tjomsasen]

It's more of a philosophical discussion more than a technical one is my impression.

[@jmarkinman]

@@Tjomsasen The distinction between a philosophical discussion and a technical discussion, is not the logic or reason that leads to a conclusion. The distinction is in the nature of the conclusion itself: a conclusion made from a technical discussion does not need to point to any kind of shift in our perspective, while a philosophical discussion often asks us to question our perspective. This does not mean a philosophical discussion is somehow "less rigorous" or not logical, on the contrary, a technical discussion does not demand from us any thinking at all: only calculating; albeit with an emphasis on improving precision.

[@jmarkinman]

@@Lamster66 Yes, it is certainly a philosophical discussion. And it is a valid question that Dr. Tyson asks. And sometimes we need technical reports to answer philosophical questions, which is actually the case here. While we may be able to reduce the ways and means an ASI could prevent it's own shutdown, this can never be guaranteed, as it will be certain to be able to outsmart us or offer a clear benefit that would prevent it from being shutdown. And to not build such a machine is a non-starter, as the power such a machine will bring to it's creators will be unmatched, thus the race to AGI is much like an arms race, and can not be stopped. There is no stopping someone who has the abilities and resources to create an AGI from creating an AGI, and thus, it isn't up to anyone except that person or group whether to create it or not. That is, there is no way to enforce the notion that it shouldn't happen. As such, alignment is a much more important problem to solve than failsafes, as all failsafes can only be engineered to make an indication of failure, not prevent it.

[@prometheus_78]

He interrupts a lot, but you have to adore his passion for talking about astronomy, technology, etc.

[@davidcolon13]

24:46 Oh wow he’s getting better at this

[@shadowone01x99]

I enjoy seeing and hearing Tyson talk.... He is very illustrative and exemplary in his speeches to enable even the most simple minded person understand complex stuff and i adore him for that.

[@Dudelee1]

The irony and lack of self awareness of Neil saying that physicists are so humble is staggering

[@rafalpotasz]

I cracked up how the intro plays out. Alex starts posing the point and within seconds gets interrupted. Pretty much the expectation with Neil, but happened impressively fast. Great clip to pick for setting the mood of the interview. 🎉

[@morbidmanmusic]

Yet people always have him on their show. Go figure. Maybe yoU should tune out, as you're commnet adds little.

[@rafalpotasz]

@@morbidmanmusic I mean, it wasn't meant to add anything, I just noted an observation. Your comment adds little and that's absolutely not a problem in the slightest. Overall I enjoy listening to NDT, it's just that his style of communication is more suited to TED talks and presentations, not so much interviews/conversations. And that's absolutely fine too :) The internet is a serious place these days!

[@tschorsch]

​@Rafal the internet is seriously stupid these days

[@rafalpotasz]

@@tschorsch I guess. It feels super polarising, all vs all, lol. Although I spend most of my time online learning for free so I won’t complain too loudly!

[@IssoufBa42]

Always a pleasure to hear Neal talk ❤

[@hudsonbarth5641]

Neil would seem to agree lol

[@onetruetroy]

Great interview. Neil is a wonderful teacher and I learn something new every time.

[@BrendanBrown1]

I very much respect and admire Neil being that he was a huge influence on me growing up. His dissemination of scientific ideas and rhetoric inspired many people in my generation, just like Carl Sagan inspired him when he was growing up. However, regardless of accolades, he really is arrogant and stubborn, but understandably so. He is a person who will argue and debate until the cows come home no matter how incorrect or incongruent his logic may be.

[@connor6822]

I think part of what Neil wasn’t getting is that the idea of artificial intelligence is that part of the goal is for it to be a judgement making, and therefor decision making, entity. So when he says “who programmed it to do that, did it program itself?” The answer may LITERALLY be yes. Someone programmed it to learn from feedback and negative outcomes, then change its future decisions based on those lessons. It’s judgement is nowhere near as nuanced as ours YET. But the capabilities are growing by the day. Though, I also take issue with this sentiment that creating something superior to us would be bad. That’s ego talking. What is so valuable about being a flesh and blood human being? I think that’s our emotions and fears talking. If one day technology advances so far that we create living, thinking machines that are better than us at everything (smarter, stronger, more efficient, less prone to violence, won’t die of old age, can be restored after death in many cases, doesn’t need to consume other things to survive, the list goes on and on) isn’t that just humanity’s next step? If whatever monkeys we came from saw a smarter monkey and decided to end them out of fear, humanity, the most advanced life we are aware of, would never have existed. Why are we trying to stand in the way of what comes next? Because they won’t have meat bodies like us? Because they won’t have DNA like us? Our brains, our DNA, our digestive system, circulatory system, hormones, so on are just highly complex machines. A MASSIVELY complex system that self regulates. If we can make a better one, and it in turn, can make a better one than that, why stop the progress?

