Analogy as the Core of Cognition

In this Presidential Lecture, cognitive scientist Douglas Hofstadter examines the role and contributions of analogy in cognition, using a variety of analogies to illustrate his points.
Stanford University:
www.stanford.edu/
Stanford Humanities Center:
shc.stanford.edu/
Stanford University Channel on KZbin:
/ stanford

[@rileyrfitzpatrick]

This video has been bookmarked for 3 years and I am finally here.

[@TheMNMmatt]

We're glad you're here :)

[@capoeirastronaut]

🤣

[@gabesmokeymartatom]

Where have you been?

[@jaykshirsagar596]

Hahahahaha doooooood SAME

[@rileyrfitzpatrick]

@@gabesmokeymartatom procrastinating

[@DanielBrownsan]

Fast-forward to 13:40 to hear the actual presentation. (13 minutes of introduction? Seriously?)

[@robbes7rh]

Thank you for that! ditto

[@CassandraDarling]

thank youuuuuuuu

[@bobpolo2964]

I love you

[@apx622]

THANK YOU

[@trdi]

The introductions are sometimes really getting out of the way. This one was almost grotesque. Nobody cares, get off the stage.

[@mowafagouli]

Read GEB twice and have to say that's not an easy thing, yet I still feel I need to check it out once more. Every time I learn so many things about myself and others, that I feel it would be a nice idea to have it as part of public literature.

[@mikepeterson78]

All high level literature should should be donation only for obvious reasons

[@patbrown454]

Yes, all should read GEB!

[@SimulationSeries]

we love you Doug! one of the greatest synthesists still alive today, inspired millions around the world with his ideas!

[@jonschmid3904]

He comments at 28:02 (in the 1:08:37 version of Apr/'20) that his thought -- which he should have described as a micro-seconds short mental picture -- was so "fleeting" that it nearly escaped his notice. As a songwriter, a key to capturing the purity and power of such a flashing mental picture is the following. Be very aware of these flashes and try (to the extent possible) to slow them, track them, and absorb them. It's a challenge.

[@chuchaichu]

WOW, the best way to enjoy my solitary Xmas eve 2020. WOW

[@shabnamsingla9566]

Yes!

[@utuonebrown2273]

Almost identical to my solo Xmas eve! I watched his lecture on Godel numbering last Xmas

[@Kevin-lj5pq]

Thanksgiving 2021 represent

[@DarkShroom]

omg Hofstadter , i am gonna have to rewatch this, it's incredible actually as I find GEB difficult at times this really helps, really thanks for existing :)

[@benjaminhamilton2338]

At 59:30 he's talking about inadvertant combinatory words and pronunciation distortions, and unwittingly says something like "at one point I must have made 500 werrors per class". Combining w(ord)+error. Hah, awesome. Great talk btw, love it.

[@thedocta_certified]

It was more of a "wor-errors", the word 'word' was not blended with 'errors', it was abruptly cut by it.

[@tyranmcgrath6871]

Portmanteau

[@parrotraiser6541]

Nice to see a presentation with transparancies again. His book on analogies is great.

[@TodaySatan]

I’m not sure if anyone else has had this experience, but I’ll be contemplating something and I’ll be looking at something and from looking at whatever it is (usually it’s a pattern), I’ll everything becomes clear and I’ll understand

[@thewealthofnations4827]

When I decided to take a break from rap music and other mindless pursuits I listened to this lecture as a developing young man. It was a turning point for me to think differently about the world. Thank you Stanford for allowing me a "seat" at the table to see something I never would otherwise.

[@DubLightning]

what are mindless pursuits

[@thewealthofnations4827]

@@DubLightning I was seduced by the allure of reality tv and more base desires. Pursiits of less substance pursuing things of litte substance or consequence. At what point in your life did you find this lecture?

[@williamkibler592]

Thought is (seeking) the highest level of abstraction. putting one's fingers on the essence of the situation, bouncing back and forth between the situation and the essence that it had reviled in its memory/memories

sounds kinda erotic when you put it like that (also yes, good analogical description...)

[@misaelrivera2347]

The essences of personal experiences play in present moments as memories as to HARD drive + central processing unit is to software.

[@uetzgenfatz]

I love how he uses an OHP. Great talk! As a linguist, I immediately thought about Hermann Paul when Hofstadter mentioned proportional analogies.

[@ozachar]

The two minutes starting at the 1h mark to 1:02 are basically describing an AI algorithm self-sensations within his brain. Amazing.

