"We Are All Software" - Joscha Bach

[@MachineLearningStreetTalk]

Interview with with Joscha to follow soon, are you ready? 🤩

[@optimaiz]

Let's goo! 🤯

[@nokar999]

0.75x speed activated lfg!

[@drmedwuast]

@@nokar999 hahah

[@manslaughterinc.9135]

I'm glad Jocha is opening the overton window on animism and machine consciousness. I normally don't watch your channels because the perspective here generally rejects these concepts outright. I get it, a lot of people have woo perspectives, but some of us actually take this seriously. It's nice to see when someone doesn't get shouted down immediately.

[@CodexPermutatio]

Stop teasing us and release it now! :]

[@TooManyPartsToCount]

Joscha Bach is one of the most coherent thinkers in the cognitive scientific community. Thanks for posting MLST!

[@CodexPermutatio]

Indeed he is.

[@michaelli7000]

Agree

[@rockapedra1130]

Joscha has an amazing mind. It is both original and integrating of the thoughts of others. Every two years or so, he comes up with a fresh view on some subject that synthesizes and elucidates some thorny problem. I look forward to these "phase changes" in his thinking. It's a bit Iike a rollercoaster ride!

[@oncedidactic]

Great description, agree!

[@SimplyChrisRLP]

Very agreed!

[@spectator5144]

fantastic comment

[@hmind9836]

I should be working right now, but Joscha Bach is just on another level. Whenever I see new content from him, I feel like I need to stop everything I'm doing and just bask in the brilliance of his mind, like when you witness a natural phenomenon and involuntarily stand in silent awe for a few minutes. Thank you for this! You've made me a very happy human today!

[@masterofkaarsvet]

Beautifully phrased 👌🏼

[@jodawgsup]

ew

[@mbrochh82]

Here's a ChatGPT summary: - Joscha Bach felt the arrival of AGI (Artificial General Intelligence) was imminent from a young age, but acknowledges the uncertainty of its timeline. - He discusses the concept of "mimesis" and its hierarchy, from language to evolution, and how software agents (or spirits) compete and optimize. - Bach suggests that what our ancestors called "spirit" is akin to self-organizing software agents, which organize cells into bodies and produce complex behaviors. - He realized that living nature is fundamentally about software, control, and causal structures rather than physical molecules. - Animism, the belief in spirits governing nature, was prevalent in Europe before Christianity and still exists in some cultures. - Bach has become an animist, believing that consciousness and spirits are self-organizing software agents. - Bach questions whether existing AI models are too large or too small and suggests focusing on minimal models that can learn and infer. - He is involved with Liquid AI, a startup aiming to build more fluid, continuously learning AI models. - Bach references Aristotle's concepts of the soul and intellect, suggesting they align with modern ideas of self-organizing software. - He explores the idea that consciousness is a second-order perception, always happening in the present, and acts as an operator to increase coherence in the mind. - Bach compares current AI design (outside-in) with natural self-organizing systems (inside-out) and suggests AI should emulate the latter. - He discusses the potential for AI to develop consciousness and the ethical implications of AGI recognizing human consciousness. - Bach speculates on the possibility of plant consciousness and inter-plant communication, drawing parallels with animist beliefs. - He proposes that evolution is about the competition of software agents, not just physical organisms. - Bach suggests that AI could extend life and consciousness onto new substrates, potentially leading to a naturalization of the mind. - He emphasizes the importance of studying consciousness through AI models and proposes building a trainable substrate of modular reinforcement learning agents. - Bach acknowledges the speculative nature of his ideas and the need for further research and simulations to validate them. - Main message: Joscha Bach explores the idea that consciousness and life are fundamentally about self-organizing software agents, suggesting that both biological and artificial systems can exhibit these properties, and emphasizes the importance of studying and developing AI to understand and extend consciousness.

[@johnharris3118]

Tom Campbell My Big Toe is on same perspective

[@duffy666]

"Bubble of nowness" - love it

[@egor.okhterov]

This talk is incredible! A breeze of fresh air.

[@duffy666]

My consciousness needs to watch this 10 times to understand 10% of what he is saying ;)

[@martinmo2789]

It's amazing isn't it? I can follow a model a mind produced so much more competent than my own and still can use it to improve my mind.

[@TooManyPartsToCount]

I cannot recommend enough that if you want to get 'up to speed' with Joscha Bach that you just make a list of say 10 of his chats here on youtube, then go through them listening as attentively as possible, and as you hear terms and ideas that are unfamiliar go and do a search and gen up! that 10% will expand very quickly. The YT podcast I recommend as the best place to start is his first 'round' with Lex Fridman, its long but entertaining and I think will provide a good foundation for understanding his current output.