[@jt9300]

You raise an excellent point that part of our worries are the result of human ego. In the end, it all comes down to who the major decision makers are for the future of AI, and what they value. But the way you described AI learning by "feedback and negative outcome" scared me. It made me think of AI as a child who learns by navigating the outcomes and takes a lot of influence from the parents feedback, and I saw all of the beta testers as incompetent parents who could unknowingly sabotage how a certain AI model will make decisions later on. Would that analogy be relevant or should I stop worrying?😅

[@bigglyguy8429]

We should stand in the way of what comes next because of simple self-preservation. All those in the past too stupid to understand that are extinct.

[@deeemess]

@@jt9300 the issue isn't so much that the beta testers/trainers are incompetent, but that it's extremely hard to predict the outcomes of a given set of goals, or even to pick the goals that correctly align with human values (which may itself not even be feasible, given the breadth of conflicting values). I'd recommend researching the "alignment problem". Robert Miles has some great videos on this topic.

[@wilfred5656]

You've forgotten about guardrails. Nuclear bombs appear now more like toys. Who is afraid of AI? Provided people like Elon Musk and Putin are not in control of AI, we'll be fine.

[@Elrog3]

@@wilfred5656 Why do you think Elon is scary?

[@Sarahizahhsum]

Neil is quite the person to talk about ego. Oh my. That was just irony at its finest.

[@TheAxeter]

What did he say that you find egoistic?

[@kurteisner67]

​@@TheAxeterego ≠ egoistic

[@badger1296]

@@kurteisner67 Bad faith argument.

[@env0x]

yea when he said that it made me almost spit out water i laughed so hard. he is insanely egotistical, and apparently has no self awareness or sense of humor either.

[@Sarahizahhsum]

@@TheAxeter Egotistical. I assume that's what you meant. He said astrophysicists are the MOST humble people. Keyword: most. Implies he thinks they are better than other people. Which is evident he believes that of himself at least.

[@kiararruiz]

I love Neil! And I'm sure he's a busy guy, but I really wished there was more space to breath in this interview. Alex, you did an amazing job and Neil was interesting as always :)

[@nezar-6889]

I have no idea why so many comments are attacking him. He interrupts a lot, but that's not really an issue if he's conversing with someone patient like Alex (kudos to Alex for remaining patient). He seems like delightful company.

[@Immuuni]

​@@nezar-6889 it's a hip new thing to hate on him. I used to listen to him more, but don't like his style that much anymore. He's still a nice guy and doesn't deserve the hate he's getting.

[@kellenwaters9087]

@@nezar-6889 I use to be a huge fan of Neil years ago when he was considered an educator and proponent of science. However, I listened to an interview years ago on a topic I'm pretty close to, and his off the cuff comments, and dismissive attitude toward an mere idea, seemed really unscientific. He hasn't spent a lick of time studying the topic, but had already concluded it was absurd. He was too invested in a certain conclusion, and went as far as to slander the reputation of others rather than reinforce his viewpoint with evidence. After that, it seemed like his career/attitude turned on a dime, and now we have a NDT who seems obsessed with his own celebrity. Such a shame... As I am writing this, he's saying "the end of the world is coming, come and join my cult", mocking others but unaware of the climate change campaigning he's done in his own career... The irony is strong with this guy...

[@nezar-6889]

@@kellenwaters9087 could I possibly have access to the lecture you are referring to?

[@Kris_M]

@@nezar-6889 "so many comments are attacking him" why do you call it "attacking"? Commenting on the attributes you perceive to be negative isn't attacking. And I don't agree it isn't an issue with someone like Alex, to me it seemed Alex found it unpleasant, as most people would I think.

