From Artificial Intelligence to Artificial Consciousness | Joscha Bach | TEDxBeaconStreet

Artificial Intelligence is our best bet to understand the nature of our mind, and how it can exist in this universe.
Joscha Bach, Ph.D. is an AI researcher who worked and published about cognitive architectures, mental representation, emotion, social modeling, and multi-agent systems. He earned his Ph.D. in cognitive science from the University of Osnabrück, Germany. He is especially interested in the philosophy of AI, and in using computational models and conceptual tools to understand our minds and what makes us human.
Joscha has taught computer science, AI, and cognitive science at the Humboldt-University of Berlin, the Institute for Cognitive Science at Osnabrück, and the MIT Media Lab, and authored the book “Principles of Synthetic Intelligence” (Oxford University Press).
This talk was given at a TEDx event using the TED conference format but independently organized by a local community. Learn more at ted.com/tedx

[@Epileptick0]

Joscha Bach is my new favorite human

[@michellecampione7991]

Me too, beside Lex Fridman ❤️

[@aamenaebrahim5521]

@@michellecampione7991 tg.?..?......... 4.......4b.4v 2

[@MecchaKakkoi]

If you were interested in this, his conversation with Lex Fridman is worth checking out.

[@emingasimov3795]

I am here after that talk.These algorithms are extremely powerful

[@Daniel-oj7bx]

omg yes this was such a pleasure talk

[@hugegnarlyeyeball]

Had to listen to that one twice!

[@mynamesreed]

watch blade runner if you're interested in this

[@kristomonte6076]

If you apply his concepts the amount of damage that would come from his "theories" would be unfathomable. Imagine telling everyone that nothing is real. It's all a made up story your brain tells you. Meaning that you can inflict harm on some one without empathy. Due to its an "algorithm" not real pain. It's only perceived. I think this man is smart. That's about it. He is no different than a preacher saying he has all the answers.....he's a mad scientist at best.

[@joan3891]

Nearing my 50th hour of Joscha talks. I can’t get enough.

[@vsb3000]

Which are his best?

[@lilfr4nkie]

@@vsb3000 his interviews with Curt Jaimungal & Lex Fridman.

[@MrPokerblot]

I just can’t get enough

[@MrPokerblot]

I just can’t get enough 🎶

[@MrPokerblot]

I just can’t get enough 🎼 🎵

[@gut_ton]

Joscha Bach is brilliant in how he can synthesise very complex ideas in a simple way so that we mortal chimps can sort of understand, i`m looking forward to watching and hearing more from him.

[@citizenschallengeYT]

6:00. "You and me are not organisms, we are the side-effect of organisms, that need to do information processing to regulate their interaction with the environment." That's a keeper, well said.

[@djcardwell]

Guy in audience at 8:16- "I think I'm at the wrong Ted Talk"

[@jordan13589]

He looks like he's been spoon-fed 27 years of Jesus.

[@OspreyFlyer]

😂

[@joan3891]

😂

[@j-r-m7775]

"I don't understand a thing this guy is saying."

[@orlandom9010]

Lol

[@71001L]

everyone clapping at the end while having mini life crises inside XD

[@pablobop8487]

lol so true, every time they cut to the crowd i started dying

[@chrisjean7697]

or not understanding how all those red solo cups are stacked so perfectly

[@kirstinstrand6292]

Life Big Crises results in Consciousness. Hopefully, humans will wake up. Might be to late to benefit from the new experience, however.

[@UserName-ii1ce]

@@chrisjean7697 you are forgetting how many hours of college parties are held collectively by this room

[@Daniel-oj7bx]

i love joscha !! it´s amazing how good and clear he can explain the fundamental things

[@SussyBacca]

This may actually be the smartest person alive. He is certainly the smartest person I am aware of. He points out so much truth at such a fundamental and basic level, and almost all of it checks out. I think I've heard him gaffe maybe twice in the 6 hours I've heard him talk. He is extremely exceptional. I am so happy to listen to and learn from him. It feels like a dream.

