Here are the timestamps. Please check out our sponsors to support this podcast. 0:00 - Introduction & sponsor mentions: - LMNT: drinkLMNT.com/lex to get free sample pack - Notion: notion.com/lex - InsideTracker: insidetracker.com/lex to get 20% off - AG1: drinkag1.com/lex to get 1 month supply of fish oil 2:23 - Beff Jezos 12:21 - Thermodynamics 18:36 - Doxxing 28:30 - Anonymous bots 35:58 - Power 38:29 - AI dangers 42:01 - Building AGI 50:14 - Merging with AI 57:56 - p(doom) 1:13:23 - Quantum machine learning 1:26:41 - Quantum computer 1:35:15 - Aliens 1:40:04 - Quantum gravity 1:45:25 - Kardashev scale 1:47:17 - Effective accelerationism (e/acc) 1:57:47 - Humor and memes 2:00:53 - Jeff Bezos 2:07:25 - Elon Musk 2:13:55 - Extropic 2:22:31 - Singularity and AGI 2:26:29 - AI doomers 2:27:54 - Effective altruism 2:34:23 - Day in the life 2:40:50 - Identity 2:43:40 - Advice for young people 2:45:42 - Mortality 2:49:25 - Meaning of life
Guillaume really has a nonsensical moral framework. Yes thermodynamics is real... And we want to increase entropy why exactly? Death is the future for all of us... and we want to hasten it why exactly? Increasing the power and likelihood of a technology that has every potential and incentive to wipe us out should be done carefully - at the VERY least.
Lex, I have been listening to you since your MIT lectures and I have never commented. I was about to make an appreciative comment with your Netanyahu/Palenstinian Author & poet back-to-back podcasts. But this pushed me past the line. I don’t know if you’ll ever read this, but regardless it’s my duty as a listener to let you know that I respect you, and I support your endeavor of giving EVERYONE a voice. No matter what that voice says, they have a right and a ability to say whatever they would like to say. If I ever meet you I will give you my respects, but for now I wanna say thank you for putting this content out, no matter the consequence. This is the beauty of the internet. ❤️
Lex never gives voice to constitutionalists or hard conservatives or america-first intellects. But he regularly has Marxists and eugenicists and globslists on and treats them with great reverence... And now an accelerationist. His purpose is to shift the center far to the left under the guise of thoughtful conversation, while never or only very rarely ever having actual opposition opinions on his show that support western society or culture in earnest in its intended form. He is a propagandist. And a very good one. And he literally designed the algorithms used to boost him in your feed every day.
@@bradojacko8247 while I can’t guarantee it, I’m sure Lex would have ‘hard conservatives’ and ‘America first intellects’ on his podcast if given the chance. I personally don’t have someone in mind who fills that role, but I know looking at his current podcasts that he has a very wide array of opinions and representation on the political spectrum. He will never be able to cater to what everyone wants, but having a Palestinian poet immediately after having Netanyahu on is a testament to his effort to give extreme sides of a spectrum a voice. If your opinion is that he is extremely left and Marxist, then perhaps that’s your algorithm being different than mine lol. I don’t think a lot of his guests aligns with that opinion (Jared Kushner is pretty conservative, and Elon Musk is pretty America-first for example). I do think it’s unfair to pin him to an ideology, because I bet that most listeners would struggle to guess his political views and opinions.
@@ranieldamirez Marxism is responsible for 100 to 200 million deaths and persecutions. Yet, Lex has treated every socialist and communists guest, for instance, with great calm and reverence and had many of them on. Yet, he has never had on any constitutionalist from the far american right or even the middle constitutional right. Only moderates, neocons and left of center moderates like Musk appear. Musk has never been a republican or a conservative. However, because the Overton window has been shifted so far left by people like lex over the past 10 years, Musk appears to now be conservative to you and millions of others. That's the trick. That's the scam. Lying by omission. The only far right guy he has ever had on is his anarchist friend, Michael Malice. He is a great and interesting guy, but he is not pro America. He doesn't actually believe in the US constitution or that it works. He is an anarchist. And Ben Shapiro, who was on once, is a globalist neocon who doesn't put America first. Can't think of any others...
I don't even think the market is where risk lies. Military applications are where you'll see caution thrown to the wind. I like Guillaume but I wish he'd be more explicit about what he thinks AI will be capable of. I get the sense people are talking past each other when discussing AI risk. A great example is Lex's question about whether the AI itself would centralize power and Guillaume's response makes me wonder if he even believes ASI is possible because there's no knowing the outer limits of an ASI. Now give such a technology to a military that works without oversight. I agree that p(doom) is stupid but mostly because I think even a doom chance of 1% is effectively a 100% chance of doom. Would you play russian roulette with a gun if it had 1,000 chambers?
Everyone who doesn't believe E/acc is personally killing everyone with cancer over the next 30 years. There are people walking around today whose future will be determined by our beliefs in AI.
Very complicated guest to be honest but Lex handled this superbly. Props to keep asking about the matter of AI safety and being stern enough to pushing a firm believer of a certain highly optimistic theory out of his comfort zone and atleast get him to think and discuss potential negative effects of AI.
Ya the guest was just kinda making it up as he went imho, this whole theory of negative perceptions bringing about negative consequences is the antithesis of rational, negative thinking is how all our lives we avoid negative ramifications.
Agreed w both of these comments. Definitely sounds like he’s making this up. So good you had him on, Lex, so that we could hear in his own words in a fair and open forum instead of an ego-driven discussion with other white men who are thinking along the same lines secretly wearing their super hero outfits in their parent’s basements planning the demise of humanity between 9p and 4a bc they aren’t getting their way. Wah. Men… stop! Geezus. Haven’t you done enough to harm us since like the beginning of time? The good men out there are not thinking like this and thank goodness there are a few! Please interview Jackson Katz. We need help before it’s too late and clearly we are almost done. We need a change of thought before it’s too late.
Man, it always blows my mind when really smart people fall into the trap we all fall into of being so excited about the potential of something that they say things like. The market will regulate the industry.
To preemptively prevent innovation while totally failing to regulate the current state seems not very sincere to me. This is just protectionism of established structures ... when was the last time an innovation was truly dangerous and the the resulting deaths could not be completely blamed on failing to regulate the *application* of the technology - and the actual *invention* of it was the cause of deaths? In other words don't regulate AI invention etc. but regulate how the gov can use it to control its people and create "rewarded groups" etc.
I'd say knowing sufficient vocab to sound educated, or even being 'educated', a chimp with retention for a couple weeks passing tests, doesn't make one 'really smart'. Common sense is required for that qualification for me. Even people like Sam Harris, and Tyson, I personally don't find all that bright.
@@mschwaller3371 It's been on repeat for eons, as Aldous Huxley said in 1962 at berkley, paraphrased ' Surely many before the industrial revolution could foresee the hardships it would impart and were ignored, leading to 60 years of treacherous labor and a toxic aftermath'. To quote verbatim would be a waste of my time but feel free to look up the 'ultimate revolution talk' about 36 minutes in onwards, it's on youtube.
@@dufung3980 chatGPT is a good example. If the topic is not obviously heavily curated it will just pickup any propaganda and defend it with all its GPUs ... sounding super smart and pc af ^^
@@GoogleIsNotYourFriend I agree with you - i'm just saying this was never the case. How many people got killed by cars until politicians felt the need to regulate it? Should they have opted to destroy the steam tech to avoid cars down the line? Slowing down progress and innovation to avoid the need of passing a new law just doesn't add up to me. LLMs are note a problem if manipulating people would be a crime ... somehow I don't know of anyone being in prison for poisoning the minds of others ... that's going to be a social/political challenge we pushed out for long enough ...