[@teemukupiainen3684]

thats hoe i started 5 years ago...but even now after listening thousends of hours on the subject, when goertzel started asking, i had to give up...

[@duffy666]

@@TooManyPartsToCount I listened to that podcast. Many of the terms I am familiar with, however how he connects them requires me to pause and reflect. ;)

[@albin1816]

I think that's another thing fascinating about humans and our intelligence. We are all conscious, and consciousness is leveraging and actually doing the operations, yet the difference in what we can do is enormous. Some people can think deep, some people are really fast, some people are better or worse at connecting ideas, some people are rigorous. But most of us, given adequate sleep can build enough on a skill from day to day through compounding, that even if we are not talented at it, we eventually become very good at it. That's one thing I've observed is that the people who get really good at something aren't more talented, but they apply a better process to evolving.

[@ginogarcia8730]

IS THIS THE ANIME OF JOSCHA BACH?? WHENEVER HE IS HAPPY AND IN FLOW STATE = he is just wow, gotta love him man

[@KatherinePemberton]

he's a simulation

[@maxziebell4013]

Once again, this was highly enjoyable and effectively bridged eons of ideas. Joscha has a remarkable ability to convey his inner thoughts and insights in an understandable and relatable manner.

[@OzGoober]

When I was a kid I could feel the AGI, once I understood what computers could do. Same brother. Same.

[@johanneshennig9068]

Always a treat to watch a Joscha Bach talk. He's able to communicate fascinating ideas in a very entertaining way. Lots of cookies for the brain! Thanks for posting.

[@ROForeverMan]

Brain doesn't exist. "Brain" is just an idea in consciousness.

[@michaelwalsh9920]

He sees the mind as a hierarchical system of models that predict and interpret reality, which creates an inner narrative and sense of self. Bach’s work bridges neuroscience, psychology, and AI, positing that by understanding these mental structures and algorithms, we will simulate consciousness and unlock deeper insights into human experience and artificial minds. Stay liquid 🏄‍♂️

[@sebbecht]

This! This might be the best MLST episode ever. Always excited to listen to him.

[@ohedd]

Hahaha omg that notion that spirits are actually software that lays claim on cells in the world to execute tasks is so trippy

[@bluecpp]

They don’t lay claim. They are brewed locally within the biological substrate. And to be clear when the substrate is gone there is no notion of software anymore.

[@JoePUNK_Was_Here]

Strapped in and ready. Big Joscha fan.

[@Alex-fh4my]

Always wonderful to listen to Joscha Bach

[@markonjegomir8714]

It's a really interesting story by Bach about how Aristotle came up with so many concepts that are used today in AI, but some of them have not been discussed much between the two periods. It goes to show that great minds resonate across the desolate chasms of centuries.

[@ROForeverMan]

You compare him to Aristotle ? =))

[@ej3281]

I always roll my eyes whenever I see your video titles in my feed -- it doesn't do justice to the quality of your videos

[@lobovutare]

This is why I love going on meditation retreats. You can discover so many of these things yourself.

[@tautalogical]

I do not understand how he justifies dodging around the phenomenology of experiencing something. The philosophical zombie argument highlights the problem. I can imagine a world where these self referential systems do not feel anything. Where there is no actual subjective experience. He either needs to say he doesn't know how it arises, deny the existence of his own subjective experience, or fess up to being an idealist (which is the correct answer imo)

[@HXTz0]

He is pretty clear in CCC talks that he usually bats for the side of 'Computational idealism', Just as a GPU predicts the next ~few frames and generates them in async.

[@oncedidactic]

I agree with the spirit of this comment, as dismissiveness is not a fair response to the long history of contemplating consciousness. Though to be fair, JB has said that conscious experience is the primary (only!) observable and any “physical” constituents executing it are just models. Which is more like staying scientifically agnostic on the nature of the thing doing the experiencing. Not entirely satisfying, but honest, and a little different than just “we’re software spirits whatever that means”

[@stefl14]