[@sealchan1]

This makes a great deal of sense. If you consider that the cerebral cortex, which houses areas dedicated to language, is cross-mapped with other cortical areas dedicated to sensory processing and also higher levels of cognitive function, then language is immersed/housed in a massively, interconnected, parallel-processed associational matrix. The sentences we speak must "crystalize" out of these inter-twining, systemically-mapped, scrambled shadows of one map onto another. Given this, then the sorts of things Hofstadter is talking about are simply measurable byproducts of that mind-brain behavior/structure.

[@johneric3886]

What a well-written paragraph Sir.

[@jsmdnq]

In neural network design the same thing happens. The logic functions that a neural network encodes and finds during training also end up having "shadow" functions that it also inherently trains for. E.g., you might train a neural network on data in a way that it discovers a hidden A & B logic function... but in doing so it will also be trained on all the various other logic functions to some degree or another with some probability of the input being "captured" by them. The brain seems to be a large network of essentially logic functions(due to the chemical basis which comes from the properties of electricity which are "logical") and hence functions nearly the same. That is, the nature/structure of the brain as being a *network* of connected "switches" has inherent structural features that exhibit a sort of "shadowing" that you describe. In some sense it is the nature of encoding information/knowledge/experience in to the "matrix" that produces these "byproducts" and, of course, which are exhibit in the output(in behavior aspects, social, experiential, and even medical such as brain damage). There is somewhat of a "holographic" aspect to these large networks as they try to learn models of their inputs and the data modifies the switches. It has both local and non-local aspects. A large part of mathematics is figuring out methods to descramble information. That is, data can have many different interpretations depending on ones "point of view". It seems that, in fact, this "shadowing" is actually due to the complexities of information itself. Basically knowledge is fractal and our brains and mathematics are effectively linear systems of very high dimensionality. This makes encoding, processing, and storing information very error prone in the same sense as trying to linearize non-linear data is error prone. It may be that the human brain has developed different functional parts precisely due to the informational aspects(sensory inputs produce electrical and chemical effects that modify the physiology over time) and these act to decypher/decode information. This then modifies the DNA so that these modifications evolve over time(so DNA is not static of of course but is constantly evolving even from day to day experiences).

[@jsmdnq]

The way to think about what I just said might be to analogize with that of acoustic modes.Sound waves produce interference and these modify the particles in the air to change the density patterns. Similar information as it enters the human mind causes physical density changes due to the patterns they produce. Through repetition of input they will modify the physical distribution and relationships of the brain matter so that a sort of "encoding" takes place. The same type of think happens with neural networks/AI where input->network->output, when repeatedly trained on the same input will reach a fixed point where the network no longer changes(f(x) = x). So in some sense the brain is just a system that can be reconfigured to represent information. Through repetition the "device" gets it's internal structure modified to represent that information more accurately. Over time it more accurately approximates the information it receives to reproduce it. This happens both at the individual level as we learn and over generations as the the "blueprint" of the brain is encoding in to the DNA. Artificial networks(and other systems) can do the same. The more connections, the more "adaptable"/"plastic", and the more logical(essentially the atomic parts(neurons) can act as simple discrete functions with repetition) the more "intelligent" it is. Humans seem to be the first to develop a symbolic language and this has allowed us to transfer knowledge through encoding it symbolically. This then has allowed for a more "abstract"(rather than just on our main senses) develop of the human brain.

[@sealchan1]

@@jsmdnq We have human culture as a higher layer to help keep our individual variations in our languaging in check. We try to keep words from morphing through our exchange of language and our socially shared sense of meaning and desire to be mutually understood. But even here, and in especially differing cultures and sub-cultures, we have a continual morphing of language and meaning. This morphing is not so quick that we can't continue to translate different cultural mappings onto each other though not in a simplistic one-to-one way but rather a complex, non-linear way approximated by our rational understanding which seek the simplicity of linear relationships (and the ideal/conceit of logical concision). Then again at the lower level we have the basic reality of biological similarity (in body and in the instinctual needs that must be met) and external physical reality which provides another objective check against both the variability of our language and against its petrification into inadequate simple, linear rationalizations. Who we are, how we know is all a product of this rich layering of systemic inter-relationships.