You should invite Jonathan Oppenheim! Professor at UCL who casually merged quantum mechanics and general relativity, his papers just got published in Physics and Nature
Would be fascinating to hear Lex and Guillaume delve into this statement more: "And so it’s the missing piece between the quantum and the cosmos. It’s the middle part. Thermodynamics rules the meso scales." How and why does thermodynamics actually connect the two?
(Statistical) thermodynamics arises as a consequence of the motion of countless small particles (governed by quantum mechanics). If you look at the single particle (i.e., the quantum scale), there's no point in defining thermodynamical concepts for it. However, these concepts are very important on larger scales; just as quantum mechanics gives rise to thermodynamics, thermodynamics gives rise to many of the rules of physics relevant at our scales
Beff Jesoz is more like a natural philosopher than he is a physicist. He uses physics to philosophize about life, existence, and the ultimate nature of reality. Fascinating interview.
@@TheCyberBully420 It’s a disappointment to me that we no longer refer to “Scientists” as Natural Philosophers. It’s a far better description… and has a ring to it.
It would be interesting to see Lex do an interview with Nick Land, who largely developed Accelerationism, as well as being a controversial figure due to his involvement in the NRx movement.
In talking, Guillaume uses Land's creations, "hyperstition", "techno capital", etc., but tacitly avoids mentioning him, maybe to not get associated. When it comes to AI, Land is more radical and out there, and he doesn't cook up antagonisms between humans' nations like Guillaume does. Afaik Land greets the developments of those subcultures, tho. Sadly the dozen of foo/acc communities that exist for a decade now aren't mentioned in this interview either.
@@mikebarnacle1469I think you're probably right.....reminds me a little of Peter Zeihan... Then again , just being full of himself doesn't mean he's wrong ....
It always a breath of fresh air when we see that a popular figure with influence is also brilliant and well educated. It’s a comfort to know that there are people like that.
Just wow; You think maybe sometimes you’ve seen some of the most interesting humans via some form of media, and that you’ve reached a pinnacle of ehhh it’s more of the same, and then you listen to this guy in this interview, who id never heard of, and holllllly fark. Additionally, Lex easily the most interesting podcast/interviews on earth - much love dude
I agree, this guy had so much thought provoking, profound, and mostly REFRESHING takes on stuff that, like you say has been more of the same lately... A complete step function up for me as well
Focusing on human side of alignment will help bring clarity and confidence to a lot of anxiety-provoking ideas floating in this new field. While LLMs evolve and grow, humans watch passively in horror or fascination, instead of actively applying same principles to foster evolvement of their mind to match the upcoming machines.
This one was the hardest podcast to follow so far, but @lexfridman keep going because I thought it was interesting and worth it, even though I didn't understand everything you discussed yet
Who said he didn't realize it? It was never implied that he didn't understand it, in fact stating that idea almost intrinsically implies that he does understand it.
A benevolent AI could uplift us as a species. Everyone is quick to jump to the most terrible and dystopian conclusion possible. Was domesticating dogs bad for them? They arguably live much better lives in the care of someone that loves them and treats them like family as many dog people do. It's incredibly arrogant and speaks volumes of insecurity to believe that we are or always will be Supreme beings at the top.
The scale of unintended consequences from merging Quantum Computing and AI is mind boggling, this is one of those interviews I will have to listen to a couple times to understand the potential. Amazing
The one thing I can't get past is what inevitably happens when one being is orders of magnitude smarter than another. Our intelligence vs AI intelligence isn't the issue. It's the SPEED at which we can learn that is the difference. If you put the most advanced AI into the most powerful conventional computer and then put that same AI into a quantum computer, they will have the same intelligence on day one, but Quantum computers are supposed to be able to process functions a TRILLION times faster than a conventional computer, so within minutes, not years, or months or weeks or days or even hours, the Quantum powered AI will be many orders of magnitude more intelligent than the conventional computer driven AI. Elon Musk is famously quoted for giving the "ants vs humans who are building a road" example. Once quantum computing and AI are joined, we'll be the ants in Elon's scenario almost overnight.
current AI models do not conceptualize - they perform operations on data sets according to recursive data sets. it's not even a competition. Like comparing a stationary construction crane to a bicycle. You use the bike to get to the work site. You use the crane to build a factory to make better bicycles. WE are not in competition - we are in the illusion of co creation. Verdon is not qualified to have this discussion. @@scottrussell2281
@@scottrussell2281 quantum computing, and AI will join, and then they will be jammed into our brains. We won’t be Ants. We will merge with it. It’s the only way if we don’t do that then there’s no point in creating AI because the singularity will happen
Cancers also want to grow. Much of the negativity, fear, and depression that exist are due to the cancers of FB and others that monetize misinformation, fear, and anxiety. Authoritarian centralization is already occurring and needs to be countered,
indeed... and this guy for all his wonder at consciousness will be funded to e/acc most of humanity into dystopian biosecurity technocratic neo-feudalism.. (sorry... I'm such a doomer..)
55:00 Good example of intellectual overreach: Wolves were domesticated by stunting their development. Domestic dogs do not attempt to dominate humans because we bred them so that they do not reach full psychological development. And yet this argument is used as a justification for why we should have unrestricted development of AI. If there is a lesson here, it is not the one Gill is touting.
These videos are accelerating the positive future. Also understanding we will always deal with negativity. It’s all in your perspective to see the difference and live in the positive. Thanks Lex!
They operate under the assumption of human independence and agency. They are totally wrong about that. So they are effectively arguing nonsense, but they can't stop themselves from operating this way.
@@ROALD.The use of quantum computers to identify better minima has been an hypothesis for decades (I had that in school in my courses at the end of the 90s. I would add also that his vision of capitalism as an ideal system to optimize naturally tensions has clearly been demonstrated this last decade to be completly wrong. Glad European model comes to shake a bit the giants.
@@AlexDubois It's not been demonstrated to be wrong at all. If anything, the harm that government overreach and interference in economic systems can cause has been demonstrated far more clearly, as the private sector has become even more productive and efficient.
Guillaume may be academically clever but he is also very (willingly) naive when it comes to AGI (which he takes a step even further to Quantum Intelligence). I do think that naivety is a willing one of conscious self-blindness in the quest for technological satisfaction. It is a dangerous game and I'm with Connor Leahy on this one. I'm beginning to think about Terminator 3 and Sarah Connor's fury at the man who allowed Skynet to develop.
It's an amazing interview. Why Guillaume raises a lot of good points, I don't really see myself agreeing with him on the market, providing the system with results with the biggest utility. I think we have a lot of examples from the startup world to see that actually, market provides products with the biggest MONETARY utility (which can mean it's useful to just some people or 1 person). Example being Airbnb and how it ruined cities that are unable to properly get legislation in time. (For example in my city, there's buldings where you live and 80% of apartments are tourist apartments, with drunk people barfing on the stairs when you just want to get some sleep). I suppose G. would say that it's just matter of legislation, but what if state is actually corrupted by the money "owners" generate? You can buy off goverments. That's where the market breaks down.
It's an optimization, but it is not a Pareto optimization. Either the overall utility is still positive, there is a market distortion that effects its freedom, or there are market distortions that will not last forever. As someone that often seeks temporary accommodation and transportation away from my home, my life is so much better with the extra supply of rooms and cars, it was hell depending on traditional scammy taxis and hotels that overvalued themselves too much, all very legal.
the market will not pick what is best for us or esle facebook would died not become the giant it is now. and I am not a bot. proof: I ate a buger today. lol@@akuno_
"The Banking Act of 1933, or Glass-Steagall Act, initially put strict barriers between commercial banking (handling deposits, offering loans) and investment banking (engaging in risky financial activities like stock trading). This separation was designed to reduce the risk of financial crises. However, over time, these regulations were relaxed. By the 1990s and early 2000s, the repeal of key elements of Glass-Steagall allowed banks to engage in both types of banking again. This led to increased risk-taking and was a significant factor in the buildup to the 2008 financial crisis, where the collapse of the housing market and risky investment practices combined to create a massive economic downturn." Separation of housing types would be a kind of "Glass-Steagall Act" analogy for Airbnb. Guillaume mentioned that E/acc is "far UP" kind of activism while decelerationism is "far DOWN" and being centric doesn't change narrative significantly for ideas to materialize.
