It’s impossible to see the world as it is, argues a neuroscientist [video](aeon.co)
aeon.co
It’s impossible to see the world as it is, argues a neuroscientist [video]
https://aeon.co/videos/its-impossible-to-see-the-world-as-it-is-argues-a-cognitive-neuroscientist
44 comments
Very interesting.
Reminds me a lot of The Idea of the World by Bernardo Kastrup, who also argues that the basic building blocks of existence is consciousness, and that idealism is a better model for the all.
Can you define consciousness in a material way? Is it an energy or a feedback system? Can it be observed or quantified?
I think any system that responds as if it is modeling another system is conscious.
But there are levels of consciousness, what differentiates “higher consciousness” is maybe the question you’re getting at.
That unfortunately is an largely ascientific question. Like “what’s a good map of Seattle?” there is no such thing as a map of Seattle, so there’s nothing for science to study. Seattle exists for sure. Some maps will help you meet different ends in Seattle. But there is no such thing as an objective map. List n things a map gets right, and I will tell you n things it gets wrong.
But if you want a starting place, you could temporarily accept, e.g., Timothy Leary’s map of consciousnesses. That one is quite materially grounded. Within that orientation you can do all kinds of science and learn many things about consciousness.
But there are levels of consciousness, what differentiates “higher consciousness” is maybe the question you’re getting at.
That unfortunately is an largely ascientific question. Like “what’s a good map of Seattle?” there is no such thing as a map of Seattle, so there’s nothing for science to study. Seattle exists for sure. Some maps will help you meet different ends in Seattle. But there is no such thing as an objective map. List n things a map gets right, and I will tell you n things it gets wrong.
But if you want a starting place, you could temporarily accept, e.g., Timothy Leary’s map of consciousnesses. That one is quite materially grounded. Within that orientation you can do all kinds of science and learn many things about consciousness.
You pointed me to https://en.m.wikipedia.org/wiki/Eight-circuit_model_of_consc...
Very interesting. I'm finding patterns in explorations of consciousness. The first four levels Leary describes could be Freud's id. The next 4 contain his super ego.
Compared to id = Dąbrowski's first level and super ego = 5th level.
Like you said these are all maps of consciousness. Psychological models that describe the actual physicality of a material mind.
I suppose I'm trying to determine how to physically recreate the material that will lead to a new consciousness.
We can't recreate Seattle from maps of Seattle to continue your analogy. Is it possible to recreate consciousness from our maps of consciousness?
It seems like that is what we have been trying to do in our pursuit of AI, when we know no more about what consciousness is than we know what Seattle is.
Very interesting. I'm finding patterns in explorations of consciousness. The first four levels Leary describes could be Freud's id. The next 4 contain his super ego.
Compared to id = Dąbrowski's first level and super ego = 5th level.
Like you said these are all maps of consciousness. Psychological models that describe the actual physicality of a material mind.
I suppose I'm trying to determine how to physically recreate the material that will lead to a new consciousness.
We can't recreate Seattle from maps of Seattle to continue your analogy. Is it possible to recreate consciousness from our maps of consciousness?
It seems like that is what we have been trying to do in our pursuit of AI, when we know no more about what consciousness is than we know what Seattle is.
There are no roads where you are going. But I can offer you some pointers:
Consciousness requires sensation and action. Without action animals don’t learn to perceive anything at all.
If I were trying to make a conscious artificial being the first thing I would do would be to sign them up for a fast paced online FPS and teach them to play.
Next give them emotions (0 dimensional precepts).
Any time your codebase starts to feel big, delete most of it. Your end goal will be maybe 1000 lines total.
Use matrices.
Consciousness requires sensation and action. Without action animals don’t learn to perceive anything at all.
If I were trying to make a conscious artificial being the first thing I would do would be to sign them up for a fast paced online FPS and teach them to play.
Next give them emotions (0 dimensional precepts).
Any time your codebase starts to feel big, delete most of it. Your end goal will be maybe 1000 lines total.
Use matrices.
Can you define material properties without consciousness observation? How would you quantify with out observation?
Not only empirically but personally?
Not only empirically but personally?
I can't observe a bacterium without a microscope and I can describe all the material properties of a microscope.