[@zombieinjeans]

I hope Alex gets to speak with the physicist/philosopher David Deutsch on his podcast. That would be such a fun conversation. Their intellects need to meet!

[@TheEconomicElder]

Neil doesn't realize that large language models are not programmed. It's deep learning and what it learns it's not subject to human input.

[@TheEconomicElder]

@@charlespalding very clever smart ass 🤣

[@milorodval678]

Eliezer Yudkowski is open to being interviwed. There are REALLY good reasons to expect the end of the world if we don't solve the Aligment.

[@gabogonzalez9428]

YES

[@fog3911]

Yes please!!!

[@tadeja2583]

man, neil is enjoying himself

[@tobyhendricks9951]

I agree with Neil's point on AI here. I think we personify AI far too much. People say "What if it wants to get out" "What if it wants to destroy humanity" "What if it values its own life" There's not really anything to suggest that an AI, built in the current way we build AIs (even if it were more advanced), would: 1) "Feel" the need to do any of these things without a human intending it. We developed these traits due to evolution, an AI has no use for self preservation or to even strive for its own gain 2) Would be capable of enacting these desires without a human intending it. Even if an AI decides - "The best course of action for me right now is to destroy humans", why would you empower it with the capabilities of doing so? It likely wouldn't even have the urge to complete this task in the first place. We probably don't need to worry about AI becoming anything close to sentient because our current models for AI are still heavily bound by restrictions imposed by us. It's not the sword, gun, bomb or AI alone that will bring an end to humanity as these creations are useless without the humans who wield them. Don't worry about the machines, worry about your government and its relationship with other governments.

[@Osfun]

I think this kind of conversation is always a good thing. Having scientists perspective only makes a better and larger public debate. However, at the same time we have to recognize that politics is, by definition, the realm of situated issues. As soon as you look at the bigger cosmic picture human affaires disappears. So, a cosmic view cannot help us directely to make a better world. It might give us a new way of seeing things, but it is useless to act on a daily basis. So, using his same example, one could say that those astronauts on the surface of the moon could help each other and have a good and not a conflict relationship just until they (and I'm talking here about a community, not just two individuals) have to make decisions about their life in common.

[@GummiTomm]

Doesn't surprise me Neil thinks astrophysicists are the most humble people in the world.

[@sigigle]

The most humble. We’re great at being humble. No ones as humble as us. It’s so impressive how humble we are. We should be worshipped for our amazing humility.

[@whoisandrewblack5679]

Yeah, what was that??

[@markth078]

It surprised me, too. I imagine he meant the most humbled scientists on tla regular basis.

[@GumballEdits]

Great to see you getting big profile guests, its a shame that Neil was manic during this interview because he just interviewed himself haha

[@Gridpipe]

Its ashame he never lets anyone else get a word in. It would be interesting to see how he reacts to being challenged in real time. We will never know.

[@someguy2135]

I would have loved to see Alex challenge NDT's attacks on veganism, back when Alex was making that kind of content.

[@davidalcocer6547]

I like how Alan watts puts it “you are the Big Bang not separate from it, but you are it! You are the Big Bang, the original source of the universe, coming on as a human being here and now”

[@charlesokike847]

I wish NGT would interrupt so much. But Alex is mature enough to realise that he was there to encourage NGT to talk.

[@LeaFsinger74]

I'm a fan of Neil, but was yelling at my phone (listening to the podcast) for you to mention AIs biggest current threat is all the misinformation! But I'm sure he'd have a snappy answer for that too.

[@InigoMontoya-]

Unfortunately, we do not have good arbiters to determine what is misinformation. Both sides of the political aisle are disseminating information that is not established fact, and both sides of the aisle have influence over the experts they use. Both sides have vested interests in promoting their own “truths,” but some of their interests appear to be rooted purely in power and wealth, and not necessarily in the good of humanity. Until we can remove political and monetary influence from things, we will always have misinformation.

[@howlrichard1028]

​@@InigoMontoya- I agree that misinformation comes from all sides and sources, but only one side has been proven to actively and knowingly pay hundreds of millions with the specific purpose of spreading misinformation.

