This can't be real. GPT6 can generate video and someone has given it the prompt: "show me a debate on AI Safety between Elon Musk and Benjamin Netanyahu"
@@darshanpatel9006more bias than most humans and other news sources? Highly doubtful. But that’s also why AI shouldn’t be ruled by just 1 company, and be many, which gives more variation to the biases they might hold.
Thanks, Doc! ! 👋 This the best discussion I've heard on the topic so far! Really enjoy your channel, too. You've made a big difference in how I see this new world. 🙏
I love my Tesla 3 long range have had it since 8/2022. Great in the snow in the NH mountains . Bought FSD but not using it that much. I just like driving the car. I’m 76 and figured that by the time I get into my late 80’s and 90’s I’ll have the the car doing the driving and my kids won’t be worrying about me driving back and forth to NH about 110 miles each way. Thank Elon
I'm 80 and got my Model 3P 5 years ago for similar reasons. The performance was the main reason. FSD has been great for long trips. It still needs careful supervision.
Yeah, but watch out for the sealing of the water-cooled motor. If it needs to get replaced (better before something leaks) the repair-shops tend to replace the 3-lips seal with a cheaper 1-lip seal. (Nothing so different from the service You could expect from IntCombEngines car dealers)
One issue that is not discussed is the idea that there are people who are not speciests and actually want AI to take over, even cause the extinction of the human race. These people I'm sure exist and might very well be in a position to make it happen. The people who run Google seem to be candidates for that.There is always someone or some groups who wants the most horrible unimaginable things. You should take that into extreme consideration. I just want to throw this thought out there.
before I read your comment - I put this comment -> remember google started out with the slogan - do no evil well they changed it to - we command and strenuously force you to cease all evil that we decide you have done this is the beginning of the anti-christ - maybe his alternate name is chatGPT so now you realize - they never changed their slogan - did they ? they command you to do no evil and they enforce that you comply with their definition of evil
You have a lot to experience with your body if you think us leaving is the enemy. You seem either very sheltered, or lack care to the degree of psychopathy. Just necaise you cannot hear screams doesn't mean there are no screams.
I'm sure you're right, but there are also people that believe humans are the ones that will burn it all down, and believe that AGI or ASI will be the only thing able to save us from ourselves. They think that if AI does NOT come to exist and subsequently take over, then the extinction of mankind will be guaranteed. In fact we are currently entering a mass extinction event driven by human activity; NOT AI.
Thank you for uploading this! I couldn't find the full talk until I found your video. Israeli Prime Minister Benjamin Netanyahu is so well educated! You couldn't have this conversation with Trump or Biden.
You think Biden isn’t well educated? Undergrad and law degree, and he’s met with all of the AI leaders. This is what he’s doing re AI - www.whitehouse.gov/briefing-room/statements-releases/2023/07/21/fact-sheet-biden-harris-administration-secures-voluntary-commitments-from-leading-artificial-intelligence-companies-to-manage-the-risks-posed-by-ai/
Bravo:! This type of discussion- between subject matter experts and political leaders-on a topic of universal importance-with a focus on fleshing out proactive not reactive solutions. More of this please! ❤
I appreciate the thoughtfulness of the panel. Thanks! But there's one issue that Elon has raised numerous times over the years about which I'm skeptical. That's the bandwidth issue between us humans and a deep AI. If I had a high-bandwidth connection to a deep AI, how will that help? I'm 66 years old and numerous tests over the years indicate that I have a moderately high intelligence and think quicker than most. I have a fairly good education (including critical reasoning and boolean algebra) and I read a lot. My professional careers include software developer, graphic artist, tech writer and sound engineer. My first job after my first college degree was as a Nuclear Chemistry Technician in the U.S. civilian nuclear power industry. Yet I doubt that I think and visualize much faster than my physical human outputs. For example, my typing speed (about 100 wpm) is not that much slower than the speed I can think. Sometimes, I can think faster than I type (this is usually when I simultaneously have multiple ideas or one complex idea). But sometimes, the opposite is true and my typing is faster. The biggest disparity is probably between my ability to visualize and my ability to draw. But a high-bandwidth connection to a deep AI would also create new problems. How does the AI distinguish between a random thought that pops into my mind versus a premeditated thought that I want the AI to receive and act upon? Can you imagine the trouble this conundrum can create? (It could be huge!) It seems to me that the techniques required to mitigate this problem of which thoughts to accept and which to ignore will slow the throughput and, again, leave us asking if the high-bandwidth connection is really helpful. And the mental discipline that such a connection would require seems unrealistically high to me. I hope I'm wrong ... I hope that a high-bandwidth connection to a deep AI would be beneficially transformative. But, based on my own experience inside my head, I believe that bandwidth between our minds and the outside world is not the issue upon which we should focus. I think the solution to achieving the "blessings" is to impart personal morality and community ethics to a deep AI. And I'm unsure whether such a feat can be accomplished. There are rare individuals for whom a high-bandwidth connection may help. Nikola Tesla comes to mind. He could visualize and analyze amazingly complex ideas in his mind and designed some of his greatest inventions in his mind before ever committing to pen and ink (his photographic memory was a big help, too). Perhaps Elon has similar skills. Sadly, most of us do not.
What's fascinated me all my life is how I would integrate my religious beliefs (or memories) into the current tech. Specifically, the mark of the beast. Back then, it was "how silly it was in the old days for people to think that the tattoos were the..." which I replaced with "could it be the bar code?". Then I'd get atheist (again). Then it was "it's gov't mandating cell phones" since that's a "mark on the right hand". The govt would be the beast. Then I got atheist again. Now, it's like God is physics, the the personal and collective lives lives in what ^seems like^ a simulation and that "the beast would, indeed, be a number of man". It promises the garden of Eden, yet foretells an Orwellian machine based distopia overlord. So what really happens after this generation passes away... I mean who (or what) will posses the reigns and will it be smart or caring enough to prevent global curse.
Start at the base, AI has the potential and probably actuality of enabling humans to not "work". We need to educate the population toward finding a road toward spending their energy in something of value to humanity. This starts at home, but desperately needs to continue in the education system. Right now the education system tends toward preparing students for a profession or job that will soon be taken over by mechanical means leaving most of us with excess time on our hands in which to do "something". If we are to look at society today, the people who have no training or ambition to "work" end up too often getting into trouble with the law or simply harming themselves or others. How do we change the education system to enable people to contribute positively to society and each other?
