Levin is the only guy I've heard who has made me see the possibility of what seems to be a genuine new and actionable paradigm for a deeper understanding of life and intelligence.
I am at the 48:07 mark, 'listening lightly', doodling Levin's insights into latticing thoughts of my own... I will listen a second time when I finish today... let it steep... and listen raggedly a third time and take notes that wander off in my mind last. After this phase I put down my notes and wait for Levin's and Lopez's ideas to comeback to me latticed in my thoughts and far more understandable.
Subscribed. I could not agree more with your guest about the current absurd concept of "alignment"; given that we don't align with ourselves ... whether at the social scale, or even individually. I've been saying exactly that in comments under dozens of videos on AI. *Alignment* is the mother of all red herrings. The whole point of it is to cover the corporate rears; they hire a few "alignment specialists" and tell them _"Try to catch up with us, as we move full speed ahead with this technology; and good luck telling us something we like to hear and is cheap to implement."_ Just so that if one day their AI takes over the world they can say *_"Sorry, your honor; we did TRY ..."_*
When I was the executor of my father’s estate I managed his 30+ properties. I had some grand ideas which incorporated an entire hillside of the city he had his properties. Redo the entire hillside with a self contained, energy efficient, living area with replaceable /interchangeable living areas entirely energized by solar, wind water and most importantly my invention of mechanical power energy production. Gardens, aquaponics, etc. one could theoretically move across country taking your living area and putting it into a different building in any location. One and done, transformative transforming. On another consideration: Why is the metamorphosis of a caterpillar to a butterfly is a power function of zeta? Protein folding: first you have to understand what is going on in a cell, what is a protein and why it is doing what it is doing. Prions first. Proteins run the show. Good luck!
Thank you for sharing your time and work Leo, Mike, and Nathan, could you include the link to your 1st meeting, in the description, please, have a great day, peace
@@CognitiveRevolutionPodcast looking at it now and I've seen it previously, I gather a lot of Mike's work and podcasts in a library I've been building for several years now, just wanted to double check I had it there, thank you Nathan, have a great day, peace
"We don't want to micromanage everything, we want to give very simple signals." Simple signals = complexity & emergence. This could apply to societal systems as well.
I wonder whether all of this can be extended to neuroscience as well. Namely treating the brain in isolation as a unit to which the meta-cognitive intelligence that Michael is talking about can be applied. For instance you have all this treasure trove of data on how the brain responds to different drugs, think clinical trials, which can be used to form a vector space along with an energy function that governs the dynamics of the different vectors that represent different people. So that upon the introduction of a new vector one has an idea of how it would evolve under the influence of some drug or drugs. And then again, why only neuroscience, everything can be treated this way.
It is a great insight that any organism will be resistant to anything we try, even if administer a potential cure. The way we must administer is by using transformer models to interpret, translate, and negotiate a solution with the cells. Comfort, convenience, and accessibility to guide the cells accepting administration. Working with the natural flow and function instead of going against it. This requires our transformer systems to be well trained on real time continuous live data of the cells and their behavior during all kinds of situations, and while giving administration.
This is fascinating in a good and scary way. I can only imagine the kinds of biological weapons of the future revolving around this research. How much of a "disruption" in the networking between cells is needed to cause just enough harm to R.I.P. someone. And how easily detectable is it.
They are. There are new clinical trials for different types of cancers constantly. Several cancers now have treatments so successful that almost nobody dies of them if it’s caught early enough. Cancer isn’t a singular disease, it’s a category of disease that contains many different diseases
@@happyday2912 google cancer clinical trials near me and you’ll find a bunch of them being run out of hospitals within a few miles of you, let alone at the biggest research hospitals across the country
I liked Data Limitations part of the talk (00:36:42), something that actually makes sense. But most other things are buzzword gibberish: LLMS, embeddings, networks, consciousness, theory of mind, bioelectricity lol. Isn't GABA a chief inhibitory neurotransmitter in the vertebrate nervous system, used in about 40% of all human neurons? If so, then when someone finds a causal link between GABA and cancer, how meaningful is that? Isn't it like saying that life and cancer are connected and we've spent N years coming to that conclusion?
especially in Nature 'action minus reward' seems to be the unifying thought behind collective intelligence as seen in symbiosis and synergistic geometries like the tetrahedral shape of amino acids.