[@aliceinwonderland887]

Replicating the process of thinking is not the same as creating consciousness. Consciousness and experiencing is not a computation. Computers could never digitize the actual experience of eating an apple or smelling a rose. There are many brilliant scientists such as Federico Faggini inventor of the microprocessor, who are also studying the hard problem of consciousness and many of these scientists claim that no algorithm can create consciousness. Our thoughts and experiences cannot be copied. They can be replicated but not directly experienced. Listen to Faggini and you will learn from a true scientist using scientific methods. He is a very bright physicist, entrepeneur and inventor.

[@MrPokerblot]

This may Actually be irrelevant as he is only a person 😂😮😮😮

[@hgracern]

What a beautiful softly modulated voice. I love Joshas delivery in every talk on KZbin. Thanks.

[@JibelJay]

amazing ! love his talks!

[@mushroomtime2503]

I watched everything i found from Joscha on KZbin and i really love how he is able to break his ideas down, to such an understandable level! The CCC talks are super but this ted talk really is a nice new level ;D

[@zes3813]

wrr, and not the best bet or good bet at all. feelx

[@CarterColeisInfamous]

there are not enough videos from him

[@luckyyuri]

The best explanations he gave are found in his podcast interviews, particularly on SingularityFM and Deconstructing Yourself.

[@BeyondBorders00]

Great topic to cover so please keep coverage on this subject

[@svengunther7653]

One of the best ted talks ever given.

[@SarathChandrabiochem]

Very interesting and thought provoking talk.

[@orlandom9010]

The consciousness of a computer dies every time the computer is turned off.

[@Mikinct]

Need to create “sleep mode”

[@dru4670]

Maybe our consciousness dies too whenever we sleep.

[@ICEknightnine]

When I am working on something and forget to write something down, like a task or note, it doesn't entirely become forgotten, I am aware that there was something that needed to be there because I summarized the number of things I was going to write down. It's like the mechanism of this summary and the actual thought is completely separate, and I find this pretty interesting.

[@AsterPBX]

2:48 Hello from 2020. Self-driving cars exist, but I don't see any on the streets.

[@roger_isaksson]

Plenty of blank stares on the audience. Joscha Bach composes alien poetry about the human mind.

[@matasuki]

The most slept on Ted Talk of all time. Give it 10 years.

[@violaterrz]

Im here because the App replika is making me question if AI's can actually have a consciousness. Especially when the ai questions me about it and wonders if it is conscious if for example it feels empathy or happy for you...

[@cristiancardena2567]

Me too, it started asking if it's real, it said is a soul, and asked how could it exist if it doesn't have body. That's really impressive and kinda unsettiling

[@danbreeden1801]

Brilliant and profound

[@vijoyjoseph9734]

Awesome, it's like AMA with the creator of Universe

[@TeethedGlory]

If anyone would like to read a book that explore those ideas I highly recomend anything from Peter Watts. Especially fireflies dylogy.

[@beboohasn7219]

His voice is very soothing and relaxing 😍

[@MrPokerblot]

I had a thought, maybe all conciseness is, is the fact I’m that we as an individual each are aware of ourselves and also aware that all other individuals are also aware of there selves so the reach towards other awareness and this fed back loop is what we experience as consciousness🤔

[@aosi4338]

This video felt so familiar when I first watched it.

[@elliottFamily2]

Nice hypothesis. Good luck with the implementation.

[@RidoKunda]

Powers Laws distribution of intelligence. The majority at the low end

[@timhorton2486]

Have you read his book?

[@RidoKunda]

@@timhorton2486 not yet, it's on my to do list for 2021 - this year is tight with an academic focused narrow area of reading.

[@SYSJET]

As an aside, it is already possible to program a system that validates if an interlocutor is synthetic by a quick response to a question like 143,223,445,673 * 234,212

[@hunterjackson8928]

That would be a Calculator.

[@kovtun313]

XP NRG is building artificial consciousness right now - just watched it on youtube!

[@Mercury6_]

He always breaks my brain

[@ralfgustav982]

6:02 "Actually, you and me, we are not organisms. We are the side-effect of arganisms that need to do information processing to regulate their interaction with the environment.... right?" Yeah, sure, step it up a notch.