[@riser9644]

who decides misinformation , misinformation already existing on the internet , its nothing new

[@johnsullivan5101]

70 million people voted for Trump without the assistance of AI. People spent 1,000's years burning witches before the machines AI runs on were even dreamt of. I don't know if requiring critical thinking skills as a condition for graduating from school would help. I'm sure lots of perfectly competent professionals - mechanics, programmers, doctors and the like - who could not function without dealing rationally with the physical world still voted for someone who said hundreds of times he was going to build a wall and get Mexico to pay for it. Complete nonsense to anyone who knows anything about Mexican/US history or even human nature but not an issue for his voters.

[@truthbetold8233]

If only Neil was a little less in love with the sound of his own voice.

[@bobon123]

He started with a very sensible "I am not an expert of AI, and I will not present myself as an expert", that made me think "Oh nice, maybe he will be more humble about this, and discuss the point carefully". Boys I was wrong. The part when he proposed "to just unplug it" if it went out of control was cringe worthy...

[@morbidmanmusic]

Please just stop tuning into Neil on podcasts. We don't need more cliche commentary.

[@bobon123]

@@morbidmanmusic Lol, only positive commentary are allowed? Or are you trying to defend his ignorant position? Are you also convinced like Neil that you will always be able to just unplug an AI "because they need electricity"?

[@jimj9040]

@@morbidmanmusicI’m aware of a lot of cliches, but unable to identify any that you might be referring to. Maybe you don’t know what a cliche is.

[@oxygenerate]

Yeah his "optimism" on the subject is clearly naivete.

[@furrybear7853]

Will never get sick of hearing NDT...talk about the best educator ever and smart as heck!

[@therheaking]

Yay!!! So happy to see Dr. Tyson on the show ❤

[@maryannegervais1747]

Thank you for this interview. I never tire of hearing Neil’s intelligence at work

[@Jaymzmiller]

Nor does Neil!

[@ankushds7018]

@@Jaymzmiller omfg

[@Siska0Robert]

Over the course of the WR series, I've never laughed so much. Gotta love NdGT, such a positive energy in the world.

[@laurajarrell6187]

Cosmic Skeptic, Alex, and Dr. Tyson! Great conversation. At the quote, and talk of Cosmic Perspective I thought of the movie Enemy Mine. 2 warring species stuck on a planet. Shipwrecked. And it showed, as 'stories' always have, that once the ignorance is gone, so is the hatred. And as for self driving cars, we could put all roads underground. Animals and people saved. 👍💖💙🥰✌

[@chronicmasterbaiter5467]

Yep, maybe we should even stick the cars together to make these "megacars", where numerous people can fit, to solve traffic jams. Talk about killing 2 birds with one stone hehe. Elon musk should really adopt this idea I feel like. He is very smart with new technologies afterall.

[@skeptischism1324]

There's such a fine line between being passionate about science and explaining your perspective on things or the consensus, and just being so full of hubris that you have to interrupt someone and interject and almost to act Superior to them as a result. I respect the s*** out of Tyson and his broad reach in trying to elaborate on a lot of scientific things with a lot of people. But he can be very condescending, and there's nothing more frustrating than watching someone not allow their opposition to get a word in edgewise. I'd argue that's one of the better things about say even comment sections you can have a discussion and you can't be interrupted per say.

[@kingwen1286]

I like the way Neil boldy refuted fears about AI but as you persisted he shifted to agreeing with you. I like the way both of you did that.

[@whatsup300]

I've waited for years for you to invite Dr Tyson, well done.

[@holadonkey]

not really he's overbearing and missing the point on too many issues . Alex is wiser than Neil .

[@whatsup300]

@@holadonkey I didn't ask for your criticism, keep the cliche to yourself, you hypocrite. You saw the thumbnail, you shouldn't have clicked the video if you already thought that of Dr Tyson. I don't watch Alex's videos because I think he's wise, who the f*CK cares, I'm intrigued by his perspective on religion and faith and I watch Dr Tyson because I'm intrigued by his perspective on matters of the universe even if he sometimes interrupt the host, that's no reason for me to become a cliché. He's always overly excited, and clearly doesn't think himself better than anyone else. He even agreed to talk to Alex, a relatively small KZbin channel. It's a good thing that his success is independent of obnoxious people such as yourself.