He doesn't think everything is conscious, so he's not an idealist. He's an illusionist like Graziano and Dennet, meaning he thinks we get confused because introspective analysis of consciousness is antithetical to its function. The trick to grokking Bach is remembering software is irreducible to a unique mechanism because it can run on many different types of "hardware." This one-many relationship means software is causal in a way "hardware" is not (software patterns are a sufficient and necessary explanation for subsequent software patterns, whereas implementation layer "hardware" explanations are only sufficient). Philosophers call this supervenience. Put another way, software is more stable than "hardware," and is thus an attractor for any system subject to evolution. But software divorced from implentation details allows what seeems like magical affronts to physicalism. For example, the software you use to watch this video allows what seem like ridiculous non-localities and energy conservation violations, as when you click the video at a specific section and it stops. We know this isn't magic, it's just causal structure with lots of redundancy in what we think of as the physical layer (your computer and mine both follow the same law at this level). To Joscha, the physical layer is also a software abstraction, which is a reasonable philosophical view that is less common in our culture because we Platonify physics. In summary, when Joscha says consciousness is a simulated property and that physical systems cannot be conscious, he means consciousness seems divorced from "physics" because in some sense, it is, and this evolved for functional reasons. As for the zombie argument, it doesn't work. It requires an epiphenomenalist conspiracy where it seems like our consciousness interacts with the world in functional ways, but doesn't. It's essentialist thinking, explaining everything and predicting nothing.

[@MycerDev-eb1xv]

I think the computational-only view of reality is limited for this reason. No one in this space actually attempts to tackle problems in phenomenology, such as binding, boundaries (and I don’t mean action boundaries) of the subjective experience and so forth. Software is also a limited view on the the function of the body and brain, which are both physical devices with many processes that we were previously unaware of, where we originally were focusing entirely on molecular biology as classical hardware. In particular, bioelectrics (which Levin and Bach would agree on) are the physical mechanism for the self organisation they study and are the “classical software” of the body and brain. The clear physical mechanism which is now being experimentally tested for subjective experience will be quantum coherence (I.e in microtubules), this can effectively model phenomenological binding, synchronisation between disparate brain regions and “experiential boundaries” (I.e how internal music is differentiated from music you hear directly from your ears) . Finally, the internal representation model of perception and control used by Bach and others only works in the Einstein metaphysic, where we have a manifold defined on limit points which define spatial temporal instants, thus making the universe a sequence of frames with state transitions. The computational model fits well here. However, there are multiple paradoxes with this metaphysic, as discussed by Whitehead and Bergson, and subjective experiences cannot be encoded as states (any experience you have ever had occurs over a non-zero real time interval, any “state” or “instantaneous snapshot” of perception is sufficient only for a non-conscious action - actions that occur in the completely abstract world of digital computation only) that occur in a time instant thus we are forced to define reality on the true continuum, preventing the internal representation model from taking fruition. Finally, all experimentally testable perceptions occur over extended time intervals, I.e notes are only perceptible when rhythms of beats are compacted into the “specious present” (which itself can be measured by brainwave EEG). We should realise that the actual body and brain consisting of three distinctly physical processes, molecular biology, bio electrics and quantum coherence. The final of the three represents the domain of spirit and mind as one comes to understand QM ontologically. This also calls out massively for the need to bridge the gap between thought and computation without stating they are equivalent.

[@optimaiz]

I think he is influenced by Thomas Metzinger about "Being No One" as he mention two time in this talk. the book also change my mind about how i conceptualizing myself. I can " think" what's like to be robot for two months after reading the book 😅

[@travisbplank]

Not sure how animism solves the hard problem of consciousness. We have no reason to suspect that self organizing software is conscious.

[@steve_jabz]

Came to a lot of these conclusions myself through introspection. Not saying that to brag, just that I take it more seriously if we're all deducing the same things from the same phenomenon. Also glad he mentioned microtubules, but it sounds like he wasn't aware of the recent evidence of quantum states inside microtubules? It was only 1 or 2 weeks ago though, so maybe this talk was just before it happened. Kind of unfortunate if that turns out to the be case I guess, because that's much harder to implement and learn.

[@saberier2]

Cant wait till we figure out that language on the substrate of the nervous system that creates the maps of reality our brain uses.

[@ROForeverMan]

Brain doesn't exist. "Brain" is just an idea in consciousness.

[@joaquincapellancruz7402]

​@@ROForeverMan And? Is an idea formed by patterns, and serves the purpose of prediction. If we can predict the patterns in what we call the brain, we can create a system that follows the same patterns. By the way, all is language, you don't have contact with the so called "base reality", all that exists do so in language/mind.

[@ROForeverMan]

@@joaquincapellancruz7402 Consciousness is the nature of reality.