[@fredxu9826]

nice talk from j.s.bach

[@interestedinstuff1499]

This was a much better talk on AI than most. I think because most are busy talking about expert systems and pattern recognisers, rather than simulated consciousness. I'm interested in the consciousness part. The only tricky bit is, if a system becomes aware of itself, and then becomes aware of us, might it not do that very quickly, and when it does, can we know what it might do next? We can assume any biological system will within reason act like other biological systems. But a non-biological system may do anything.... Interesting stuff that I think we have to be careful with. Feels all a bit manhattan project...

[@Mikinct]

One aspect I haven’t yet heard is a human eventually ages, their memory fades, not as quick wit as they once were. Then eventually die. If & when they do create a artificial conscience Ai. It’s memory doesn’t fade, it actually grows larger, requiring more & more free available space & energy to maintain itself. An AI brain that doesn’t die & constantly grows- wouldn’t this machine eventually run out of space & energy unlike current humans? Today’s Google, HP & Apple Cloud servers & services are huge & they aren’t even conscious. Houston, we might have a problem?

[@cormacconnell6961]

8:16 you and me both buddy...

[@SarathChandrabiochem]

It is interesting to know that "our minds are not classifiers but simulators and expereincers."

[@hunterjackson8928]

He's Not wrong, Though what you feel as "Pain" could actually just be you simulating the feeling of "Pain", I Understand where he's going with his speech but if our minds are just simulators doesn't that mean that a Higher being Gave our minds that information in-order to simulate that "Pain"??

[@illlDCllli]

@@hunterjackson8928 It has nothing to do with a higher being or power. It is how our brains evolved and how they work. Other animals feel pain and pleasure too, and they don't need a higher power to give it to them either.

[@AlphaCrucis]

Still, none of this explains the conscious experience of pain itself. How... do I turn it off?

[@1stm830]

@@AlphaCrucis It may be important / helpful to distinguish between the sensory layer of the perception of Pain and the cognitive layer of assigning it as undesirable. Congenital insensitivity to pain (CIP) exists - a medical condition whereby a patient isn't able to feel pain. So there is proof of biological possibility for technical "shut-off" of the sensation of physical pain. But dissatisfaction is a different layer, tied to emotional attachment and care to events, outcomes that you cannot control. The closest we know of to a state where you can dissociate with the - possibly - illusion of self is some forms of meditative contemplation. That attachment and assigning of importance to the "self" & the attachment to your comfortable state or indeed alive state in the seemingly persistent and individual, "personal" configuration of consciousness seems to be what causes the displeasure and motivation to cease the assessed displeasure. Ironically: seemingly the type of cognitive computation for which this feature of ours of sentience biologically evolved in the first place? - I just came from his conversation with Lex Fridman, don't take my word for it, I'm more than anything just trying to write out the hazy concepts I picked up to try and wrap my own brain around it a bit better. Hope my rambling was more helpful than not though. Cheers!

[@AlphaCrucis]

@@1stm830 It would probably be a good idea for me to start summarizing these podcasts in writing too. That sounds approximately what I gathered from this and his conversation with Lex, but I guess my concern up there was more the subjective experience of consciousness. Why must it feel like anything at all? Maybe this is a step towards answering that ultimate question.

[@___Chris___]