[@jsmdnq]

@@sealchan1 Yes, I agree but I do not think that is something out of the ordinary. How else is it suppose to work? I would also say the "morphing" of language you speak of isn't something we have some awareness of and try to prevent it from morphing. It happens the way it happens because that is the only real way it can happen. Maybe it is due to linearity or maybe it is simply due to no other mechanisms working. E.g., we can say "The human brain has evolved to be highly optimized for social data processing" or, the human brain actually has simply evolved with human society. They are one in the same. You can't separate them. Our language too morphs as it does and it is not repeatable from what we have evolved to. If it evolved in a different way then we would be different. It evolved the way it has because of the entire state of the universe and how it has interfaced with everything else. While things may seem complex and chaotic things are precisely the way they are because it could be no other way. E.g., when a billiards player hits the cue ball all the balls are set in to motion precisely based on the state of the universe at that point and evolve according to the rules of the universe(which is mainly that of physics). But even the player himself is just part of the evolution and state of the universe. I think it is our ignorance of the universe that causes us to be amazed by what we experience. I imagine though that if we knew every detail we would realize everything is necessarily just a causal outcome and fits perfectly as we would expect. E.g., the way the human brain has evolved is necessarily the way it should have evolved by all the evolutionary events that have shaped it. It couldn't be anything else just as the cue ball hitting the 8 ball could not have done anything else but what it did. Things would seem far less mysterious if we could see the universe working and comprehend it all. Since we can't even come close we approximate things(no choice) and linearization is the means of approximation(no choice, at least by mathematical standards). In some sense our own ignorance of the universe is reflected back in to our consciousness and everything we do. One can't actually unravel it until at the every end of evolution... maybe. Most of what people describe about the structure of human evolution is really just differential topology. E.g., things evolve differentially(smoothly) but what else could it be? E.g., suppose our brains would have evolved differently... well, we would be different but almost surely still asking the same types of questions. Some of the specifics might change but overall nothing would really have changed(everything and nothing).

[@hillarysemails1615]

This guy makes such clever quips. He is a delightful speaker. We could all learn from his use of humor to illustrate his points and keep the audience's attention.

[@0olong]

Thanks man, that is some seriously voluminous introducing they got in there.

[@SomethingEternal]

Sometimes these transient word blends become so ingrained, you intentionally memorize and carry them forward. One of my personal favorites as I've grown older and math has become more integrated into my daily life: Probablemistic. Probably problematic. But I don't have the statistics. It's a blend of 3 words and sums up an entire paragraph explaining that I don't see a problem, but I expect one when scaling, and that I'd like to have more proof before believing in the suggestion. One day this word came out, completely by accident. The crazy part is that my coworker naturally understood exactly what I meant without it needing explained. And I caught it and realized that it was a really good term in a sense. Fits this talk nicely. It's also a good measure of group cohesion when you find people using these word blends naturally amongst each other: If they understand and can naturally, without great hesitation, use the same localized terminologies.. They probably think in ways that they find easy to communicate to those around them. My older brother, who was my practical machining tutor of a sort at a young age, often uses the term "force and intent." But at times, when a high pressure system is leaking and he needs to communicate that I have to close this with "force and intent" will instead just shout the word "FORTENT!" This is his way of quickly yelling "just close the dang thing as hard as you can, it's not going to break but you don't have much time." That one word is a safety mechanism that has prevented damage and injury, resulting from one time he was too worried and panicked to communicate the whole phrase. Ironically, as we're both former military.. We find the formal language and syntax of communications in the military to be lacking in this regard: Concepts that can be communicated in half of a second take 6 seconds to express.. While they'll use acronyms like OPSEC to shorten terms in a briefing, in the field they won't shout something like "Bear 3, 4!" Which is something he and I use in online gaming: 4 targets visibly centered at bearing of roughly 30 degrees. Response time is key, and finding shorthand is actually a way to react faster... But because it sounds so informal, you'd never hear that on a radio in operations.. Even though for us, it's become an effective communication device for expressing the most important details as quickly as possible.

[@IIllytch321nonadinfinitum]

The "fight" to come to a word or phrase I find really interesting. Not only that but the implications it has on communication. All these subtle dynamics sometimes are not so subtle and may completely change the direction of a conversation without either party planning it. I always found this an interesting limitation in human cognition. Nice to have someone else think so and lay it out a bit more thoroughly, although I wish it was even more thorough still, heh.

[@RogerBarraud]

+IIllytch321 It's not purely a *limitation* though - sometimes it's an enhancement; How boring would it be if all conversations were completely linear? :-/

[@IIllytch321nonadinfinitum]

Roger Barraud Mind you it's been a while since I've seen this video. It's only a limitation in that humans have finite mental faculties and must succumb to the limitations it brings about when having a dialogue with someone. If there were two gods that knew everything it seems to me that there would be no need to conversation between them in the first place. I'm not saying it makes conversation completely useless or anything. We can be limited, but still have interesting conversations, heh. Sometimes it might bum me out, but such things are a part of being human. And I wouldn't say it's about conversations being linear or not (unless this term needs to be disambiguated), but rather conversations have more or less an ability to be controlled and guided by the participants, all things being considered. I think we're kind of a slave to the factors governing the progress of the conversation and take it for a kind of real-time ride, heh. I have no problem with this. I have a high external locus of control and have made plenty of peace with this. I respect your positivity though.

