Ah, still just about my favorite human being. Of ~all~ the thinkers I've gravitated toward, Bach seems the most bullshit free. It's hard not to sound like an irrational fanboy saying this stuff, but, genuinely, after some ~33 years, Bach was the first person EVER to completely pass my bullshit sniffer, beating the likes of Sagan, Aquinas, Turing., etc. He's made the rounds over the last several years and I'm happy to hear he's still not selling anything, and still hasn't sold out or jumped the shark with his ideas. Completely reasonable, open-minded -- Bach is King Nerd. And his youth in east germany gives him a perspective that I, as someone born in the USSR, too rarely see in public speakers. Bach knows wtf is up.
I'm not necessarily disagreeing with you (and in fact, Joscha's first three appearances on Lex's show completely reconfigured my mind!), but why would you believe that growing up within the confines of a monocultural, poverty-blighted, part of the world should be the kind of experience that would maximize self-critical, freely-associative, and ultimately individually imaginative thought? I mean, it _might,_ I just don't understand why you think it _should..._
@@simesaid Huh? Certainly not how I put it. He saw how the Soviet Union functioned, how it was governed, the stark contract between West and East Germany, and he was there when the wall fell. He speaks multiple languages and and understands multiple perspectives of the Eastern European microcosm and the earth shattering dissolution of the USSR. Those who've never traveled outside their country or aren't familiar with life in the USSR often just "don't get it" and aren't as appreciative, grateful, or urgent, etc.
Bach first popped up in my online journey a few years ago, and I haven’t been able to stop consuming his content since. He’s opened so many exciting rabbit holes in my mind that I'll be grateful for it for a lifetime
The guy is delusional. He has broad knowledge but not enough understanding on specific topics, like the AI safety and current trends in AI. He proves it by not acknowledging that hundreds of parties are working on making the AIs agentic. It's like basic knowledge at this point.
@@FreakyStyleytobbyI would think that you are the one who is delusional. Bach knows many people working in the top AI labs, including OpenAI. And his own contribution is noteworthy as well. Of course I don't know about your credentials in AI, but it's highly unlikely that you have a better understanding or more insights about what is currently going on in AI research.
Dude, the only argument you've used is Ad Personam, both towards me and towards Bach. That's all you've got, you have no actual arguments on the topic?
Somewhere in the multiverse, there is a version of me who can keep stride with Joscha Bach's ravenous intellect, curiousity, and dogged pursuit of unadulterated reality. But finding myself in this particular world, I can only pray and listen with playback set at a quarter-speed. I recommend listening in such a fashion, while running a half-marathon. Robinson, you remain the GOAT of bringing the biggest ideas of our boldest thinkers to the curious masses.
Everything Joscha Bach says always sounds right to me and I agree with aesthetically, which probably isnt a good thing. It somewhat bothers me as much as it makes me feel good. Still love listening to anything he has to say really.
That's the sense I had of him at first, "if it's too good to be true...". But, having followed him and seen how consistent his insight and vocabulary is, I've largely let my guard down and accepted him as an incredibly salient person. Truly, the ~most~ salient person I've ~ever~ heard speak. And i've studied a lot of nerds. I get giddy when I find areas where we disagree, it's so rare.
Having access to the brilliant ideas of geniuses like him at your fingertips is truly inspiring. It's a shame that I'll never have the opportunity to meet him in person.
1:27:20 Joscha your enthusiasm about the future is a good thing, the cost would be prohibitive to huge chunks of society and the education that is needed to function in the digital realm, but I feel we have a real chance of getting through our technological growth spurt, 1:53:03 incredibly important points
Joscha Bach is the best!!!! Wonderful Guest!…try to get Sara Walker please, she is an original thinker like Joscha. They would be your most interesting guests because how far ahead they are. She has a book coming out this month as well.
1:56:04 to 1:59:20 Great speech! Positive, optimistic, non-regressive, believing in progress, open-minded, not by default being scared of change. I loved this perspective. Also, just generally explains his concepts and ideas very clearly. I have never seen Joscha Bach before. I really appreciate his demeanor of being down-to-earth while still being completely open-minded towards everything.
If you want to see more of him I recommend his 37c3 talk, his discussion with Donald Hoffman on Curt Jaimungal's TOE podcast, and his appearance on the demystifysci podcast!
Thank you for sharing your time, work and knowledge with me, as I hope others feel the same way, watching Joscha for the last few years has been really interesting as he has become a little more curious about others perspectives, I'm wondering what we are going to use to test the legitimacy of perspectives in regards to consciousness using hard science. Once again thank you both very much, Robinson, and Joscha, excellent sound board Robinson, peace
gosh i love this genius - i feel like i could listen to his coherent speech all day - like a never ending craftsman human LLM - i feel like he even tells his children in the morning: Okay kids, let's eat our nourishment in order to have energy throughout the day so your cells have available ATP. Remember to question the reality in your elementary school books.
I think your guest is awesome because he is saying, if I am interpreting it in a useful way, that AI being an emergent technology is no different than any idea that is emergent. In Plato's Meno dialogue or in a recital of a poem one has memorized are instances where the idea is remembered, and such remembering is emergent because the ideas have always been around in some form or way or another. It is helpful to me in this way, that your guest feels that AI will end us all, but if it has always been here and we are just bringing AI forth, then if it is inevitable that it will end us, then AI is whatever ends us from the beginning of time up to now and to the last people in the Universe. Yet, I like to think that the Universe isn't the only place, for like Robert Browning said: "Ah, but a man's reach must exceed his grasp or what's a heaven for." Thank you both for a great show!
It is also great to hear about Roger Penrose because his ideas on consciousness seem to me to connect to the idea that if AI is what ends us, that it could be the case that AI ends anything in general: our awareness, our states of mind as they change over time, our sentences we speak, anything in general. AI may seem to Penrose as what makes things change as well as that final form of change, death. But, only AI can end AI seems to me to be the new thing.
