Neil deGrasse Tyson Is Not Afraid of Artificial Intelligence | Within Reason #28
ฝัง
- เผยแพร่เมื่อ 12 พ.ค. 2024
- To support me on Patreon (thank you): / cosmicskeptic
To donate to my PayPal (thank you): www.paypal.me/cosmicskeptic
- VIDEO NOTES
Neil deGrasse Tyson is an astrophysicist, director of the Hayden Planetarium, and one of the most recognisable communicators of science in the world.
- LINKS
Starry Messenger: Cosmic Perspectives on Civilisation: amzn.to/40Cdkgi (affiliate link)
- TIMESTAMPS
0:00 Introduction
1:11 Feeling small in the universe
9:55 Is the world ending?
14:35 Artificial Intelligence
32:30 What is the greatest threat to Earth from space?
36:02 Dr Tyson's "Forbidden Twitter File"
45:28 Outro
- SPECIAL THANKS
As always, I would like to direct extra gratitude to my top-tier patrons:
Itamar Lev
Evan Allen
John Early
Dmitry C.
Seth Balodi
James Davis
g8speedy
James Davis
Mouthy Buddha
Audio production by Charlie Shan: www.shanmusic.co.uk
- CONNECT
My Website/Blog: www.cosmicskeptic.com
SOCIAL LINKS:
Twitter: / cosmicskeptic
Facebook: / cosmicskeptic
Instagram: / cosmicskeptic
Snapchat: cosmicskeptic
The Within Reason Podcast: podcasts.apple.com/gb/podcast...
- CONTACT
Business email: contact@cosmicskeptic.com
Or send me something:
Alex O'Connor
Po Box 1610
OXFORD
OX4 9LL
ENGLAND
------------------------------------------
So nice of you to let Neil interview himself
🤣🤣😂
Lol Alex trying to butt in
You're a cliché
U just hate him cause u r rasist and he's black. Neil has accomplish in science u and your generation will never get.
I have seen Cosmic skeptic interrupt before. Did he forget how to do it?
"Astrophysicists are the most humble scientists" - Neil deGrasse Tyson, an astrophysicist
His lack of self awareness is hilarious. Such a bellend
The irony
This is an underrated comment lol
Lmaooo, I get a feeling he's thinking of his peers and his idols when he says that, not himself
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.
Neil is probably the easiest person to interview: just ask him a question and watch him go.
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
@@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.
Robin Williams used to do that, take over the interview, but it was much more entertaining.
@@Anjalena #1. Robin Williams was a comedian/actor/ entertainer. #2. He was white so he didn't trigger your racism.
Me, literally me
Great video alex! I hope you get to speak to Neil deGrasse Tyson soon!
Hahaha
😂😂😂
underrated comment :D
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.
Motivation. Computers Don't Care About Anything.
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.
Which is pretty goddamn impressive, to say the least.
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.
@@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:)
Neil is the kind of guy that will wake you up to tell you he's going to sleep.
😂😂😂
Wake up Alex, it's time for your sleeping pill.
Lmao
What does that mean?
@@reinforcedpenisstem he is saying he finds NDT very annoying and obnoxious
Neil is the type to ask you to help him with something, then just make you watch him do all the work.
🤣🤣🤣
Hahahahahah 😂 this is so true im like that too tho so it was a moment where you held a mirror up to me friend
I'd be asking questions the whole way and probably learn allot
I’ve gotten paid for doing that kind of “work” before, so I wouldn’t actually mind🤣
Neil’s passion and excitement for these subjects have reached conversationally uncontainable levels 😂 and I love it!
He’s a shill
exactly lol
@@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.
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
@@ljubomirjuraj9658 raid shadow legends?
I've never seen someone whose skill at delivering one-sided speeches is so vastly different than their ability to hold 2-way conversations.
Can you elaborate that further? I don't quite understand
@@Dann-md9eq Neil is great at giving speeches, and really sucks at having conversations.
@@garywood97 he's great at conversations 🙄 watch the show.
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.
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.
Alex : It seems experts in the field feel that AI could be an existential threat to humans.
Neil: We'll just turn it off.
Terminator hunting you? Just say no.
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.
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
@@mordin999
Daily reminder that AGI doesn't exist yet, not one single line of working code has been written.
ChatGPT is just a server some where. You think you can't turn of Google, but you can
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.
I can't believe that NGT actually apologised for interrupting at 24:48 😳😳😳
😳
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.
That's a huge overstatement.
@@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.
@@MrSnootsdumb
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.
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
@@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.
Absolutely, NDT is the worst at communicating ideas.
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.
@@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.
Awesome collab! Alex’s channel and brand is growing I love it
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
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* TH-cam 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.
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
Alex should try to interview someone like Robert Miles, who has been creating educational TH-cam videos about AI risks for years. Even appeared on Computerphile a couple times.
@@cookiecan10 Great suggestion! Huge fan of Robert Miles; just edited my original comment.