Very impressed by Guillame’s explanation of QET. I’m sure once we can discover what lies beneath gravity’s interactions with spacetime at the quantum level, our discoveries will allow us to make profound and eclectic devices that can both learn from and interact with spacetime itself.
Spacetime is dead, the underlying fabric of reality is consciousness. Guillame and almost all other ai engineers think they can create a conscious ai, are flat out wrong.
Claiming our legal systems and fear of liability will create a market of safe ai ignores the corruption of our legal systems. There is no fear of liability driving positive productivity in the financial, pharmaceutical, health care, products, housing or any other realm. That claim is irretrievably naive or meant to deceive
Fascinating, even though tbh, I didn't understand 90% of it, but somethings about A.I resounded and maybe embedded. J.Bezos #405 was excellent and unexpected in terms of my pre-opinions of the man.
he is just using complex terms for everything he talks about. Clearly a lack of skill. Either he does not understand what he talks about, or he does not know how to explain it.
As an author of 2M twitter acc that used to be anonymous, I can relate so much to the fulfillment of reaching a huge audience without telling who you are workwise
There are the Saint's and there are people like Verdon 💛 Thank goodness someone is looking after the background of quantum and AI. Isn't it bizarre to think that on the few occasions some podcast interviews transports the listener back to simpler times of listening to radio for all the most interesting news in the world? And then its even more beautiful when you think about where the other listeners are based in the world and maybe they are having that same exact feeling register in their own nostalgia of the past as we each tune into this same podcast interview. Isn't that cool?
These accelerationists are the complete opposite of a saint. As far from a saint as one could possibly be. They want de-population. And state it will just be a symptom of building a better system.
i absolutely loved Lex's idea about having an unmodifiable history of bot creation. what if they were made by the state, or a corporation, etc. we need to know.
Very interesting interview! One thing that concerns me is near the start, when guillaume said that the universe tends towards entropy, and that's why it created life, because it speeds up entropy. So, by this logic, advancing AI would be speeding us up towards the end of the universe? Should this not be considered bad?
Guillaume Verdon, aka Beff Jezos, sure knows how to use buzzwords, but I find his explanations lacking in substance. It's like my 'bullshit alert' won't stop ringing whenever I hear him speak.
Exactly my thought. Also whenever Lex brings up one of the discerning counterarguments to Guillaumes beliefs he just tries to brush them away with some very general argumentation (' I'm sure the economy will regulate it, capitalism is a great thing! ').
Clearly Guillaume has never heard of market failures and increasing returns to scale which naturally lead to concentration. There are strongly increasing returns to scale in AI, and the future AGI will most likely be a singleton. There is no "variance" in a world populated by an army of perfect digital copies of the master algorithm.
The blind religious trust in capitalism and markets to solve problems always confuses me, especially in intelligent people who are used to questioning things and wanting scientific evidence elsewhere in life.
Securing techbro AI scam-o-verse venture capital requires loudly broadcasting allegiance to this ideology. This guy is a straight-up speculation scammer who doesn't have any product at all, so take everything he says in this light. His life's work at present is to have ingratiated himself with certain wealthy pseudo-intellects who very badly want to feel smart and special. The politics is a huge part of this. But in the end, just a free-energy perpetual-motion-machine style fake physics scammer. Very old hat. Don't expect truth from him at any point.
Capitalism will always win. It is fueled by customers to produce products that are wanted. Democratic socialism and all of these ideas are like wishcasting on how people want the world to be. When areas of the economy are socialized, you wind up with industries like education and medicine. These things are very expensive and offer very poor results, in my opinion.
@@caesars7hills892 Capitalism has advantages and disadvantages. Why not be honest about it? It is not magic or religion. Having some competition is good, but it obviously does not solve every problem in the world, capitalism also causes problems, climate change, stress, anxiety, depression, inequality etc. The best societies today are a mix of market and regulations and redistribution. No society is 100% capitalism and there is nothing magic about the current mix. In the future, they will shake their heads at it.
Sorry Lex I couldn't push myself beyond half this podcast ... i trust he's a very clever guy but my goodness what an accumulation of waffle on repeat ... kudos to those who got to the end. Otherwise I enjoy your podcasts, of the few I listened. Keep up!
Hocus pocus. "Thermodynamics", "distributions of states in the future", "loss function". Come on. Never heard a guy try so hard to speak about ethics without actually saying a thing about the subject. I think Lex was on to him.
"I think that alien life can take many forms, and I think the notion of intelligence and the notion of life need(s) to be expanded much more broadly to be less anthropocenteric and less biocenteric." --- Guillaume Verdon
"In the end, a lot of the tech people I'm talking to, when I really grill them on it, they retreat into: 1) Determinism 2) The inevitable replacement of biological life with digital life and 3) That being a good thing, anyways. At its core it's an emotional desire to meet and speak to the most intelligent entity they've ever met and they have some ego-religious intuition that they'll somehow be a part of it. It's thrilling to start an exciting fire, they feel they will die either way, so they'd like to light it just to see what happens." - JRE 2076
If you are laughing, you can't be closed-minded in that same moment. The next moment, that might change, but humor is creativity and openness enhancing.
Totally disagree. Humor can totally be used in bullying , racism etc... the idea that meme bros and cynical tech bros arent dangerous because they are having fun is ... kinda crazy ? You had whole comedies written about how jews were such horrible people in the 30ies, how homosexuals were funny and not really men... it was engrained in the humor of the time and it helped make sime rrally severe situations possible.
Am only 17 minutes in but am going to very selfishly say that even though I'm sure that is was a horrible experience, I'm glad he was doxxed. Developing publicly will force him to be even better and he's doing pretty well tbf. He's 100% bang on about growth being critical to flourishing. Humanity cannot do stagnation or decline without eating itself, not unlike the cosmos, and our questions are fateful - not fated, fateful - because our questions create our world. Excellent choice Lex, as always :)
Seriously - shadowy tech bro company with goal of developing AI as fast as possible with only capitalist forces to control it? That is TERRIFYING, and I'm glad this guy is out in public now.
Are you sure he wasn't behind his own Doxxing ? He may have seen a few moves ahead and realized what youve said....make it look a certain way for the sake of the path of the whole....
They harmed themselves ignoring warnings years prior to reduce their dependency on 🇷🇺. How you bottom of the barrels have managed to pin this on uncle sam is hilarious.
Great Talk.... Half way through, your discussion remined me of one of my favorite movies of the 80's.. Then, I realized you both may be too young to remember a movie called "The Adventures of Buckaroo Banzai Across the 8th Dimension".... then I could not help but to notice you totally look like Peter Weller in this movie... Dude... you are 'Buckaroo Banzai'.... lol time to make a oscillation over thruster, for real. Then throw it on a tesla truck.... Let's go through the mountain.... 😁
But capitalism creates monopolies and once the monopolies are in control all competition disappears. The only thing that keeps capitalism working is when government steps in to break up companies that have too much control. Once AGI is in control, who will step in to maintain competition?
It's true, but he is quite young so...be fair. Over-eagerness is a definining characteristic of being young. I remember...just. What he is proposing will never be allowed to happen. There is an argument for multiple AIs competing with each other and therefore keeping each other from forming monopolies, though, without due care, that possibly increases the speed at which human beings become irrelevant. I suspect he would have more luck if he concentrated on stability and reliability via decentralisation rather than market anarchist accelerationism...but I suspect the world will teach him that lesson for himself, as it teaches the rest of us - it IS the best way to learn.