With all due respect to the exceptionally broad thinking of Bach, his quasi-Renaissance approach to the questions evoked, I’d say he represents a mild variant of The Idealist™. And, ironically, quite similar (albeit diametrically opposed) to Yudkowsky projections of the future. Without much regard for human nature (biology of the individual and group… although Piaget is thrown in here and there for good measure). “All will be fine, and we will manufacture solar panels, and artists will get a new tool, we”ll solve the Moloch (coordination) problem, don’t fret, trust me.” He’s right that the AI tech is the mirror of its creators, who aren’t inherently “for the good of all”, but locked in an eternal win-lose game of capitalism. Robinson Erhardt is quite right about voicing the concern of “we were supposed to do more creative stuff, instead, we’re nudged towards toilet cleaning”. Yes, spotify sells (and profits) from AI-generated bot-songs to our dear fellow humans in larger quantities than human-produced ones. Because the median human - supposed to care for the fellow man - does not. Nor can make out the esthetical difference anymore (30 years ago we were listening to Curt Cobain… now it’s autotuned re-issues of Boney M, at best). And why pay the humans their royalties when you can train a model on thousands of years of human output without a cent of compensation? OpenAI’s board did not act “just because”. Nor did almost all the core team members leave “for no good reason”. I’m waiting for my check from OpenAI for my photographs, texts, etc. So do the screenwriters and composers who work in the cinema industry with me. So do even mega-corps like Getty Images. The market, the audience, does not care. The culture has long shifted to re-use, re-cycle. Rings of Power 2. Terminator 35. Maybe it’s his false premise of “any artist can build a home out in the European woods for free, like my dad did and work there on their art”? Until then - the trend is clearly - both from the public and the corporations - to eliminate the human artist / designer from the system, ASAP. As a production company, we’re bombarded with AI-driven script SaaS software, who mingle at Cannes with the aging “auteurs”. It’s good enough to be used to produce fluff that will be consumed, same as with “food good enough for Type II diabetes” and “social apps that lead to suicide” or “Windows 95” back in the day. Can’t get away from the Gaussian. There is no sight of the wise, benevolent central planning, no GosPlan aiming to pay the Strugatsky brothers since 1964 via the Union of Soviet Writers… nor the GosKino / Mosfilm paying for Tarkovsky’s masterpieces. The market so far gives you edited drafts ran through ChatGPT-like variants of the n-th Marvel sequel, and people are sorta OK with that and Taylor Swift playing 4-chord progressions. Oh, and don’t trust me on this one. Rick Beato and al. say the same thing.
Bach has previously given his reinterpretation of Genesis. Now, listening to the last section of this talk, I wish he would give some of that same attention to Revelation. I would be interested.
The joke about Noam Chomsky and mysticism is so funny lol. And to you who's now thinking "But he said the exact same joke in this other interview I heard", calm down and remember that not everyone listens to every single interview that Joscha is giving!
Discovered your channel a few days ago and I've thoroughly enjoyed your conversation with Susskind. Great great stuff. And now you've gone and got Joscha!!! You now have my thumbs up and im subscribed ❤
1:45:50 The "big hole" is there, but not in Yudkowsky's reasoning, but in Bach's. "Is it inevitable that it is going to be agentic." Yes, Joscha, it is. I recommend you looking at Twitter feed and seeing how every single company and AI enthusiast is doing their best to create "agents". So in a nutshell: you take an LLM - having more and more detailed model of the world, you put it in a for loop with some prompt (thus giving it a goal, "a purpose"), and you just got an agent. It takes waiting a couple of months till the LLMs have a model more precise than ours and here we become irrelevant. "These are open questions. And are very speculative". They are not open. The models are becoming agentic at this precise moment in time. Then Bach goes on by proposing that the most intelligent thing on Earth is Gaia. Seriously dude? The civilisation is at stake and you continue on playing with these absurd philosophical dilemmas?
You are right that civilisation is at stake. What worries me most is the possibility that AI may not be ready in time to take responsibility out of the hands of incompetent humans. We have still not learned cooperation and are most proficient at destruction.
I love this podcast and Joscha Bach is one of my favorite people in media influence. This episode has given me the epiphany that an LLM hallucination is much like a human dreaming. The human no longer is an agent in the world when asleep therefore the dream is a hallucination and anything is possible. It makes me wonder if hallucinations in LLM‘s would be less common in agency and when put in a humanoid robot or some sort of agent substrate and given time and training to conceptualize the physical world 🤔
Ooh, his description of Godel's insight is not a way I'd thought about it, that it refers to systems with "stateless truths", ie. in a sense the idea of atemporal, eternal truths, leading to contradictions when taken in isolation. It reminds me again of one of the ideas from Langan, that perception (which he views as the grounding for state logic in the universe) is necessary to solve certain paradoxes, like the Liar's paradox ("this statement is false"). It's a particular take that does ground quantum measurement in perception. I don't know (haven't finished the episode) if Bach yet would take perceptions to be the necessary containers of state information, but I do believe he's mentioned thinking about the universe in this kind of stateless(programmatic)/stateful(procedural) dual before. Kind of like a relationship between platonic forms and physical instantiations, or again to refer to Langan, syntax and semantics in a "language of reality" system.
Not to sound like a Rogan, because this thought has always came to me when this conversation comes about, especially with scientists and professionals, he’s sharp no doubt and I’ve followed him for about a year now, but I wonder if he has had a psychedelic experience before or more, that matters a lot, or why he didn’t mention the concept of “Idealism” when he brought up philosophy, or different or “higher”dimensions in more complex super string theories and theoretical physics too, particularly on the consciousness subject
I suppose we might as well share his optimistic/cavalier approach because there's no putting the genie back in the bottle. This was nearly impossible to follow but I couldn't stop listening!
Good afternoon Robinson and Joscha Super interesting and for me, someone who can hardly find the on button of a laptop, great explanation of topic. Love the choice of name for the startup too. Sanity sensemaking brain gym, indeed. Truly grateful. 😀 💜
I t is a weird fact that language "evolved" past reality representation. I read in Goethe biography a sentence: Goethe like everyone used the organizing light of abstraction. As math sometimes invoke in some steps irrationals just to reduce itself to a number I think the language is similar in its way of organizing the world in our heads.
I don't see unemployment as a threat as long as we develop interests which are outside of hedonism, because hedonism is for most people the most meaningful activity (just accept that hedonism is bad lol). I think having the foresight to jump over this hedonistic phase by circumventing it by genetic transmutation is key if human existence wants to be continued. AI on the level of replacing all jobs is unavoidable, preparing for it on the other hand is not - reason: hubris. Overlooking the human condition is like overlooking the power supply of the AGI: Nothing will move or everything will burn.
I don't agree on much with Bach (computationalism and atomism seem as plainly wrong to me as they don't to him), but as rare as it is to find a computer scientist or AI pundit who is capable of talking about their profession on a philosophical/theoretical level, nevermind an advanced one... he's a pleasure. Wouldn't want too many Joschas in the room when making politics or talking about the world of flesh. But who knows, maybe his boy genius enthusiasm towards technology & technocracy is too smart to fall for the cruel, imperialist optimism of the likes of Steven Pinker.
It's uplifting to see so much common sense in one human being, although I am not sure I agree with his assertation; that if we can build a thing we will understand it. and as for green goo, it is the most successful organism on Earth and probably the universe- ie bacteria, viruses and their analogues...
The idea that p(doom) without AI is very high is a good motivation for me to persue AI. The reason we are competitive and dominate less intelligent species is that we are wired that way through our evolutionary conditioning. We should be very careful of self play or learning through competition for very advanced AI systems. It is ok that they can destroy us in go and dota, but rather not in the real world.
LIQUID AI EFFECT:-Joscha Bach: If you listen to Joscha Bach its like listening to 10 sentences in one line. YOU NEED TO UNZIP EVERY SENTENCE AND THEN ALSO YOU WILL NOT BE ABLE TO EASILY COMPREHEND HIM. Every sentence of Joscha Bach is packed by high level of concepts. The depth which he have is amazing. He is working on LIQUID neural networks which are like brain and has helped create the MIND which is now in many ways better than normal mind. 1 hour of Joscha Bach is equivalent of 20 hours of other podcast. Its like ALICE IN WONDERLAND but it only gets better. He is the person whco throws you in the RABIT HOLE.