I suspect the thing we shouldn’t do is basically Krang from TMNT. Organoid interfaced with an LLM. Need reference frames, continuous feedback, and integration of world model. I also agree we shouldn’t do it.
Omg, I have been saying this for a while now. AI super power is seeing patterns in massive amounts of information. We are all information processors, but with AI, we have an elite pattern and trend finder that can handle large amounts of data. Data that range from multiple different topics and fields, where we could not possible comprehend existing connections. We gotta stop trying to make AI like us and start maximizing it for what it is good at.
You are correct. Look up Wolfram Cellular Automaton Rule 30. It is the template for how we comprehend this universe/ framework. When calculated to about 1000 steps a strange pattern is revealed that looks like the golden ratio. To the left of the angle is the visually coherent pattern, and to the right it looks random, when in fact it is just a pattern we can't recognize and we named it "random".
Prof. Michael Levin when he is discussing "collective intelligence" of cells having the same goals, he is describing "stopgaps" that each cell has! There is no "collective intelligence", it just appears that way!
Lmao Take plants that rotate leaves to maximize sunlight energy from window. To understand in what direction to rotate they either have to have an optical system (no) Or be able to calculate based on distributed signals from many different cells Or have memory while making rotation tryouts (?) Also for the specific mass form and shape of organs all cells have to have information about all other cells… Enormous amount of data.
54:11 Hmmm ... Very curious to know what this secret sauce is that Dr. Levin is afraid to share. Would be sad if it is something very useful and whatever it is he fears is not as terrible as he thinks it is. Hopefully he will change his mind and help advance our knowledge. It's hard to tell if it's just a philosophical or religious shackle or if it is truly dangerous in some respect. Or could be a big yawner. Don't usually get stuff like this Dr. Levin. Maybe we should talk it out with Joscha Bach to figure out how dangerous (?) it really is? It sounds ominous ... 😳
Aside from being interviewed by one of the Little Rascals or someone's 'cool grandma' -- that hat style gotta go, my dude. ffs. Nevertheless, excellent interviews!
Your thumbnail is bad. It looks like a statsquest video, but it's a podcast... very confusing. I didn't click this video for weeks despite youtube consistently showing it to me.
Your beginning monologue is way too long if you're going to show us hot clips from the podcast at the beginning. It sounds like you're reading from a script, which makes what you're saying sound canned and boring. Introducing the people whilst they're on the call is much more fluid and pleasing.
Michael Levin is one of the smartest people out there
True.
Levin is the only guy I've heard who has made me see the possibility of what seems to be a genuine new and actionable paradigm for a deeper understanding of life and intelligence.
Stunning as always!🎉 I'm waiting for the dream collab: Wolfram + Levin !
I deeply believe that if the Weinstein Bros, Wolfram and Levin got together with a solid science model they could break the current paradigm.
Always great to see Professor Michael Levin pop up. 🙏👍
🙏🏽Amen 😊
Thanks. Prof Levin is one of a kind. Always a learning experience and a to do list for days after listening to him.
Michael is always insightful and pithy in his responses and was very 'kind' in his reply on the question about a 'wellness' paradigm.
I am at the 48:07 mark, 'listening lightly', doodling Levin's insights into latticing thoughts of my own... I will listen a second time when I finish today... let it steep... and listen raggedly a third time and take notes that wander off in my mind last. After this phase I put down my notes and wait for Levin's and Lopez's ideas to comeback to me latticed in my thoughts and far more understandable.