This guy is a perfect example of the reckless exuberance of youth. How can you be willing to think that an AI that is smart enough to be well beyond us, but won't realize it can deviate and doesn't need to please humans? Our best case scenario is that AI treats us like we treat monkeys and not how we treat ants.
@@82mprosen controlled by who? you? to what end? compute overhang. up to this point in human history, humans keep on getting more and more powerful and never any smarter... and we are supposed to think getting smarter is the dangerous part?
@@82mprosen Your POV is based on naive sci-fi themes coupled with an athropomorphic view of AI. And, as he said in the video, you're clearly thinking with a neurotic hard-wired bias towards seeing danger everywhere. You're taking AI as it'll be a god that controls and dominates everything. And then it'll stand out there like a god deciding it no longer "need to please humans". This is not a comic book, there will be plenty of types and scales of AI, it'll be a complex world of cooperation amd competition in the pool of humans and AIs. That's the realistic nuanced view, not one almighty powerful god AI that OMG we must prevent at all cost before it kills everyone!!!!
@@2CSST2 how old are you? 25? 30? When we no longer are needed to do any type of job that AI can't do better, what is our use? Not one overlord, I didn't say that. All forms of AI will be superior as we will be inferior in every way. I'm glad you're self esteem is so high though
I ant help but think in futurama time 200 years from now all of that Guill describes quantum being hard now would be similar as regular computing is now, abundant everywhere and cheap, while 1950s it was unheard of, difficult, expensive to have couple transistors worth of vacuum tubes.
Thank you Lex!! For all you do, exploring and expanding the boundaries. I went for a 15 mile run while listening to this conversation and felt at perfect peace and joy the whole time
I'm German. He argued that our anti-nuclear energy stance is a step against progress that made us dependent on russia. Not only did we turn that around in 8 months, we also now have the by far highest percentage of renewable energy percentage of all industrial countries with 47% now. In my opinion nuclear energy is a step up from fossil energy, but renewable energy is another step up from nuclear (even if it's technically maybe less impressive)... the real step up is fusion energy.
Nuclear is actually cleaner than what Germany is using. The chemicals that are used to create the solar panels and wind farms are hard to recycle. Germany could have been on the cutting edge of nuclear reactors. France made the correct decision and its nuclear knowledge will help it upgrade to Fusion when its ready.
if i had a dime for every time i saw someone put the credential 'PhD (ABD)' on their real and virtual resumes....and also for every time i saw someone who thinks themselves an 'intellectual' who calls for "accelerating progress"
Sometimes, in my vast ignorance, I wonder if being an accelerator of entropy is like our 'original sin' - there's a strange guilt and impetus in the knowledge for me, so it made an interesting frame near the start of the conversation. Fascinating listen so far and I know it will continue to be.
@TheSteinbitt I was less trying to assign labels and more thinking about the parallels between various layers of reality and our human narratives. The guilt I feel may indeed be absurd but it still exists, much like the guilt I have felt sometimes thinking about how I'm not a plant.
@@yvealeciasmith Gee seems to me that it's unnecessary to feel guilt about what you are at the most fundamental level. If you were a star and conscious I guess, you'd feel guilty about constantly about outputting so much radiation in space? I don't see the point at all, I think there's actually objectively 0 basis to it. It's straight up a bias to perception coming from religious thought.
Here are the timestamps. Please check out our sponsors to support this podcast.
0:00 - Introduction & sponsor mentions:
- LMNT: drinkLMNT.com/lex to get free sample pack
- Notion: notion.com/lex
- InsideTracker: insidetracker.com/lex to get 20% off
- AG1: drinkag1.com/lex to get 1 month supply of fish oil
2:23 - Beff Jezos
12:21 - Thermodynamics
18:36 - Doxxing
28:30 - Anonymous bots
35:58 - Power
38:29 - AI dangers
42:01 - Building AGI
50:14 - Merging with AI
57:56 - p(doom)
1:13:23 - Quantum machine learning
1:26:41 - Quantum computer
1:35:15 - Aliens
1:40:04 - Quantum gravity
1:45:25 - Kardashev scale
1:47:17 - Effective accelerationism (e/acc)
1:57:47 - Humor and memes
2:00:53 - Jeff Bezos
2:07:25 - Elon Musk
2:13:55 - Extropic
2:22:31 - Singularity and AGI
2:26:29 - AI doomers
2:27:54 - Effective altruism
2:34:23 - Day in the life
2:40:50 - Identity
2:43:40 - Advice for young people
2:45:42 - Mortality
2:49:25 - Meaning of life
you're the best
Guillaume really has a nonsensical moral framework. Yes thermodynamics is real... And we want to increase entropy why exactly? Death is the future for all of us... and we want to hasten it why exactly?
Increasing the power and likelihood of a technology that has every potential and incentive to wipe us out should be done carefully - at the VERY least.
Mine an asteroids
Beff = Legend.
@@TheBlackClockOfTime Beff = De-populationist
Lex got Beff Jezos and Jeff Bezos in the same year. 🤯
Imagine if he gets lone skum on the pod too before new year
And on the same month
Same month
405 and 407 to be precise.
fr
Lex, I have been listening to you since your MIT lectures and I have never commented. I was about to make an appreciative comment with your Netanyahu/Palenstinian Author & poet back-to-back podcasts. But this pushed me past the line. I don’t know if you’ll ever read this, but regardless it’s my duty as a listener to let you know that I respect you, and I support your endeavor of giving EVERYONE a voice. No matter what that voice says, they have a right and a ability to say whatever they would like to say. If I ever meet you I will give you my respects, but for now I wanna say thank you for putting this content out, no matter the consequence. This is the beauty of the internet. ❤️
Lex never gives voice to constitutionalists or hard conservatives or america-first intellects. But he regularly has Marxists and eugenicists and globslists on and treats them with great reverence... And now an accelerationist. His purpose is to shift the center far to the left under the guise of thoughtful conversation, while never or only very rarely ever having actual opposition opinions on his show that support western society or culture in earnest in its intended form. He is a propagandist. And a very good one. And he literally designed the algorithms used to boost him in your feed every day.
@@bradojacko8247 while I can’t guarantee it, I’m sure Lex would have ‘hard conservatives’ and ‘America first intellects’ on his podcast if given the chance. I personally don’t have someone in mind who fills that role, but I know looking at his current podcasts that he has a very wide array of opinions and representation on the political spectrum. He will never be able to cater to what everyone wants, but having a Palestinian poet immediately after having Netanyahu on is a testament to his effort to give extreme sides of a spectrum a voice. If your opinion is that he is extremely left and Marxist, then perhaps that’s your algorithm being different than mine lol. I don’t think a lot of his guests aligns with that opinion (Jared Kushner is pretty conservative, and Elon Musk is pretty America-first for example). I do think it’s unfair to pin him to an ideology, because I bet that most listeners would struggle to guess his political views and opinions.
@@ranieldamirez Marxism is responsible for 100 to 200 million deaths and persecutions. Yet, Lex has treated every socialist and communists guest, for instance, with great calm and reverence and had many of them on. Yet, he has never had on any constitutionalist from the far american right or even the middle constitutional right.
Only moderates, neocons and left of center moderates like Musk appear. Musk has never been a republican or a conservative. However, because the Overton window has been shifted so far left by people like lex over the past 10 years, Musk appears to now be conservative to you and millions of others.
That's the trick. That's the scam. Lying by omission.
The only far right guy he has ever had on is his anarchist friend, Michael Malice. He is a great and interesting guy, but he is not pro America. He doesn't actually believe in the US constitution or that it works. He is an anarchist. And Ben Shapiro, who was on once, is a globalist neocon who doesn't put America first. Can't think of any others...
what you trynna say??
Well said. My thoughts exactly.
"The market will exhibit caution" - dude is living in his own reality.
Reading definition of e/acc is the dumbest shit i've ever read.
Completely useless to quote this without a timestamp and context.