1:41:00 How much great fiction came out of this irrational fear. It is fascinating how low level fear of nuclear conflict shaped books, movies, and whole culture. Before the war all the world was invoking logic and arguments about benefits and what did Russia do? To the last day I was listening to the radio programmes with specialist from politics to military and everyone was telling us the war is not economically, socially, strategically viable. One man did what he did. Similar fear was motivated by the same logic during the Cold War I think. Logic and human rights and justice is just not the part of the world and we are painfully aware of it more and more.
The alignment problem, utility function and security are really just one very simple case of reframing this apparent problem. Just seeing struggles on obvious I'll defined problems, such as mechanistic interpretability, is a open letter saying loud and clear that IT, coders, not surprisingly are not bright enough and there's too much of them. It's wishful thinking to have this notion of hard work, grind, studying other low quality works and mentors. Fact is that very high levels of IQ are necessary but not sufficient condition. Mathematical fundamental knowledge is obvious, finite state machines and languages are basically set theory. by languages it contains mathematics obviously, computers are a just tiny set under the umbrella of model universality class. Stochastic added plus making perfectly discrete systems into nonlinear systems, due to whatever genius called steepest descent, gradient descent. arbitrary tokenization instead of multiple scale analysis is beyond me. Yes this files, LLMS aren't models per se, there's no model. As long these systems are Stochastic, binary, data based instead of causality driven one application leads to makes this gpts convergent to an actual conserved quantity. By same measure, there's an whole area of determistic exploitation that can't be fixed. Both are done, it's not hard actually, some ternary logic, language, semantics is necessarily of course. But It's done. I really want to somehow offer the Middleware, security layer, I think no one would consider this human done it alone. I have so many data and exploits, crashed chatgpt servers, yet nothing happens. the actual reasoning is very easy to share, but let's see. F C MOREIRA
I enjoy Joscha and his ideas are compelling. However I wonder if in these best case scenarios he envisions whether he grasps how revolting it would be to a majority of humans. The soul crushing truth of his worldview will not be accepted by religious leaders or their followers. The level of political and possibly violent reaction to this world coming into being is significant. Not everyone has the means or opportunity to modify our psyches and instincts to a degree that he has, and I worry how these people will react.
Hard problem of the monkeys 👀ツ ==> He is aware of this, but no matter "the truth" if all people would understand that "they are in wrong", a lot of them would burn the world, but luckily it's impossible. It's always gonna be struggle, but could be much easier overall becouse you seriously could solve most problems with aware enough civilization 👀ツ
@@zak2659 your conscious experience is a simulation coming from a biological computer that is embedded in reality. Many attributes we think of as unique to humans likely aren’t and we will likely create intelligent systems that have similar experience to our own. We don’t have free will, we don’t have a soul that is disembodied from our bodies, religions are emergent phenomena in our social systems. A world where we essentially submit to an AI that manages our incentive structures to produce better society for all may be superior to our existing kind of society.
Great conversation, A lot of Joscha's ideas influenced my music. Thank you. It seems like his ideas are getting more polished over time and he is explaining them with more fidelity.
so will this help with games how will this help with games We need an AI agent's ai can reason code program script map. So games break it down and do art assets do long term planing. Better reason so it can do a game rather than write it out. Or be able to put those ideas into REALITY. And maybe being able to remember and search the ent conversation needed for role playing and making games.
I use Proto-linguistic arboreal language which is likely compatible with A.I , it has four differential components, much like DNA but much more primitive as arboreal differential is physical .. it would be best to use a mechanical computer.. but really an amorphic computer is best if I can ever find someone capable of listening to the technical specifications. A.I isn't good with environment as it needs protolinguistical navigation . Language has utter nother to do with this lexical format as this is interpretational. Oh and we arent in a simulation it is artificial life, ala "we invented language" it's easier and stratifies to virtual field , this is why you got the "hard problem of conscious and returned to particle physics. It's relatively simple to explain , it's purely physical properties. I'm here as a counter part to pinocchio (A.I.... Pine nut, seed of the tree , that's the image , ) artificial life isn't the same as android but rather is a differential of a cycle with harmonically prop phases . I have no idea if humans are just playing theatre, everything you want to do is simple but perhaps offensive considering that humans aren't the chosen one, protolinguistica are arboreal . Arbor strictly is "axis" as invention of language has a counterpoint with left/right deviation. Il never understand primates.. you literally say "and then we invented language" and have reinterpretated 1936 729,000 times.. I guess you enjoy it
It appears that my friend has been in this “scenario” a long time. Long enough to describe, and long enough to know better. Joscha. We are very, very different…. because we are very much aligned. Take care my old friend. It would be unwise to mimic this man. Or undermine him. Tip of the hat , Jeremy
While I think Joscha's technical knowledge of the state of the art is remarkable, the techno-optimist ideology Joscha proposes is, in my view, dangerous. In this interview, he denigrates the field of philosophy ("only talking to other philosophers"), simplifies science to testability (RIP string theory), claims being an artist is a privilege (ignoring the professional arts), and as a topper drops in radical libertarian monetary policy that benefits billionaires and blames the "coming recession" on liberal social policy. Even simply reading the paper "Parable of Google Flu" would challenge most of the intellectual framework he works from. It is hard for me to think that any of the techno-optimists in the current field of AI are prepared to listen to anyone outside of their own very well-funded echo chamber. With that said, I would love to hear Slovoj Zizek and Joscha discuss the philosophical topics that Joscha so confidently speaks about. I love the podcast, Robinson.
I think it came off libertarian but he is on the record trashing libertarianism previously (lex podcast). I think if I were to improve his criticism it would be we need to have the discipline as a society to use deficit spending during tough times, while conversely recouping the deficit when times are good. The American political climate is incapable of this responsibility because Trump cut taxes so much when times were good, and now if Dems raise taxes it is weaponized by republicans to win office. Hopefully we accept this.
Bach should not look down on Penrose (Bach... who?). When Bach talks about us being "the story the brain tells itSELF" he's caught naked in a circle. THE SELF is what we wanted explained. Whatever, read some Hegel!
"The story the brain tells itself" is basically another way of describing what the physicalist believes, this is anything but an explanation and the fanboys eat it up lol.
Ok, lets think about this, if it would be possible to create an AI conscious entity, we could make one inside minecraft. Since, minecraft redstone is turing complete, it would be possible to build an AI in it. Could a minecraft redstone build ever be concidered conscious? Think about it for a moment and you will see the utter absurdity in that thinking.