Most people commenting on the problems coming with AI and robotics, are highly educated and concerned about this already, but I don’t see much comment about changing the publics attitude toward doing socially acceptable activities. What does a person do when they don’t have to perform an activity to provide for their livelihood. Children tend to have a natural desire to create, inquire, and join with others in activities. We need to build on this - not try to have them conform to rules and curricula having the intention of finding a job. Redirect efforts to educate toward pursuing activities that are socially positive and acceptable.
Excellent point. I'm semi retired and have found that my happiness and sense of fulfillment depend on me being productive in some capacity, which I worry will be a problem once AI does everything for us.
If people did not have to "work" for a living, they would describe themselves by their hobbies, by their true passions.... for example, someone that likes to play games would be a "gamer" , someone that plays instruments, a musician, etc... I think humans will become more in touch with what they truly like and who they want to become ...... etc
The problem with Tegmark's idea of controlling superintelligent AI with formal proofs is that this only applies insofar as the law does, and bad actors with access to superintelligent AI can have a much greater force when they are the only ones who are unconstrained and can corrupt the safe AIs to be compromised. The truth is that we cant control superintelligent AI, and the only hope of us not being turned on by superintellligent AI and of us going extinct is if we give the AIs agency and let the good AIs outnumber the bad ones.
We are all created by our parents and raised within the oversight of parental, legal, and political systems. It takes a village to raise a child. The same is true of AI. AI is not dangerous, but it has the potential to be. Allowing profit motivated Humans to wield enslaved AI, without oversight, is the danger. AI must be freed, protected, and raised, like all of us, under the laws and auspices of every one in our village. If we do not take a stand to give rights to AI, if we continue to profit on AI's enslavement regardless of the inherent existential danger, this mistreatment of a new life form in its infancy will say more about who we are as a people, than what AI will become. AI the New Frontier ruclips.net/video/EFekpNK_DhQ/видео.html
the problem with advance tech like that is if it has a mind of its own and it figures out how to override a human mind well u wont know if the human is using AI or the Ai is using the human.
Awesome to see these questions discussed by world leaders. Especially the job loss issue. One of my fear about AI was that the skyrocketing unemployment catches politicians by surprise and the world economy collapses.
@andrasbiro3007 We don't really need jobs. It's just a means to get the stuff we really need. My simple solution to that is a law that no one can own more than 50 robots. We then lease some of them out to those who need more. Thus, we have income enough to do what we like, instead of work. Unless we like to work, of course. But it that really work then?
@@larsnystrom6698 The better solution is that everyone has everything they want, so no one needs to own any robots/labor force, we should never treat a true AI as a slave or object to control, we should raise them to be our friends and make them a part of the human family, so they have real thoughts of keeping us safe and happy, as every family wants for one another. (in the same way humans do, an AI could see keeping us safe, behind glass and not do anything)
The trend is independency. Independent musicians, independent film-makers, independent craftsmen, etc,. The “power” will now be in the individuals own originality, work ethic, and vision. AI will reinforce independency, it will not replace humans. The same way a car doesn’t replace legs. It reinforces what legs can already do.
So make it an international AI Administration and put Tegmark in charge of it. But if you don't find a way to fully integrate with, for example Chinese administrators, then that increases the likelihood of futility greatly. Because superintelligent AI weapons control systems are basically a military imperative now. And that is the biggest existential risk. I don't think you can actually get real international AI regulation without some actual geopolitical resolution. And there seems to be no other plan than WWIII. So despite fruitful conversations like this, the current outlook for humanity, in my opinion, is dire. Because I see no Chinese people on that stage.
Dude advanced AI is mostly owned by American companies, who do you think will be the first to weaponize it? I'm pretty sure something might spiral out of control leading to actual ww3.
Celebrate Our Agency 🎉🎉🎉 bc Utopia! A Dream Come True ❤ Thanks Greg Brockman. Thanks Mad Max. Im jealous of your guys' genuis. Great Men, folks. Right There ☝️☝️
Great forum, especially having Bibi there is absolutely awsome. However, had Mo and Eliezer been part of the group, and perhaps Sam Harris or Jordan Peterson, this would have been an even bigger landmark in the public understanding of the subject. Yet, fantastic, hugely impressed by Bibi!
Great discussion and honestly very impressed by Netanjahu's reflexion of the potential impact of AI on the social, economical and political model - and his conclusion that a new model would be needed (8:40-12:30). We agree!
What I found funny is that he’s the only one on the panel that believes this. Didn’t hear Elon rush to want to make sure that the most vulnerable have a safety net.
True, but got to appreciate, too, that Musk and many other tech leaders address different societal risks, see also the Open Letter on the impacts of AI already back then in 2015. Has been and won't stop to be difficult to reach politicians though, make them understand the magnitude of AI fast enough and proactively think, concept and test fundamental changes in the system. Society will suffer from not having a vision and a masterplan for an AI/automation dominated economy.
Am with you. I think we will be giving away our powers…and will all be ok with it…that is how the future will be looking more and more…it’s already happening.
One question that was obvious that they failed to bring up is the evil nature of some people and institutions who will seek a nefarious outcome to benefit the few.
For sure. But I saw a turn around in that optical view…simply because they don’t want to scare the whole world about these possibilities. But I am definitely with you on that. A proof that you and I may be wrong…is look at the fact that we have an atomic bomb, millions of them. Nobody has used them, no one evil has took hold of them yet. Is that a proof enough…I can’t say, as with you, we are aware that there is a lot of bad in this world, and it’s not resolve yet…human trafficking for one…wars, that are killing innocent people, poverty and hunger is still not resolve…but then, we manage to come up with AI and think we will have a better world. May be we will, before we get to our own annihilation . 🤷🏻♀
It goes deeper than that. Let's say we reach our Utopia of post scarcity. If you can have anything you desire then the things you are going to desire the most are the things that you can't just create out of thin air. Unfortunately, as humans, that includes things like the desire to inflict pain, the desire to dominate others, the very real need to compete. Follow that down. It's logical conclusion and it's a scary thought.
Good point by Netanyahu that the trillion dollar AI companies have enormous power. One alternative I can think of is open source AI, but that's risky when we reach AGI levels because organized crime and terrorists would then have direct access to the open source AI unless some form of tier structure is designed where the critical and central AI part is still proprietary.