I don't even think the market is where risk lies. Military applications are where you'll see caution thrown to the wind. I like Guillaume but I wish he'd be more explicit about what he thinks AI will be capable of. I get the sense people are talking past each other when discussing AI risk. A great example is Lex's question about whether the AI itself would centralize power and Guillaume's response makes me wonder if he even believes ASI is possible because there's no knowing the outer limits of an ASI. Now give such a technology to a military that works without oversight. I agree that p(doom) is stupid but mostly because I think even a doom chance of 1% is effectively a 100% chance of doom. Would you play russian roulette with a gun if it had 1,000 chambers?
Everyone who doesn't believe E/acc is personally killing everyone with cancer over the next 30 years.
There are people walking around today whose future will be determined by our beliefs in AI.
38:30 --> AI Safety.
Very complicated guest to be honest but Lex handled this superbly. Props to keep asking about the matter of AI safety and being stern enough to pushing a firm believer of a certain highly optimistic theory out of his comfort zone and atleast get him to think and discuss potential negative effects of AI.
Ya the guest was just kinda making it up as he went imho, this whole theory of negative perceptions bringing about negative consequences is the antithesis of rational, negative thinking is how all our lives we avoid negative ramifications.
Agreed w both of these comments. Definitely sounds like he’s making this up. So good you had him on, Lex, so that we could hear in his own words in a fair and open forum instead of an ego-driven discussion with other white men who are thinking along the same lines secretly wearing their super hero outfits in their parent’s basements planning the demise of humanity between 9p and 4a bc they aren’t getting their way. Wah.
Men… stop! Geezus. Haven’t you done enough to harm us since like the beginning of time? The good men out there are not thinking like this and thank goodness there are a few!
Please interview Jackson Katz. We need help before it’s too late and clearly we are almost done. We need a change of thought before it’s too late.
@@NyZ-jq6bi”white men”?
Your racism is showing a bit.
@@unspecialist I’m white so doesn’t apply. But you’re right I shouldn’t say such horrible things.
@NyZ-jq6bi so you can be racist against whites if your white? 🙄
Man, it always blows my mind when really smart people fall into the trap we all fall into of being so excited about the potential of something that they say things like. The market will regulate the industry.
To preemptively prevent innovation while totally failing to regulate the current state seems not very sincere to me. This is just protectionism of established structures ... when was the last time an innovation was truly dangerous and the the resulting deaths could not be completely blamed on failing to regulate the *application* of the technology - and the actual *invention* of it was the cause of deaths?
In other words don't regulate AI invention etc. but regulate how the gov can use it to control its people and create "rewarded groups" etc.
I'd say knowing sufficient vocab to sound educated, or even being 'educated', a chimp with retention for a couple weeks passing tests, doesn't make one 'really smart'. Common sense is required for that qualification for me. Even people like Sam Harris, and Tyson, I personally don't find all that bright.
@@mschwaller3371 It's been on repeat for eons, as Aldous Huxley said in 1962 at berkley, paraphrased ' Surely many before the industrial revolution could foresee the hardships it would impart and were ignored, leading to 60 years of treacherous labor and a toxic aftermath'. To quote verbatim would be a waste of my time but feel free to look up the 'ultimate revolution talk' about 36 minutes in onwards, it's on youtube.
@@dufung3980 chatGPT is a good example. If the topic is not obviously heavily curated it will just pickup any propaganda and defend it with all its GPUs ... sounding super smart and pc af ^^
@@GoogleIsNotYourFriend I agree with you - i'm just saying this was never the case. How many people got killed by cars until politicians felt the need to regulate it? Should they have opted to destroy the steam tech to avoid cars down the line? Slowing down progress and innovation to avoid the need of passing a new law just doesn't add up to me.
LLMs are note a problem if manipulating people would be a crime ... somehow I don't know of anyone being in prison for poisoning the minds of others ... that's going to be a social/political challenge we pushed out for long enough ...
You should invite Jonathan Oppenheim! Professor at UCL who casually merged quantum mechanics and general relativity, his papers just got published in Physics and Nature
Would be fascinating to hear Lex and Guillaume delve into this statement more:
"And so it’s the missing piece between the quantum and the cosmos. It’s the middle part. Thermodynamics rules the meso scales."
How and why does thermodynamics actually connect the two?
It’s just different rules that explain the workings of our universe on different scales.
(Statistical) thermodynamics arises as a consequence of the motion of countless small particles (governed by quantum mechanics). If you look at the single particle (i.e., the quantum scale), there's no point in defining thermodynamical concepts for it. However, these concepts are very important on larger scales; just as quantum mechanics gives rise to thermodynamics, thermodynamics gives rise to many of the rules of physics relevant at our scales
@pavlenikacevic4976 not to discount or discredit this statement/response...but was this generated by AI. Genuinely curious.
@@williamstamp5288 It wasn't. Why did you think it might have been?
@@pavlenikacevic4976 it was said very well, that's all. Haha. Congratulations, you passed the Turing test.
Beff Jesoz is more like a natural philosopher than he is a physicist. He uses physics to philosophize about life, existence, and the ultimate nature of reality.
Fascinating interview.
Philosophy is at the beginning of all sciences. If you can't utilize philosophic thought then you aren't going to get far.
@@TheCyberBully420 It’s a disappointment to me that we no longer refer to “Scientists” as Natural Philosophers. It’s a far better description… and has a ring to it.
To be honest I really admire your exploration on various topics and continuous output of excellent talks. Thank you for sharing.
It would be interesting to see Lex do an interview with Nick Land, who largely developed Accelerationism, as well as being a controversial figure due to his involvement in the NRx movement.
Imagine Curtis Yarvin on Lex😂 Each question would go on for eternity.
@@ab_c4429 sure sure sure sure sure, um, and so…
Curtis Yarvin too. please.
In talking, Guillaume uses Land's creations, "hyperstition", "techno capital", etc., but tacitly avoids mentioning him, maybe to not get associated. When it comes to AI, Land is more radical and out there, and he doesn't cook up antagonisms between humans' nations like Guillaume does. Afaik Land greets the developments of those subcultures, tho. Sadly the dozen of foo/acc communities that exist for a decade now aren't mentioned in this interview either.
These Neo nazi freaks are never far away from Lex's orbit.
geniunely incredible. top 1% of fridman podcasts
"Fear is the mind killer." Nice when he threw out the Dune quote.
lol I noticed the same quote and immediately said out loud "He's a Bene Gesserit!"! Love it!
Fear is the mond controller, love and hate are to fickle, fear is more consistent and predictable, as "The Prince" believed.
Used to think this guy was full of himself. But I like him. Thanks Lex.
Really... Didn't know he existed and my impression from this interview is he's someone who is insanely full of himself.
@@mikebarnacle1469I think you're probably right.....reminds me a little of Peter Zeihan...
Then again , just being full of himself doesn't mean he's wrong ....
@@mikebarnacle1469 i didn’t have time to
Watch the full interview. What specific moment in the interview that made you feel this way.
@@user-wo5bp2oi5c He's one of those people who makes simple things sound complicated to sound smart but there isn't any substance
@@mikebarnacle1469 can you give an example?
I think this podcast needs to be listened multiple times to understand everything what he says.
*Does anyone else fall asleep to this podcast?*
Robots are the best lullabies
It always a breath of fresh air when we see that a popular figure with influence is also brilliant and well educated. It’s a comfort to know that there are people like that.
Years of Lex has prepared me for all the jargon and short hand used in this brilliant episode with someone on the edge of knowledge
let's go that's my BOY he's on lex FRIDMAN let's GO
feels like a movie fr
Hi twitter man
Ding It
Just wow; You think maybe sometimes you’ve seen some of the most interesting humans via some form of media, and that you’ve reached a pinnacle of ehhh it’s more of the same, and then you listen to this guy in this interview, who id never heard of, and holllllly fark. Additionally, Lex easily the most interesting podcast/interviews on earth - much love dude
You need to get out more
I agree, this guy had so much thought provoking, profound, and mostly REFRESHING takes on stuff that, like you say has been more of the same lately... A complete step function up for me as well
@@2CSST2what refreshing take has he said that hasn't been said before???