[@IdolKiller]

Love this. Reminds me of my young daughter using "moviater", blending movie & theater

[@approved7397]

loved the applause after “sorry, I don’t use Powerpoint” lmao

[@thenewtowncryer]

Very interesting and helpful- thank you.

[@Jan96106]

Creativity is the ability to see analogies in disparate things, as evidenced by the metaphysical poets who, as described by Johnson, yoke together heterogeneous images by violence in metaphysical conceits. When we nurture our ability to analogize, we develop our intellect and our creativity in whatever area of study we apply it to. Even the way science sees the world, as Kuhn has pointed out, is with ever-shifting analogy.

[@Al-himathy]

Can also be called aphorenia hahaha but the metaphysical poets were wild

[@Brickzot]

"Doug being disillusioned by his father about subscripts being analogous to superscripts in mathematics" would have been a good title too.

[@timgeoghegan7458]

‘Why is a raven like a writing desk?’ Because each is analogous to the other. In the context and circumstance. But is the how the same as the why? And if we solve for why, and understand why it is, is that the same as how we solve for how? Or is 'it' something else entirely, without analogy? Hopefully there will be some more lectures to help clarify this.

[@LeeGee]

Superb. Thank you for posting.

[@ts4gv]

douglas enters the stage at 13:30 first word leaves his mouth at 13:43 giving a 13 minute intro to a 60 minute speech is a surprising move

[@YawnGod]

Inflation is more than just a currency phenomenon.

[@Chris.4345]

The Presidential lecture and others like it are also meant to celebrate the person and their ideas.

[@raydot]

I've spent a lot of my life struggling with GEB. One thing that strikes me about GEB, SL, and this vid is this: While Hofstadter is clearly an amazing thinker, the whole never quite seems to be the sum of its parts. The MU theorem, for instance. It seems that throughout that entire analogy he never quite makes the point he's making. He always almost makes it, but never quite makes it. And that's the whole problem I have with his work. Having said that, I'd love to work with him!

[@capoeirastronaut]

Strange loops are a massive contribution, on their own.

[@Robert_baiguan]

That may just be the nature of truth right? In Chinese philosophy there is a saying, the truth, once spoken, is no longer the truth

[@pocket83squared]

@@Robert_baiguan _The truth, once spoken, is no longer the truth._ Hmm. Now that you mention it, I don't think that's quite right.

[@Robert_baiguan]

@@pocket83squared my statement must be false :P

[@bkalnoky]

​@@pocket83squared the truth is not compounded by everything that is correct. The truth is bigger, with more categories, some qualitatively distinct from one another.

[@saraessex8124]

Very interesting discussion on cognition

[@weweblog]