@@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.
Or interview Yoshua Benugio.
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.
Alex, thanks for this episode. For one who loves Astronomy, I enjoyed it. You're doing a great job, Man.
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.
Brief flash of self-awareness
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.
Super cool that you were able to do this!
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.
Loved this conversation!! Wish it took longer!
Alex speaks 7.13mins in total of this 44.45min podcast.
That's 16% of Alex; 84% of NDG.
How did you count?
I salute your diligence
Sounds like you had a very fruitful bout of boredom😂
Well technically he's doing an interview. He's not the one being interviewed.
I still think NGT kept interrupting Alex more than I find desirable.
*Great Interview! Cant wait for more videos like this!*
Amazing interview as always Cosmic!
Great chat. NGT can sure talk...and I love it. Lovely discussion thank you.
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.
Absolutely
He’s unquestionably an arrogant person.
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.
@@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 "
@@willashland4597 writing a million lines of code must have fried your brain because he literally said he wasn’t a qualified expert after that.
The last 10 minutes of this podcast are the most illuminating - Neil for the win!
Thanks for the interview! Very fascinating to get this perspective.
44:30 Total highlight of the conversation. So true. Includes almost everyone.
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.
I really like Neal and his insight is fascinating. I just wish he would let others talk and direct the conversation more.
I mean, aren't we here to hear from the interviewee, not the interviewer?
@@Firehazard159 isn’t the interviewee there to be entertaining and not threatening?
Your complaint cuts both ways
@@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
@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.
@@whoisandrewblack5679 threatening??
Some logical oversights from Neil off the top of my head:
- If the tweet can be interpreted in multiple ways, then someone who disagrees with any of those ways is valid in criticizing it - no matter if their critique is solid or not.
- If there are positive things that we're potentially missing about AI, then there might also be negative things that we're missing about AI. So don't paint too rosy of a picture.
- Trying to wait for the public to mature to a point where they're able to handle the contents of a certain tweet/message without exposing them to said content to learn from doesn't work.
- I find it odd that Neil says his biggest worry is the threat from asteroids and then follows it up by saying that the biggest impact we need to fear isn't big enough to be an extinction event. So, does he then think that there are no threats to humanity that are at level of an extinction event?
That was a great episode. I love how contagious Neil's enthusiasm is
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
It doesn't help that he talking to someone who is clearly not trained in science.
@@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.
@@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.
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
@@masonmurphy1941
NGT tries to explain complex scientific Ideas in simple and comedic language that most people should understand. His answers seem low bar due to the fact that it's usually the level of those he is talking with.
NGT is extremely confident and knows that when he is being interviewed that he "IS" the centre of attention.and the interview is about him and he generally conducts interviews on his terms. In otherwords he is controlling the agenda not the person asking him the questions. That confidence may come across as arrogance or narcissism but it's more likely his way of not getting tied up in pointless small talk.
very very intresting guest Alex. Thanks for all of the value you are providing to us.
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.
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!
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.
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.
He's pationate, I'm no genius but whenever someone asks me about starwars lore I go on a 30 minute explainer lol
If only Neil was a little less in love with the sound of his own voice.
Dr. Tyson on your podcast wow! That’s awesome.
Wow! You got Dr. Tyson on your channel! You've truly arrived! Kudos to you!
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.
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!
thankyou this was very enjoyable. Its good to see science educators being represented.
Yay!!! So happy to see Dr. Tyson on the show ❤
Neil is often almost unbearable to listen to in these long form interviews. I think he should really try to reflect and realise how petutlant he can sometimes be. He constantly cuts people off and steamrolls the entire discussion, prevents a good back and forth (even with points of agreement), and almost shouts in a shrill manner for the duration. Obviously, this is not a critism of any of his thoughts, merely how he presents them.
might be only to your perspective but not everyone's. "I could do this all day"
Oh really!,Actually, I’d really love to listen to him all day
@@yohannesgetachew5869 Weirdo.
I wonder how many thought provoking perspectives NDT misses out on because he doesn't let anyone else speak and engage with them in a proper back and forth conversation.
What a load of rubbish. He makes the most of his time available to him to share his educational experience he has with us. It's all limited time.
@@roberthereandthere4366 ah yes, that's why he talks about mice in basements in his book instead of poverty (which he actively didn't choose to put in there because "there wasn't enough to write about"). Neil is maybe smart, but he's also egotistical, illogical, ideologically blended and not that great of a guy. I would summarise him as an average but well educated redditor.
@@roberthereandthere4366 Hi teacher I'd like to learn about "a"and how it relates to "b" and "c"
Teacher: "I'm so glad you asked, I'm going to exclusively talk about 'b' the next 45 minutes and maybe mention those other two if it happens to come to my mind randomly"
@@NotLordAsshat Yeah, that sounds right for an Asshat
@@roberthereandthere4366 that's hilarious. NdGT is insufferable.