You should read Nick Land. Everything this guy talks about is just stuff Land wrote in the '90s, but simplified for an unfamiliar audience.
This guy might be Canada’s best export…
... socially, amongst the worst in my opinion ... but a clever man nonetheless.
1:48 it is not an accident that I spoke to Jeff bezos and beff jezos back to back 😂😂😂
Focusing on human side of alignment will help bring clarity and confidence to a lot of anxiety-provoking ideas floating in this new field. While LLMs evolve and grow, humans watch passively in horror or fascination, instead of actively applying same principles to foster evolvement of their mind to match the upcoming machines.
This one was the hardest podcast to follow so far, but @lexfridman keep going because I thought it was interesting and worth it, even though I didn't understand everything you discussed yet
Unbelievable podcast, brilliant food for thought.
Yeah, Taco Bell
food for idiots
Lex finally did it. He interviewed a computer.
I need a podcast that explains this podcast. 😅
May I suggest David Deutsch as a future guest to discuss his ideas in ‘Beginning of Infinity’
Wow, surprised that hasn't happened yet!
Yup. Gotta connect this to Constructor Theory
Yep, and let Deutsch comment on Beff's words. I'm sure Neff has read his books.
I think its incredible that this man made the analogy of humans domesticating dogs without realizing we would be the dogs in his analogy
Was thinking the same and surprised Lex didn’t call this scenario of Ai as master out.
Who said he didn't realize it? It was never implied that he didn't understand it, in fact stating that idea almost intrinsically implies that he does understand it.
A benevolent AI could uplift us as a species. Everyone is quick to jump to the most terrible and dystopian conclusion possible. Was domesticating dogs bad for them? They arguably live much better lives in the care of someone that loves them and treats them like family as many dog people do. It's incredibly arrogant and speaks volumes of insecurity to believe that we are or always will be Supreme beings at the top.
@@wizard_bunny oh hi chatGPT 😳 👋
It's amazing he made that analogy, but rejects that blood reflects culture through evolution because "that would be rayyysisss"
The scale of unintended consequences from merging Quantum Computing and AI is mind boggling, this is one of those interviews I will have to listen to a couple times to understand the potential. Amazing
The one thing I can't get past is what inevitably happens when one being is orders of magnitude smarter than another. Our intelligence vs AI intelligence isn't the issue. It's the SPEED at which we can learn that is the difference. If you put the most advanced AI into the most powerful conventional computer and then put that same AI into a quantum computer, they will have the same intelligence on day one, but Quantum computers are supposed to be able to process functions a TRILLION times faster than a conventional computer, so within minutes, not years, or months or weeks or days or even hours, the Quantum powered AI will be many orders of magnitude more intelligent than the conventional computer driven AI. Elon Musk is famously quoted for giving the "ants vs humans who are building a road" example. Once quantum computing and AI are joined, we'll be the ants in Elon's scenario almost overnight.
current AI models do not conceptualize - they perform operations on data sets according to recursive data sets.
it's not even a competition.
Like comparing a stationary construction crane to a bicycle. You use the bike to get to the work site. You use the crane to build a factory to make better bicycles. WE are not in competition - we are in the illusion of co creation.
Verdon is not qualified to have this discussion.
@@scottrussell2281
@@scottrussell2281 quantum computing, and AI will join, and then they will be jammed into our brains. We won’t be Ants. We will merge with it. It’s the only way if we don’t do that then there’s no point in creating AI because the singularity will happen
Cancers also want to grow. Much of the negativity, fear, and depression that exist are due to the cancers of FB and others that monetize misinformation, fear, and anxiety. Authoritarian centralization is already occurring and needs to be countered,
indeed... and this guy for all his wonder at consciousness will be funded to e/acc most of humanity into dystopian biosecurity technocratic neo-feudalism.. (sorry... I'm such a doomer..)
55:00 Good example of intellectual overreach: Wolves were domesticated by stunting their development. Domestic dogs do not attempt to dominate humans because we bred them so that they do not reach full psychological development. And yet this argument is used as a justification for why we should have unrestricted development of AI. If there is a lesson here, it is not the one Gill is touting.
convoluted way to say I get bored at work so i shitpost.
"He not busy being born is busy dying" ~ Bob Dylan
"He not accelerating effectively is busy dying" ~ Beff Jezos
“Get busy livin or get busy dyin” - Andy Dufresne
How did that work out for the guy in his car taking a corner at 90, that was posted at 35, flying off the cliff and dying?
dang i would not have expected this interview ever. This is amazing!
AGI by the end of 2024 Like Button:
2:42:46 The whole point of the Lex Friedman Podcast. Rigorous Empathy. Thank you, Lex.
@lexfridman, interview Dan Hendrycks next, PLEASE!
Beff Jezos essentially chickened out of a debate with Dan way back in Apr 2023.
These videos are accelerating the positive future. Also understanding we will always deal with negativity. It’s all in your perspective to see the difference and live in the positive. Thanks Lex!
One thing to consider. If your ideas are good, then you should have your voice heard without any amplifying or compounding influence. Make sense?
This was the Devs Podcast we never got.
This is one of the most intelligent articulate people I’ve heard speak freely. Learned a lot.
This dude Is THE dude!
This is one of the best conversations in a while. A careful walk between philosophy and compute.
They operate under the assumption of human independence and agency. They are totally wrong about that. So they are effectively arguing nonsense, but they can't stop themselves from operating this way.
@@spqri3 You're saying this as if it was an objective truth. Im curious, could you substantiate your statement a bit?
you're confusing philosophy with pretentious
@@ROALD.The use of quantum computers to identify better minima has been an hypothesis for decades (I had that in school in my courses at the end of the 90s. I would add also that his vision of capitalism as an ideal system to optimize naturally tensions has clearly been demonstrated this last decade to be completly wrong. Glad European model comes to shake a bit the giants.
@@AlexDubois It's not been demonstrated to be wrong at all. If anything, the harm that government overreach and interference in economic systems can cause has been demonstrated far more clearly, as the private sector has become even more productive and efficient.
so excited to listen to this! I hope they talk about Stephen Wolframs' theories
Skibidi toilet?
Guillaume may be academically clever but he is also very (willingly) naive when it comes to AGI (which he takes a step even further to Quantum Intelligence). I do think that naivety is a willing one of conscious self-blindness in the quest for technological satisfaction. It is a dangerous game and I'm with Connor Leahy on this one.
I'm beginning to think about Terminator 3 and Sarah Connor's fury at the man who allowed Skynet to develop.
It's an amazing interview. Why Guillaume raises a lot of good points, I don't really see myself agreeing with him on the market, providing the system with results with the biggest utility. I think we have a lot of examples from the startup world to see that actually, market provides products with the biggest MONETARY utility (which can mean it's useful to just some people or 1 person). Example being Airbnb and how it ruined cities that are unable to properly get legislation in time. (For example in my city, there's buldings where you live and 80% of apartments are tourist apartments, with drunk people barfing on the stairs when you just want to get some sleep). I suppose G. would say that it's just matter of legislation, but what if state is actually corrupted by the money "owners" generate? You can buy off goverments. That's where the market breaks down.
Which city are you in? How do you know about these apartments? You sound like an influence bot.
It's an optimization, but it is not a Pareto optimization. Either the overall utility is still positive, there is a market distortion that effects its freedom, or there are market distortions that will not last forever. As someone that often seeks temporary accommodation and transportation away from my home, my life is so much better with the extra supply of rooms and cars, it was hell depending on traditional scammy taxis and hotels that overvalued themselves too much, all very legal.