I know a microscope exists. I can see it.
I know a microscope exists. I can see it.
Discussions about perception and reality are always good, even if this article seems a little tame compared to what philosophers have argued about for centuries. But it's nice to see such a discussion coming from scientists, and not just philosophers. Physicists have especially given new life to these ideas though. You can go down an existential wormhole if you want to, where the implications of spacetime readily suggest the passage of time itself is an illusion (brief intro of one idea: https://www.preposterousuniverse.com/blog/2015/04/03/the-rea...). One implication of a timeless universe (there are many labels/forms: Laplace's Demon, eternalism, block universe) is that there are no privileged moments in time. So why do we find ourselves localized and perceiving this coordinate in the 4D spacetime block and not any other? Conceivably, in an infinte universe which ends in maximum entropy, there are many more observers where entropy is higher than it is now or was near the big bang. (Assuming observers can't form in the extreme densities of the initial singularity, there is an infinite time in comparison for Botlzmann brains to fluctuate into existence in the high entropy, low density "future" - ours in an in between point, but infinitely tiny when compared to an eternal heat death). About the only solution from physicists to avoid such Boltzmann brain problems is to invoke the multiverse (where some verses recollapse and can never form observers).
Disclaimer: only a layman but I believe what I've said is an accurate synopsis of the debate
Disclaimer: only a layman but I believe what I've said is an accurate synopsis of the debate
> why do we find ourselves localized and perceiving this coordinate in the 4D spacetime block and not any other?
Well we started out perceiving a 0-dimensional universe and 4 dimensions is as far as we’ve gotten!
We’re certainly at work on the next one.
Well we started out perceiving a 0-dimensional universe and 4 dimensions is as far as we’ve gotten!
We’re certainly at work on the next one.
We see things in the optimal way for us to navigate the world we evolved in. "As it is" has no meaning outside our perceptual affordances.
What evidence is there to suggest evolution has resulted in optimal perception?
My use of 'optimal' is more of 'sufficient to keep reproducing given current conditions' than anything.
If you're suggesting that evolution isn't the right vehicle for this process, then we'll have to disagree.
If you're suggesting that evolution isn't the right vehicle for this process, then we'll have to disagree.
Ah, the 'sufficient' definition of 'optimal' ;)
Yeah, sure, I can agree with that.
Evolution does appear to be the most likely candidate, yes.
Yeah, sure, I can agree with that.
Evolution does appear to be the most likely candidate, yes.
Tautology.
Every person sees the world through their own way. Therefore there is no way to see it as it truly is. The sum of the parts can never equal the whole, no matter how much effort is done to reduce bias.
Worthless to discuss further.
Every person sees the world through their own way. Therefore there is no way to see it as it truly is. The sum of the parts can never equal the whole, no matter how much effort is done to reduce bias.
Worthless to discuss further.
Any good reads/books on this?
Cameras have to filter out infrared.
How Can Mirrors Be Real If Our Eyes Aren't Real
This whole idea is called autopoiesis, it's in the book The Tree of Knowledge by Maturana and Varela. Those guys "scientifically proved it", to whatever extent science can prove something like that.
Could you provide more of an explanation and details/resources?
The guy has been doing the TED circuit for ages. I once looked closer at his claims and his supposed proof by simulation and came to agree with the more negative takes at https://news.ycombinator.com/item?id=11588698
I think the devil is in the details when it comes to this argument. Some philosophy speculation ahead.
1) It's unclear which parts of our perception reflects the real world and which don't, and also how to measure how accurate they are. We obviously see objects as much more than what the photons representing them that hit our retina tells us. The brain creates all kinds of backstory and assumptions based on its model of the world, however, I would argue that the basic shapes we see are accurate.
So basically the geometry/shape of a face, of a chair, of <pick any object> is real, while the extended cultural, practical and personal associations the brain attaches to those are "not real" (or rather, we see a chair as a utility we can sit on, so we see both the shape of a chair but also potential uses for it, so it becomes an extended model of ourselves+chair, but that doesn't mean the shape isn't real). Also even though we can change our extended model of objects, we don't normally change our base perception, e.g. a chair always looks the same shape even though we may see different uses for it in different contexts. I would argue we really see the shape of the chair. This then leads to difficulty knowing how to define what is real/accurate and what isn't.