I remember the first time objective information was taken that way in my observation I had the same mind blown experience at that realization.
I think your podcast works much better when you have an actual conversation.
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!
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
He expanded his arguments more fully in "Shadows of the Mind" (1994)
@@charbelbejjani5541 very interesting.
Thx for the Info!
@@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
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.
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.
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.
Love the interview
Great interview. Neil is a wonderful teacher and I learn something new every time.
Awesome podcast. Two people I admire deeply. Thanks, Alex!
Neil: I'm no AI expert.
Also Neil: [talks for 30 minutes about how AI is perfectly safe, it's all fine.]
no no no. Listen again
He's saying he's a science expert and you don't need to be afraid of science so long as there are regulations
he asked for his opinion , he gave it with a disclaimer , don't hate because his opinion isn't same as yours
@@pooglechen3251 and how do you regulate China or Russia?
@@nakkadu
You secretly build it, just in case!
And leave it switched off until some other idiot turns theirs on.
And now I will thank you for posting the full video on the clip. Even better..🎉
I would suggest Robert Miles for the next talk on AI 😊
Or Eliezer Yudkowsky, he's been on a fair few interviews lately!
💯
@@vanderkarl3927 💯💯
👌
Robert Miles is one of the best. And if you've watched computerphile, you've probably seen him, but not known his name.
1:22 Neil's smile to frown scared me
I enjoyed watching this episode, thank you.
Great interview 🙏
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!
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
@@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
Everyone and their mother's dog overuses "cancel culture" just to make the blandst point ever.
@@DarthVaderfr Go back in your crib with your sippy cup, adults are talking. You're literally crying about nothing, you unintelligent clown.
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!
Great Video!!
Fuck Yeah! That was awesome Alex. I think basically, at the end, Neil was saying he wants to come on and chat again. I want to hear more!
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. 🎉
Yet people always have him on their show.
Go figure. Maybe yoU should tune out, as you're commnet adds little.
@@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!
@Rafal the internet is seriously stupid these days
@@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!
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.
YES
Yes please!!!
Thank you for being you .
Specifically, I'm wondering about the effects of the rate of job displacement on society because of AI and what that would entail as it is a completely different ballgame in that the pace of this is faster than humans can adjust if 80% of the workforce can be replaced for example.
Love his insight on things.
Someday I wish to find someone who loves me as much as Neil loves Neil.
I don't think you have a realistic chance on that seeing how much Neil loves Neil
I think you are conflating "Neil loving Neil" with "Neil loving to be Neil." I'd love to be him too, at least for a while.
I've always loved you...but here you are ogling astrophysicists😞
@@feedback1204 very true. He's quite the specimen.
i think the way he talks is 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
hell yes. I love seeing this. Two great minds coming together.
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.
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!
The irony and lack of self awareness of Neil saying that physicists are so humble is staggering
"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.
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.
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."
@@orangestapler8729
I think the point NGT was making was metaphorical.
The suggestion of pulling the plug wasn't so much his lack of understanding more over that the developers of AI ensure that there is a plug to pull. Hence the comment "Why would you do that?" The implication being that those that develop AI must be aware of the risks and unless they are absolutely nuts they are not going to relinquish control to something that has the potential to outsmart them and end humanity.
It's more of a philosophical discussion more than a technical one is my impression.
@@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.
Great episode ❤😊
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...
He could just talk it to death
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. 👍💖💙🥰✌
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.
A very interesting podcast indeed !! Love Neil Tyson
Nice oneversation! Alex even holds every belief he references: Alex: “Some say …”, NDT: “Ah, you think …”
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?
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?😅
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.
@@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.
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.
@@wilfred5656 Why do you think Elon is scary?
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.
Great video
I like Neil's point on how new technology makes lives easier and gives us more space to solve other problems. Similar to how creatives utilize image generators like Bluewillow AI, they use them to produce and they do anything post-production.
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.
@@charlespalding very clever smart ass 🤣
What an incredible conversation to listen in on.
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 :)
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.
@@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.
@@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...
@@kellenwaters9087 could I possibly have access to the lecture you are referring to?
@@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.
There’d have been a good segue into asking about regulatory changes regarding AI and how mitigating potential for harm would be beneficial in the field of AI just as in his example NDT of aeronautical engineering.
24:50 i did not have NDGT apologizing for interrupting on my 2023 bingo card
Always a pleasure to hear Neal talk ❤
Neil would seem to agree lol
I thought Alex steered the ship on this interview just fine. Just Neil being Neil. I enjoyed the interview.
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.
Great show
This was exhausting. I feel like I just got incoherently ranted at for an hour.
Doesn't surprise me Neil thinks astrophysicists are the most humble people in the world.
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.
Yeah, what was that??
It surprised me, too. I imagine he meant the most humbled scientists on tla regular basis.
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.
Good stream
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...
Please just stop tuning into Neil on podcasts. We don't need more cliche commentary.
@@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"?
@@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.