🎯 Key Takeaways for quick navigation: [17:18 URL Analogy is the Interstate Freeway System of Cognition 🚗] [Section key takeaway: Analogy is the most important part of cognition, connecting all the different pieces together.] - Analogy is not just a small part of reasoning, it is the central mechanism of cognition. - Analogy is like the interstate freeway system of cognition, connecting all the different parts together. [18:27 URL Analogy is the Motor of the Car of Thought 🧠] [Section key takeaway: Analogy is the driving force of thought.] - Analogy is like the motor of a car, it is the driving force of thought. - Analogy allows us to categorize things and understand their essences. - Analogy is not just a way to compare two things, it is a way to connect them at a deep level. [19:56 URL Analogy Making is the Perception of Common Essence Between Two Things 💡] [Section key takeaway: Analogy making is the ability to see the common essence between two things.] - Analogy making is the ability to see the common essence between two things. - Analogy making is what allows us to learn and to understand new things. - Analogy making is not just a way to come up with creative ideas, it is a way to deepen our understanding of the world around us. [24:20 URL Analogy is a Reminding Event 💡] [Section key takeaway: Analogy is a reminding event.] - Analogy is often a reminding event, where one thing reminds us of another. - Analogies can be fleeting and transient, and may not have any purpose. - Analogies are a natural part of human thought, and we are constantly making them. [26:47 URL Analogies Happen All the Time 💡] [Section key takeaway: Analogies happen all the time, even if we don't notice them.] - Analogies are all around us, and we are constantly making them, even if we don't realize it. - Analogies can be simple or complex, and they can be found in all aspects of our lives. - Analogies can help us to understand new things, to solve problems, and to create new ideas. [28:41 URL Snow Shadow 🌨️] [Section key takeaway: Analogy can help us to expand our understanding of the world.] - Analogies can help us to see the world in new ways and to make new connections. - Analogies can lead to new insights and discoveries. - Analogies are a powerful tool for learning and creativity. [31:36 URL Analogy Expands Concepts 💡] [Section key takeaway: Analogy expands concepts.] - Analogy helps us to see new instances of concepts, which expands our understanding of those concepts. - Analogy can also lead us to create new concepts, by combining existing concepts in new ways. - Analogy is a powerful tool for learning and creativity. [33:51 URL Pluralization 💡] [Section key takeaway: Pluralization is a way in which a single entity becomes a wide category.] - Pluralization is the process of creating a new concept by generalizing from a single instance. - Pluralization can be used to create new concepts in any domain, from mathematics to art to science. - Pluralization is a powerful tool for expanding our understanding of the world. [36:34 URL Chunking 💡] [Section key takeaway: Chunking is the process of grouping concepts together to create new, more complex concepts.] - Chunking is a fundamental process in human cognition. - Chunking allows us to learn and understand complex information. - Chunking is also essential for creativity and problem-solving. [39:04 URL Analogy Helps Us Think with Complex Concepts 💡] [Section key takeaway: Analogy helps us to think with complex concepts.] - Analogy allows us to chunk complex concepts into smaller, more manageable pieces. - This makes it easier for us to understand and reason about complex concepts. - Analogy is a powerful tool for learning and creativity. [40:55 URL Examples of Complex Concepts 💡] [Section key takeaway: Analogy can be used to understand complex concepts from a variety of domains.] - Hofstadter provides a list of complex concepts from a variety of domains, including science, technology, politics, and culture. - He argues that we are able to understand these concepts because we can make analogies to simpler concepts that we already understand. - Analogy is a powerful tool for expanding our understanding of the world. [44:29 URL Analogy Making Retrieves Concepts from Different Levels 💡] [Section key takeaway: Analogy making can retrieve concepts from different levels of our conceptual hierarchy.] - Hofstadter argues that analogy making is a powerful tool for retrieving concepts from different levels of our conceptual hierarchy. - This allows us to make connections between seemingly unrelated concepts. - Analogy is a powerful tool for creativity and problem-solving. [46:29 URL Analogy Is Not Just About Visual Categories 💡] [Section key takeaway: Analogy is not just about visual categories.] - Analogy can also be used to understand and reason about abstract concepts, such as emotions, relationships, and situations. - Analogy is a powerful tool for expanding our understanding of the world beyond the concrete and visual. [48:46 URL Compound Words, Phrases, and Proverbs Are Also Categories 💡] [Section key takeaway: Compound words, phrases, and proverbs are also categories.] - Analogy can be used to understand and reason about compound words, phrases, and proverbs. - Analogy is a powerful tool for understanding and using language. [50:52 URL Proverbs Are Situation Labels 💡] [Section key takeaway: Proverbs are situation labels.] - Proverbs can be used to categorize and understand situations. - Proverbs are a powerful tool for navigating the social world. Analogies are the subterranean fight going on inside every word. Hesitations, distortions, and funny intonations are revelatory of the fight between words and the analogies that are struggling to take over. Native speakers often use analogies without even noticing. Examples include "count my lucky stars" and "hit the stack." Analogies can be funny, but Hofstadter collects them to understand the human mind. Every effortless category assignment is actually a seething sub-terrain battle of analogies. When the battle is a landslide, there's no evidence. When the battle is close, there's evidence galore. Thought is the highest level of abstraction seeking, and analogies help us bounce back and forth between the actual situation and the essence of that situation. Einstein used analogies to create the concept of the photon. He saw a similarity between the bell-shaped curve of the ideal gas and the blackbody spectrum. He hypothesized that light, like an ideal gas, is made of particles. Hofstadter concludes by comparing analogy to the shadows in the snow. Just as the shadows can reveal the hidden shape of an object, analogies can reveal the hidden essence of a situation. Analogies are a powerful tool for thought, discovery, and creativity. Made with HARPA AI

[@zengfong4666]

your majesty, this is a hell of a new clothe

[@LimeyLassen]

Reminds me of how a lightning bolt branches before committing to a path.