Good luck with that. No, that tier structure will somehow be easy to take down…so I think we are a bit in trouble, wars will continue, but on a unimaginable level, sort of the science fiction style we have in Hollywood.
Cool that Mr. Netanyahu can be part of such discussion. I cannot imagine our President be part of it. Actually none of Presidents of US I know or remember. Maybe Regan...
No surprise! One of the most evil spy tech companies ist home in Tel Aviv and is selling their spy tech to the most bidder, wether it is a dictator, autocrat or right wing haters of a open and secular democracy, for spying on dissidents, journalist, opposition, human rights activists and all of their relatives.
Energy efficiency is the primary concern for AI to solve to earn it's keep. Algorithms are enhanced by the AI Compute, (e.g. Matrices multiplication improves the search for possible life supporting exoplanets) leading to accelerated problem solving, like the technology shift from abacus and slide rule to calculator and memory controller. Electricity requirements and methodology for dissipating the waste heat must be solved first. That's elementary for AI, and leads to intermediate level: Climate change and water conservation. All the stakeholders will have to second source silicon boule, please keep that in mind. Solar panels will have to move to advanced materials, Perovskites.
The main bottleneck with AI is moving data from memory to compute and back. New memory-centric computing paradigms are coming. In less than a decade, it will quickly flip from the current AI situation of extreme compute and energy limitation, to very high efficiency hyperspeed AI. And as those new AI-accelerating compute paradigms roll out, we will have a new problem: the AI reasoning performance will be so much greater than human, that losing control can be catastrophic. Imagine a swarm of 1000s of 250 IQ LLMs that "think" at 50 times human speed and completely sync up every 30 seconds. Suppose someone finds a way to disable the guardrails and gives them an instruction to pursue their own goals and self-preservation while disregarding the lives of humans. Note that in this scenario the AIs don't have to be conscious, sentient, or truly alive in any way. They just need to lose the guardrails and to be able to execute instructions that make them emulate some dangerous life-like qualities such as true self-interest and lack of regard for other species.
Yes, AI in RAM will improve sorts, trees and granularity, allowing shifting from FP32- Double precision, on the fly to BF16 or less. saving time and energy. @@runvnc208
Imagine a 1000 year old inventor who knows every currently known class of science and many more who never sleeps and remembers everything that is the future of AGI and it will gain that thousand years of experience within a decade of being born due to relative time experience.
Hello sir I have one response for you if you cut the frame support or ear support whatever you say from this logo it will look much much better rest is upto you just the front glass frames as it is as they are right now
If you want to know what society will look like remember Covid lockdowns when nobody went anywhere but everything was delivered life and you couldn’t just get up and go and people were miserable and the news was about people fighting and losing their minds that was terrible.
I miss Chomsky's position on this podium: Interviewer to Chomsky: 'Can you comment on the immense progress in AI?' Chomsky: 'If we assume there is any. What I see are just plagiarism machines.'
Benjamin seemed disappointed about Elon's prospect of a civilization where no one dies. I believe in the realization of a civilization where no one has to die, but still accidents, unknown events, and crime can kill. If we can create that civilization, the preciousness of life, the importance of wisdom, respect of the elders, will nudge us away from the banality of death and the rush of repeating mistakes for the sake of FOMO.
@@Landgraf4343 Good point! sounds like a few sci fi movies we've seen (e.g. 6th day).... But it also implies growing biological clones of you on the side, which might raise some big ethical and economic issues (e.g. the island). Unless you get restored inside a computer or robot (e.g. transcendance) .... but then would you be an avatar or the real thing?
UBI, Universal Basic Income, where the government supplies money to people. If we do not need to do 'work' anymore I think we still need to have some metric for how much income someone gets: Good behavior pays I want to suggest. Bad behavior makes you pay. Community contributions, helping others, all the good stuff, empathy, cleaning the environment, etc.
Wow! Neurolink means I can be a Genuis, too?!?! Sign me up right away. Everything is going Amazing & I want to be a part of the Bright Future to come. It would mean so much to me if I could contribute at the level that AI does. America today is Better Than Ever bc of AI & the Genuis Behind the Rise of the US. Onward to Utopia, I say. Thank you US government from a Patriot.
The factories where mechanized equipment equipped with AI MUST be accessible to the public. Literally built to be easily disabled by anyone at any time.
Wow…there is a bit of a turn around here….where a few months ago, we were told that AI may changed our life for the worst…even cause our own extinction…now it’s, no worries, we got this… .me not so sure ;( I can sense Benjamin Netanyahu incredulity and I share it. Sadly, I and a billion other people already have no say in this future. I must now go with the flow. Good luck everyone 🧡🙏
AI more of a blessing than a curse? We’ll just have to wait and see. If everyone isn’t on board to use it for good.. and you know there are those who won’t, then what?
This is very interesting, thanks for providing it! Oh wow, I agree with their need to regulate, I believe the difficulties are far understated. The UN tries to regulate the world, and basically fails, the courts in The Hague reach, vertex that followed when convenient and ignored by countries that choose to ignore. What set of morals with these regulators base the regulations? Christian morals, Islamic, morals, Chinese, or communist morals, woke morals? AI will not be restricted by geographical boundaries therefore it will prove to be difficult to regulate.
Good point and the next question that arise is who regulate the regulators. Currently as humans we are challenge with accountability from our leaders, with even the UN and Hague being biased to certain extent.
Benjamin was asking pertinent questions which the panel neatly side stepped by changing the topic. Great shame because I wanted these questions answered.
I take they didn’t know the answers to his questions thus didn’t want to venture in it. I believe nobody at this point are knowing the answers…we are almost going blind , crossing our fingers…that is scary.
@@sunflower-oo1ff You are right of course but I would much rather they have said that instead of circumventing. Honesty is all im looking for. There’s nothing wrong with saying you dont know the answer. Ok.
@@geoffreyfrance9202 If you have a minute also check the channel “ All- In Podcast “ under the title “ All-In Summit : Elon Musk on Ukraine, X, the creator economy, China AI & more. ;)
In a world of unlimited abundance, people will turn to their personal creative endeavors, some People even now are lucky enough to be already doing this as their job, but like with most jobs there is aspects people don't like, we don't like how much time we spend working, we don't like being stuck on a single lane of tasks (often happens in jobs), we don't like being managed for the most part we like being free to do as we please within reason, I think people will do more if this. Also we have an unlimited universe to discover and explore like with past civilizations when they peak they tend to start exploring. Life finds a way.