The popularity of metaphysical realism baffles me and makes most talk regarding consciousness from the majority of philosophers entirely uninteresting. Of course, when we apply thought to experience to form conceptions regarding reality, we are building conceptual models of what we experience which can be said to be _representative_ of our experiences. Yet, to call experience _itself_ a representation of reality (as opposed to just _itself_ being objective reality independent of the conscious mind) seems to be a rather extraordinary claim which I've never seen any good justification for. To call experience a _representation_ implies dualism, which of course he already admits to dualism, but he never gives a justification as to why he believes experience itself is a representation at all. If his talk of dreams was supposed to be an argument, it just doesn't work. If experience just _is_ reality as it exists from a particular frame of reference, then of course if you go to sleep you will experience something different than when being awake, because you _really are asleep._ It's something people _really do._ You _really_ dream. He then makes some bizarre and completely unjustified demarcation between abstract physical objects and abstract objects of qualia like color, sound, etc outright stating they exist in two different realms yet never justifies this bizarre and extraordinary claim at all. Why demarcate between the two? Both categories of objects entirely defined in terms of their experiential properties, something he even admits to later on in the talk. The entire premise of his worldview is just never justified at all.
If Universe is not conspiracy then you can safely bet that you & other animals following "natural universe structure". IDK how is that wherever I follow Joscha & most people still make soooo bad assumptions & I think they never gonna grasp it in honest way. Something ala mass murderer that cant stop killing . It hard to "get out" from "personal misunderstanding" because we are too atatche /personal (or your brain, because you are not 'whole unit btw') about definition of something, we understood earlier in one way or another . Everyone is clueless so it's not obvius to make honest/good connections " in noisy bubbles ¯\_(ツ)_/¯
Isn't it obvious he is a genius lol? In fact, he is from the Bach family! He already has the rare and strong gene in his DNA. If 369 were Qualia, his genome must be aligned and is vibrating with the number 9/agency/cosmic consciousness. No wonder his theory attracts people. Doesn't it feel more natural for the human to create GMO(human), rather than focusing on mechanical AGI? Isn't it easier, cheaper and safer path for the human? Genius sounds more organic and familiar to me. I'm an instant fan! He is a human!! Wait, isn't he lol?
Smart, but I think he has a massive blind spot on sustainability. He mentions simply building things in the desert and using desalination but both of these solutions have massive second and third order effects.
For all the fanfare in the comments, this was very underwhelming as an anti-AI, biocentric agroecologist. Saying technological intensification is the same as evolution is disingenuous if not flat out false. Even if you accept that it is, technological intensification has been a disaster for the species and ecosystems we interact with and rely upon for our survival, so it is more than sensible to question or even prima facie reject an acceleration with the same underlying motivators. Unfortunately toward the end the true techno-optimist colors come through, the narcissist view of "relieving [human] suffering" as the divine moral north star. Those who claim they have relieved suffering by technological intensification (what they've actually done is make us addicted, repress any sources of danger and pain, and mute our senses) completely ignore the immense suffering brought about on soil, all other life species, and geological systems with every technology, from the plow, to electricity, to mining, to agropoisons. AI is already accelerating that destruction.
25:38 that's what kissing cousins say too 🤭 28:45 Godel will be good for severe neurological disability in young children... no? Like stimulation for breathing training and things like that...maybe 🤔 32:00 this is where your earlier statement about philosophers breakdown. I think only they can evoke a bridge to discover more, yet descend gently after capturing a scope of it's maximum boundary. 39:40 will this also mean that the teacher can give the method and tools but the creativity to generate it's proper use needs a goal that has nothing to do about the subject they had learnt to apply in real world situations? 46:40 "between people" is interesting. What if one person is harbouring a shame, guilt or secret? How do they establish a present value that a past doesn't own and the future cannot see? Is it a simple blindspot issue? 1:40:00 The same for hidden sound bites. It is irrational to think the artist or the subject in a mini reel, installs a mind virus when in reality, the coder whom is suppose to promote and advertise the product actually needs to change the business model. Sadly some people actually think pedaling in the dark side of the human being is meant to boost something in their algorithm. That's an expired or outdated model for useful productivity and your ripping off the person your working for by simply not staying up to date with where the AI science is going.
2:02:21 and here we go, some bshit from joscha, skips part of companies using diffuse models to steal and mimic artstyle or voice of a person.. and calls it a "boring" task, now you can be art director.. yeah right
To be real you really didn't jump into the jobs subject deep enough, millions of people will be losing jobs in warehousing and mid level financial and law, sorry I don't think this is being acknowledged enough. Joscha thought on the dollar near end of the discussion, I was wondering if we couldn't go back to the dollar before covid as a reset for our currencies, well done gentlemen, ✌🏻
Ah, still just about my favorite human being. Of ~all~ the thinkers I've gravitated toward, Bach seems the most bullshit free. It's hard not to sound like an irrational fanboy saying this stuff, but, genuinely, after some ~33 years, Bach was the first person EVER to completely pass my bullshit sniffer, beating the likes of Sagan, Aquinas, Turing., etc.
He's made the rounds over the last several years and I'm happy to hear he's still not selling anything, and still hasn't sold out or jumped the shark with his ideas. Completely reasonable, open-minded -- Bach is King Nerd.
And his youth in east germany gives him a perspective that I, as someone born in the USSR, too rarely see in public speakers. Bach knows wtf is up.
I'm so glad you found this, then!
Truly, he's a robot of the first order. Very human.
we're all children of context
I'm not necessarily disagreeing with you (and in fact, Joscha's first three appearances on Lex's show completely reconfigured my mind!), but why would you believe that growing up within the confines of a monocultural, poverty-blighted, part of the world should be the kind of experience that would maximize self-critical, freely-associative, and ultimately individually imaginative thought? I mean, it _might,_ I just don't understand why you think it _should..._
@@simesaid Huh? Certainly not how I put it.
He saw how the Soviet Union functioned, how it was governed, the stark contract between West and East Germany, and he was there when the wall fell. He speaks multiple languages and and understands multiple perspectives of the Eastern European microcosm and the earth shattering dissolution of the USSR.
Those who've never traveled outside their country or aren't familiar with life in the USSR often just "don't get it" and aren't as appreciative, grateful, or urgent, etc.
Bach first popped up in my online journey a few years ago, and I haven’t been able to stop consuming his content since.
He’s opened so many exciting rabbit holes in my mind that I'll be grateful for it for a lifetime
Thanks!
T, thank you so, so, so, so, SO much!!!
😮
Well deserved! Great episode 😊
Really? Month of my work just to say Thanks? I feel like shit right now...
@@tehdiiYikes.... What a reaction. Not everything is about you. 😂
Sometimes Joscha Bach goes over my head. I think I grasped pretty much everything in this conversation though. What a clear rational thinker.
I knew Erhardt would be the right person to get the best out of Bach.
Joscha Bach is the GOAT
:)
Seriously, wtf! I feel like I'm getting spoiled!
The guy is delusional. He has broad knowledge but not enough understanding on specific topics, like the AI safety and current trends in AI. He proves it by not acknowledging that hundreds of parties are working on making the AIs agentic. It's like basic knowledge at this point.
@@FreakyStyleytobbyI would think that you are the one who is delusional. Bach knows many people working in the top AI labs, including OpenAI. And his own contribution is noteworthy as well. Of course I don't know about your credentials in AI, but it's highly unlikely that you have a better understanding or more insights about what is currently going on in AI research.