[@FuKungGrip]

Forbes has a great article on what Hofstadter has been up to. He's a purist, wanting to understand how the mind works.

[@tokotokotoko3]

I would also say that its kind of obvious that the basis of our thinking must be associations, so I think you're right, the idea cant be really new (I bet the Greeks already came up with it - they were always first ;)). But analogies are one step further I think - and its an excellent talk. I enjoyed it a lot.

[@tomerbauer]

It seems to me that the analogies in the vacuum case with DH’s daughter and in the case of the airport and the woman with the suitcase have a pragmatic purpose. He became aware of these analogies *after* the deed, but they may have guided his behavior *towards* the deed subconsciously. That is, he chose to show his daughter that the other button doesn’t do anything because that’s what his brain knew to do from an analogous situation in his past; he chose to anticipate the lady’s movement at the airport and adjust his own lateral movement while walking because his brain had just learnt a new coping behavior and had to test it to store it.

[@bkalnoky]

i think that's too obvious. It's more like.... storing mental states and revoking them when we see something analogous, because our brain is just like that.... pattern seeking. So we can in general conceptualize future. And anticipate events. but it does is on an abstract level. It is very obvious that the analogies mentioned are totally incomplete and rudimentary. They are just an example to showcase how are brain is doing this all the time with everything ... if it doesn't do it too much it probably means you're thinking less and this can mean you're less concious. Scary stuff. I dunno when i drink like 8 beers i literally stop thinking, but i am still conscious. i guess there's lot of very deeply ingrained analogies going on in my head still, so firmly asserted, they are the ground blocks of anything else going on in my head.

[@IsaacDarcheMusic]

Hofstadter did not come up with the analogy as cognition-core hypothesis. Ludwig Wittgenstein proposed the same idea in his "Philosophical Investigations," published 1953.

[@pulse2781]

Cool story

[@ReallyLee]

I work with severely disabled non verbal kids. A much simpler mechanism than analogy I see is a matching process between incoming sensory data and a memory image (think of it as a plane or stack of layers). A shiny object in the field of view results in the brain turning on muscles to move the eyes and head to face the shiny object. The matching process can result in an OK feeling. If the incoming data exceeds the size of the memory area the child has sensory overload, and may avert their gaze. The brain is controlling the muscles. It is a great privilege to watch how the brain works. After seeing this in kids, I sometimes see similar actions in everyday interactions.

[@Debunker246]

I missed that

[@bluberrybrit]

this vid made my day!

[@RekMone]

Isn't analogy more fittingly described as re-cognition (=positive match in some aspect between a new impression/pattern and an older one), what'd put perception (=taking in a 1st impression) at the core of cognition?

[@Adhil_parammel]

Random neurons

[@ToddAndelin]

Great video.

[@bdeaner]

Freud's ideas are more relevant and validated by evidence today than ever before. This is just another example of that.

[@ratnamvadlamudi8620]

Very interesting concept👌👌👏👏👍👍

[@gabriellucena6291]

as I was having some food the analogy between the structures of concepts and the carbon structures that make life possible just ocurred to me..

[@DogStarAstrology]

I’m here now too. Thank you 🙏

[@pietrolunz]

I'm totally in love with this guy

[@Silly.Old.Sisyphus]

.. but it was worth the wait. thanks, Doug.

[@thepenguin114]

@ZeusDeusEx I just meant that -he- is not wasting -his- time. He's doing what he loves. That's what makes it not a waste to him, in his frame of reference. Maybe not in yours, but that doesn't matter to him.

[@ericarogers555]

Congnitive science is really neat!

[@DoubleBassX2]

"Pull no stops unturned" will be used ironically every time I can from this moment forward.

[@chrispaquette7513]

I think a more interesting (and pertinent) example of conceptual "chunking" would have been exponentiation in formal mathematics. This is because, for example, addition and multiplication (and by extension, exponentiation) can be shown to relate to each other *analogously*; that is, 5 x 4 is just a way of writing 5 + 5 + 5 + 5; but 5 x 4 may be understood *formally*, as something which is learned by rote. This would show that a "concept"--a conjunctive idea which works by analogy--is in fact very unlike a "black box", and that, in the formal sense, there is no difference between "primordial" and "complex" ideas. This in turn would give broader context to a discussion of how analogy works in thinking, since evidently not all processes in learning and computation relate to analogy (if this were the case, how would it be possible to make a "bad" analogy?). Hofstadter is very clear that "there are no purposes in analogy"; that is to say, the question "why does the brain make analogies?" is not relevant to their study and description. But to me what has always been interesting about Hofstadter's work is just this "why" which seems to beckon behind the formulation of his ideas, whether he likes it or not!