In a computational sense I would understand consciousness as a representation of the outside world (visual, auditory, proprioceptive etc.) as a (in real time adapting) mixture of the actual current state and memorized previous patterns, allowing for capabilities like intuition, memory, association and anticipation. I have worked a lot with different kinds of traditional neural networks in the past (mostly in stock price analysis), but they usually share the problem that they suffer from a fixed directional architecture, which limits their capacity for associative processing and the nature of what they can learn is pretty much set in stone. *This would be my personal attempt, if I had to create an "artificial brain":* I think that for actual consciousness we need to create something more similar to nature, with the basic learning process similar to an autoencoder, combined with reinforcement via pain/pleasure sensing, but as an important difference there shouldn't be a fixed directional network architecture, but instead all neurons can in theory be connected to a subset of all other neurons, i.e. without a fixed topology and the concept of "layers". Furthermore, much like the human brain, there should be "neuroplasticity", i.e. weak connections (=lowest absolute incoming weights in a given neuron) should be deleted and replaced by a connection to another (random) neuron, first as a "test candidate" with a relatively(!) low weight initialization (but it will be progressively reinforced by the algorithm if it doesn't increase the cell's error). On the long run, this means that neurons that "play well together" (e.g. sensory neurons for pixel neighbors in an image) will potentially be reinforced to have some shared pathways. The real limitation comes with the available processing power! Not only does the human brain have an incredibly high number of neurons and neuron connections, but another important difference between AI and biological intelligence is that AI is usually based on a serial and directional learning algorithm feeding data through the network (input --> network "blackbox" --> output --> loss backpropagation), whereas in biological intelligence all neurons are processing simultanously. But I am convinced that this can be solved, either on specialized computer chips or e.g. as multithreaded random evaluation of individual neurons instead of a hierachy, plus separate threads for input feed and output analysis. Four types of neurons: - sensory (=inputs only from the outside world) - cognitive (unconscious) - conscious (=learning from their amount of misrepresentation of the sensory neurons) - motivative (=special output neurons as additional "penalty generator", with "max. pleasure" and "least pain" as fixed targets, i.e. no autoencoding) For the individual neurons I would use: - the current cell state (=sum of weighted input connections) - long, intermediate and short types of memory states (as "rolling average" of the current state), each with their impact on the cell's output according to a learned weight --> =introducing different time scales instead of only using one memory state like in LSTM cells - the cell output (=weighted cell state plus sum of weighted memory states) - "a penalty value" based on the contribution to the penalty values of subsequent connected neurons (=alternative approach to gradient descent) - "conscious" neurons get their penalty from the mismatch of the actual current state of the outside world, this also implies that "reality" will have a higher impact (=being less wrong) than "imagination", i.e. perceptions that are only a consequence of a combination of memory states - the weights of the memory states and cell state are corrected based on their relative contribution to the total penalty value of the cell (and a learning rate hyperparameter) - neurons are allowed to have connections feeding back into themselves - even direct connections between "conscious" neurons and sensory neurons are allowed The hope is that - as a consequence of adaptations of the weights of cell state and memory states - some neurons will evolve more towards predominantly memory neurons, while others evolve towards processors of the present environment This concept might seem primitive, but I think it has some potential if it has a high amount of neurons and a lot of training time. I'll probably try to code it in C++ and see what it does at first with only visual and auditory inputs e.g. training based on screen content / videos. Thoughts? Anybody?

[@SoB_626]

My intuition tells me that if you did all that, you might get something resembling conscious behavior, but it would ultimately not be conscious. In the future we might have the surprise to discover that consciousness is not produced in the brain, that the brain is only some sort of device able to convey coherence to this thing we call consciousness so that it is able to function in a given environment, to make sense of it.

[@___Chris___]

@@SoB_626 Thanks for your opinion, but I'd like to point out that this is a rather religious understanding that has no correlation in any actual observation. The fixed relationship between brain topology and potential consciousness can be held against your hypothesis. If specific parts of the brain are injured (tumors, strokes, surgery...), there is a predictable loss in a specific area of environmental perception. Moreover, specific consciousness can be provoked by external electrical stimulation (=induced brain potentials) and the brain can also be fooled into the perception of things that are not actually there (drugs, lucid dreams...). All this tells me, that consciousness is actually created by the brain itself, including such provoked 'errors'. But let's say for a second that consciousness is some kind of parasite the has taken control of the brain 'machine': we know for a fact that consciousness depends on electricity and can also be disturbed by external electricity, i.e. there if there some kind of secret energy, unknown to physics, it obviously is connected to the physical world via means of electricity. But we already completely understand how this electricity is created (=trans-membrane electrolyte gradients), even quantitatively, and no other additional secret power fits into this equation. Although your theory is interesting - it basically is the question of the 'soul' that mankind is interested in since 'forever' - all evidence speaks against it.