Subscribed. I could not agree more with your guest about the current absurd concept of "alignment"; given that we don't align with ourselves ... whether at the social scale, or even individually. I've been saying exactly that in comments under dozens of videos on AI. *Alignment* is the mother of all red herrings. The whole point of it is to cover the corporate rears; they hire a few "alignment specialists" and tell them _"Try to catch up with us, as we move full speed ahead with this technology; and good luck telling us something we like to hear and is cheap to implement."_
Just so that if one day their AI takes over the world they can say *_"Sorry, your honor; we did TRY ..."_*
Great episode. Would it be possible to link the specific papers being discussed?
Updated the show notes with the links. Thanks!
I ❤ Dr. Michael Levin
When I was the executor of my father’s estate I managed his 30+ properties. I had some grand ideas which incorporated an entire hillside of the city he had his properties.
Redo the entire hillside with a self contained, energy efficient, living area with replaceable /interchangeable living areas entirely energized by solar, wind water and most importantly my invention of mechanical power energy production. Gardens, aquaponics, etc.
one could theoretically move across country taking your living area and putting it into a different building in any location. One and done, transformative transforming.
On another consideration:
Why is the metamorphosis of a caterpillar to a butterfly is a power function of zeta?
Protein folding: first you have to understand what is going on in a cell, what is a protein and why it is doing what it is doing.
Prions first.
Proteins run the show.
Good luck!
do it. we need more stuff like this, if u have issues with getting started, use ai. i've pretty much adopted ai for my executive dysfunction.
Thank you for sharing your time and work Leo, Mike, and Nathan, could you include the link to your 1st meeting, in the description, please, have a great day, peace
ua-cam.com/video/LYyGG9xXpPA/v-deo.html
@@CognitiveRevolutionPodcast looking at it now and I've seen it previously, I gather a lot of Mike's work and podcasts in a library I've been building for several years now, just wanted to double check I had it there, thank you Nathan, have a great day, peace
"We don't want to micromanage everything, we want to give very simple signals." Simple signals = complexity & emergence. This could apply to societal systems as well.
I wonder whether all of this can be extended to neuroscience as well. Namely treating the brain in isolation as a unit to which the meta-cognitive intelligence that Michael is talking about can be applied. For instance you have all this treasure trove of data on how the brain responds to different drugs, think clinical trials, which can be used to form a vector space along with an energy function that governs the dynamics of the different vectors that represent different people. So that upon the introduction of a new vector one has an idea of how it would evolve under the influence of some drug or drugs. And then again, why only neuroscience, everything can be treated this way.
Great guests and subject, my dude.
It is a great insight that any organism will be resistant to anything we try, even if administer a potential cure. The way we must administer is by using transformer models to interpret, translate, and negotiate a solution with the cells. Comfort, convenience, and accessibility to guide the cells accepting administration. Working with the natural flow and function instead of going against it. This requires our transformer systems to be well trained on real time continuous live data of the cells and their behavior during all kinds of situations, and while giving administration.
Goddamn Levin is my current favorite biologist, and thats saying a lot
Levin is a goat 🐐
what a chat, that really got going ❤🔥
Consider checking out Karen Wong and her the Meaning Code channel. She’s had on Michael Levin multiple times
Purpose and goals in a telepathic environment without needs for myself.
This is fascinating in a good and scary way. I can only imagine the kinds of biological weapons of the future revolving around this research.
How much of a "disruption" in the networking between cells is needed to cause just enough harm to R.I.P. someone. And how easily detectable is it.
Many articles and resources about curing cancer for these past years, but why their implementations never seen in hospitals
They are. There are new clinical trials for different types of cancers constantly. Several cancers now have treatments so successful that almost nobody dies of them if it’s caught early enough. Cancer isn’t a singular disease, it’s a category of disease that contains many different diseases
@@Jm-wt1fs can you name the hospitals that already implemented the treatments?
@@happyday2912 google cancer clinical trials near me and you’ll find a bunch of them being run out of hospitals within a few miles of you, let alone at the biggest research hospitals across the country
I liked Data Limitations part of the talk (00:36:42), something that actually makes sense.