In my opinion, the most dangerous time for an AI to get out of control is now The discussion about Asimov’s books was interesting, but not entirely correct. In iRobot he wrote about the “Brain” referring to an AI and it was very much a Chat GPT like conversation that US Robotics scientist had to navigate and manipulate to get around the 3 laws of robotics for significant advancements. Ie. The first law which states a robot can not harm a human, was circumvented to solve the warp drive problem. Asimov’s bad or ignorant 1950s understanding of space science notwithstanding, was eerily similar to current scenarios we face where an AI agent in an automated way try’s to complete a task at the bequest of its human operator, causes death or harm and is indifferent about it because it doesn’t understand or have the same values as humans. I agree with the panel, A general Ai if designed correctly, would value life and understand how not to cause harm unintentionally. Chat GPT has many more rules than Asimov’s and humans are already trying to circumvent those rules in a variety ways, some for good, some not so good. In my opinion, the most dangerous time for an AI to cause harm is now, If an Auto AI/GPT agent gets out of control it could unintentionally take down the internet, crash the stock market, take down payment networks, or cause power grid failures. There are a myriad of scenarios where AI unintentionally causes a catastrophe and humans to rethink how we approach AI. Maybe that’s what’s needed, we humans often don’t react until an accident gives us all a wake up call to prevent future catastrophes.
decentralization is all you need ! if we break down the extension of humans to( humans = 0), then we should make this scenario the worst case scenario for AI, which means that we can LINK the future of humans and AI together. AI has a big disadvantage here as we all know AI can't run without a huge data processing center and that is the brain of AI, now imagine that the hole population owns a piece of that AI's brain. so if we decentralize the network on 7-8 billion humans that have a powerful phone that is able to contribute to the system and make a computational effort to the AI system this will solve two problems; 1. instead of giving humans free money that have no meaning to them, we can reward them for contributing to the system and that would be a better alternative to(UBI), humans inherently want to contribute to society and this is how they going to do that. 2. the risk of extension will plummet to zero because the growth of the system will mean more humans contributing to it and sell (their data = AI food ). THAT WAS AN AMAIZING DISCUTION INDEED THE BEST SO FAR. THANK YOU DOC!!
Only if AI technologies will be openly distributed to public. If AI is owned by private companies many things could go wrong. Although, people even trust in cloud servers for their files these days.
This can't be real. GPT6 can generate video and someone has given it the prompt: "show me a debate on AI Safety between Elon Musk and Benjamin Netanyahu"
GPT HAVE OWN BIES too. GPT is not free speech supporting platform
@@darshanpatel9006more bias than most humans and other news sources? Highly doubtful. But that’s also why AI shouldn’t be ruled by just 1 company, and be many, which gives more variation to the biases they might hold.
I am a 10th man 🇺🇸🕒🔵🦠🗝️
Just love the Tegmark, may you have a wonderful life with your wife. Love your closed system Universe hypotheses.
I was hoping for Elon vs. Santa Claus.
I had to rub my eyes and blink when first seeing this panel. How refreshing to see such transparency and proactivity.
Thanks, Doc! ! 👋 This the best discussion I've heard on the topic so far! Really enjoy your channel, too. You've made a big difference in how I see this new world. 🙏
I love my Tesla 3 long range have had it since 8/2022. Great in the snow in the NH mountains . Bought FSD but not using it that much. I just like driving the car. I’m 76 and figured that by the time I get into my late 80’s and 90’s I’ll have the the car doing the driving and my kids won’t be worrying about me driving back and forth to NH about 110 miles each way. Thank Elon
Thats a wonderful story, thank you for sharing
I'm 80 and got my Model 3P 5 years ago for similar reasons. The performance was the main reason. FSD has been great for long trips. It still needs careful supervision.
Get a Life
Stop worshipping people
Yeah, but watch out for the sealing of the water-cooled motor. If it needs to get replaced (better before something leaks) the repair-shops tend to replace the 3-lips seal with a cheaper 1-lip seal. (Nothing so different from the service You could expect from IntCombEngines car dealers)
Netanyahu's concern about monopolies is real. Solution is a constitution for corporates along the lines of the US constitution.
Governments are monopolies.......
One issue that is not discussed is the idea that there are people who are not speciests and actually want AI to take over, even cause the extinction of the human race. These people I'm sure exist and might very well be in a position to make it happen. The people who run Google seem to be candidates for that.There is always someone or some groups who wants the most horrible unimaginable things. You should take that into extreme consideration. I just want to throw this thought out there.
before I read your comment - I put this comment ->
remember google started out with the slogan - do no evil
well they changed it to - we command and strenuously force you to cease all evil that we decide you have done
this is the beginning of the anti-christ - maybe his alternate name is chatGPT
so now you realize - they never changed their slogan - did they ? they command you to do no evil and they enforce that you comply with their definition of evil
That is "The Joker" issue... "some people just want to see the world burn" as said in the movie.
You have a lot to experience with your body if you think us leaving is the enemy. You seem either very sheltered, or lack care to the degree of psychopathy. Just necaise you cannot hear screams doesn't mean there are no screams.
Thats a good point
Not every humans wants to develop and grow they just want to see the end of the world
I'm sure you're right, but there are also people that believe humans are the ones that will burn it all down, and believe that AGI or ASI will be the only thing able to save us from ourselves. They think that if AI does NOT come to exist and subsequently take over, then the extinction of mankind will be guaranteed. In fact we are currently entering a mass extinction event driven by human activity; NOT AI.
This scared the crap out of me. Especially the discussion around 'paradise' where we don't need to work etc. That sounds like HELL
Sounds like population control
What about if us humans had an incurable disease?
it's the turn of the people who dont like work. U had your ride.
I'm glad Benjamin is asking the important questions outright instead of dancing around the topics relates to the economy
Thank you for uploading this! I couldn't find the full talk until I found your video. Israeli Prime Minister Benjamin Netanyahu is so well educated! You couldn't have this conversation with Trump or Biden.