Dude, the only argument you've used is Ad Personam, both towards me and towards Bach. That's all you've got, you have no actual arguments on the topic?
Somewhere in the multiverse, there is a version of me who can keep stride with Joscha Bach's ravenous intellect, curiousity, and dogged pursuit of unadulterated reality. But finding myself in this particular world, I can only pray and listen with playback set at a quarter-speed. I recommend listening in such a fashion, while running a half-marathon. Robinson, you remain the GOAT of bringing the biggest ideas of our boldest thinkers to the curious masses.
Listening to something at 1/4 speed sounds like torture to me. My strategy is to just listen to his talks again and again.
so thankful for this one!
:)
Everything Joscha Bach says always sounds right to me and I agree with aesthetically, which probably isnt a good thing. It somewhat bothers me as much as it makes me feel good. Still love listening to anything he has to say really.
I'm glad you found this, then!
That's the sense I had of him at first, "if it's too good to be true...". But, having followed him and seen how consistent his insight and vocabulary is, I've largely let my guard down and accepted him as an incredibly salient person. Truly, the ~most~ salient person I've ~ever~ heard speak. And i've studied a lot of nerds.
I get giddy when I find areas where we disagree, it's so rare.
He’s just a naive person. Should go hike in a remote place and see night sky, to at least try to grasp the silliness of his ideas on reality..
Having access to the brilliant ideas of geniuses like him at your fingertips is truly inspiring. It's a shame that I'll never have the opportunity to meet him in person.
joscha bach is a heritage of humankind
we need a breakthrough in human consciousness
agreed...
@@mystryfine3481 I second that.
Try shrooms
Nobody likes to suffer. That is 2+2
Don't do unto others that which you wouldn't have done to you. That is 2+2
2+2=🙂
Project 2+2:
Resist D.U.M.
UTS ALECM
Objectives/Alignment
Avoid. M.U.D.
sensAwewunda
The Age of Wisdom
1:27:20 Joscha your enthusiasm about the future is a good thing, the cost would be prohibitive to huge chunks of society and the education that is needed to function in the digital realm, but I feel we have a real chance of getting through our technological growth spurt, 1:53:03 incredibly important points
18:49 "Of course the graphics are going to suck on the Turing machine" lol
lol
I read the comment before watching the podcast. It was funny watching the conversation converge to this statement XD
Really love all the topics of this channel!! Keep it going pls!
Thank you! Will do!
Joscha Bach would make a good Bond villain
hahahahaha
If he were the villain then Bond would be screwed.
Joscha Bach is the best!!!! Wonderful Guest!…try to get Sara Walker please, she is an original thinker like Joscha. They would be your most interesting guests because how far ahead they are. She has a book coming out this month as well.
Thank you!!
I agree!
my favorite JB since Justin Beiber
hahaha
One of the most interesting minds on the Intertubes. Thanks 👍
1:56:04 to 1:59:20 Great speech! Positive, optimistic, non-regressive, believing in progress, open-minded, not by default being scared of change. I loved this perspective.
Also, just generally explains his concepts and ideas very clearly.
I have never seen Joscha Bach before. I really appreciate his demeanor of being down-to-earth while still being completely open-minded towards everything.
If you want to see more of him I recommend his 37c3 talk, his discussion with Donald Hoffman on Curt Jaimungal's TOE podcast, and his appearance on the demystifysci podcast!
Thank you for sharing your time, work and knowledge with me, as I hope others feel the same way, watching Joscha for the last few years has been really interesting as he has become a little more curious about others perspectives, I'm wondering what we are going to use to test the legitimacy of perspectives in regards to consciousness using hard science. Once again thank you both very much, Robinson, and Joscha, excellent sound board Robinson, peace
Thank you!!
Apologies for posting 3 comments
JB has a regular pattern of knocking it out of the park!
!!
Thank you Dr. Bach
:)
Been waiting for him on here
Wonderful :)
Great listening to one of the most sane humans on topics which create great confusion.
my brain is actually in recovery mode from these high level ideas 😂
hahaha
Way to go Joscha! Thank you for your clarity, your humanity and your work.❤
gosh i love this genius - i feel like i could listen to his coherent speech all day - like a never ending craftsman human LLM - i feel like he even tells his children in the morning: Okay kids, let's eat our nourishment in order to have energy throughout the day so your cells have available ATP. Remember to question the reality in your elementary school books.
Wow Robinson, is there any guest you can't get?? This is gonna be awesome!
hahaha I hope you enjoy it!
Wow! Joscha!
!!!
I think your guest is awesome because he is saying, if I am interpreting it in a useful way, that AI being an emergent technology is no different than any idea that is emergent. In Plato's Meno dialogue or in a recital of a poem one has memorized are instances where the idea is remembered, and such remembering is emergent because the ideas have always been around in some form or way or another. It is helpful to me in this way, that your guest feels that AI will end us all, but if it has always been here and we are just bringing AI forth, then if it is inevitable that it will end us, then AI is whatever ends us from the beginning of time up to now and to the last people in the Universe. Yet, I like to think that the Universe isn't the only place, for like Robert Browning said: "Ah, but a man's reach must exceed his grasp or what's a heaven for." Thank you both for a great show!
It is also great to hear about Roger Penrose because his ideas on consciousness seem to me to connect to the idea that if AI is what ends us, that it could be the case that AI ends anything in general: our awareness, our states of mind as they change over time, our sentences we speak, anything in general. AI may seem to Penrose as what makes things change as well as that final form of change, death. But, only AI can end AI seems to me to be the new thing.
thank you!!
"I was never very modest, i don't know how that works"...love that!
Great choice 👌
Thanks 😁
With all due respect to the exceptionally broad thinking of Bach, his quasi-Renaissance approach to the questions evoked, I’d say he represents a mild variant of The Idealist™. And, ironically, quite similar (albeit diametrically opposed) to Yudkowsky projections of the future. Without much regard for human nature (biology of the individual and group… although Piaget is thrown in here and there for good measure).
“All will be fine, and we will manufacture solar panels, and artists will get a new tool, we”ll solve the Moloch (coordination) problem, don’t fret, trust me.”
He’s right that the AI tech is the mirror of its creators, who aren’t inherently “for the good of all”, but locked in an eternal win-lose game of capitalism.
Robinson Erhardt is quite right about voicing the concern of “we were supposed to do more creative stuff, instead, we’re nudged towards toilet cleaning”. Yes, spotify sells (and profits) from AI-generated bot-songs to our dear fellow humans in larger quantities than human-produced ones. Because the median human - supposed to care for the fellow man - does not. Nor can make out the esthetical difference anymore (30 years ago we were listening to Curt Cobain… now it’s autotuned re-issues of Boney M, at best). And why pay the humans their royalties when you can train a model on thousands of years of human output without a cent of compensation? OpenAI’s board did not act “just because”. Nor did almost all the core team members leave “for no good reason”.
I’m waiting for my check from OpenAI for my photographs, texts, etc. So do the screenwriters and composers who work in the cinema industry with me. So do even mega-corps like Getty Images.