But most other things are buzzword gibberish: LLMS, embeddings, networks, consciousness, theory of mind, bioelectricity lol.
Isn't GABA a chief inhibitory neurotransmitter in the vertebrate nervous system, used in about 40% of all human neurons? If so, then when someone finds a causal link between GABA and cancer, how meaningful is that? Isn't it like saying that life and cancer are connected and we've spent N years coming to that conclusion?
especially in Nature 'action minus reward' seems to be the unifying thought behind collective intelligence as seen in symbiosis and synergistic geometries like the tetrahedral shape of amino acids.
I suspect the thing we shouldn’t do is basically Krang from TMNT. Organoid interfaced with an LLM. Need reference frames, continuous feedback, and integration of world model. I also agree we shouldn’t do it.
This is huge!
Whatever future you'd like to imagine, I strongly suggest one first have deep self/ no self buddhist-type understanding of who you even are.
Did that guy talk about how giving sells muscimol makes them forget what they're supposed to do? I take muskimol allot. And that freaked me out.
One learns from mistakes,,,,through our own agency,,,,use it or loose it,,,,
Omg, I have been saying this for a while now. AI super power is seeing patterns in massive amounts of information. We are all information processors, but with AI, we have an elite pattern and trend finder that can handle large amounts of data. Data that range from multiple different topics and fields, where we could not possible comprehend existing connections. We gotta stop trying to make AI like us and start maximizing it for what it is good at.
You are correct. Look up Wolfram Cellular Automaton Rule 30. It is the template for how we comprehend this universe/ framework. When calculated to about 1000 steps a strange pattern is revealed that looks like the golden ratio. To the left of the angle is the visually coherent pattern, and to the right it looks random, when in fact it is just a pattern we can't recognize and we named it "random".
& so there probably isn't such a thing as 'random' . Very true.@merodobson
Prof. Michael Levin when he is discussing "collective intelligence" of cells having the same goals, he is describing "stopgaps" that each cell has! There is no "collective intelligence", it just appears that way!
Lmao
Take plants that rotate leaves to maximize sunlight energy from window.
To understand in what direction to rotate they either have to have an optical system (no)
Or be able to calculate based on distributed signals from many different cells
Or have memory while making rotation tryouts (?)
Also for the specific mass form and shape of organs all cells have to have information about all other cells…
Enormous amount of data.
@@extremumone Correct! Plants move slowly and move towards light and then stops. No enormous amount of data!
Tell the guy in hat to keep it concise
fair
Ooops ..Slip of the toungue at ..31: 46
54:11 Hmmm ... Very curious to know what this secret sauce is that Dr. Levin is afraid to share. Would be sad if it is something very useful and whatever it is he fears is not as terrible as he thinks it is. Hopefully he will change his mind and help advance our knowledge. It's hard to tell if it's just a philosophical or religious shackle or if it is truly dangerous in some respect. Or could be a big yawner. Don't usually get stuff like this Dr. Levin. Maybe we should talk it out with Joscha Bach to figure out how dangerous (?) it really is? It sounds ominous ... 😳
Aside from being interviewed by one of the Little Rascals or someone's 'cool grandma' -- that hat style gotta go, my dude. ffs. Nevertheless, excellent interviews!
AI development is truly the Holy thing to do.
so many neologisms, so little hard knowledge
Your thumbnail is bad. It looks like a statsquest video, but it's a podcast... very confusing. I didn't click this video for weeks despite youtube consistently showing it to me.
Creepy most podcasters upload video too. Boomer
Your beginning monologue is way too long if you're going to show us hot clips from the podcast at the beginning. It sounds like you're reading from a script, which makes what you're saying sound canned and boring.
Introducing the people whilst they're on the call is much more fluid and pleasing.
Deceptive bait title
Video would have been a lot better if ya edited out all the guy with hat scenes
I fast forward through his dribble quibble
Quite ridiculously boring
Bot or a huge weirdo lol. You forgot to switch profiles.
UnIncredible