[@btanonymous]

Chris Paquette I agree with you about the “why” of analogies. Hof seems to write this question off as being uninteresting or unimportant. The power of being able to extract principles via analogy and graft them onto other areas of thought and perception seems like an important part of this topic. Especially considering how lighting fast and automatic this process is.

[@REDPUMPERNICKEL]

@@btanonymous "In 1981, an Indian man named Rajan Mahadevan accurately recited 31,811 digits of pi from memory. In 1989, Japan's Hideaki Tomoyori recited 40,000 digits. The current Guinness World Record is held by Lu Chao of China, who, in 2005, recited 67,890 digits of pi". I would like to automatize the formula that generates the numbers of pi and commit it to my unconscious such that, supplied with suitable initialization input, whenever I felt inclined to recite a sequence, the numbers would simply flow into my conscious mind. Seems to me this would take a lot less time, effort and storage space then memorization. It seems to me this should be possible. Or would we need something like a genetic upgrade designed to grow the necessary neural mechanism? Sorry, just felt like typing something.

[@clearmenser]

The Hof comes in at 13:31

[@LeoWhiteockhamsbeard]

At 12:38 introduction ends, talk begins soon afterward

[@blenderpanzi]

As a none-native speaker some of the blends go over my head.

[@calinwfraser]

In his publications (i.e., GEB and more recently "I am a Strange Loop") he does give some credit claiming there is a "similarity in spirit" between his and Freud's work.

[@kianatehrani]

Touché.

[@fabioarruda3184]

Really nice talk. I've came to the same conclusions independently. And I guess we can use this knowledge to go much further.

[@tschulz412]

Nice analogy.

[@pikiwiki]

thank you, thank you, thank you

[@therealMrBEAM]

awesome!

[@johneric3886]

Good stuff excellent used to humor.

[@DarkShroom]

so do we get to see the less formal next day?

[@BGasperov]

A couple of concepts that would be interesting to consider (and even study) from this perspective: a) the feeling of deja-vu, b) (internet) memes, c) Jungian archetypes

[@goldmineaustralia]

Hi. I have a mild hunch it's all in the Data Structures. As you say this too, a good data representation goes a long way :)

[@borntobemild-]

What I believe is, transformers in large language models are doing. They are adding an integer to analogies. Subscripts are analogous to analogies.

[@johnmartin2813]

Metaphor and simile are based on analogy. And language itself is based on metaphor. So yes, analogy does go pretty deep and is pretty pervasive.

[@daraghaznavi7171]

Your friend made an analogy between categorization and feature detection

[@99dynasty]

Dejavu would have interesting to hear discussed in this light.

[@fackarov9412]

i thinked the same thing

[@Tanthango]

Could someone tell me the title of the opening music? Quite a beautiful interaction between a clarinet and a synth sequencer. Oh, and there's a piano. Hmm, wonder if the piano itself is the sound that's also being sequenced chaoticly in 1/32

[@FOSology]

Yolo D Hopefully you grew up by now...

[@edwigcarol4888]

48:00 others responses than words to analogies we become aware of: an emotional state makes the link.

[@FirstnameLastname-kn5sw]

Analogy is the analogy of analogy

[@benhayes7985]

What does Mark Twain say in Huckleberry Finn about consciousness when the King and Duke are tarred and feathered?

[@nelsonwhite1463]

You have got to admit that the brain above all things seeks shelter, a common core analogy.

[@RogerBarraud]

+nelson white What's your ref for that? Dylan?

[@danielebonadiman4378]

"Ciao" from Trento :)

[@deadman746]

I couldn't help noticing that Hofstadter's language throughout is different from that of cognitive scientists who say the same thing only more comprehensively.

[@ericlifee7485]

Wow, just wow.

[@luizhenriquemoraismazzucco4526]

thanks

[@alauc]

With his father he has used red colour, with Monica green lights. I see analogy with traffic lights. Does he see it? Do you see what I want to say?

[@Diachron]

GEB has been a lifelong companion.

[@Tsnore]

His father was a Nobel laureate in Physics.

[@Nisstyre56]

@bdeaner Maybe the two theories are analogous?