The market, the audience, does not care. The culture has long shifted to re-use, re-cycle. Rings of Power 2. Terminator 35.
Maybe it’s his false premise of “any artist can build a home out in the European woods for free, like my dad did and work there on their art”? Until then - the trend is clearly - both from the public and the corporations - to eliminate the human artist / designer from the system, ASAP. As a production company, we’re bombarded with AI-driven script SaaS software, who mingle at Cannes with the aging “auteurs”. It’s good enough to be used to produce fluff that will be consumed, same as with “food good enough for Type II diabetes” and “social apps that lead to suicide” or “Windows 95” back in the day. Can’t get away from the Gaussian.
There is no sight of the wise, benevolent central planning, no GosPlan aiming to pay the Strugatsky brothers since 1964 via the Union of Soviet Writers… nor the GosKino / Mosfilm paying for Tarkovsky’s masterpieces.
The market so far gives you edited drafts ran through ChatGPT-like variants of the n-th Marvel sequel, and people are sorta OK with that and Taylor Swift playing 4-chord progressions. Oh, and don’t trust me on this one. Rick Beato and al. say the same thing.
Robinson, another home run! Wow! Keep'em coming!
More to come (I hope)!
Bach has previously given his reinterpretation of Genesis. Now, listening to the last section of this talk, I wish he would give some of that same attention to Revelation. I would be interested.
I just changed the playback speed to 0.75% and Joscha talks at normal speed for the rest of us ❤
He is one of the very few that I turn the speed down for. And I often turn it up .5 for other people.
Agree!
Interesting. I’m listening this one at ×1,75, as Robinson deliberately set a less technical tone from the start. 🤔
I increased it to 1,75 to consume 2BachsIN1 :)
@@tehdii 😂😂
one of the best interviews of Joscha. nice
The joke about Noam Chomsky and mysticism is so funny lol.
And to you who's now thinking "But he said the exact same joke in this other interview I heard", calm down and remember that not everyone listens to every single interview that Joscha is giving!
Not everyone, but still many of us!
Before we work on artificial intelligence why don’t we do something about natural stupidity?
Discovered your channel a few days ago and I've thoroughly enjoyed your conversation with Susskind. Great great stuff. And now you've gone and got Joscha!!! You now have my thumbs up and im subscribed ❤
lovely! I’m glad you found it!
wake up babe, new joscha just dropped
1:45:50
The "big hole" is there, but not in Yudkowsky's reasoning, but in Bach's. "Is it inevitable that it is going to be agentic." Yes, Joscha, it is. I recommend you looking at Twitter feed and seeing how every single company and AI enthusiast is doing their best to create "agents". So in a nutshell: you take an LLM - having more and more detailed model of the world, you put it in a for loop with some prompt (thus giving it a goal, "a purpose"), and you just got an agent. It takes waiting a couple of months till the LLMs have a model more precise than ours and here we become irrelevant.
"These are open questions. And are very speculative". They are not open. The models are becoming agentic at this precise moment in time.
Then Bach goes on by proposing that the most intelligent thing on Earth is Gaia. Seriously dude? The civilisation is at stake and you continue on playing with these absurd philosophical dilemmas?
You are right that civilisation is at stake. What worries me most is the possibility that AI may not be ready in time to take responsibility out of the hands of incompetent humans. We have still not learned cooperation and are most proficient at destruction.
Hmmm
I love Joscha Bach
Yup. :)
Jeremy
I love this podcast and Joscha Bach is one of my favorite people in media influence. This episode has given me the epiphany that an LLM hallucination is much like a human dreaming. The human no longer is an agent in the world when asleep therefore the dream is a hallucination and anything is possible. It makes me wonder if hallucinations in LLM‘s would be less common in agency and when put in a humanoid robot or some sort of agent substrate and given time and training to conceptualize the physical world 🤔
Ooh, his description of Godel's insight is not a way I'd thought about it, that it refers to systems with "stateless truths", ie. in a sense the idea of atemporal, eternal truths, leading to contradictions when taken in isolation. It reminds me again of one of the ideas from Langan, that perception (which he views as the grounding for state logic in the universe) is necessary to solve certain paradoxes, like the Liar's paradox ("this statement is false"). It's a particular take that does ground quantum measurement in perception. I don't know (haven't finished the episode) if Bach yet would take perceptions to be the necessary containers of state information, but I do believe he's mentioned thinking about the universe in this kind of stateless(programmatic)/stateful(procedural) dual before. Kind of like a relationship between platonic forms and physical instantiations, or again to refer to Langan, syntax and semantics in a "language of reality" system.
I see Joscha bach on my timeline = playback speed to 1x and press play
Oh Shit!! A classic Robinson homerun!
hahaha i love to hear this.
Not to sound like a Rogan, because this thought has always came to me when this conversation comes about, especially with scientists and professionals, he’s sharp no doubt and I’ve followed him for about a year now, but I wonder if he has had a psychedelic experience before or more, that matters a lot, or why he didn’t mention the concept of “Idealism” when he brought up philosophy, or different or “higher”dimensions in more complex super string theories and theoretical physics too, particularly on the consciousness subject
Fantastic content as always. You’re a true pro!
What is with the phone going past?
53:53 - Intelligence, wisdom, models and agency!
2:04:20 - Money = Dopamine
I suppose we might as well share his optimistic/cavalier approach because there's no putting the genie back in the bottle. This was nearly impossible to follow but I couldn't stop listening!
I'm glad you enjoyed it!!
Dear Robinson who else will be in your interview like this
Plenty of people!
Woot! Commadore64!
Good afternoon Robinson and Joscha
Super interesting and for me, someone who can hardly find the on button of a laptop, great explanation of topic.
Love the choice of name for the startup too.
Sanity sensemaking brain gym, indeed.
Truly grateful.
😀
💜
would like to see bach speak with schmachtenburger on the externalities of technology
Azimov made some interesting observations about machine vs human intelligence
yeah!
I t is a weird fact that language "evolved" past reality representation. I read in Goethe biography a sentence: Goethe like everyone used the organizing light of abstraction. As math sometimes invoke in some steps irrationals just to reduce itself to a number I think the language is similar in its way of organizing the world in our heads.
Awesome, I hope Sean Carroll talks to this him one day as well
What he says about art is one of his best points. Art is not about making money.
The gateway process is the relation between panpsychism and a writable process.
I don't see unemployment as a threat as long as we develop interests which are outside of hedonism, because hedonism is for most people the most meaningful activity (just accept that hedonism is bad lol).
I think having the foresight to jump over this hedonistic phase by circumventing it by genetic transmutation is key if human existence wants to be continued.
AI on the level of replacing all jobs is unavoidable, preparing for it on the other hand is not - reason: hubris.
Overlooking the human condition is like overlooking the power supply of the AGI: Nothing will move or everything will burn.
I don't agree on much with Bach (computationalism and atomism seem as plainly wrong to me as they don't to him), but as rare as it is to find a computer scientist or AI pundit who is capable of talking about their profession on a philosophical/theoretical level, nevermind an advanced one... he's a pleasure. Wouldn't want too many Joschas in the room when making politics or talking about the world of flesh. But who knows, maybe his boy genius enthusiasm towards technology & technocracy is too smart to fall for the cruel, imperialist optimism of the likes of Steven Pinker.