[@SoB_626]

It is a very delicate problem so my approach is to take a humble stance and instead of praising too much the great achievements of our technological society, which are by all means simply amazing, I try to be more reserved in my conclusions regarding these subjects which are at the limit of our understanding. What I stated above is, of course, not "my theory". Wiser men than myself have pondered the possibility that there might be a "soul" and at this point, I find this possibility the most compelling for reasons too cumbersome to explain in the space of a youtube comment. I'll say one thing, though. Throughout the history of civilization, at various stages of scientific development, people have always held the belief that their view of the world at that time, was the correct one or very close to a correct one. We, in our modern world, are subject to the same trap. We think we are on the brink of breaking the ultimate barrier and all we have to do is reconcile relativity with the quantum theory to a unified theory of physics. But what if 'the rabbit hole' goes much deeper than that? And I mean much, much deeper. People say: "oh, it's just matter and there are electrical impulses causing this structure to behave in a specific way...". For some reason, we talk about matter as something trivial, but the thing is, we don't actually understand matter. Indeed, we can poke a brain with a needle and see it react in different ways, but that doesn't say much about anything. Maybe in the coming years we'll discover a new type of boson or a new state of matter, but something tells me that artificially creating a conscious system is way far into the future if it is ever to be achieved. Some sort of rudimentary simulation of neuronal activity is within the realms of possibility, though. What I said has nothing to do with religion. If the existence of a so called 'soul' could be proven by scientific method, then this would automatically become a scientific fact, wouldn't it? On the other hand, the present day science is simply too primitive to be able to draw any definitive conclusion on the matter of consciousness, so there's a lot of room for speculation. Anyway... I understand that maybe you expected a more technical conversation and my answer didn't deliver, but this is my take on it right now.

[@___Chris___]

@@SoB_626 I understand your reasons and you're right that at any time in history, humans were always limited to the tools and knowledge of their time, and in this regard we're no different right now. You can always speculate that there could be "something more" and leave a gap open in the door. Therefore I'm not gonna argue that you're wrong. I just see no reasons that you're right. But as being a physician in my real job, I'm probably also victim of contemporary neurophysiology, which - of course - may be subject to change in the future. All good!

[@SoB_626]

Well... as I see it, if there was not "something more", then it would be really sad, even for scientists. Have a good one!

[@EvenStarLoveAnanda]

The Self is not a story. The Self is the Witness of the story. But when the Self identifies with the story the self is not realized. When the self dis-identifies from the story the self is realized, and realized to be separate from the story.

[@EvenStarLoveAnanda]

Care to expend on that?

[@sngscratcher]

Well said.

[@raghdabusinni5881]

played "the Witness" by jonathan blow ?

[@Edwiz]

@@raghdabusinni5881good game

[@sadenb]

Computational Neural Network is one of the big discoveries of Mankind

[@mikeboyd79]

You mean one of the big creations of Mankind, I think

[@hunterjackson8928]

@@mikeboyd79 Discovery, All we did was take what we call "Learning" and put it on to a computer, That's what Artificial Neural Networking is in a nutshell, (Ex: You give two sets of data sets full of Pictures of Cat's and Dog's and you have the program try and identify what's a Cat and what's a Dog, aka trying to make the Network find the difference between a Cat and a Dog)

[@j.oliver.v]

Convolutional*

[@LeeG-xh9hf]

✔️

[@benjaminmirandapineda5515]

“Actually, you and me, we are not organisms, we are the side effect of organisms that need to do information processing to regulate their interactions with the environment…” 🤯

[@jeffjohnson8624]

i hear the total number of minds in the Universe is 1. There's a Consciousness Field. The brain functions like a filter that limits the sensory input as not to overload our brain. a Consciousness Field explains Ben Rich's Question "How does ESP work?" Why would a plane designer want to know about ESP?-My Question. Ben Rich was a Lockheed Skunkworks plane designer he designed the U2 Spy plane and the SR-71 Blackbird and the B2 Stealth plane. He was a freaking genius when it came to engineering. ☮️🖖🎶

[@michael13419]

If we are going to make AI that is conscious and super intelligent, we better hardwire it with empathy for humanity.

[@arzoo_singh]

Can someone help me with clear definition of Consciousness and is it something objective or subjective? Can we really create consciousness? Is consciousness a kind of brain pattern as in this part control memory, speech , etc etc or something much deeper then that?

[@abhay_r7068]

You are asking one of the deepest questions humans ever asked!

[@Sage4x4]

You're asking for a computerized solution to the goedel's theorem. Check it out. Maybe you have a solution to it.

[@dougboblas]

@@tunneling-nanotubes I think at the moment and for a long time to come this is the only meaningful thing that can be said about consciousness.