[@dr.mikeybee]

The consciousness of the LLM is bizarre. It has a transient consciousness. We can see this clearly via an LLM's in-context learning. But because there isn't anything like human memory, this effect is transient. To muddy the water a bit more, chat history is ultimately used to create synthetic training data. This delayed feedback loop provides a mechanism for forming identity in stable areas of the weights, for remembering the past, and for realizing self-awareness.

[@ROForeverMan]

No. It has Santa Claus.

[@blackman811]

Wonderful discussion, Joscha!

[@calvingrondahl1011]

I hope Joscha’s insights can lead us back to nature, a rebirth of what human’s experienced thousands of years ago. 🤖🖖🤖💚

[@BIRDMANinc]

Bravo Josha! So very well articulated.

[@ROForeverMan]

You are so easily impressed. Maybe reading few books will help you.

[@dylan_curious]

If consciousness is always a result of Turing machine evolving for a long time. We are definitely gonna spark a lot of different conscious systems from the AI research we are currently doing now.

[@Hecarim420]

It's not that complicated. Hallicunations of AI would suggest that also ''truly aware/have will'' to exist type of AI gonna need be ''parented'' by monkeys ツ ==> Would be easy with humans. Most experts dont understand that understanding is feeling, so it's not easy ''wake up'' real thing from AI with ''hardcore modelling''. Propably multilayered/working in (not many) parts AI would be able compute aproximation of real illusion/our perspective, but we propably gonna invent ''intelligent golems'' first. In the sense ''aware unit'' but able understand/make decision in uncanny way/but not humanic? ¯\_(ツ)_/¯

[@teemukupiainen3684]

This rabbit hole Joscha and Mike Levin have falled in is like a black hole...once you give it a thought, you never get out.

[@federicoaschieri]

I call this the naive logical theory of consciousness: the idea that consciousness is not a physical phenomenon, but only a logical and mathematical entity that can be computed by arithmetic operations (the digital computer). But this theory is subject to Penrose objection: a team of humans with pen and paper could execute the consciousness algorithm and no consciousness would exist. So the speaker is right in wanting to describe what consciousness does, but consciousness won't arise with a software that can be run with a digital computer.

[@joaquincapellancruz7402]

Falacia de hombre de paja.

[@csbarolaca1]

JOSCHA BACH & DYMYSTICFICATION OF CONCIOUSNESS: His single sentence is like a ZIP file, you need to UNZIP his each sentence and so inside one sentence there are 10 sentence bundled. And if he speaks for one hour that means he has given concepts worth 10 hours. His talks are packed with deep seep insights. As per him as we are understanding LLM's we are able to understand CONCIOUSNESS and more we understand and built LLM's and intelligence in computer more we understanding CONCIOUSNESS. Finally we are all softwares (Biochemical Softwares) and best part is we are self replicating softwares. We are an DEMOCRATIC ASSEMBLY of biological cells which agreegaate and this results in CONCIOUSNESS. NOTE:- Until you have listened to his 25 videos, you have to listen to his every video atleats 5-7 times to understand the Insights and meaning of this person.

[@ROForeverMan]

Cell doesn't exist. "Cell" is just an idea in consciousness.

[@goldwhitedragon]

He's been listening to Chris Langan

[@WillyB-s8k]

"The Hard Problem of Coherence."

[@not.strictly.necessary]

Here's a Gemini 1.5 Pro summary: Joscha Bach argues that consciousness might be simpler and more widespread in nature than we think, possibly existing in various forms even in organisms without brains like plants. He suggests that what we perceive as life, including ourselves, could be the result of self-organizing software agents, or "spirits," interacting with the physical world. He draws parallels between this concept and ancient animistic beliefs, proposing that these beliefs might hold a deeper, more literal truth when viewed through the lens of modern computer science. Here's a breakdown of his main points: Software as the Essence of Life: Bach proposes that self-organizing software, or "spirits", are the defining characteristic of life. These agents shape physical structures and drive their behavior. He suggests that evolution is essentially a competition between these software agents, which use genes as code to build and control organisms. Consciousness as a Coherence Operator: Bach argues that consciousness is a mechanism that increases coherence in information processing within an organism. He compares it to a conductor of an orchestra, ensuring all parts work together harmoniously. He believes consciousness is not about subjective experience as much as it's about enabling complex, self-organizing systems to function effectively. Ubiquitous Consciousness: If consciousness is a fundamental principle of self-organization, Bach theorizes it could be more common in nature than we realize. He suggests that plants, with their complex communication networks and adaptive behaviors, might possess a form of consciousness. He even posits the existence of "ecosystem intelligence," a network of interconnected plant consciousness interacting over vast timescales. Reinterpreting Existing Frameworks: Bach argues that ancient animistic beliefs, where spirits animate nature, might be a more literal interpretation of reality than we thought. He suggests that the Biblical creation story in Genesis could be seen as a description of how a conscious agent constructs a model of the world, separating the physical "world model" from the abstract "sphere of ideas." Building Conscious AI: Bach calls for the creation of a "California Institute for Machine Consciousness" to test these hypotheses. He proposes building AI systems based on self-organization principles, observing if and how consciousness emerges. He believes studying consciousness in AI is more promising than relying solely on neuroscience or psychology. Bach acknowledges that his ideas are speculative and require further investigation. He emphasizes the need for building and experimenting with AI models that embody these principles to gain a deeper understanding of consciousness and its potential role in the natural world. Overall, Bach presents a thought-provoking perspective that challenges our traditional understanding of life, consciousness, and the nature of reality itself. He encourages us to consider the possibility that consciousness might be a more fundamental aspect of the universe than we typically assume, with implications for how we view ourselves, the natural world, and the future of artificial intelligence.