[@davidwilkie9551]

similiar or similya = familiar + similar (old favorite) If the reality of the QM-TIME Universe is the dynamic picture of history here-now.., the superimposed cause-effect of complex events built of quantum chemistry qualities, then analogy is a component of the picture fitted to possible future events. It's a topical identification of mental processing of the deeper mechanism., and the mechanism is identified as "Holographic", the ultimate connection of all information at an Analog Computational Singularity in Eternity-now. It's the only model from which the phenomena of the Observable Universe can be derived in self-defining elemental statements of logic. "Ideal Gas".., Corpuscule of Light.., Embodiment of Quanta.. analogous procession. Thank you for the concept. Which, to be elaborately clear, makes .dt, e-Pi-i infinitesimal function shaping a phase-locked Eternity-now wave-particle Reciproction-recirculation, and Fractal Flash Interval holographic wave-packaging composition in/of coherence-cohesion sync-duration information, the superimposed Fluxion-Integral In-form-ation wave-envelope containment state of ONE-INFINITY Singularity-> sync-duration probability condensation. WYSIWYG here-now-forever continuous creation IS instantaneous cause-effect. Analogy => "word-picture" of accumulated memory associations.

[@Sanakreon]

I want to hear how he explains the act of planning with his "all of cognition is analogy" theory.

[@FirstnameLastname-kn5sw]

Try and plan without using language or symbol, try and use language or symbol without picking a word or sign, try and pick a word or sign without comparing the concept you want to communicate to yourself and others by comparison with the concept and other words or signs Finding the best word or sign to describe something is analogizing and that's on the micro level On the macro level then planning is trying to make something happen in the future by adapting analogus actions from the past that have resulted in analogus results to what you want to occur in the future

[@vs7927]

does anyone know what's the song at the beginning

[@Megacaesar]

Heh. 0:44:12. A particularly interesting explanation on human thought.

[@Jy3pr6]

Break down of logical reasoning at 24:40- 25:00

[@bdeaner]

Also -- just did a PsycINFO search of peer-reviewed publications over the past decade. Freud was listed as a keyword in over 7800 peer-reviewed articles in psychology between 2011-2012. Hardly a "historical artifact."

[@Nintencrow]

@TheMidnighters34 bad troll, no cookie.It's a university lecture by a clever dude who talks about analogies and cognition, what's NOT to like?

[@sampathgunasena3781]

Old man Hofstadter should have told his son about the tensor notation. That's a good way to bring some consistency between superscripts and subscripts.

[@ManNamedBrUce]

Is analogy the same thing as isomorphism for him?

[@Kobe29261]

Here's a man who at 12 probably had no need of the world because he could keep himself endlessly entertained by the quirks and glitches in the fabric of time/space. This is like Game of Thrones for a nerd!

[@manthasagittarius1]

Thanks for that observation -- it is is a big, nontrivial distinction.

[@ljanardhan]

Great Talk! But I wonder why he didn't chose to call them patterns?!

[@eravulgachris]

Functionality ain't everything - think how clinical and just darned homogenous the visual part of this presentation would have been with Powerpoint. Praise be to the gods of analogue, say I, to the gods of uniqueness, to the gods of the personal touch! By ANALOGY with our own cognitive processes for hand writing we "know" Hofstadter much better after this presentation thanks to his use of slides rather than computer aided/personality diluted software.

[@OnlyNewAgeMusic]

One of the most interesting fields is Analogy of Astrology, which is by the way the true ancient science of Astrology. In any event there is no cause and effect, only synchronicity of Analogies.

[@qigong1001]

@holycow818181 Thats interesting. Its almost like a status thing to Not use powerpoint by science/math guys. And your professor is right! Computer geeks often fall into the powerpoint trap. They (okay, not all) use too many slides and don't really get to the "point". Doctors too! Come to think of it, most professional groups overuse it! I guess doing it by hand, or on the fly prevents that.

[@theinfrawolf6774]

"I wanna point out here..." - Points out here 45:50

[@haraldwolte3745]

The Semantic space discussed at 56:12 is very similar to the process of creating word "embeddings" in Deep Learning. 2 minute explanation here: kzbin.info/www/bejne/Z2mZeYiKd7ahprs

[@NeoShameMan]

The whole talk is really an explanation of how neural network works really.

[@DavidComdico]

The photographer Kenneth Josephson has made some interesting pictures of snow shadows.

[@krttd]

The 'semantic space' and 'word blend' concepts are especially poignant in the age of the transformer model (or self-attention in general). The idea of collapsing an abstract semantic concept into natural language by guessing the next word/token at each step seems closely analogous.

[@sercy-ws7ve]

chat, this one gets it

[@dreamiestseabean]

He'd be so fun to do the word association game with. He'd like. Take you places. Like. Make you see things man.

[@JCResDoc94]

45:00 not sensory information.