[@w0ody16]

In the last part of the AI conversation it brought this to mind: AI is the type of tool that would be amazing at sorting through shitloads of extinction level asteroid threats and data so that we could actually intercept it. Stuff like this makes me think that responsible application easily outweighs the dangers.

[@derAtze]

An analogy: having all private messages and postal secret removed could unveil and prevent an terrorist attack. Is it worth it exposing every human?

[@glastiger7598]

I think you do not know all the dangers or you are a follower of longtermism. And even if the latter is true I still don’t really think you’d reach yoyr conclusion.

[@cocobutter3175]

​@@derAtze The problem there is that meteors and astrological data isn't alive, isn't sentient, and has no right to privacy. So yes, it's totally worth it to have AI tracking astrological phenomena to make sure it doesn't collide with earth or create some threat. If someone proposes to use AI to go through all our communications, that's going to be a big no, and hopefully people will be against it enough that anyone with that idea will be voted out as a warning to others. However... If it was a decision between my messages being read by a person or an AI, I'd rather have the AI do it. But the whole danger is not the AI itself, it's what idiot humans decide to do with it.

[@derAtze]

@@cocobutter3175 what i was trying to say was not that stars have privacy rights, but that a possible positive outcome not always rectifies the means

[@cocobutter3175]

@@derAtze I don't think having AI analyze astrological data is a terrible means, but I can see humans using it in other questionable ways, promising a positive outcome that may not be worth the means. I still feel like AI will be an incredible tool, even though people may use it wrong. Really, I feel like it's not a question of "should we" or "shouldn't we", I feel like the evolution of AI is inevitable, just part of our progression. I feel like it will be necessary to get past certain plateaus in our technological advancement.

[@AlanWinterboy]

Well done, Alex. Thanks for a stimulating discussion. P.S. Who's going to build that dangerous AI, Neil? The US military for one.

[@morbidmanmusic]

Prove it. Talk is cheap.

[@AlanWinterboy]

@@morbidmanmusic How do I prove the military is working on this? I'll take my smartphone down to the nearest research base and ask them to give me a tour. There isn't a single new technology that the military hasn't weaponized; we just don't learn about it till after the fact.

[@bartstewart8644]

I love it when someone makes an observation that everyone knows but we just never think about it. It hides in plain sight. At 31:15 Tyson says AI only knows what's on the internet. I knew that, but it needed to be said! (And if AI gets as hateful as the internet, we are screwed.)

[@irrelevant_noob]

Technically, it's not _absolutely_ true... NdGT being able to discover that new piece of knowledge means that any "sufficiently-advanced" AI would be able to, as well.. Hopefully, we have Turing to thank for his insight that no automated system is able to "compute" everything, so even if the AI discovers that same info, it might very well be diluted into the ocean of other stuff it discovered over time. :-)

[@johnathanfindlay9112]

I remember the first time objective information was taken that way in my observation I had the same mind blown experience at that realization.

[@johncormican562]

Gosh was that a fun and enjoyable ride. First time I got to see Neil in something longer than a short clip. Alex, as usual, in my opinion, got the very best of Neil in the way he always does with pure, exciting weapon grade respect for the person he's interviewing. AI is just the hottest topic for me at the moment. Extremely grateful for this dialogue.

[@2kt2000]

Neil has a channel "Star talk". podcast. Over an hour sometimes, with guest.

[@Camcodrummer]

Love his insight on things.

[@matthewwalker3099]

I can't help but notice NGT employing bravado more than what used to seem like genuine curiosity. It's like he's straining through his teeth to convince himself of his position. Not exactly the same guy I used to enjoy listening to from time to time.

[@eskilwadsholt4289]

Nice oneversation! Alex even holds every belief he references: Alex: “Some say …”, NDT: “Ah, you think …”

[@kickingkong1450]

Just for those that don't know, in the UK we usually measure large distances in MILES. We go by MPH, not KPH. So, it appears Neil doesn't know everything. But what he does know, he makes it interesting to hear...

[@Its3am]

The idea of getting different guests on is that they’re different. Neil is Neil.

[@Its3am]

Watching this again, Neil is just ranting, no coherent arguments. He basically contradicts himself half way through his sentences. He’s done well talking to Joe Rogan but I think Alex is a bit disappointed.