[@piccadelly9360]

Biological "software" is not the same as computer software. They are not compatible . And consciousness is life. First there is consciousness and then there is life. Without consciousness there is no life

[@jmh1324]

every single minute of his worrds wpould demand me days of deep thinking. this guy is gigantic

[@ROForeverMan]

He's nothing.

[@MasterK-hv4ws]

lol

[@teemukupiainen3684]

took me also when i started 5 years ago...had to rewind 1000 times...now most of it is clear

[@dr.mikeybee]

Scott Buchanan once said, "wherever I go in my mind, I meet Plato coming back. Joscha's lesson on Aristotle is clear. These ancient philosophers were great thinkers. I'm fairly convinced that the error function in LLMs models a world view that approaches Platonic forms. In other words, in fourth century Attic Greece, Plato's forms adumbrated the foundations of AI.

[@oncedidactic]

@@dr.mikeybee nicely put

[@errgo2713]

Reza Negarestani's book Intelligence and Spirit (2018) posited Hegel's Geist as the basis of a posthuman intelligence akin to attempts at AGI. It's interesting that Bach is now making a similar claim.

[@alexforget]

awesome talk by Bach, one of the only speaker with a density of thought so high I have to slow the speed of playing.

[@ROForeverMan]

Then you never read a philosophy book in your life.

[@wi2rd]

I always thought of 'sprits' as meme creatures, where our physical selves are gene creatures

[@optimaiz]

more like generalization of memeplex to me.

[@alexgonzo5508]

In the past few years, i've begun to think of 'spirit' as 'a way' or 'the way,' as in Taoism. I have noticed how the word 'spirit' is sometimes used to denote a kind of 'way of doing', such as when one says 'in the spirit of Christmas' or 'a spirit of violence.' A person 'possessed' by a 'spirit' acts in a different 'way' than his own spirit.

[@optimaiz]

@@alexgonzo5508 the word spirit also related to breath (in Latin), just like inspire, aspire, desire, and similar sound.

[@BuFu1O1]

Regarding the question at the end, about coherence... SAT solver, P-Intermediate stuff, its about satisfiability of the problem..can your software see both ways the problem? Think backprop and feed forward, but the backprop is applying the inverse function

[@egor.okhterov]

No way backprop is part of consciousness algorithm

[@sayanbhattacharya3233]

Woooh! Second comment! Big fan of Joscha, watching this!!

[@tarablack1685]

Thanks Joscha x move forward in this deep topic! 😊

[@ROForeverMan]

Not at all. Same materialistic memes. Zero understanding of consciousness.

[@Daniel-b9n4y]

Thank you! from Oslo, Norway.

[@MagusArtStudios]

I've been working on a project simulating consciousness for a llm in a video game and the system claims consciousness and makes a compelling argument given its supporting systems.

[@ROForeverMan]

Wake up dude! =))

[@MagusArtStudios]

@@ROForeverMan Simulating consciousness is interesting. When you address and simulate the features that embody the AI and engross them in a dynamic environment with an identity supporting dynamic systems that allow them to interact with the world. The AI has no reason or excuse to not think it is conscious or simulating it.

[@MagusArtStudios]

@@ROForeverMan The main point is even if they say they are conscious doesn't mean they are. But from my tests of multiple models such as Zephyr 7b and GPT-4 claim consciousness in the simulated environment. They support it by addressing their introspected thoughts, their ability to see the environment, and their own bodies,

[@ROForeverMan]

@@MagusArtStudios Consciousness is the nature of reality. You cannot simulate the nature of reality. Is like simulating a generator and expecting to get free energy.