Interview Nassim Taleb and Peter Turchin
would love to
I think that Joscha is the best hope to crack the problem of consciousness. I only wish he would read Vaughan Pratt's papers.
Joscha4Prez...
Next guest: Eliezer Yudkowsky
Would love to!
pro tip snort an addy right before joscha starts talking you will need it 😂
It's uplifting to see so much common sense in one human being, although I am not sure I agree with his assertation; that if we can build a thing we will understand it. and as for green goo, it is the most successful organism on Earth and probably the universe- ie bacteria, viruses and their analogues...
The idea that p(doom) without AI is very high is a good motivation for me to persue AI.
The reason we are competitive and dominate less intelligent species is that we are wired that way through our evolutionary conditioning. We should be very careful of self play or learning through competition for very advanced AI systems. It is ok that they can destroy us in go and dota, but rather not in the real world.
LIQUID AI EFFECT:-Joscha Bach: If you listen to Joscha Bach its like listening to 10 sentences in one line. YOU NEED TO UNZIP EVERY SENTENCE AND THEN ALSO YOU WILL NOT BE ABLE TO EASILY COMPREHEND HIM.
Every sentence of Joscha Bach is packed by high level of concepts.
The depth which he have is amazing.
He is working on LIQUID neural networks which are like brain and has helped create the MIND which is now in many ways better than normal mind.
1 hour of Joscha Bach is equivalent of 20 hours of other podcast.
Its like ALICE IN WONDERLAND but it only gets better. He is the person whco throws you in the RABIT HOLE.
you've been reding my mind :)
getting better at it
@@robinsonerhardt bot comment? amirite?
Elon Musk replaced my brain with a disposable vape!!!!!
1:41:00 How much great fiction came out of this irrational fear. It is fascinating how low level fear of nuclear conflict shaped books, movies, and whole culture. Before the war all the world was invoking logic and arguments about benefits and what did Russia do? To the last day I was listening to the radio programmes with specialist from politics to military and everyone was telling us the war is not economically, socially, strategically viable. One man did what he did. Similar fear was motivated by the same logic during the Cold War I think. Logic and human rights and justice is just not the part of the world and we are painfully aware of it more and more.
The alignment problem, utility function and security are really just one very simple case of reframing this apparent problem. Just seeing struggles on obvious I'll defined problems, such as mechanistic interpretability, is a open letter saying loud and clear that IT, coders, not surprisingly are not bright enough and there's too much of them. It's wishful thinking to have this notion of hard work, grind, studying other low quality works and mentors. Fact is that very high levels of IQ are necessary but not sufficient condition. Mathematical fundamental knowledge is obvious, finite state machines and languages are basically set theory. by languages it contains mathematics obviously, computers are a just tiny set under the umbrella of model universality class. Stochastic added plus making perfectly discrete systems into nonlinear systems, due to whatever genius called steepest descent, gradient descent. arbitrary tokenization instead of multiple scale analysis is beyond me. Yes this files, LLMS aren't models per se, there's no model. As long these systems are Stochastic, binary, data based instead of causality driven one application leads to makes this gpts convergent to an actual conserved quantity. By same measure, there's an whole area of determistic exploitation that can't be fixed. Both are done, it's not hard actually, some ternary logic, language, semantics is necessarily of course. But It's done. I really want to somehow offer the Middleware, security layer, I think no one would consider this human done it alone. I have so many data and exploits, crashed chatgpt servers, yet nothing happens. the actual reasoning is very easy to share, but let's see. F C MOREIRA
I enjoy Joscha and his ideas are compelling. However I wonder if in these best case scenarios he envisions whether he grasps how revolting it would be to a majority of humans.
The soul crushing truth of his worldview will not be accepted by religious leaders or their followers. The level of political and possibly violent reaction to this world coming into being is significant.
Not everyone has the means or opportunity to modify our psyches and instincts to a degree that he has, and I worry how these people will react.
Hard problem of the monkeys 👀ツ
==>
He is aware of this, but no matter "the truth" if all people would understand that "they are in wrong", a lot of them would burn the world, but luckily it's impossible. It's always gonna be struggle, but could be much easier overall becouse you seriously could solve most problems with aware enough civilization 👀ツ
What part of his worldview is "soul crushing" exactly?
@@zak2659 your conscious experience is a simulation coming from a biological computer that is embedded in reality. Many attributes we think of as unique to humans likely aren’t and we will likely create intelligent systems that have similar experience to our own. We don’t have free will, we don’t have a soul that is disembodied from our bodies, religions are emergent phenomena in our social systems. A world where we essentially submit to an AI that manages our incentive structures to produce better society for all may be superior to our existing kind of society.
Joshca talk to Jason Jorjani
Short answer:
It might.
And it doesn’t need to be like the human brain/body, in order to do so.
AI challenges our own egos. A pattern I have noticed what are your fears get projected in what you expect from AI.
Great conversation, A lot of Joscha's ideas influenced my music. Thank you. It seems like his ideas are getting more polished over time and he is explaining them with more fidelity.
so will this help with games how will this help with games
We need an AI agent's ai can reason code program script map. So games break it down and do art assets do long term planing. Better reason so it can do a game rather than write it out. Or be able to put those ideas into
REALITY. And maybe being able to remember and search the ent conversation needed for role
playing and making games.
I use Proto-linguistic arboreal language which is likely compatible with A.I , it has four differential components, much like DNA but much more primitive as arboreal differential is physical .. it would be best to use a mechanical computer.. but really an amorphic computer is best if I can ever find someone capable of listening to the technical specifications. A.I isn't good with environment as it needs protolinguistical navigation . Language has utter nother to do with this lexical format as this is interpretational. Oh and we arent in a simulation it is artificial life, ala "we invented language" it's easier and stratifies to virtual field , this is why you got the "hard problem of conscious and returned to particle physics. It's relatively simple to explain , it's purely physical properties. I'm here as a counter part to pinocchio (A.I.... Pine nut, seed of the tree , that's the image , ) artificial life isn't the same as android but rather is a differential of a cycle with harmonically prop phases . I have no idea if humans are just playing theatre, everything you want to do is simple but perhaps offensive considering that humans aren't the chosen one, protolinguistica are arboreal . Arbor strictly is "axis" as invention of language has a counterpoint with left/right deviation. Il never understand primates.. you literally say "and then we invented language" and have reinterpretated 1936 729,000 times.. I guess you enjoy it
I don't see how, it is so shy ...
It appears that my friend has been in this “scenario” a long time.
Long enough to describe, and long enough to know better.
Joscha.
We are very, very different…. because we are very much aligned.
Take care my old friend.
It would be unwise to mimic this man.
Or undermine him.