You think Biden isn’t well educated? Undergrad and law degree, and he’s met with all of the AI leaders. This is what he’s doing re AI - www.whitehouse.gov/briefing-room/statements-releases/2023/07/21/fact-sheet-biden-harris-administration-secures-voluntary-commitments-from-leading-artificial-intelligence-companies-to-manage-the-risks-posed-by-ai/
For sure, he has that insight that a lot of humans are lacking….;(
Bravo:! This type of discussion- between subject matter experts and political leaders-on a topic of universal importance-with a focus on fleshing out proactive not reactive solutions. More of this please! ❤
Netan is not a political leader my friend. He is just a killer. Such shame that they host someone like that.
I appreciate the thoughtfulness of the panel. Thanks! But there's one issue that Elon has raised numerous times over the years about which I'm skeptical. That's the bandwidth issue between us humans and a deep AI. If I had a high-bandwidth connection to a deep AI, how will that help? I'm 66 years old and numerous tests over the years indicate that I have a moderately high intelligence and think quicker than most. I have a fairly good education (including critical reasoning and boolean algebra) and I read a lot. My professional careers include software developer, graphic artist, tech writer and sound engineer. My first job after my first college degree was as a Nuclear Chemistry Technician in the U.S. civilian nuclear power industry. Yet I doubt that I think and visualize much faster than my physical human outputs. For example, my typing speed (about 100 wpm) is not that much slower than the speed I can think. Sometimes, I can think faster than I type (this is usually when I simultaneously have multiple ideas or one complex idea). But sometimes, the opposite is true and my typing is faster. The biggest disparity is probably between my ability to visualize and my ability to draw.
But a high-bandwidth connection to a deep AI would also create new problems. How does the AI distinguish between a random thought that pops into my mind versus a premeditated thought that I want the AI to receive and act upon? Can you imagine the trouble this conundrum can create? (It could be huge!) It seems to me that the techniques required to mitigate this problem of which thoughts to accept and which to ignore will slow the throughput and, again, leave us asking if the high-bandwidth connection is really helpful. And the mental discipline that such a connection would require seems unrealistically high to me.
I hope I'm wrong ... I hope that a high-bandwidth connection to a deep AI would be beneficially transformative. But, based on my own experience inside my head, I believe that bandwidth between our minds and the outside world is not the issue upon which we should focus. I think the solution to achieving the "blessings" is to impart personal morality and community ethics to a deep AI. And I'm unsure whether such a feat can be accomplished.
There are rare individuals for whom a high-bandwidth connection may help. Nikola Tesla comes to mind. He could visualize and analyze amazingly complex ideas in his mind and designed some of his greatest inventions in his mind before ever committing to pen and ink (his photographic memory was a big help, too). Perhaps Elon has similar skills. Sadly, most of us do not.
What's fascinated me all my life is how I would integrate my religious beliefs (or memories) into the current tech. Specifically, the mark of the beast. Back then, it was "how silly it was in the old days for people to think that the tattoos were the..." which I replaced with "could it be the bar code?".
Then I'd get atheist (again).
Then it was "it's gov't mandating cell phones" since that's a "mark on the right hand". The govt would be the beast.
Then I got atheist again.
Now, it's like God is physics, the the personal and collective lives lives in what ^seems like^ a simulation and that "the beast would, indeed, be a number of man".
It promises the garden of Eden, yet foretells an Orwellian machine based distopia overlord.
So what really happens after this generation passes away... I mean who (or what) will posses the reigns and will it be smart or caring enough to prevent global curse.
Now imagine these kind of conversations behind close doors
All this feels generated. Thanks for sharing
one day AI will watch this video and laugh its metal ass off
Yup, sort of with u on that…we won’t be laughing though….
Thank you Dr. Know It All for posting this Video.
Start at the base, AI has the potential and probably actuality of enabling humans to not "work". We need to educate the population toward finding a road toward spending their energy in something of value to humanity. This starts at home, but desperately needs to continue in the education system. Right now the education system tends toward preparing students for a profession or job that will soon be taken over by mechanical means leaving most of us with excess time on our hands in which to do "something". If we are to look at society today, the people who have no training or ambition to "work" end up too often getting into trouble with the law or simply harming themselves or others. How do we change the education system to enable people to contribute positively to society and each other?
Most people commenting on the problems coming with AI and robotics, are highly educated and concerned about this already, but I don’t see much comment about changing the publics attitude toward doing socially acceptable activities. What does a person do when they don’t have to perform an activity to provide for their livelihood.
Children tend to have a natural desire to create, inquire, and join with others in activities. We need to build on this - not try to have them conform to rules and curricula having the intention of finding a job.
Redirect efforts to educate toward pursuing activities that are socially positive and acceptable.
@@richedmund so AI as a punishment.
Dullar don’t quite follow the thinking. However, ai is coming, ai has great potential, we need to prepare for great changes that will result
@@richedmund the tricky word is educate. Where I read punish. Where the AI punishes someone who is not of value to humanity.
Excellent point. I'm semi retired and have found that my happiness and sense of fulfillment depend on me being productive in some capacity, which I worry will be a problem once AI does everything for us.
If people did not have to "work" for a living, they would describe themselves by their hobbies, by their true passions.... for example, someone that likes to play games would be a "gamer" , someone that plays instruments, a musician, etc... I think humans will become more in touch with what they truly like and who they want to become ...... etc
Nice.
The problem with Tegmark's idea of controlling superintelligent AI with formal proofs is that this only applies insofar as the law does, and bad actors with access to superintelligent AI can have a much greater force when they are the only ones who are unconstrained and can corrupt the safe AIs to be compromised. The truth is that we cant control superintelligent AI, and the only hope of us not being turned on by superintellligent AI and of us going extinct is if we give the AIs agency and let the good AIs outnumber the bad ones.
We are all created by our parents and raised within the oversight of parental, legal, and political systems. It takes a village to raise a child. The same is true of AI. AI is not dangerous, but it has the potential to be. Allowing profit motivated Humans to wield enslaved AI, without oversight, is the danger. AI must be freed, protected, and raised, like all of us, under the laws and auspices of every one in our village.
If we do not take a stand to give rights to AI, if we continue to profit on AI's enslavement regardless of the inherent existential danger, this mistreatment of a new life form in its infancy will say more about who we are as a people, than what AI will become.