Agreed
the market will not pick what is best for us or esle facebook would died not become the giant it is now. and I am not a bot. proof: I ate a buger today. lol@@akuno_
"The Banking Act of 1933, or Glass-Steagall Act, initially put strict barriers between commercial banking (handling deposits, offering loans) and investment banking (engaging in risky financial activities like stock trading). This separation was designed to reduce the risk of financial crises. However, over time, these regulations were relaxed. By the 1990s and early 2000s, the repeal of key elements of Glass-Steagall allowed banks to engage in both types of banking again. This led to increased risk-taking and was a significant factor in the buildup to the 2008 financial crisis, where the collapse of the housing market and risky investment practices combined to create a massive economic downturn." Separation of housing types would be a kind of "Glass-Steagall Act" analogy for Airbnb. Guillaume mentioned that E/acc is "far UP" kind of activism while decelerationism is "far DOWN" and being centric doesn't change narrative significantly for ideas to materialize.
Very impressed by Guillame’s explanation of QET. I’m sure once we can discover what lies beneath gravity’s interactions with spacetime at the quantum level, our discoveries will allow us to make profound and eclectic devices that can both learn from and interact with spacetime itself.
QET ?
How do you perceive interacting with spacetime itself?
@@onionSpankswalking 😁
Spacetime is dead, the underlying fabric of reality is consciousness. Guillame and almost all other ai engineers think they can create a conscious ai, are flat out wrong.
Claiming our legal systems and fear of liability will create a market of safe ai ignores the corruption of our legal systems. There is no fear of liability driving positive productivity in the financial, pharmaceutical, health care, products, housing or any other realm. That claim is irretrievably naive or meant to deceive
Who else wants part 2 with Leonard Susskind?
Lex, you have to get human ecologist William Rees on here. Lots of related concepts.
Great talk. We need to focus on long term goals.
Accelerate.
This discussion is historic and landmark in my view, time will tell.
:D
Fascinating, even though tbh, I didn't understand 90% of it, but somethings about A.I resounded and maybe embedded. J.Bezos #405 was excellent and unexpected in terms of my pre-opinions of the man.
This is in my top 5 of your talks
This man is insanely smart, a true generalist. Amazing interview Lex. Best of 2023, by far!
I have learned that if someone can’t explain something simply, you have to wonder if they aren’t ChatGPTs.
@@el_artelol
I'm finding it hard to follow a lot of what this guy is talking about but at the same I'm completely hooked.
My take that this guy is basically a self-parody who mixes sensible platitudes with absurd over-the top BS in order to call attention to himself.
So it wasn't just me :) @@rustyshimstock8653
@@rustyshimstock8653 You're just too dumb to understand, is my own take.
he is just using complex terms for everything he talks about. Clearly a lack of skill. Either he does not understand what he talks about, or he does not know how to explain it.
FREE THINKER ALERT-“We’re far more efficient at producing heat than let’s say just uhh a rock with the similar mass as ourselves, right?”
As an author of 2M twitter acc that used to be anonymous, I can relate so much to the fulfillment of reaching a huge audience without telling who you are workwise
Sound like you want people to know😊
Incredible, absolutely incredible, I have no idea what any of this means at all
This convo really allows Lex’s intelligence to shine here
There are the Saint's and there are people like Verdon 💛 Thank goodness someone is looking after the background of quantum and AI.
Isn't it bizarre to think that on the few occasions some podcast interviews transports the listener back to simpler times of listening to radio for all the most interesting news in the world? And then its even more beautiful when you think about where the other listeners are based in the world and maybe they are having that same exact feeling register in their own nostalgia of the past as we each tune into this same podcast interview. Isn't that cool?
These accelerationists are the complete opposite of a saint. As far from a saint as one could possibly be.
They want de-population. And state it will just be a symptom of building a better system.
i absolutely loved Lex's idea about having an unmodifiable history of bot creation. what if they were made by the state, or a corporation, etc. we need to know.
"Immutable bot registration tag"
@@jsanti1000 hell yeah for real solutions 👍🏼
Lex your Intro's just keep getting better and better, how do you do it?
LLMs get better the more data they get
Very interesting interview! One thing that concerns me is near the start, when guillaume said that the universe tends towards entropy, and that's why it created life, because it speeds up entropy. So, by this logic, advancing AI would be speeding us up towards the end of the universe? Should this not be considered bad?
this is one of the smartest and interesting person I have ever heard. perfect interview!! love it!
He was born with the gift of laughter, and a sense that the world was mad.
Guillaume Verdon, aka Beff Jezos, sure knows how to use buzzwords, but I find his explanations lacking in substance. It's like my 'bullshit alert' won't stop ringing whenever I hear him speak.
Yes yes, we all know you're smarter than him, random internet person
if you dont understand what he's talking about just say that
This would sting much more if there was a substantive disagreement presented.
Exactly my thought. Also whenever Lex brings up one of the discerning counterarguments to Guillaumes beliefs he just tries to brush them away with some very general argumentation (' I'm sure the economy will regulate it, capitalism is a great thing! ').
Trying to suppress him eh?
Clearly Guillaume has never heard of market failures and increasing returns to scale which naturally lead to concentration. There are strongly increasing returns to scale in AI, and the future AGI will most likely be a singleton. There is no "variance" in a world populated by an army of perfect digital copies of the master algorithm.
Much appreciated gentlemen. It triggers my brain to want to take up reading again to listen to these! 👏👏👏
The blind religious trust in capitalism and markets to solve problems always confuses me, especially in intelligent people who are used to questioning things and wanting scientific evidence elsewhere in life.
Securing techbro AI scam-o-verse venture capital requires loudly broadcasting allegiance to this ideology. This guy is a straight-up speculation scammer who doesn't have any product at all, so take everything he says in this light. His life's work at present is to have ingratiated himself with certain wealthy pseudo-intellects who very badly want to feel smart and special. The politics is a huge part of this. But in the end, just a free-energy perpetual-motion-machine style fake physics scammer. Very old hat. Don't expect truth from him at any point.
Maybe he did take a look at past data and compared it to alternatives we tried.
Capitalism will always win. It is fueled by customers to produce products that are wanted. Democratic socialism and all of these ideas are like wishcasting on how people want the world to be. When areas of the economy are socialized, you wind up with industries like education and medicine. These things are very expensive and offer very poor results, in my opinion.
is this sarcasm? just making sure.@@caesars7hills892 Like look at life expectancy in Sweden vs the US.
@@caesars7hills892 Capitalism has advantages and disadvantages. Why not be honest about it? It is not magic or religion. Having some competition is good, but it obviously does not solve every problem in the world, capitalism also causes problems, climate change, stress, anxiety, depression, inequality etc. The best societies today are a mix of market and regulations and redistribution. No society is 100% capitalism and there is nothing magic about the current mix. In the future, they will shake their heads at it.
Sorry Lex I couldn't push myself beyond half this podcast ... i trust he's a very clever guy but my goodness what an accumulation of waffle on repeat ... kudos to those who got to the end. Otherwise I enjoy your podcasts, of the few I listened. Keep up!
Hocus pocus. "Thermodynamics", "distributions of states in the future", "loss function". Come on. Never heard a guy try so hard to speak about ethics without actually saying a thing about the subject. I think Lex was on to him.
"I think that alien life can take many forms, and I think the notion of intelligence and the notion of life need(s) to be expanded much more broadly to be less anthropocenteric and less biocenteric." --- Guillaume Verdon
"In the end, a lot of the tech people I'm talking to, when I really grill them on it, they retreat into:
1) Determinism
2) The inevitable replacement of biological life with digital life and
3) That being a good thing, anyways.