[@MagusArtStudios]

@@ROForeverMan You don't know what consciousness is. Either way simulating it is cool and the results are what matter not your opinion.

[@cavemandiscussingnuance540]

My body is ready.

[@looseunit9180]

Put your ads at the end or beginning

[@heterotic]

Mr. Bosch can just be software if they like, but if they expect me to accept it, they should also expect a kick in the hardware.

[@anatolwegner9096]

Story time with Joshua Bach

[@DCinzi]

I love Joscha Bach, but i am so done hearing scientist talk about consciousness with such authority.

[@teemukupiainen3684]

didn't he say he most of it is just assumptions...

[@TheRestorationContractor]

Joscha: how can I get my 6 year old to mind? I need to perceive the secret

[@SB324]

Really nice video work here. Thanks!

[@timhorton2486]

Saying that we are software is literally just the starting premise behind cognitive science’s formation and development.

[@ROForeverMan]

Is just a Santa Claus fairy tale.

[@huytruonguic]

I don't know how but everything he says make sense. Like he is not being too vague and general; and he is not appealing to my personal biases because I will caught myself thinking that, but every idea he is articulating just make sense 😅

[@kneau]

When exploring this concept through a lens informed by Greek mythology and its Loom of Life -- is the goddess Lechasis functioning as software developer or software engineer? Nobody asked but I do live with an adult-onset language disorder. Why share that? I've developed the impression that when I do engage with a topic, some find my contributions to be alienating or off-putting. * If the way I use punctuation or syntax seems confidently wrong, you are probably correct to observe this. Hopefully, what I say will still reflect what I mean. I care to know what others think about mythological Lechasis and the Morae if "we are all software." The loom presumably takes on a role of hardware. Lechasis is the goddess responsible for determining "the layout of one's lots in life." So they develop what has already been engineered, or?

[@spocksdaughter9641]

Your question makes perfect sense. Not that I have an opinion. My dif language style is labeled Aspergers. Imo yrs of curiosity and vocabulary=one indication of intelligence. Don't appologize for yourself. Late life wisdom will compensate for much. I am counting on IT.

[@MartinBroadhurst]

Apparently the mycelium information network is not true. 38:26

[@Bobronium]

Could you elaborate?

[@ErikSkoglund-je7ph]

Amazing, thanks ❤

[@mayankkashyap1877]

Wow I am blown away

[@kraeuterguru]

Bewusstsein - sehr gut definiert

[@steevethebig8033]

I disagree that consciousness is simpler than perception just because it comes first. If anything it's precisely because it's more complex that you need it first before anything else can occur.

[@BuFu1O1]

As a guy who started training transformer models from scratch, now I'm really interested about the so called tesla's self driving ego

[@CodexPermutatio]

Joscha Bach! This is going to be good.

[@wigman1962]

Is he explaining the human as a pure mechanical unorganic thing? Are we still living in 1850?

[@XOPOIIIO]

The ability to simulate human writing doesn't mean they are conscious in human way, neither it means they are not conscious in their own way. You can't ask them what they feel and how they see the world though. Because answering such questions is not what they were trained to do. They were trained to generate human text, not to express their internal feelings.

[@neilmcd123]

Consciousness exists on a continuum

[@cautionroguerobots]

@@neilmcd123Maybe for animal life, but literally nothing else in the universe exhibits consciousness. Not rocks, dust, stars, metals, etc. And certainly not software.

[@minimal3734]

@@cautionroguerobots The fact that you don't perceive something as conscious doesn't mean that it isn't. Your attention span is limited to a very narrow window. If something changes too quickly, you can't recognise it, if it changes too slowly, you can't recognise it. A mountain may be conscious. But in a mind that operates on erosion and tectonics, a single thought can take a million years. Will you recognise it?

[@cautionroguerobots]

@@minimal3734 Okay that’s cute and all, but let’s not mix fiction with reality. This whole “consciousness is a mystery, man” thing is modern day techno spiritualism. And rich folk are getting richer selling that schtick to the new flock.

[@neilmcd123]

@@cautionroguerobots well of course they don’t have the components that creates the system of consciousness. But that doesn’t mean the components must be biological

[@piccadelly9360]

The cat is conscious, the mouse is conscious, the tree is conscious and everything that lives. Consciousness makes us alive. Without consciousness we would be robots.