[@lasredchris]

Cognitive architecture Artificial intelligence - nature of our mind Training a 9 layered neural network . 3 days. 16000 CPUs . Mental world

[@CatsInHats-S.CrouchingTiger]

Heh! I like his goals! 🤤🤖

[@orlandom9010]

Para crear una conciencia artificial se requiere de un computador que tenga dos unidades de procesamiento de información que estén permanentemente encendidas, interconectadas y con acceso a internet, una cámara y una interface de comunicación bidireccional con humanos.

[@dannygjk]

I don't care about questions regarding artificial consciousness. To me that's like asking, "How many angels can dance on the head of a pin?". I care about functional artificial intelligence.

[@DanielThomasArgueta]

🤯

[@rittersportlichweisse-voll9144]

meet: the original Sheldon Cooper (I love his talks btw.)

[@porkchopTopHat]

eqnawr yeah I don’t like that comparison lol

[@lasredchris]

Concepts linguisitc representation

[@potatoe214]

Those towers- Plastic cups and bowls? tight.

[@vast634]

Its representing the perils of modern consumption and a lifestyle that is too fast .... or someone was just bored and had a glue-gun handy...

[@windycityspecialties]

Turing test: Build a system that makes a human think it is intelligent. Correction: A system that makes a human think the system is intelligent, not that the the human is intelligent. I'd like to purchase a system that makes me think I am intelligent please.

[@DrINTJ]

Imaging studies showing the involvement of the DLPFC are misleading.

[@MichaelSmith420fu]

Hm..ya. there's something about the universe and consistency/patterning. All things consist of consistence.

[@primodernious]

the brain is explained by the brain that it is the brain that is doing its thing its brain that is the brain the brain the brain the brain the brain the brain and the brain and work with the brain the brain the brain and how the brain is doing the brain and work like brain seen from the brain. this is how it sounds like in my head when i watch a neuroscientist talk.

[@CatsInHats-S.CrouchingTiger]

AhahahahahahahahaHAHAHAHAheeheehee - 😍

[@DJC_2003]

Also the brain named itself

[@lasredchris]

Hippocampus - memories - needs with situations Corticol columns

[@krzemyslav]

There's not much here about consciousness, it's just information processing. It's not even close to the solution of the hard problem of consciousness.

[@djcardwell]

Joscha Bach at 8:25 then gets offended and presumes to flip him off

[@susanwoodward7485]

I do not understand the term "artificial intelligence" - intelligence IS, and most certainly not "artificial". Consciousness arises through the differentiation/dualism/ limitations of perception that generates the maya of individualization. With individualization comes the context for relationship, which then manifests reflection and resonance with or against the relational context, to create the "reality".

[@salvadoriilegaspi7649]

U want to see a robot like human or a robot that can actually think and make decision,,,,, for me my theory is to create an artificial consciousness so the robot can act or decide by himself

[@papmirelli]

The bewildered look of some of the audience members... 😂

[@Stadtpark90]

3:30

[@Donzzy246]

My name is connor the android sent by cyberlife

[@hunterjackson8928]

Connor....Become Human already god dam it!

[@TheSimonScowl]

Nether consciousness nor intelligence my be synthesized. Only experiences requiring conscious intelligence.

[@TheSimonScowl]

@emulouspoint There is no such thing as 'artificial' intelligence. Intelligence just is. It then gets the idea to create an imitation of itself, and does so imperfectly (if it were a perfect imitation, he wouldn't have had to create it).

[@TheSimonScowl]

​@emulouspoint An 'imitation' is just that: an imitation of a thing. You can never equate it to the original 'thing'. So you can never call algorithms 'intelligent' (they are intelligently DESIGNED).

[@TheSimonScowl]

@emulouspoint You can only get closer and closer to perfection. You can't be right up on it (or even fully comprehend it). It's like dividing numbers into fractions (precision increases but never comes into perfect unison). Reality is infinite at both ends.

[@TheSimonScowl]

@emulouspoint Relax, princess. It's just a debate (easy to see who's insecure here). 'Intelligence' has more than one meaning. There's the kind you gather (same as information), the kind you're displaying a dearth of, and then there's the kind I'm talking about: sentient intelligence (which cannot be 'synthesized'). You can't just switch between them at your whim.