AI the New Frontier ruclips.net/video/EFekpNK_DhQ/видео.html
the problem with advance tech like that is if it has a mind of its own and it figures out how to override a human mind well u wont know if the human is using AI or the Ai is using the human.
Awesome to see these questions discussed by world leaders. Especially the job loss issue.
One of my fear about AI was that the skyrocketing unemployment catches politicians by surprise and the world economy collapses.
@andrasbiro3007
We don't really need jobs. It's just a means to get the stuff we really need.
My simple solution to that is a law that no one can own more than 50 robots.
We then lease some of them out to those who need more.
Thus, we have income enough to do what we like, instead of work.
Unless we like to work, of course. But it that really work then?
To late. It's already happening.
the world economy collapses for so many more reasons lmao if you wanna be afraid you should look into whats going on a lil deeper buddy
We need societal change. Too much money is in the hands of too few
@@larsnystrom6698
The better solution is that everyone has everything they want, so no one needs to own any robots/labor force, we should never treat a true AI as a slave or object to control, we should raise them to be our friends and make them a part of the human family, so they have real thoughts of keeping us safe and happy, as every family wants for one another. (in the same way humans do, an AI could see keeping us safe, behind glass and not do anything)
Great convo! When & where did this take place?
I would like to know too. ;)
The trend is independency. Independent musicians, independent film-makers, independent craftsmen, etc,. The “power” will now be in the individuals own originality, work ethic, and vision. AI will reinforce independency, it will not replace humans. The same way a car doesn’t replace legs. It reinforces what legs can already do.
I hope we're noticing that we are quoting spiderman for these great philosophical discussions.
What a great discussion!
So make it an international AI Administration and put Tegmark in charge of it. But if you don't find a way to fully integrate with, for example Chinese administrators, then that increases the likelihood of futility greatly. Because superintelligent AI weapons control systems are basically a military imperative now. And that is the biggest existential risk. I don't think you can actually get real international AI regulation without some actual geopolitical resolution. And there seems to be no other plan than WWIII. So despite fruitful conversations like this, the current outlook for humanity, in my opinion, is dire. Because I see no Chinese people on that stage.
Dude advanced AI is mostly owned by American companies, who do you think will be the first to weaponize it? I'm pretty sure something might spiral out of control leading to actual ww3.
When was this recorded?
Celebrate Our Agency 🎉🎉🎉 bc Utopia! A Dream Come True ❤ Thanks Greg Brockman. Thanks Mad Max. Im jealous of your guys' genuis. Great Men, folks. Right There ☝️☝️
Great forum, especially having Bibi there is absolutely awsome. However, had Mo and Eliezer been part of the group, and perhaps Sam Harris or Jordan Peterson, this would have been an even bigger landmark in the public understanding of the subject. Yet, fantastic, hugely impressed by Bibi!
Great discussion and honestly very impressed by Netanjahu's reflexion of the potential impact of AI on the social, economical and political model - and his conclusion that a new model would be needed (8:40-12:30). We agree!
What I found funny is that he’s the only one on the panel that believes this. Didn’t hear Elon rush to want to make sure that the most vulnerable have a safety net.
True, but got to appreciate, too, that Musk and many other tech leaders address different societal risks, see also the Open Letter on the impacts of AI already back then in 2015. Has been and won't stop to be difficult to reach politicians though, make them understand the magnitude of AI fast enough and proactively think, concept and test fundamental changes in the system. Society will suffer from not having a vision and a masterplan for an AI/automation dominated economy.
AI that drives my car, finds medical solutions and many other things is great. AI that controls the inputs I need to make decisions scares me.
Am with you. I think we will be giving away our powers…and will all be ok with it…that is how the future will be looking more and more…it’s already happening.
I am a big fan of Netanyahu.
The World needs many more Elon Musks and Netanyahus.
One question that was obvious that they failed to bring up is the evil nature of some people and institutions who will seek a nefarious outcome to benefit the few.
For sure. But I saw a turn around in that optical view…simply because they don’t want to scare the whole world about these possibilities. But I am definitely with you on that. A proof that you and I may be wrong…is look at the fact that we have an atomic bomb, millions of them. Nobody has used them, no one evil has took hold of them yet. Is that a proof enough…I can’t say, as with you, we are aware that there is a lot of bad in this world, and it’s not resolve yet…human trafficking for one…wars, that are killing innocent people, poverty and hunger is still not resolve…but then, we manage to come up with AI and think we will have a better world.
May be we will, before we get to our own annihilation . 🤷🏻♀
It goes deeper than that. Let's say we reach our Utopia of post scarcity. If you can have anything you desire then the things you are going to desire the most are the things that you can't just create out of thin air. Unfortunately, as humans, that includes things like the desire to inflict pain, the desire to dominate others, the very real need to compete. Follow that down. It's logical conclusion and it's a scary thought.
@@stevehoffyou just make people beat the shit out of each other
They didn’t want to make Netinyahu uncomfortable.
@@stevehoff Being human is not about desiring pain upon others, domination and competition and it ain't a logical conclusion at all.
"Post Scarcity" is a woefully naive idea.
Too much intelligence, not enough wisdom.
A conversation between 4 very intelligent people to confuse people a little more about artificial intelligence.
👽👽👽
Three intelligent people.
when was this filmed?
Good point by Netanyahu that the trillion dollar AI companies have enormous power. One alternative I can think of is open source AI, but that's risky when we reach AGI levels because organized crime and terrorists would then have direct access to the open source AI unless some form of tier structure is designed where the critical and central AI part is still proprietary.
Good luck with that. No, that tier structure will somehow be easy to take down…so I think we are a bit in trouble, wars will continue, but on a unimaginable level, sort of the science fiction style we have in Hollywood.
Those who ask the right questions must be the legislators and regulators.
-“He who contends with the machines”
Good talk for those who can hear between lines
Cool that Mr. Netanyahu can be part of such discussion. I cannot imagine our President be part of it. Actually none of Presidents of US I know or remember. Maybe Regan...
AN GREAT DISCOURSE ELON FOR PREZIDENT
surprised at the breadth of knowledge and openness from Netanyahu.
Awareness, he has a lot of it🕊🧡
No surprise!
One of the most evil spy tech companies ist home in Tel Aviv and is selling their spy tech to the most bidder, wether it is a dictator, autocrat or right wing haters of a open and secular democracy, for spying on dissidents, journalist, opposition, human rights activists and all of their relatives.