2) It's unclear whether the extended models we have are reflective of the real world or not, and if so how much. Just because I see a chair as a utility I can sit on, doesn't mean that it doesn't reflect reality. In fact you could say the model reflects a more abstract truth about how my body fits into a chair, so it becomes a more encompassing truth about bodies+chairs. Hell you could even include that it includes some information about how my muscles tire and need rest and so on.
3) I think the basic idea is true and interesting. We clearly do not see the world like a camera does, but I wouldn't go as far as saying it doesn't reflect reality at all. I haven't looked at this authors papers, but I saw an interview where he mentioned some statistical measurement of how organisms who saw the world as it actually was had a lower survival rate than those who do, and I'd be curious on how sound that argument is. The devil is in the details again and it sounds to me like such an argument could be interpreted several ways.
1) It's unclear which parts of our perception reflects the real world and which don't, and also how to measure how accurate they are. We obviously see objects as much more than what the photons representing them that hit our retina tells us. The brain creates all kinds of backstory and assumptions based on its model of the world, however, I would argue that the basic shapes we see are accurate.
So basically the geometry/shape of a face, of a chair, of <pick any object> is real, while the extended cultural, practical and personal associations the brain attaches to those are "not real" (or rather, we see a chair as a utility we can sit on, so we see both the shape of a chair but also potential uses for it, so it becomes an extended model of ourselves+chair, but that doesn't mean the shape isn't real). Also even though we can change our extended model of objects, we don't normally change our base perception, e.g. a chair always looks the same shape even though we may see different uses for it in different contexts. I would argue we really see the shape of the chair. This then leads to difficulty knowing how to define what is real/accurate and what isn't.
2) It's unclear whether the extended models we have are reflective of the real world or not, and if so how much. Just because I see a chair as a utility I can sit on, doesn't mean that it doesn't reflect reality. In fact you could say the model reflects a more abstract truth about how my body fits into a chair, so it becomes a more encompassing truth about bodies+chairs. Hell you could even include that it includes some information about how my muscles tire and need rest and so on.
3) I think the basic idea is true and interesting. We clearly do not see the world like a camera does, but I wouldn't go as far as saying it doesn't reflect reality at all. I haven't looked at this authors papers, but I saw an interview where he mentioned some statistical measurement of how organisms who saw the world as it actually was had a lower survival rate than those who do, and I'd be curious on how sound that argument is. The devil is in the details again and it sounds to me like such an argument could be interpreted several ways.
That fits with my observations matching the decreasing propensity to procreate with IQ. At some point we become literally too smart for our own good.
It's my understanding that this correlation disappears once you correct for level of education.
In other words, it isn't intelligence per se which lowers TFR, it's college. Which makes sense, adding at least four years before someone is earning a living will decrease the number of children they have, ceteris paribus.
Student loans aren't helping either.
In other words, it isn't intelligence per se which lowers TFR, it's college. Which makes sense, adding at least four years before someone is earning a living will decrease the number of children they have, ceteris paribus.
Student loans aren't helping either.
The limitations of empirical knowledge have been known for... a few thousand years now. It is a little disconcerting that (some) neuroscientists appear to be completely ignorant of basic philosophical problems.
https://en.wikipedia.org/wiki/Ontology
https://en.wikipedia.org/wiki/Metaphysics
https://en.wikipedia.org/wiki/Epistemology
https://en.wikipedia.org/wiki/Ontology
https://en.wikipedia.org/wiki/Metaphysics
https://en.wikipedia.org/wiki/Epistemology
Knowing something to be from both an epistemological perspective and from a neuroscience perspective lends further credence in some (many) circles, even if it’s less fun to dump on people that way.
One way of avoiding confirmation bias is to simply have no prior understanding to confirm though.
I which case you'll never have a second chance to make a first impression (tm)
In Academia, as long as you do the work to finish a manuscript and submit it, you will get another chance to make another impression.