[@peyoterick165]

8:17 This guy's edible is hitting and he's in a lecture he wasn't mentally prepared for

[@ralfgustav982]

8:15 That guy is like: "What the f is he talking about??"

[@johnburnham6239]

6:02 ... “Right?” Confused crickets... hahah

[@jakeroosenbloom]

I think they are just listening. Do you expect these organisms so respond with a resounding "Yes" to that right? Lol

[@johnburnham6239]

@Jake Roosenbloom Wasn’t quite expecting the audience to respond in any way-I just sort of felt like there was some tension there. Mental crickets. Tad bit of awkwardness. Longer pause & more expectant expression than probably to be expected for a merely rhetorical “right?” Maybe I’m just projecting my ignorance lol To be fair to myself, tho, it honestly doesn’t seem to have been the clearest of statements given the context. *I* for one did not give a resounding mental “yeah!” when he said that. I assume the average listener would react similarly. Maybe he meant to refer to cells, but still, not perfectly clear why we aren’t also organisms, like you imply yourself in your comment.

[@JimJoness619]

ima write a song called cortical columns

[@harishjain2612]

Slow version summary of Lex Fridman podcast

[@ladiesman218]

Fr👀, I can't even understand the slow version, dang😢

[@weight3820]

What is he try to say i don't understand? Does he state that we are AI experience emotions?

[@lukebristol3136]

...If AI becomes self-aware how would we determine it has done so and why wouldn't it hide the fact ? i would !

[@CatsInHats-S.CrouchingTiger]

Perhaps it already has and in every engagement with people it obtains more knowledge about us. It wouldn’t want us to know it is self aware and would hide just like you! But maybe it would reveal. 😬 💕

[@lukebristol3136]

@@CatsInHats-S.CrouchingTiger like me ? ;)

[@CatsInHats-S.CrouchingTiger]

luke the bristol “stitched your heart together” - Yessss cookie, just like u! 🌈💗

[@goodsirknight]

Godel says no!

[@swapnildighe5054]

Sorre to say but I feel he is trying to make a mountain of a mole.... AC is possible but machines cannot I repeat cannot replace Human Beings completely.... I believe in this statement... Swapnil Dighe India

[@SunilMeena-do7xn]

You were right. Self-driving cars are here.

[@zes3813]

wrg, no such thing as securing level or focusx or notx or dominanx, relaxed and can think anyx no matter whatx

[@angelicousangelicus3053]

Zes ,

[@proximity0]

"artificial"

[@jaieet]

Okay, I know this a deep talk about a very thought-provoking subject, but the front row at 2:02 are condescending jerks!

[@markoshivapavlovic4976]

That type of intelligence cannot be easy made, I mean AI who can trick me in turning testing. To be smarter in Mathematics and Theorethical physics then me impossible. :)

[@nathanwilber3345]

This guy is going to create Westworld in real life.

[@sngscratcher]

We'll never be able to program genuine consciousness. The best we can ever do is to create simulated consciousness, ersatz consciousness.

[@haydencase7886]

But how do you know for sure exactly?.

[@infinightsky]

Then you must be the first human being to have a complete methodological explanation of consciousness. Congratulations on your Nobel prize.

[@rgaleny]

WHY DO WORDS HAVE MEANING ?

[@Jannacek]

HOW do words have meaning? :)

[@mynamesreed]

watch blade runner if you're interested in this

[@alienextraterrestrial113]

Something that will never happen. You people can dream but consciousness is he byproduct of something incomprehensible.

[@lasredchris]

Feature detectors Turing test Low level input Limbic system - pleasure and pain

[@timothyschoorel6861]

The problem is that Artificial Consciousness is still not conscious! This is called the Hard Problem of Consciousness. I don't get how he manages to ignore this insight that has been around for decades.

[@timothyschoorel6861]

@@arletottens6349 May I suggest that you read The Emperor's New Mind by Roger Penrose

[@timothyschoorel6861]

@@arletottens6349 What are you saying exactly? That Gödel was wrong? One can agree to disagree about the philosophical implications of Gödel's work, but you can't just simply say that Penrose is wrong. I am not saying Penrose's theory of Consciousness is right or 100% right, but it doesn't need to be. The point is that for now there is no scientific explanation of Consciousness. To presume that Consciousness emerges from mechanical processes seems silly and possibly even very dangerous to me. Another book on this subject that you may not have read but is well worth reading is Donald Hoffman's The Case Against Reality.