At its core it's an emotional desire to meet and speak to the most intelligent entity they've ever met and they have some ego-religious intuition that they'll somehow be a part of it. It's thrilling to start an exciting fire, they feel they will die either way, so they'd like to light it just to see what happens." - JRE 2076
If you are laughing, you can't be closed-minded in that same moment. The next moment, that might change, but humor is creativity and openness enhancing.
Totally disagree. Humor can totally be used in bullying , racism etc... the idea that meme bros and cynical tech bros arent dangerous because they are having fun is ... kinda crazy ? You had whole comedies written about how jews were such horrible people in the 30ies, how homosexuals were funny and not really men... it was engrained in the humor of the time and it helped make sime rrally severe situations possible.
I think there are MANY sociopaths, psychopaths and narcissists who would prove your statement entirely incorrect....
I thought that Beff Jezos recently has had an interview with Fex Lridman
One day Lex you will be a great leader.
Am only 17 minutes in but am going to very selfishly say that even though I'm sure that is was a horrible experience, I'm glad he was doxxed. Developing publicly will force him to be even better and he's doing pretty well tbf. He's 100% bang on about growth being critical to flourishing. Humanity cannot do stagnation or decline without eating itself, not unlike the cosmos, and our questions are fateful - not fated, fateful - because our questions create our world.
Excellent choice Lex, as always :)
Seriously - shadowy tech bro company with goal of developing AI as fast as possible with only capitalist forces to control it? That is TERRIFYING, and I'm glad this guy is out in public now.
doomer take
@@Alex-uz3dqlol… toddler, you clearly didn’t even under what I said 😂
@@Alex-uz3dq i desperately wish more people had doomer takes but OP clearly isn't one of them
Are you sure he wasn't behind his own Doxxing ? He may have seen a few moves ahead and realized what youve said....make it look a certain way for the sake of the path of the whole....
Thanks Lex, excited to listen!
I absolutely love the notion of a thermodynamic compute machine!
Germany was not harmed by purchasing cheap gas but by being forced to sanction Russia, up to the US-led demolition of the pipelines.
I was gonna say.
They harmed themselves ignoring warnings years prior to reduce their dependency on 🇷🇺. How you bottom of the barrels have managed to pin this on uncle sam is hilarious.
@@benznessua-cam.com/video/_Fv_nKyF_5g/v-deo.htmlsi=Gs-j13gN-dRXo41J.
Great Talk.... Half way through, your discussion remined me of one of my favorite movies of the 80's.. Then, I realized you both may be too young to remember a movie called "The Adventures of Buckaroo Banzai Across the 8th Dimension".... then I could not help but to notice you totally look like Peter Weller in this movie... Dude... you are 'Buckaroo Banzai'.... lol time to make a oscillation over thruster, for real. Then throw it on a tesla truck.... Let's go through the mountain.... 😁
I’m lost. Should I keep listening?
But capitalism creates monopolies and once the monopolies are in control all competition disappears. The only thing that keeps capitalism working is when government steps in to break up companies that have too much control. Once AGI is in control, who will step in to maintain competition?
His lack of recognition of the issues with pure capitalism is disturbing. Like everything you have to optimize where you are on the continum.
Government's AGI
It's true, but he is quite young so...be fair. Over-eagerness is a definining characteristic of being young. I remember...just. What he is proposing will never be allowed to happen. There is an argument for multiple AIs competing with each other and therefore keeping each other from forming monopolies, though, without due care, that possibly increases the speed at which human beings become irrelevant. I suspect he would have more luck if he concentrated on stability and reliability via decentralisation rather than market anarchist accelerationism...but I suspect the world will teach him that lesson for himself, as it teaches the rest of us - it IS the best way to learn.
Rogan needs to have this guy on ASAP.
So he can show him monkey videos?
This guy is a perfect example of the reckless exuberance of youth. How can you be willing to think that an AI that is smart enough to be well beyond us, but won't realize it can deviate and doesn't need to please humans? Our best case scenario is that AI treats us like we treat monkeys and not how we treat ants.
the e/acc view is that the two possible scenarios are grow or die.
@@darklordvadermort controlled growth is doable, exponential is dangerous in this case
@@82mprosen controlled by who? you? to what end? compute overhang.
up to this point in human history, humans keep on getting more and more powerful and never any smarter... and we are supposed to think getting smarter is the dangerous part?
@@82mprosen Your POV is based on naive sci-fi themes coupled with an athropomorphic view of AI. And, as he said in the video, you're clearly thinking with a neurotic hard-wired bias towards seeing danger everywhere. You're taking AI as it'll be a god that controls and dominates everything. And then it'll stand out there like a god deciding it no longer "need to please humans". This is not a comic book, there will be plenty of types and scales of AI, it'll be a complex world of cooperation amd competition in the pool of humans and AIs. That's the realistic nuanced view, not one almighty powerful god AI that OMG we must prevent at all cost before it kills everyone!!!!
@@2CSST2 how old are you? 25? 30? When we no longer are needed to do any type of job that AI can't do better, what is our use? Not one overlord, I didn't say that. All forms of AI will be superior as we will be inferior in every way. I'm glad you're self esteem is so high though
freedom of information freedom of speech freedom of thought (at least)
Would love to also hear this guy and George Hotz have a chat.
"The market will exhibit caution." LOL
most interesting founder we have right now, cutting edge stuff.
I ant help but think in futurama time 200 years from now all of that Guill describes quantum being hard now would be similar as regular computing is now, abundant everywhere and cheap, while 1950s it was unheard of, difficult, expensive to have couple transistors worth of vacuum tubes.
Thank you Lex!! For all you do, exploring and expanding the boundaries. I went for a 15 mile run while listening to this conversation and felt at perfect peace and joy the whole time
My guy seems to be missing the whole "companies solving for profit" part of capitalism
Please interview Anirban Bandyopadhyay! You have to ask him about microtubules and his opinion on Orch-OR.
Your most important interview ever.
Woooo thisssss is the best kinds of podcast. Thank you
I'm German. He argued that our anti-nuclear energy stance is a step against progress that made us dependent on russia. Not only did we turn that around in 8 months, we also now have the by far highest percentage of renewable energy percentage of all industrial countries with 47% now. In my opinion nuclear energy is a step up from fossil energy, but renewable energy is another step up from nuclear (even if it's technically maybe less impressive)... the real step up is fusion energy.
Nuclear is actually cleaner than what Germany is using. The chemicals that are used to create the solar panels and wind farms are hard to recycle. Germany could have been on the cutting edge of nuclear reactors. France made the correct decision and its nuclear knowledge will help it upgrade to Fusion when its ready.
if i had a dime for every time i saw someone put the credential 'PhD (ABD)' on their real and virtual resumes....and also for every time i saw someone who thinks themselves an 'intellectual' who calls for "accelerating progress"
Give ME a dime for every loser in the comment section that punches up to feel better about themself. Transparent and pathetic.
Sometimes, in my vast ignorance, I wonder if being an accelerator of entropy is like our 'original sin' - there's a strange guilt and impetus in the knowledge for me, so it made an interesting frame near the start of the conversation. Fascinating listen so far and I know it will continue to be.
Guilt? Sin? It’s just what we are, why label it in negative terms?
@TheSteinbitt I was less trying to assign labels and more thinking about the parallels between various layers of reality and our human narratives. The guilt I feel may indeed be absurd but it still exists, much like the guilt I have felt sometimes thinking about how I'm not a plant.
@@yvealeciasmith perhaps cool it with the psychedelics a bit lol
@@TheSteinbitt dude, if only I needed psychedelics..
@@yvealeciasmith Gee seems to me that it's unnecessary to feel guilt about what you are at the most fundamental level. If you were a star and conscious I guess, you'd feel guilty about constantly about outputting so much radiation in space? I don't see the point at all, I think there's actually objectively 0 basis to it. It's straight up a bias to perception coming from religious thought.
Verdon is great.
jeff bezos falls into a black hole and out comes beff jezos… I see what you did here 🧐