The fact that Plato and Lao Tzu articulated that the noumenon is fundamentally unknowable does not make the fact that evolution results in processes which actively obscure raw data about the world uninteresting; it could theoretically be the case that evolution tended towards more accurate data over time in all processes while still being the case that base reality is unknowable. I think it's been obvious that this isn't the case for some time now; if you couldn't figure it out from the basic principles of evolution, it should have been obvious in light of chronostasis illusions that hack our perception of casaulity into hearing a sound before we press the button which triggers that sound by first conditioning us to expect the sound to come after an artificial delay and then removing that delay. But this doesn't make the modelling done by the researcher something done out of ignorance, I would assume the researcher is probably familiar with that line of prior work and is creating models that help us better understand the process. I do feel like the article is misleading, and the first two sentences are phrased as if the two views presented (accurate data v.s. inaccurate data) are opposed to one another rather than working together towards the same end (survival/procreation). I haven't watched the video, so I don't know how in-depth it gets, but I did upvote your comment despite my reservations because I felt like it added commentary and context that the linked text was missing.
I came here to say the same thing. It's very amusing seeing modern scientists try to speak about things like this while totally unaware of the historical conversations that preceded them
I feel like anyone drawing a line from deep learning to general AI also has this problem. And that appears to be a large part of Silicon Valley.
While I wasn't entirely convinced by a lot of arguments in the philosophy of mind as an undergrad 20 years ago, it definitely addresses problems "in the zeitgiest" now, e.g. with self-driving cars, etc.
And the conversation there is of a higher level than I see now.
I think taking a tiny bit of undergrad psychology, philosophy, and cognitive science, would help a lot of people think more clearly about AI. I didn't necessarily think that was true when I took it, but with hindsight I think there is a big hole in the conversation.
Gary Marcus is an AI practitioner with a psychology background: https://en.wikipedia.org/wiki/Gary_Marcus
he has a decidely different perspective that I think is a lot more grounded and reasoned: https://arxiv.org/abs/1801.00631
Also related: https://idlewords.com/talks/superintelligence.htm
While I wasn't entirely convinced by a lot of arguments in the philosophy of mind as an undergrad 20 years ago, it definitely addresses problems "in the zeitgiest" now, e.g. with self-driving cars, etc.
And the conversation there is of a higher level than I see now.
I think taking a tiny bit of undergrad psychology, philosophy, and cognitive science, would help a lot of people think more clearly about AI. I didn't necessarily think that was true when I took it, but with hindsight I think there is a big hole in the conversation.
Gary Marcus is an AI practitioner with a psychology background: https://en.wikipedia.org/wiki/Gary_Marcus
he has a decidely different perspective that I think is a lot more grounded and reasoned: https://arxiv.org/abs/1801.00631
Also related: https://idlewords.com/talks/superintelligence.htm
Agreed 100%. The media's conflation of "artificial intelligence" with neural networks is particularly irritating.
It would definitely be good for more "AI"-interested tech people to grapple with Searle, Nagel, Chalmers, even if you don't fully buy all their arguments.
It would definitely be good for more "AI"-interested tech people to grapple with Searle, Nagel, Chalmers, even if you don't fully buy all their arguments.
> It would definitely be good for more "AI"-interested tech people to grapple with Searle, Nagel, Chalmers, even if you don't fully buy all their arguments.
No, it wouldn't, especially Searle whose best known argument in this area is entirely for a distinction which can by definition make no difference in the material universe. It might have theological/spiritual applications (though beyond holding open the possibility than certain material facts do not demand the naively obvious spiritual conclusions without itself rebutting those conclusions, it doesn't likely to bear much load even there), but from a material perspective it's pure intellectual autoeroticism that has already been granted far too much attention by people who could contribute more to human progress by spending that time on (almost literally) any other pursuit.
No, it wouldn't, especially Searle whose best known argument in this area is entirely for a distinction which can by definition make no difference in the material universe. It might have theological/spiritual applications (though beyond holding open the possibility than certain material facts do not demand the naively obvious spiritual conclusions without itself rebutting those conclusions, it doesn't likely to bear much load even there), but from a material perspective it's pure intellectual autoeroticism that has already been granted far too much attention by people who could contribute more to human progress by spending that time on (almost literally) any other pursuit.
It also matters what problems you work on at the beginning of your career and who you are surrounded by.
People who work on things that nobody in their group has good answers for, become much more aware of other disciplines and the histories, as they go looking for better answers.