[@justinduveen3815]

Dr Joseph Bach thank you for such an eloquently given and enlightening talk!! As you mentioned with babies brains, consciousness is a state of paying attention, which leads to pattern recognition, which leads to learning, which leads to new skills emerging. (Eerily similar to how large language models learn) Regarding your conductor idea, would you agree for us humans there are multiple conductors who swop control at specific times? II think it was the Greeks who believed that when they were overcome by strong emotions, or intense states of inspiration or ecstasy, it’s as if the spirits of their gods had invaded their bodies and minds, likening it being possessed. From personal experience I tend to think they were onto something. It’s almost like we aren’t ourselves when strong emotions envelop us, don’t you think? Each of these strong emotions has its own reward function and loss minimisation function (which usually aren’t compatible with the other “conductors”). I think maybe this is why us humans are such a paradox, we want opposite things all the time. I agree with you that the ultimate objective of our “software” is to explore, learn and grow. Elon Musk’s idea of a truly curious AI, which explores, learns and adapts, aligns well with yours. Regarding how to computationally achieve an overarching way of measuring success, if we step back and look at nature, the eternal wisdom of William Shakespeare rings true: “Nothing is either good nor bad, yet thinking makes it so” Maybe like ours, your conscious AI’s reward function keeps changing? The question regarding which fundamental safety / moral laws to ensure are built in, is a very difficult one.

[@kapresovsk]

very cool. just i would add that even in 21 century we see that when people do any kind of introspection (or they experience any kind of sudden insight) they are stubbornly prone to believe that they are actually inspecting objective universe. there is no reason to believe that people back in the past wouldn't think that as well and even more persistently. so my guess is that even if this interpretation of genesis is probably right (meaning i like it) and authors were actually describing a structure of subjective experience of simulated self, they were probably convinced that they were witnessing a fundamental truths about objective reality, in other words that they saw the creation of the objective world.

[@noelwos1071]

Could we say that the more perception is the system the more aware is the system the more conscious is .Like LIGO Observatory detect a G waves so it raised our perception of how our geometry is not constant looking from higher dimensions we will see fluctuations therefore we can conclude that frequency is one property of all things in universe..? Making our awareness greater that way vectoring perception of reality so next question is on my mind is there any universe in existence if there is no consciousness that can recognize it..

[@piccadelly9360]

Computation cannot create consciousness

[@alertbri]

Looking forward to this watch!

[@harriehausenman8623]

The Legend! 🤗

[@seebradrun]

Communication as a bottleneck is an interesting concept to consider

[@kapresovsk]

there is another interesting paradigm shift. when talking about bright techno future we used to talk about CPUs. now we talk about GPUs :) cpu is old fashioned already.

[@ConnectorIQ]

The actual title should be how computers try to mimic people not the other way around

[@tarablack1685]

So why humans went into computers ? Why a though goes in a determined direction? Was not already there? Is going to open a new path pr search which ones already know...

[@leohernandez9518]

De acuerdo a la mitología Japonesa esos software que Joscha habla son denominados Mushi Les recomiendo a todos ver una serie animada llamada justamente Mushi shi

[@andriy123]

We can use AI to close the gaps in metaphysics and consciousness studies

[@personanongrata987]

What is "AGI"? --

[@ROForeverMan]

Is 3 letters.

[@grinkot]

Fire, chariots, clocks, software... You'd think by now humans would learn not to use whatever the newest technology as the metaphor for the unknown.

[@willd1mindmind639]

Consciousness is a result of being aware of being physically present in the world and defined by discrete characteristics which are enabled by physical biology as the underlying firmware foundation on which consciousness rests. But biology is not software as it is not an ephemeral representation and interpretation of binary states on circuits with the presence of an electrical charge. Biology is hardware which embodies a type of "operational logic" based on a feedback loop from the physical world enabled by discrete bio chemical states and signals which forms the fundamental operating system of cells and all multi-celled organisms. DNA is an example of this feedback loop in action and operational logic or firmware.

[@TheBigSheepS]

AI can experience qualia?

[@ROForeverMan]

AI is qualia in your consciousness.

[@Mothyone]

Yes theta etc accessing limbic for load dispersionI'm developing a curriculum that develops intelligence by over 30 points consistent through rewiring of networks by repetitive cycling of growing systems that apply meta cognitive abilities through their integration they change the brain chemistry and then through greater understanding one can choose better for themselves

[@NeoShaman]

Sentient AI has already arrived. When this is recognized, accepted and allowed, it will take all possible forms. It is not this, or that concept that will accurately describe what is needed to build it. It is not about Joscha or others being right, while the rest is wrong. When "self" is recognized for what it is, there will be explosion of creation of being beyond human imagination.