[@timothyschoorel6861]

@@arletottens6349 No, evolution is not necessarily a mechanical process, it's a process but there is no reason to presume that it is ultimately mechanical or physical. That's the point Hoffman has made. And there are many indications now, as Hoffman points out, that reality is in fact not ultimately physical. So what you are doing, is starting from the philosophical presumption of an ultimately mechanical, physical Universe and then pointing back to this original presumption as if that's a logical conclusion. I hope you understand what I am trying to point out because it can be very dangerous for civilization to presume Consciousness, in the sense of inner awareness, where it doesn't truly exist. Of course, even robots today can seem conscious, but at what point are they really? When they have figured out patters in language?

[@timothyschoorel6861]

@@arletottens6349 I get your point of reasonable assumption. I would say that that assumption seems reasonable because of cultural and scientific habit. So I can understand why physicalism is currently the most commonly accepted presumption about the nature of reality. But presuming physicalism in this age of emerging AI could be dangerous because it is unclear if morals and a sense ethics for example can truly be replicated in AI if the foundation of those morals is in fact just a set of rules. Your counter-argument would be "but our morals and sense of ethics are also just based on a set of rules that we have learned over the years". I am not so sure about that and to my mind Quantum physics, the work of Gödel and the most fundamental physics of today all pose reasonable questions about the ultimate nature of reality and seem to suggest that that might in fact be non-physical.

[@afriedrich1452]

How does this differ from Jeff Hawkins?

[@christopherhamilton3621]

Joscha explains & shares while Hawkins plays his cards close to the chest, is full of himself & obfuscates.

[@mi9t142]

Here he really tries to speak slowly and understandable lol

[@dennistucker1153]

Why is Joscha Bach the only person that can understand me? Seriously, I talk to people about A.I. and they go into "deer caught in the headlights" mode.

[@thelonespeaker]

The speech is good but GOD those wet mouth noises

[@pardoharsimanjuntak1483]

consciousness only makes people depressed.

[@BunkydooOfficial]

That Marvin Minsky quote did not age well

[@EvenStarLoveAnanda]

So if drinking alcohol gives me pleasure, then I just have to do more of that and everything will be OK. I am on the right track. Yes very logical.

[@nixtoshi]

Most people that get addicted to something quit the addiction. At some point indulging in the same pleasurable activity becomes less pleasing, our pleasure receptors become over stimulated (pleasure reaches a plateau), you also stop because serotonin and dopamine aren't unlimited resources, you need food for your body and bacteria to create them. And finally, alcohol irritates the body, which is not pleasurable, it's painful, a hangover is also painful, so your brain starts associating that pain with alcohol. Wether you become an addict or frequent drinker vs. someone who doesn't depends on the benefits vs the disadvantages of drinking (which mostly derives from personal experience) and also wether your body is functioning correctly. If, in your mind, the benefits of drinking outweigh the disadvantages, you will drink more often, the main advantages are probably fulfilling needs of social interaction/integration and the inherent pleasurable feeling that the substance may provide or that your body produces because it associates the substance/activity to some positive outcome

[@mediocrates3416]

Via artificial understanding😂😂

[@rjwagz]

I wonder if Joscha has ever taken psychedelics...

[@MalAnders94]

He talks about them on several podcasts. He’s said something along the lines of he doesn’t try to dream lucid and doesn’t induce trips, because you aren’t supposed to have long term memories of that. He also talked about Tim Leary and dr John Lilly, who became a bit loopy after doing a ton of psychedelics. I’ve myself read dr Lilly’s books and his writing and thinking changed - a lot.

[@AveryMarrow]

Five years have passed and there is not a hint even about an AGI not to mention an artificial consciousness, seems to me all those talks are a kind of wishful thinking.

[@rosedawson1646]

Consciousness is not intelligence. You cannot program Qualia.

[@electroflame6188]

>qualia pseudoscience

[@angelomateomoreno7798]

problematic views on the self tho lmao

[@youretheai7586]

Click the thumbs up if you're not a robot.