But if entering a group working on known paths, or existing ideas (being tested), then wandering too far from base camp is going to look unfocused. The existing path has to hit a dead end. Which in many fields/professions takes decades.
Herbert Simon started off with Political Science. But look where the path takes him as he doesn't find answers on what he is working on - https://en.wikipedia.org/wiki/Herbert_A._Simon
People who work on things that nobody in their group has good answers for, become much more aware of other disciplines and the histories, as they go looking for better answers.
But if entering a group working on known paths, or existing ideas (being tested), then wandering too far from base camp is going to look unfocused. The existing path has to hit a dead end. Which in many fields/professions takes decades.
Herbert Simon started off with Political Science. But look where the path takes him as he doesn't find answers on what he is working on - https://en.wikipedia.org/wiki/Herbert_A._Simon
> “These people haven’t fully accepted the intricacies of certain well known abstract philosophies? Dummies. They are obviously not contributing useful scholarship to their field.”
In all fairness, speculative philosophy and evolutionary neuroscience are completely different angles on the problem. Also the problem at hand is not arguing about the philosophical properties of empirical knowledge, but that evolution may favor an increasing level of "ignorance" in order to maximize survival. Which may be surprising to, say, some hegelians who believe that as time passes species get more enlightened.
You are completely right, I guess "Good artists copy; great artists steal".
Here's a related, although sort of tangent, NPR episode about originality: https://www.npr.org/programs/ted-radio-hour/321797073/what-i...
Also, even in neuroscience this is not new at all, Maturana and Varela researched and wrote about this subject extensively, their book The Tree of Knowledge is a good starter.
Here's a related, although sort of tangent, NPR episode about originality: https://www.npr.org/programs/ted-radio-hour/321797073/what-i...
Also, even in neuroscience this is not new at all, Maturana and Varela researched and wrote about this subject extensively, their book The Tree of Knowledge is a good starter.
I don't think it relates to originality at all. If you're interested in solving a problem, it helps to understand or at least be aware of the underlying foundations of it. A whole lot of modern scientists don't seem to understand the basics of empiricism (what science is) or philosophy of science in general. Thus you get situations like this one, where an apparently brilliant neuroscientist reaches a conclusion that a philosopher in Ancient Greece did 2,300 years ago. It's a waste of time and resources that could have been avoided by simply educating scientists on the history of philosophy.
Why is this the current state of affairs? My guess is because Anglo-American academic philosophy [1] has gone a little too far down the language analysis hole and subsequently isn't perceived as relevant by scientists. It's a tragedy and unfortunately the supposed irrelevance of philosophy continues to be perpetuated by science popularizers.
[1] The academic philosophy traditions in continental Europe and the English-speaking world have drifted apart over the past ±150 years. You see less of this blatant ignorance of philosophy in somewhere like France. https://en.wikipedia.org/wiki/Continental_philosophy
Why is this the current state of affairs? My guess is because Anglo-American academic philosophy [1] has gone a little too far down the language analysis hole and subsequently isn't perceived as relevant by scientists. It's a tragedy and unfortunately the supposed irrelevance of philosophy continues to be perpetuated by science popularizers.
[1] The academic philosophy traditions in continental Europe and the English-speaking world have drifted apart over the past ±150 years. You see less of this blatant ignorance of philosophy in somewhere like France. https://en.wikipedia.org/wiki/Continental_philosophy
> Thus you get situations like this one, where an apparently brilliant neuroscientist reaches a conclusion that a philosopher in Ancient Greece did 2,300 years ago.
Or, this scientist did the research, learned about what philosophers did 2300 years ago and then adapted/repackaged the message/knowledge in a way that fits his audience.
Also, in your statement you are essentially saying the scientist was not original because he reached the same conclusion as someone else before him.
Then you conclude that being unoriginal is a waste of resources.
So of course this relates to originality.
Or, this scientist did the research, learned about what philosophers did 2300 years ago and then adapted/repackaged the message/knowledge in a way that fits his audience.
Also, in your statement you are essentially saying the scientist was not original because he reached the same conclusion as someone else before him.
Then you conclude that being unoriginal is a waste of resources.
So of course this relates to originality.