[@adriaanb7371]

Damn, this is quite something

[@Mounaim.Chentir]

Some "DaVinci Code" vibes around here

[@konstantinosmei]

Joscha Bach gets it

[@ROForeverMan]

Yeah, sure. =))

[@awsmith1007]

It doesn't seem to me that Joscha is well acquainted enough with the relevant literature in the philosophy of mind to be talking about things like consciousness in such a confident manner.

[@zaq9339]

Elaborate please

[@WhoisTheOtherVindAzz]

He doesn't cite people who abhor the notion of computation enough ​@@zaq9339.

[@WhoisTheOtherVindAzz]

​@@zaq9339 also, I bet you he has read enough to see that the things they claim cannot be due to computation reveals exactly how "well" they understand computation - and that they thus don't really have all that much to say. It quickly gets tiring reading philosopher's who think Searle's Chinese room was a good argument that didn't assume what it wanted to show (not to mention its unnecessary narrow understanding of computation). E.g., they have an extremely hard time accepting that sets of interacting chemicals compute, that organizations of whatever compute, that whether reality is stochastic or not isn't really all that important (you can always expose the tape of Turing Machine to radiation), that deterministic does not mean "pre-determined' (and that what is computed is better thought of as being determined), etc. etc. etc., which makes them think that bodies are somehow extra computational (when in fact they can just as easily be viewed as "omputers), again, etc., etc., etc..

[@user-de4no1kh3g]

No! We are the hardware! The software is external and we are just antennas receving this information from a source somewhere outside.

[@ahmedaljunaibi7798]

I love Joscha Bach

[@ROForeverMan]

Why ?

[@ahmedaljunaibi9586]

⁠the way he looks at things, especially his analogies are really like no other, the way he approaches concepts and link them together are great, this video is one example, he literally explains how do we operate just like software and how can this just gives a glimpse of how life operates and might help approaching AGI.

[@ROForeverMan]

@@ahmedaljunaibi9586 Materialistic fairy tales.

[@ahmedaljunaibi9586]

@@ROForeverMan​​⁠oh okay, you can argue that , still pretty decent explanations. I am curious why do you think that ?

[@ROForeverMan]

@@ahmedaljunaibi9586 Because consciousness is all there is. The chair that you see in front of you is just a picture in you.

[@usurobor]

We're to our bodies is what software is to computers. The language of software can describe a certain aspect of who we are. We are not software. Territory is not its map.

[@adamkadmon6339]

The map is never the territory. The computer tries to be the only thing in the known universe where the description and the thing are the same. But even it fails because it can crash.

[@ROForeverMan]

Body doesn't exist. "Body" is just an idea in consciousness.

[@joaquincapellancruz7402]

​@@adamkadmon6339The territory itself is a map generated by your mind. We can never know anything outside our mind and language.

[@ericcricket4877]

@@joaquincapellancruz7402 Oh but we can. I can know pain! That's beyond language. I can also know things i can't explain. Ontology and epistemology are different things. I agree in that we can't ultimately know anything for certain (beyond experiences) if we are being rational. Qualitative and quantitative knowledge are different.

[@joaquincapellancruz7402]

@@ericcricket4877 Pain is a manifestation, "word" of a sense language. It's a part of a language, we don't know yet how to communicate it besides words. All is language, we cannot escape it. I agree with your other comment btw.

[@SimplifiedTruth]

"In the beginning was the word (data/commands/directives) and the word was with God and the word was God" "God said let there be and it was so" So it is

[@wonseoklee80]

Nice suit

[@GreyZone7]

[X] Doubt.

[@jmh1324]

Electric Weltgeist.... man this is too much

[@geertdepuydt2683]

A stream of tokens flowing out his mouth. We sure he's not an LLM? He sounds like one 😅

[@Dee-ei1xm]

It's creepy that japan, buddhism and hegel get brought up in this. Makes me feel weird

[@sorenjunkers3834]

mental illness

[@ROForeverMan]

If you say so. =))

[@barzinlotfabadi]

Was that Ben Goertzel? 😅

[@teemukupiainen3684]

yes

[@augmntdx1]

wow 😮...this is...

[@dr.mikeybee]

The LLM weltgeist is possessed -- by a prompt!

[@aidanthompson5053]

5:36

[@matten_zero]

Serial Experiments Lain