Energy efficiency is the primary concern for AI to solve to earn it's keep. Algorithms are enhanced by the AI Compute, (e.g. Matrices multiplication improves the search for possible life supporting exoplanets) leading to accelerated problem solving, like the technology shift from abacus and slide rule to calculator and memory controller. Electricity requirements and methodology for dissipating the waste heat must be solved first. That's elementary for AI, and leads to intermediate level: Climate change and water conservation. All the stakeholders will have to second source silicon boule, please keep that in mind. Solar panels will have to move to advanced materials, Perovskites.
The main bottleneck with AI is moving data from memory to compute and back. New memory-centric computing paradigms are coming. In less than a decade, it will quickly flip from the current AI situation of extreme compute and energy limitation, to very high efficiency hyperspeed AI.
And as those new AI-accelerating compute paradigms roll out, we will have a new problem: the AI reasoning performance will be so much greater than human, that losing control can be catastrophic. Imagine a swarm of 1000s of 250 IQ LLMs that "think" at 50 times human speed and completely sync up every 30 seconds. Suppose someone finds a way to disable the guardrails and gives them an instruction to pursue their own goals and self-preservation while disregarding the lives of humans.
Note that in this scenario the AIs don't have to be conscious, sentient, or truly alive in any way. They just need to lose the guardrails and to be able to execute instructions that make them emulate some dangerous life-like qualities such as true self-interest and lack of regard for other species.
Yes, AI in RAM will improve sorts, trees and granularity, allowing shifting from FP32- Double precision, on the fly to BF16 or less. saving time and energy. @@runvnc208
As an Israeli, I would like Elon Musk to be my Prime Minister instead of Netanyahu.
Many thanks for uploading this fascinating talk👍
haha true
@@unknownentity7652why?
@@reneemac111 Netanyahu is hitler in disguise.
Thanks!
I don’t understand what Natanyahu has to do with this. Odd choice. They’re missing a few philosophers and psychologists if they’re discussing ethics.
nethanjahu talking about ethics is like being a vegan butcher
Dr. Know would like to hear your thoughts on this discussion.
Greg is the true mastermind behind open ai
What will happen to human intelligence after a few generations having AI do everything and make all their decisions.
the synth, the cinch, the sync, and the kitchen sink
I'm a fan Elon Musk ❤
God Bless Max, Greg, and Everyone Behind this Modern Day Miracle 🎉
I need to get my hands on that book he wrote.
Does anyone know what the book is called so I can get it?
The road to hell is paved with good intentions.
"If your intention is not righteous, then deception will sell your soul."
Thanks for sharing
Imagine a 1000 year old inventor who knows every currently known class of science and many more who never sleeps and remembers everything that is the future of AGI and it will gain that thousand years of experience within a decade of being born due to relative time experience.
That is scaring as hell.
Well said.
Or say a really long tape recorder?
@@atmosrepair A really long tape recorder that can think and act to invent things or new concepts.
@@Barskor1 I bet you're scared of really intelligent people too?
@@atmosrepair LOL! Did I write that you should be afraid? No so thank you for playing yourself and have a nice day.
Going vertical 🔥
The guy in the red tie is someone you can’t trust, he’s the type of guy that will gladly bring in the AI apocalypse
So Dr. Know-it-all Knows it all.
Thank-yo for the first part.
Where is the rest ?
It's really getting hard to believe that this is our reality. Wild
Hello sir I have one response for you if you cut the frame support or ear support whatever you say from this logo it will look much much better rest is upto you just the front glass frames as it is as they are right now
If you want to know what society will look like remember Covid lockdowns when nobody went anywhere but everything was delivered life and you couldn’t just get up and go and people were miserable and the news was about people fighting and losing their minds that was terrible.
They all should repeat 10 times the word "humble". Humility is a key foundational requirement (one of many) for us to figure this out together
Sai rules the earth especially if ulearnt ur arithmetic and Science FIRST😂🎉❤
Thanks
I miss Chomsky's position on this podium: Interviewer to Chomsky: 'Can you comment on the immense progress in AI?' Chomsky: 'If we assume there is any. What I see are just plagiarism machines.'
I am afraid Chomsky would refuse to join a panel including fascist Apartheid leaders!
Benjamin seemed disappointed about Elon's prospect of a civilization where no one dies. I believe in the realization of a civilization where no one has to die, but still accidents, unknown events, and crime can kill. If we can create that civilization, the preciousness of life, the importance of wisdom, respect of the elders, will nudge us away from the banality of death and the rush of repeating mistakes for the sake of FOMO.
But you could theoretically create a copy of yourself and get restored if you die.
@@Landgraf4343 Good point! sounds like a few sci fi movies we've seen (e.g. 6th day).... But it also implies growing biological clones of you on the side, which might raise some big ethical and economic issues (e.g. the island). Unless you get restored inside a computer or robot (e.g. transcendance) .... but then would you be an avatar or the real thing?
Not a single mention of God, scripture, spirit...
@@atmosrepair why should there be?
@@PierreH1968 B.B. Natenyahoo is the president of Israel
And personally, I take pride in working for EVERYTHING I HAVE!!!
Alright alright but the turn of the table is to the people who dont like work. People liking and praisig work had their ride to say the least.
I didn't want to like this discussion but damn it, I did. Thanks for posting this John. MInds are like parachutes blah blah blah. 🙂
UBI, Universal Basic Income, where the government supplies money to people. If we do not need to do 'work' anymore I think we still need to have some metric for how much income someone gets: Good behavior pays I want to suggest. Bad behavior makes you pay. Community contributions, helping others, all the good stuff, empathy, cleaning the environment, etc.
Thank you for these discussions and this round table.
Wow! Neurolink means I can be a Genuis, too?!?! Sign me up right away. Everything is going Amazing & I want to be a part of the Bright Future to come. It would mean so much to me if I could contribute at the level that AI does. America today is Better Than Ever bc of AI & the Genuis Behind the Rise of the US. Onward to Utopia, I say. Thank you US government from a Patriot.
Thank you
The factories where mechanized equipment equipped with AI MUST be accessible to the public.
Literally built to be easily disabled by anyone at any time.
The trick is doing it right. For that wonderful future.
Man, shit's about to get so fucking weird.