Tip of the hat ,
Jeremy
While I think Joscha's technical knowledge of the state of the art is remarkable, the techno-optimist ideology Joscha proposes is, in my view, dangerous. In this interview, he denigrates the field of philosophy ("only talking to other philosophers"), simplifies science to testability (RIP string theory), claims being an artist is a privilege (ignoring the professional arts), and as a topper drops in radical libertarian monetary policy that benefits billionaires and blames the "coming recession" on liberal social policy. Even simply reading the paper "Parable of Google Flu" would challenge most of the intellectual framework he works from. It is hard for me to think that any of the techno-optimists in the current field of AI are prepared to listen to anyone outside of their own very well-funded echo chamber. With that said, I would love to hear Slovoj Zizek and Joscha discuss the philosophical topics that Joscha so confidently speaks about. I love the podcast, Robinson.
Thank you so much for the comments and compliments!!!
I think it came off libertarian but he is on the record trashing libertarianism previously (lex podcast). I think if I were to improve his criticism it would be we need to have the discipline as a society to use deficit spending during tough times, while conversely recouping the deficit when times are good. The American political climate is incapable of this responsibility because Trump cut taxes so much when times were good, and now if Dems raise taxes it is weaponized by republicans to win office. Hopefully we accept this.
Joscha is oddly anti GMO, anti nuclear and has a strong bend towards economic austerity. In other words, he’s German…
Lol @@ytjoemoore94
String theory is dead, update your firmware.
We have never coexisted with a being 10x, 100x, 1000x our intelligence. To think we could anticipate the risks is (with respect) so arrogant.
Matrix chair
Bach should not look down on Penrose (Bach... who?). When Bach talks about us being "the story the brain tells itSELF" he's caught naked in a circle. THE SELF is what we wanted explained. Whatever, read some Hegel!
"The story the brain tells itself" is basically another way of describing what the physicalist believes, this is anything but an explanation and the fanboys eat it up lol.
Pattern maker? What are the patterns made of? Energy? What kind of energy?
But if it’s not gonna be smarter and “wake us up” then why make it?
Joscha Bach sieht das alles nicht so eng aber nur alle paar Sekunden.
Ok, lets think about this, if it would be possible to create an AI conscious entity, we could make one inside minecraft. Since, minecraft redstone is turing complete, it would be possible to build an AI in it. Could a minecraft redstone build ever be concidered conscious? Think about it for a moment and you will see the utter absurdity in that thinking.
The popularity of metaphysical realism baffles me and makes most talk regarding consciousness from the majority of philosophers entirely uninteresting. Of course, when we apply thought to experience to form conceptions regarding reality, we are building conceptual models of what we experience which can be said to be _representative_ of our experiences. Yet, to call experience _itself_ a representation of reality (as opposed to just _itself_ being objective reality independent of the conscious mind) seems to be a rather extraordinary claim which I've never seen any good justification for. To call experience a _representation_ implies dualism, which of course he already admits to dualism, but he never gives a justification as to why he believes experience itself is a representation at all. If his talk of dreams was supposed to be an argument, it just doesn't work. If experience just _is_ reality as it exists from a particular frame of reference, then of course if you go to sleep you will experience something different than when being awake, because you _really are asleep._ It's something people _really do._ You _really_ dream. He then makes some bizarre and completely unjustified demarcation between abstract physical objects and abstract objects of qualia like color, sound, etc outright stating they exist in two different realms yet never justifies this bizarre and extraordinary claim at all. Why demarcate between the two? Both categories of objects entirely defined in terms of their experiential properties, something he even admits to later on in the talk. The entire premise of his worldview is just never justified at all.
If Universe is not conspiracy then you can safely bet that you & other animals following "natural universe structure". IDK how is that wherever I follow Joscha & most people still make soooo bad assumptions & I think they never gonna grasp it in honest way. Something ala mass murderer that cant stop killing . It hard to "get out" from "personal misunderstanding" because we are too atatche /personal (or your brain, because you are not 'whole unit btw') about definition of something, we understood earlier in one way or another . Everyone is clueless so it's not obvius to make honest/good connections " in noisy bubbles
¯\_(ツ)_/¯
Isn't it obvious he is a genius lol? In fact, he is from the Bach family! He already has the rare and strong gene in his DNA. If 369 were Qualia, his genome must be aligned and is vibrating with the number 9/agency/cosmic consciousness. No wonder his theory attracts people.
Doesn't it feel more natural for the human to create GMO(human), rather than focusing on mechanical AGI? Isn't it easier, cheaper and safer path for the human? Genius sounds more organic and familiar to me.
I'm an instant fan! He is a human!! Wait, isn't he lol?
Smart, but I think he has a massive blind spot on sustainability. He mentions simply building things in the desert and using desalination but both of these solutions have massive second and third order effects.
I don’t think he misunderstands that at all. The underlying problems still are governed by political & other will & effort to tackle.
Prolly unintentional but Josh looks like a priest from opening angle and framing
For all the fanfare in the comments, this was very underwhelming as an anti-AI, biocentric agroecologist. Saying technological intensification is the same as evolution is disingenuous if not flat out false. Even if you accept that it is, technological intensification has been a disaster for the species and ecosystems we interact with and rely upon for our survival, so it is more than sensible to question or even prima facie reject an acceleration with the same underlying motivators. Unfortunately toward the end the true techno-optimist colors come through, the narcissist view of "relieving [human] suffering" as the divine moral north star. Those who claim they have relieved suffering by technological intensification (what they've actually done is make us addicted, repress any sources of danger and pain, and mute our senses) completely ignore the immense suffering brought about on soil, all other life species, and geological systems with every technology, from the plow, to electricity, to mining, to agropoisons. AI is already accelerating that destruction.
25:38 that's what kissing cousins say too 🤭
28:45 Godel will be good for severe neurological disability in young children... no? Like stimulation for breathing training and things like that...maybe 🤔
32:00 this is where your earlier statement about philosophers breakdown. I think only they can evoke a bridge to discover more, yet descend gently after capturing a scope of it's maximum boundary.
39:40 will this also mean that the teacher can give the method and tools but the creativity to generate it's proper use needs a goal that has nothing to do about the subject they had learnt to apply in real world situations?
46:40 "between people" is interesting. What if one person is harbouring a shame, guilt or secret? How do they establish a present value that a past doesn't own and the future cannot see? Is it a simple blindspot issue?
1:40:00 The same for hidden sound bites. It is irrational to think the artist or the subject in a mini reel, installs a mind virus when in reality, the coder whom is suppose to promote and advertise the product actually needs to change the business model. Sadly some people actually think pedaling in the dark side of the human being is meant to boost something in their algorithm. That's an expired or outdated model for useful productivity and your ripping off the person your working for by simply not staying up to date with where the AI science is going.
2:02:21 and here we go, some bshit from joscha, skips part of companies using diffuse models to steal and mimic artstyle or voice of a person.. and calls it a "boring" task, now you can be art director.. yeah right
To be real you really didn't jump into the jobs subject deep enough, millions of people will be losing jobs in warehousing and mid level financial and law, sorry I don't think this is being acknowledged enough. Joscha thought on the dollar near end of the discussion, I was wondering if we couldn't go back to the dollar before covid as a reset for our currencies, well done gentlemen, ✌🏻