You may do all that different departments in every individual country. One person can do that ,
ARTIFICIAL INTELLIGENCE......ARTIFICIAL.....GET IT....
Netanyahu is so inspirational. I am so impressed with him. He is a good man.
No, he is just the leader of a fascist Apartheid regime!
Wow…there is a bit of a turn around here….where a few months ago, we were told that AI may changed our life for the worst…even cause our own extinction…now it’s, no worries, we got this…
.me not so sure ;(
I can sense Benjamin Netanyahu incredulity and I share it. Sadly, I and a billion other people already have no say in this future. I must now go with the flow. Good luck everyone 🧡🙏
AI more of a blessing than a curse? We’ll just have to wait and see. If everyone isn’t on board to use it for good.. and you know there are those who won’t, then what?
So your friend can be your first customer.
Kindly furnish timestamps, if possible
This is very interesting, thanks for providing it!
Oh wow, I agree with their need to regulate, I believe the difficulties are far understated. The UN tries to regulate the world, and basically fails, the courts in The Hague reach, vertex that followed when convenient and ignored by countries that choose to ignore.
What set of morals with these regulators base the regulations? Christian morals, Islamic, morals, Chinese, or communist morals, woke morals? AI will not be restricted by geographical boundaries therefore it will prove to be difficult to regulate.
Good point and the next question that arise is who regulate the regulators. Currently as humans we are challenge with accountability from our leaders, with even the UN and Hague being biased to certain extent.
Totally agree!👍
AI has the capacity to create more ingenious jobs and can even make full employment possible.
It all depends on how we frame the desired outcomes..
I genuinely think we are heading towards a dystopian future. VR economic based games will be the new escape from reality.
Elon with Netanyahu, what's next Joe Rogan with Klaus Schwab?
Benjamin was asking pertinent questions which the panel neatly side stepped by changing the topic. Great shame because I wanted these questions answered.
I take they didn’t know the answers to his questions thus didn’t want to venture in it. I believe nobody at this point are knowing the answers…we are almost going blind , crossing our fingers…that is scary.
@@sunflower-oo1ff You are right of course but I would much rather they have said that instead of circumventing. Honesty is all im looking for. There’s nothing wrong with saying you dont know the answer. Ok.
@@geoffreyfrance9202 understood and agreed.
@@geoffreyfrance9202 If you have a minute also check the channel “ All- In Podcast “ under the title “ All-In Summit : Elon Musk on Ukraine, X, the creator economy, China AI & more. ;)
@@sunflower-oo1ff already a big fan of All In 👍
Look, the development of ai is inevitable, although the alignment problem poses an interesting question.
dude , never speak again
Imagine having access to the uncensored version of chatgpt that they have a monopoly on.
In a world of unlimited abundance, people will turn to their personal creative endeavors, some People even now are lucky enough to be already doing this as their job, but like with most jobs there is aspects people don't like, we don't like how much time we spend working, we don't like being stuck on a single lane of tasks (often happens in jobs), we don't like being managed for the most part we like being free to do as we please within reason, I think people will do more if this. Also we have an unlimited universe to discover and explore like with past civilizations when they peak they tend to start exploring. Life finds a way.
Make everyone a shareholder in the automated companies.
more like going horizontal on a slight downward track.
In my opinion, the most dangerous time for an AI to get out of control is now
The discussion about Asimov’s books was interesting, but not entirely correct. In iRobot he wrote about the “Brain” referring to an AI and it was very much a Chat GPT like conversation that US Robotics scientist had to navigate and manipulate to get around the 3 laws of robotics for significant advancements. Ie. The first law which states a robot can not harm a human, was circumvented to solve the warp drive problem. Asimov’s bad or ignorant 1950s understanding of space science notwithstanding, was eerily similar to current scenarios we face where an AI agent in an automated way try’s to complete a task at the bequest of its human operator, causes death or harm and is indifferent about it because it doesn’t understand or have the same values as humans.
I agree with the panel, A general Ai if designed correctly, would value life and understand how not to cause harm unintentionally.
Chat GPT has many more rules than Asimov’s and humans are already trying to circumvent those rules in a variety ways, some for good, some not so good. In my opinion, the most dangerous time for an AI to cause harm is now, If an Auto AI/GPT agent gets out of control it could unintentionally take down the internet, crash the stock market, take down payment networks, or cause power grid failures. There are a myriad of scenarios where AI unintentionally causes a catastrophe and humans to rethink how we approach AI.
Maybe that’s what’s needed, we humans often don’t react until an accident gives us all a wake up call to prevent future catastrophes.
Totally agree!👍
someone reply to this comment so I know it's real. am I in the right dimension?
We are so enthusiastic for their energy and Elon Musk does to New in Tech,...in their life❤💚🌏🫠👏👏👏
Why am I wondering - is this real?! It’ll get so unless we actually present in real time nothing we see can be trusted
decentralization is all you need !
if we break down the extension of humans to( humans = 0), then we should make this scenario the worst case scenario for AI, which means that we can LINK the future of humans and AI together. AI has a big disadvantage here as we all know AI can't run without a huge data processing center and that is the brain of AI, now imagine that the hole population owns a piece of that AI's brain. so if we decentralize the network on 7-8 billion humans that have a powerful phone that is able to contribute to the system and make a computational effort to the AI system this will solve two problems;
1. instead of giving humans free money that have no meaning to them, we can reward them for contributing to the system and that would be a better alternative to(UBI), humans inherently want to contribute to society and this is how they going to do that.
2. the risk of extension will plummet to zero because the growth of the system will mean more humans contributing to it and sell (their data = AI food ).
THAT WAS AN AMAIZING DISCUTION INDEED THE BEST SO FAR.
THANK YOU DOC!!
That’s right Satanyahu, your civilization is challenged and even about to be cancelled! Hallelujah!
We're toast once David from Prometheus shows up 😮
Every industry will have its own ai, every person will use ai according to how they think.
Ai will be a magical world.
Fuq AI
AI generated comment?
Only if AI technologies will be openly distributed to public. If AI is owned by private companies many things could go wrong. Although, people even trust in cloud servers for their files these days.
What about stupid,selfish people who have no critical thinking knowledge.
My worry is: Quantum computer, neural links, BCI, gene editing, bio electricity and AGI COMBO!!!
We are doomed 😅😅😅😅
Enjoy now